Katie Jolly | January 21, 2021

New Year, Great Data: The Best Ookla Open Data Projects We’ve Seen So Far

When we announced Ookla® Open Datasets from Ookla For Good™ in October, we were hoping to see exciting projects that raise the bar on the conversation about internet speeds and accessibility — and you delivered. From analyses of internet inequity in the United States to measures of data affluence in India, today we’re highlighting four projects that really show what this data can do. We also have a new, simpler tutorial on how you can use this data for your own efforts to improve the state of networks worldwide.

Highlighting the digital divide in the U.S.

Jamie Saxon with the Center for Data and Computing at the University of Chicago married Ookla data on broadband performance with data from the American Community Survey to create interactive maps of the digital divide in 20 U.S. cities. These maps provide views into many variables that contribute to internet inequities.

Ookla_open_datasets_James_Saxon_0121-1

Building a data affluence map

Raj Bhagat P shows how different variables can be combined with this map of data affluence that combines data on internet speeds and device counts in India.

Ookla_open_datasets_Raj-Bhagat-P_0121-1

Internet speeds are beautiful

This map of fixed broadband speeds across Europe from Boris Mericskay shows that internet performance can be as visually stunning as a map of city lights.

Ookla_open_datasets_Boris-Mericskay_0121-1

Topi Tjunakov created a similar image of internet speeds in and around Japan.

Ookla_open_datasets_Topi-Tjunakov_0121-1

Use Ookla Open Datasets to make your own maps

This section will demonstrate a few possible ways to use Ookla Open Datasets using the United Kingdom as an example. The ideas can be adapted for any area around the world. This tutorial uses the R programming language, but there are also Python tutorials available in the Ookla Open Data GitHub repository.

library(tidyverse)
library(patchwork)
library(janitor)
library(ggrepel)
library(usethis)
library(lubridate)
library(colorspace)
library(scales)
library(kableExtra)
library(knitr)
library(sf)

# colors for plots
purple <- "#A244DA"
light_purple <- colorspace::lighten("#A244DA", 0.5)
green <- colorspace::desaturate("#2DE5D1", 0.2)
blue_gray <- "#464a62"
mid_gray <- "#ccd0dd"
light_gray <- "#f9f9fd"

# set some global theme defaults
theme_set(theme_minimal())
theme_update(text = element_text(family = "sans", color = "#464a62"))
theme_update(plot.title = element_text(hjust = 0.5, face = "bold"))
theme_update(plot.subtitle = element_text(hjust = 0.5))

Ookla Open Datasets include quarterly performance and test count data for both mobile networks and fixed broadband aggregated over all providers. The tests are binned into global zoom level 16 tiles which can be thought of as roughly a few football fields. As of today, all four quarters of 2020 are available and subsequent quarters will be added as they complete.

Administrative unit data

I chose to analyse the mobile data at the Nomenclature of Territorial Units for Statistics (NUTS) 3 level (1:1 million). These administrative units are maintained by the European Union to allow for comparable analysis across member states. NUTS 3 areas mean:

In England, upper tier authorities and groups of unitary authorities and districts
In Wales, groups of Principal Areas
In Scotland, groups of Council Areas or Islands Areas
In Northern Ireland, groups of districts

To make a comparison to the U.S. administrative structure, these can be roughly thought of as the size of counties. Here is the code you’ll want to use to download the NUTS shapefiles from the Eurostat site. Once the zipfile is downloaded you will need to unzip it again in order to read it into your R environment:

# create a directory called “data”
dir.create("data")
use_zip("https://gisco-services.ec.europa.eu/distribution/v2/nuts/download/ref-nuts-2021-01m.shp.zip", destdir = "data")

uk_nuts_3 <- read_sf("data/ref-nuts-2021-01m.shp/NUTS_RG_01M_2021_3857_LEVL_3.shp/NUTS_RG_01M_2021_3857_LEVL_3.shp") %>%
  filter(CNTR_CODE == "UK") %>%
  st_transform(4326) %>%
  clean_names() %>%
  mutate(urbn_desc = case_when( # add more descriptive labels for urban variable
    urbn_type == 1 ~ "Urban",
    urbn_type == 2 ~ "Intermediate",
    urbn_type == 3 ~ "Rural"
  ),
  urbn_desc = factor(urbn_desc, levels = c("Urban", "Intermediate", "Rural")))

# contextual city data
uk_cities <- read_sf("https://opendata.arcgis.com/datasets/6996f03a1b364dbab4008d99380370ed_0.geojson") %>%
  clean_names() %>%
  filter(fips_cntry == "UK", pop_rank <= 5)

ggplot(uk_nuts_3) +
  geom_sf(color = mid_gray, fill = light_gray, lwd = 0.08) +
  geom_text_repel(data = uk_cities, 
                           aes(label = city_name, geometry = geometry), 
                           family = "sans", 
                           color = blue_gray, 
                           size = 2.2, 
                           stat = "sf_coordinates",
                           min.segment.length = 2) +
  labs(title = "United Kingdom",
       subtitle = "NUTS 3 Areas") +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank(),
        axis.text = element_blank(),
        axis.title = element_blank())

plot_uk-1-1

Adding data from Ookla Open Datasets

You’ll want to crop the global dataset to the bounding box of the U.K. This will include some extra tiles (within the box but not within the country, i.e. some of western Ireland), but it makes the data much easier to work with later on.

uk_bbox <- uk_nuts_3 %>%
  st_union() %>% # otherwise would be calculating the bounding box of each individual area
  st_bbox()

Each of the quarters are stored in separate shapefiles. You can read them in one-by-one and crop them to the U.K. box in the same pipeline.

# download the data with the following code:

use_zip("https://ookla-open-data.s3.amazonaws.com/shapefiles/performance/type=mobile/year=2020/quarter=1/2020-01-01_performance_mobile_tiles.zip", destdir = "data")
use_zip("https://ookla-open-data.s3.amazonaws.com/shapefiles/performance/type=mobile/year=2020/quarter=2/2020-04-01_performance_mobile_tiles.zip", destdir = "data")
use_zip("https://ookla-open-data.s3.amazonaws.com/shapefiles/performance/type=mobile/year=2020/quarter=3/2020-07-01_performance_mobile_tiles.zip", destdir = "data")
use_zip("https://ookla-open-data.s3.amazonaws.com/shapefiles/performance/type=mobile/year=2020/quarter=4/2020-10-01_performance_mobile_tiles.zip", destdir = "data")

# and then read in those downloaded files
mobile_tiles_q1 <- read_sf("data/2020-01-01_performance_mobile_tiles/gps_mobile_tiles.shp") %>%
  st_crop(uk_bbox)
mobile_tiles_q2 <- read_sf("data/2020-04-01_performance_mobile_tiles/gps_mobile_tiles.shp") %>%
  st_crop(uk_bbox)
mobile_tiles_q3 <- read_sf("data/2020-07-01_performance_mobile_tiles/gps_mobile_tiles.shp") %>%
  st_crop(uk_bbox)
mobile_tiles_q4 <- read_sf("data/2020-10-01_performance_mobile_tiles/gps_mobile_tiles.shp") %>%
  st_crop(uk_bbox)

As you see, the tiles cover most of the area, with more tiles in more densely populated areas. (And note that you still have tiles included that are outside the boundary of the area but within the bounding box.)

ggplot(uk_nuts_3) +
  geom_sf(color = mid_gray, fill = light_gray, lwd = 0.08) +
  geom_sf(data = mobile_tiles_q4, fill = purple, color = NA) +
  geom_text_repel(data = uk_cities, 
                           aes(label = city_name, geometry = geometry), 
                           family = "sans", 
                           color = blue_gray, 
                           size = 2.2, 
                           stat = "sf_coordinates",
                           min.segment.length = 2) +
  labs(title = "United Kingdom",
       subtitle = "Ookla® Open Data Mobile Tiles, NUTS 3 Areas") +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank(),
        axis.text = element_blank(),
        axis.title = element_blank())

tile_map-1-3

Now that the cropped tiles are read in, you’ll use a spatial join to determine which NUTS 3 area each tile is in. In this step, I am also reprojecting the data to the British National Grid (meters). I’ve also added a variable to identify the time period (quarter).

tiles_q1_nuts <- uk_nuts_3 %>%
  st_transform(27700) %>% # British National Grid
  st_join(mobile_tiles_q1 %>% st_transform(27700), left = FALSE) %>%
  mutate(quarter_start = "2020-01-01")

tiles_q2_nuts <- uk_nuts_3 %>%
  st_transform(27700) %>%
  st_join(mobile_tiles_q2 %>% st_transform(27700), left = FALSE) %>%
  mutate(quarter_start = "2020-04-01")

tiles_q3_nuts <- uk_nuts_3 %>%
  st_transform(27700) %>%
  st_join(mobile_tiles_q3 %>% st_transform(27700), left = FALSE) %>%
  mutate(quarter_start = "2020-07-01")

tiles_q4_nuts <- uk_nuts_3 %>%
  st_transform(27700) %>%
  st_join(mobile_tiles_q4 %>% st_transform(27700), left = FALSE) %>%
  mutate(quarter_start = "2020-10-01")

In order to make the data easier to work with, combine the tiles into a long dataframe with each row representing one tile in one quarter. The geometry now represents the NUTS region, not the original tile shape.

tiles_all <- tiles_q1_nuts %>%
  rbind(tiles_q2_nuts) %>%
  rbind(tiles_q3_nuts) %>%
  rbind(tiles_q4_nuts) %>%
  mutate(quarter_start = ymd(quarter_start)) # convert to date format

With this dataframe, you can start to generate some aggregates. In this table you’ll include the tile count, test count, quarter and average download and upload speeds.

Exploratory data analysis

aggs_quarter <- tiles_all %>%
  st_set_geometry(NULL) %>%
  group_by(quarter_start) %>%
  summarise(tiles = n(),
            avg_d_mbps = weighted.mean(avg_d_kbps / 1000, tests), # I find Mbps easier to work with
            avg_u_mbps = weighted.mean(avg_u_kbps / 1000, tests),
            tests = sum(tests)) %>%
  ungroup()


knitr::kable(aggs_quarter) %>%
  kable_styling()

aggregates_table_kj

We can see from this table that both download and upload speeds increased throughout the year, with a small dip in upload speeds in Q2. Next, you’ll want to plot this data.

ggplot(aggs_quarter, aes(x = quarter_start)) +
  geom_point(aes(y = avg_d_mbps), color = purple) +
  geom_line(aes(y = avg_d_mbps), color = purple, lwd = 0.5) +
  geom_text(aes(y = avg_d_mbps - 2, label = round(avg_d_mbps, 1)), color = purple, size = 3, family = "sans") +
  geom_text(data = NULL, x = ymd("2020-02-01"), y = 47, label = "Download speed", color = purple, size = 3, family = "sans") +
  geom_point(aes(y = avg_u_mbps), color = light_purple) +
  geom_line(aes(y = avg_u_mbps), color = light_purple, lwd = 0.5) +
  geom_text(aes(y = avg_u_mbps - 2, label = round(avg_u_mbps, 1)), color = light_purple, size = 3, family = "sans") +
  geom_text(data = NULL, x = ymd("2020-02-05"), y = 14, label = "Upload speed", color = light_purple, size = 3, family = "sans") +
  labs(y = "", x = "Quarter start date",
       title = "Mobile Network Performance, U.K.",
       subtitle = "Ookla® Open Datasets | 2020") +
  theme(panel.grid.minor = element_blank(),
        panel.grid.major = element_blank(),
        axis.title.x = element_text(hjust=1)) +
  scale_y_continuous(labels = label_number(suffix = " Mbps", scale = 1, accuracy = 1)) +
  scale_x_date(date_labels = "%b %d")

line_up_down-1

Examining test counts

We also saw above that the number of tests decreased between Q1 and Q2 and then peaked in Q3 at a little over 700,000 before coming back down. The increase likely followed resulted from interest in network performance during COVID-19 when more people started working from home. This spike is even more obvious in chart form.

ggplot(aggs_quarter, aes(x = quarter_start)) +
  geom_point(aes(y = tests), color = purple) +
  geom_line(aes(y = tests), color = purple, lwd = 0.5) +
  geom_text(aes(y = tests - 6000, label = comma(tests), x= quarter_start + 5), size = 3, color = purple) +
  labs(y = "", x = "Quarter start date",
       title = "Mobile Test Count, U.K.",
       subtitle = "Ookla® Open Datasets | 2020") +
  theme(panel.grid.minor = element_blank(),
        panel.grid.major = element_blank(),
        axis.title.x = element_text(hjust=1),
        axis.text = element_text(color = blue_gray)) +
  scale_y_continuous(labels = comma) +
  scale_x_date(date_labels = "%b %d")

line_tests-1-1

Data distribution

Next, I wanted to check the distribution of average download speeds.

ggplot(tiles_all) + 
  geom_histogram(aes(x = avg_d_kbps / 1000, group = quarter_start), size = 0.3, color = light_gray, fill = green) + 
  scale_x_continuous(labels = label_number(suffix = " Mbps", accuracy = 1)) +
  scale_y_continuous(labels = comma) +
  facet_grid(quarter_start ~ .) +
  theme(panel.grid.minor = element_blank(), 
        panel.grid.major = element_blank(), 
        axis.title.x = element_text(hjust=1),
        axis.text = element_text(color = blue_gray),
        strip.text.y = element_text(angle = 0, color = blue_gray)) + 
  labs(y = "", x = "", title = "Mobile Download Speed Distribution by Tile, U.K.", 
       subtitle = "Ookla® Open Datasets | 2020")

histogram-1-1

The underlying distribution of average download speeds across the tiles has stayed fairly stable.

Mapping average speed

Making a quick map of the average download speed in each region across the U.K. is relatively simple.

# generate aggregates table
nuts_3_aggs <- tiles_all %>%
  group_by(quarter_start, nuts_id, nuts_name, urbn_desc, urbn_type) %>%
  summarise(tiles = n(),
            avg_d_mbps = weighted.mean(avg_d_kbps / 1000, tests), # I find Mbps easier to work with
            avg_u_mbps = weighted.mean(avg_u_kbps / 1000, tests),
            tests = sum(tests)) %>%
  ungroup()

ggplot(nuts_3_aggs %>% filter(quarter_start == "2020-10-01")) +
  geom_sf(aes(fill = avg_d_mbps), color = blue_gray, lwd = 0.08) +
  scale_fill_stepsn(colors = RColorBrewer::brewer.pal(n = 5, name = "BuPu"), labels = label_number(suffix = " Mbps"), n.breaks = 4, guide = guide_colorsteps(title = "")) +
  theme(panel.grid.minor = element_blank(),
        panel.grid.major = element_blank(),
        axis.title.x = element_text(hjust=1),
        legend.text = element_text(color = blue_gray),
        axis.text = element_blank()) +
  labs(title = "Mobile Download Speed, U.K.", subtitle = "Ookla® Open Datasets | Q4 2020")

choropleth-1-1

As you can see, the areas around large cities have faster download speeds on average and the lowest average download speeds are typically in more rural areas.

Rural and urban analysis

People are often interested in the difference between mobile networks in urban and rural areas. The Eurostat NUTS data includes an urban indicator with three levels: rural, intermediate and urban. This typology is determined primarily by population density and proximity to a population center.

ggplot(uk_nuts_3) +
  geom_sf(aes(fill = urbn_desc), color = light_gray, lwd = 0.08) +
  geom_text_repel(data = uk_cities, 
                           aes(label = city_name, geometry = geometry), 
                           family = "sans", 
                           color = "#1a1b2e", 
                           size = 2.2, 
                           stat = "sf_coordinates",
                           min.segment.length = 2) +
  scale_fill_manual(values = c(purple, light_purple, green), name = "", guide = guide_legend(direction = "horizontal", label.position = "top", keywidth = 3, keyheight = 0.5)) +
  labs(title = "U.K., NUTS 3 Areas") +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank(),
        axis.text = element_blank(),
        axis.title = element_blank(),
        legend.position = "top")

rural_urban_reference-1

Data distribution overall and over time

When you aggregate by the urban indicator variable different patterns come up in the data.

# generate aggregates table
rural_urban_aggs <- tiles_all %>%
  st_set_geometry(NULL) %>%
  group_by(quarter_start, urbn_desc, urbn_type) %>%
  summarise(tiles = n(),
            avg_d_mbps = weighted.mean(avg_d_kbps / 1000, tests), # I find Mbps easier to work with
            avg_u_mbps = weighted.mean(avg_u_kbps / 1000, tests),
            tests = sum(tests)) %>%
  ungroup()

As you might expect, the download speeds during Q4 are faster in urban areas than in rural areas – with the intermediate ones somewhere in between. This pattern holds for other quarters as well.

ggplot(rural_urban_aggs %>% filter(quarter_start == "2020-10-01"), aes(x = avg_d_mbps, y = urbn_desc, fill = urbn_desc)) +
  geom_col(width = .3, show.legend = FALSE) +
  geom_jitter(data = nuts_3_aggs, aes(x = avg_d_mbps, y = urbn_desc, color = urbn_desc), size = 0.7) + 
  geom_text(aes(x = avg_d_mbps - 4, label = round(avg_d_mbps, 1)), family = "sans",  size = 3.5, color = blue_gray) +
  scale_fill_manual(values = c(purple, light_purple, green)) +
  scale_color_manual(values = darken(c(purple, light_purple, green))) +
  scale_x_continuous(labels = label_number(suffix = " Mbps", scale = 1, accuracy = 1)) +
  theme(panel.grid.minor = element_blank(),
        panel.grid.major = element_blank(),
        axis.title.x = element_text(hjust=1),
        legend.position = "none",
        axis.text = element_text(color = blue_gray)) +
  labs(y = "", x = "", 
       title = "Mobile Download Speed Distribution by NUTS 3 Area, U.K.", 
       subtitle = "Ookla® Open Datasets | 2020")

rural_urban_bar-1-2
Interestingly though, the patterns differ when you look at a time series plot. Urban mobile networks steadily improve, while the intermediate and rural areas saw slower average download speeds starting in Q2 before going back up after Q3. This is likely the result of increased pressure on the networks during stay-at-home orders (although this graph is not conclusive evidence of that).

ggplot(rural_urban_aggs) +
  geom_line(aes(x = quarter_start, y = avg_d_mbps, color = urbn_desc)) +
  geom_point(aes(x = quarter_start, y = avg_d_mbps, color = urbn_desc)) +
  # urban label
  geom_text(data = NULL, x = ymd("2020-02-01"), y = 50, label = "Urban", color = purple, family = "sans", size = 3) +
  # intermediate label
  geom_text(data = NULL, x = ymd("2020-02-15"), y = 35, label = "Intermediate", color = light_purple, family = "sans", size = 3) +
  # rural label
  geom_text(data = NULL, x = ymd("2020-01-15"), y = 26, label = "Rural", color = green, family = "sans", size = 3) +
  scale_color_manual(values = c(purple, light_purple, green)) +
  scale_x_date(date_labels = "%b %d") +
  scale_y_continuous(labels = label_number(suffix = " Mbps", scale = 1, accuracy = 1)) +
  theme(panel.grid.minor = element_blank(),
        panel.grid.major = element_blank(),
        axis.title.x = element_text(hjust=1),
        legend.position = "none",
        axis.text = element_text(color = blue_gray)) +
  labs(y = "", x = "Quarter start date", 
       title = "Mobile Download Speed by NUTS 3 Urban-Rural Type, U.K.", 
       subtitle = "Ookla® Open Datasets | 2020")

rural_urban_line-1-1

When you repeat the same plot but map the test count to the site of the point, you can see why the overall download speed increased steadily. The number of tests in urban areas is much higher than in intermediate and rural areas, thus pulling up the overall average.

ggplot(rural_urban_aggs) +
  geom_line(aes(x = quarter_start, y = avg_d_mbps, color = urbn_desc)) +
  geom_point(aes(x = quarter_start, y = avg_d_mbps, color = urbn_desc, size = tests)) +
  # urban label
  geom_text(data = NULL, x = ymd("2020-02-01"), y = 50, label = "Urban", color = purple, family = "sans", size = 3) +
  # intermediate label
  geom_text(data = NULL, x = ymd("2020-02-15"), y = 35, label = "Intermediate", color = light_purple, family = "sans", size = 3) +
  # rural label
  geom_text(data = NULL, x = ymd("2020-01-15"), y = 26, label = "Rural", color = green, family = "sans", size = 3) +
  scale_color_manual(values = c(purple, light_purple, green)) +
  scale_x_date(date_labels = "%b %d") +
  scale_y_continuous(labels = label_number(suffix = " Mbps", scale = 1, accuracy = 1)) +
  theme(panel.grid.minor = element_blank(),
        panel.grid.major = element_blank(),
        axis.title.x = element_text(hjust=1),
        legend.position = "none",
        axis.text = element_text(color = blue_gray)) +
  labs(y = "", x = "Quarter start date", 
       title = ("Mobile Download Speed by NUTS 3 Urban-Rural Type, U.K."), 
       subtitle = "Ookla® Open Datasets | 2020",
       caption = "Circle size indicates test count")

rural_urban_line_size-1-1

Spotlighting regional variances

Parsing the data by specific geographies can reveal additional information.

bottom_20_q4 <- nuts_3_aggs %>% 
  filter(quarter_start == "2020-10-01") %>% 
  top_n(n = -20, wt = avg_d_mbps) %>%
  mutate(nuts_name = fct_reorder(factor(nuts_name), -avg_d_mbps))

map <- ggplot() +
  geom_sf(data = uk_nuts_3, fill = light_gray, color = mid_gray, lwd = 0.08) +
  geom_sf(data = bottom_20_q4, aes(fill = urbn_desc), color = mid_gray, lwd = 0.08, show.legend = FALSE) +
  geom_text_repel(data = uk_cities, 
                           aes(label = city_name, geometry = geometry), 
                           family = "sans", 
                           color = blue_gray, 
                           size = 2.2, 
                           stat = "sf_coordinates",
                           min.segment.length = 2) +
  scale_fill_manual(values = c(purple, light_purple, green), name = "", guide = guide_legend(direction = "horizontal", label.position = "top", keywidth = 3, keyheight = 0.5)) +
  labs(title = NULL,
       subtitle = NULL) +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank(),
        axis.text = element_blank(),
        axis.title = element_blank(),
        legend.position = "top")

barplot <- ggplot(data = bottom_20_q4, aes(x = avg_d_mbps, y = nuts_name, fill = urbn_desc)) +
  geom_col(width = .5) +
  scale_fill_manual(values = c(purple, light_purple, green), guide = guide_legend(direction = "horizontal", label.position = "top", keywidth = 3, keyheight = 0.5, title = NULL)) +
  scale_x_continuous(labels = label_number(suffix = " Mbps", scale = 1, accuracy = 1)) +
  theme(panel.grid.minor = element_blank(),
        panel.grid.major = element_blank(),
        axis.title.x = element_text(hjust=1),
        legend.position = "top",
        axis.text = element_text(color = blue_gray)) +
  labs(y = "", x = "", 
       title = ("Slowest 20 NUTS 3 Areas by Download Speed, U.K."), 
       subtitle = "Ookla® Open Datasets | Q4 2020")

# use patchwork to put it all together
barplot + map

bottom_20-1-2
Among the 20 areas with the lowest average download speed in Q4 2020 there were three urban areas and six intermediate. The rest were rural.

top_20_q4 <- nuts_3_aggs %>% 
  filter(quarter_start == "2020-10-01") %>% 
  top_n(n = 20, wt = avg_d_mbps) %>%
  mutate(nuts_name = fct_reorder(factor(nuts_name), avg_d_mbps))

top_map <- ggplot() +
  geom_sf(data = uk_nuts_3, fill = light_gray, color = mid_gray, lwd = 0.08) +
  geom_sf(data = top_20_q4, aes(fill = urbn_desc), color = mid_gray, lwd = 0.08, show.legend = FALSE) +
  geom_text_repel(data = uk_cities, 
                           aes(label = city_name, geometry = geometry), 
                           family = "sans", 
                           color = blue_gray, 
                           size = 2.2, 
                           stat = "sf_coordinates",
                           min.segment.length = 2) +
  scale_fill_manual(values = c(purple, light_purple, green), name = "", guide = guide_legend(direction = "horizontal", label.position = "top", keywidth = 3, keyheight = 0.5)) +
  labs(title = NULL,
       subtitle = NULL) +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank(),
        axis.text = element_blank(),
        axis.title = element_blank(),
        legend.position = "top")

top_barplot <- ggplot(data = top_20_q4, aes(x = avg_d_mbps, y = nuts_name, fill = urbn_desc)) +
  geom_col(width = .5) +
  scale_fill_manual(values = c(purple, light_purple, green), guide = guide_legend(direction = "horizontal", label.position = "top", keywidth = 3, keyheight = 0.5, title = NULL)) +
  scale_x_continuous(labels = label_number(suffix = " Mbps", scale = 1, accuracy = 1), breaks = c(50, 100)) +
  theme(panel.grid.minor = element_blank(),
        panel.grid.major = element_blank(),
        axis.title.x = element_text(hjust=1),
        legend.position = "top",
        axis.text = element_text(color = blue_gray)) +
  labs(y = "", x = "", 
       title = "Fastest 20 NUTS 3 Areas by Mobile Download Speed, U.K.", 
       subtitle = "Ookla® Open Datasets | Q4 2020")

top_london <- ggplot() +
  geom_sf(data = uk_nuts_3 %>% filter(str_detect(fid, "UKI")), fill = light_gray, color = mid_gray, lwd = 0.08) +
  geom_sf(data = top_20_q4 %>% filter(str_detect(nuts_id, "UKI")), aes(fill = urbn_desc), color = mid_gray, lwd = 0.08, show.legend = FALSE) +
  geom_text_repel(data = uk_cities %>% filter(city_name == "London"), 
                           aes(label = city_name, geometry = geometry), 
                           family = "sans", 
                           color = "black", 
                           size = 2.2, 
                           stat = "sf_coordinates",
                           min.segment.length = 2) +
  scale_fill_manual(values = c(purple, light_purple, green), name = "", guide = guide_legend(direction = "horizontal", label.position = "top", keywidth = 3, keyheight = 0.5)) +
  labs(title = NULL,
       subtitle = NULL) +
  theme(panel.grid.major = element_blank(),
        panel.grid.minor = element_blank(),
        axis.text = element_blank(),
        axis.title = element_blank(),
        legend.position = "top",
        panel.border = element_rect(colour = blue_gray, fill=NA, size=0.5))

top_map_comp <- top_map + inset_element(top_london, left = 0.6, bottom = 0.6, right = 1, top = 1)

top_barplot + top_map_comp

top_20-1-1
Meanwhile, all of the fastest 20 NUTS 3 areas were urban.

What else you can do with this data

Don’t forget there are also more tutorials with examples written in Python and R. Aside from what I showed here, you could do an interesting analysis looking at clustering patterns, sociodemographic variables and other types of administrative units like legislative or school districts.

We hope this tutorial will help you use Ookla’s open data for your own projects. Please tag us if you share your projects on social media using the hashtag #OoklaForGood so we can learn from your analyses.

Ookla retains ownership of this article including all of the intellectual property rights, data, content graphs and analysis. This article may not be quoted, reproduced, distributed or published for any commercial purpose without prior consent. Members of the press and others using the findings in this article for non-commercial purposes are welcome to publicly share and link to report information with attribution to Ookla.

About the Author

Katie Jolly

| More Articles

Ookla Research™ Articles

Isla McKetta | March 14, 2019

Ditch the Lag: Cities with Great Gaming Culture and Low Ping

Yes, you can game from anywhere with an internet connection. But if you’re at all competitive, it’s nice to play from somewhere with low ping and fast internet speeds. Plus when you need to leave the house, it’s extra nice to know you’re also surrounded by gamer culture. We’ve examined February 2019 Speedtest results in 35 cities that are known for their esports events, gaming conferences, game companies and more to find out who has the advantage and ranked them based on their ping.

The top contenders

Eleven_Gaming_Cities_0219

First place Bucharest, Romania is home to super-low ping, a lightning fast download speed and a thriving gaming culture. From Bucharest Gaming Week (which includes the CS:GO Southeast Europe Championship and the FIFA National Tournament) to their numerous local game studios, Bucharest is a great place to be a gamer whether you’re online or out and about.

The next five gaming cities with the lowest pings are all in Asia. Hangzhou, China comes in second overall with a fast ping and world-class download speeds. This city is so devoted to its gamers that it opened a $280 million gaming “city” in 2018 and plans 14 new esports arenas before 2022. Coming in third, Chengdu, China has an equally low ping to our first two contenders and serves as one of two host locations in China for the Global Mobile Game Confederation (GMGC). Both Hangzhou and Chengdu are also franchise holders in the Overwatch League, giving local gaming fans something to cheer about. Fourth place Singapore, host of the 5th Annual GameStart Convention in October 2018, had only a slightly slower ping than the first four cities and the fastest download speed of any of the cities we considered.

South Korea is home to the fifth and sixth best cities for gamers. A satellite city of Seoul, Seongnam-si boasts the Pangyo Techno Valley (a.k.a. the Silicon Valley of Korea) and numerous game development companies. Perfect for a city with a 9 ms ping. Though Incheon’s ping was a little slower at 12 ms, gamers there can console themselves with the city’s gamer cred — the 2018 League of Legends World Championship was held in Incheon’s Munhak Stadium.

Coming in at number seven, Budapest, Hungary is an emerging game city, having hosted its first big esports event (the V4 Future Sports Festival) in 2018, but a 12 ms ping makes them a strong contender. More established Malmö, Sweden is number eight with a slightly slower average download speed but the city is headquarters to Massive Entertainment, creators of Tom Clancy’s The Division series, Far Cry 3, Assassin’s Creed: Revelations and many more.

Vancouver, Canada, North America’s only qualifier for the top gaming cities list, comes in at number nine with a 12 ms ping and many gaming companies including the Canadian arms of Nintendo of Canada and EA (Electronic Arts). We included both Shanghai, China and Moscow, Russia on the top gamer cities list as both had a 12 ms ping as well, though the internet speeds in Shanghai are superior. Shanghai will also host the International Dota 2 in 2019 while Moscow is known for Epicenter.

The rest of the pack

Notably absent from the list above is most of the western hemisphere. Cities in North America were held back by their high pings. Cities in South America suffered from high pings and also slow internet speeds — something that esports leagues have complained is a barrier to investment.

Our full list of gaming cities provides wider geographical representation, even if the internet performance is not always as stellar. You’ll find Los Angeles in 27th place, behind Seattle, Boston and Las Vegas. And São Paulo, Brazil has the best showing in Latin America at 23rd.

**Internet Performance in 35 Cities with a Gaming Culture**
Speedtest Results | February 2019
City	Ping (ms)	Mean Download (Mbps)	Mean Upload (Mbps)
Bucharest, Romania	8	172.13	126.57
Hangzhou, China	8	125.93	29.54
Chengdu, China	8	101.92	33.80
Singapore	9	196.43	200.08
Seongnam-si, South Korea	9	155.25	114.83
Incheon, South Korea	12	139.84	102.91
Budapest, Hungary	12	132.72	54.46
Malmö, Sweden	12	126.28	105.67
Vancouver, Canada	12	117.55	50.23
Shanghai, China	12	75.14	30.06
Moscow, Russia	12	64.56	63.59
Oslo, Norway	13	115.46	69.03
Hong Kong, Hong Kong (SAR)	14	167.59	161.14
Zürich, Switzerland	14	144.36	109.39
Seattle, United States	15	138.50	79.88
Stockholm, Sweden	15	134.16	93.83
Auckland, New Zealand	15	92.05	53.30
Toronto, Canada	16	134.75	67.42
Boston, United States	17	152.42	60.87
Las Vegas, United States	17	141.69	41.22
Chennai, India	17	48.40	42.93
Cologne, Germany	18	63.77	18.36
São Paulo, Brazil	18	46.43	21.57
Jakarta, Indonesia	18	17.88	10.21
Mumbai, India	19	23.40	19.26
Paris, France	20	161.04	93.68
Los Angeles, United States	20	121.00	23.57
London, United Kingdom	20	63.58	23.18
Rio de Janeiro, Brazil	20	36.50	13.33
Buenos Aires, Argentina	21	34.31	6.40
Katowice, Poland	22	83.99	20.91
Mexico City, Mexico	25	37.66	15.39
Sydney, Australia	25	34.20	9.61
Santiago, Chile	26	56.13	18.49
Tokyo, Japan	28	99.24	101.90

Of course, die-hard gamers will know that a low ping in your city won’t necessarily save you if you’re playing on a distant server.

What’s the ping like in your city? Take a Speedtest and see if your connection is hurting your gameplay.

About the Author

Isla McKetta

| More Articles

Ookla Research™ Articles

Mark Giles | July 17, 2023

48 New Ookla Market Reports Available for Q2 2023

Ookla® Market Reports™ identify key data about internet performance in countries across the world. This quarter we’ve provided updated analyses for 48 markets using Speedtest Intelligence® and summarized a few top takeaways below. Click through to the market report to see more details and charts about the countries you’re interested in, including the fastest fixed broadband providers and mobile operators, who had the most consistent service, and 5G and device performance in select countries during Q2 2023. Jump forward to a continent using these links:

Africa | Americas | Asia | Europe | Oceania

Africa

Cameroon: Speedtest Intelligence data showed no winner for fastest mobile operator in Cameroon during Q2 2023. blue had the lowest median mobile multi-server latency at 191 ms, while Douala had the fastest median mobile download speed among Cameroon’s most populous cities at 15.51 Mbps.
Ethiopia: Safaricom had the fastest median mobile download speed at 35.19 Mbps during Q2 2023. Safaricom also recorded the lowest median mobile multi-server latency at 42 ms, and highest Consistency of 89.4%. Of Ethiopia’s most populous cities, Gondar had the fastest median mobile download speed of 61.22 Mbps.
Tanzania: There were no winners over fastest mobile or fixed broadband in Tanzania during Q2 2023. Maisha Broadband registered the lowest median multi-server latency in Tanzania at 14 ms. Of Tanzania’s most populous cities, Dar es Salaam had the fastest median mobile download speed of 26.33 Mbps, while Mbeya had the fastest median fixed download speed of 21.32 Mbps.

Americas

Argentina: Personal had the fastest median download speed over mobile (35.05 Mbps) and lowest mobile multi-server latency (38 ms) during Q2 2023. In the fixed broadband market, Movistar recorded the fastest median download speed (98.37 Mbps) and lowest multi-server latency (12 ms). Among Argentina’s most populous cities, Buenos Aires recorded the fastest download speeds across mobile and fixed broadband networks.
Belize: Digi had the fastest median mobile download and upload speeds of 17.61 Mbps and 9.88 Mbps respectively during Q2 2023. It also recorded the highest Consistency of 79.8%. smart! recorded the lowest median mobile multi-server latency, of 67 ms. NEXGEN had the fastest median download and upload performance over fixed broadband in Belize at 48.65 Mbps and 47.38 Mbps respectively.
Canada: Bell was the fastest mobile operator in Canada with a median download speed of 116.59 Mbps in Q2 2023. Bell also had the fastest median 5G download speed at 208.05 Mbps. Rogers had the fastest median mobile upload speed of 13.29 Mbps, and the highest Consistency of 84.7%. Bell pure fibre was fastest for fixed broadband across both download (277.24 Mbps) and upload (235.27 Mbps) speeds. Of Canada’s most populous cities, St. John’s recorded the fastest median mobile download speed (214.29 Mbps) and Fredericton recorded the fastest median fixed download speed (239.28 Mbps).
Colombia: Movistar was fastest for fixed broadband with a median download speed of 161.28 Mbps in Q2 2023. ETB had the lowest median multi-server latency over fixed broadband at 8 ms. Of Colombia’s most populous cities, Cartagena recorded the fastest median fixed download speed of 109.01 Mbps.
Costa Rica: Claro had the fastest median download and upload speeds among mobile operators at 51.88 Mbps and 12.56 Mbps respectively. Liberty had the lowest mobile multi-server latency at 34 ms, and the highest Consistency at 79.7%. Metrocom was fastest for fixed broadband download and upload performance, at 192.00 Mbps and 143.94 Mbps respectively.
Dominican Republic: Claro had the fastest median download and upload speeds among mobile operators at 30.60 Mbps and 8.70 Mbps respectively. Viva had the lowest mobile multi-server latency at 44 ms. SpaceX’s Starlink was fastest for fixed broadband at 57.31 Mbps.
Ecuador: CNT was the fastest mobile operator in Ecuador with a median download speed of 28.45 Mbps in Q2 2023. It also recorded the highest Consistency of 81.5%. Movistar registered the lowest median multi-server latency in Ecuador at 39 ms. Netlife was fastest for fixed broadband, at 78.36 Mbps.
El Salvador: Claro had the fastest median download and upload speeds among mobile operators at 42.00 Mbps and 15.42 Mbps respectively. Movistar registered the lowest median multi-server latency in El Salvador at 65 ms. Cable Color recorded the fastest median fixed download speed (51.14 Mbps), upload speed (47.58 Mbps), and lowest median multi-server latency (35 ms).
Guatemala: Claro was the fastest mobile operator in Guatemala with a median download speed of 34.67 Mbps and median upload speed of 20.68 Mbps. Claro also had the highest Consistency with 84.4% of results showing at least a 5 Mbps minimum download speed and 1 Mbps minimum upload speed. Claro was also fastest for median fixed download performance, at 40.60 Mbps, while Cable Color was fastest for fixed upload performance, at 26.85 Mbps, and had the lowest median multi-server latency, of 35 ms.
Guyana: ENet was the top performing operator in the market, recording a median mobile download and upload speed of 67.58 Mbps and 20.92 Mbps respectively, and a median fixed download and upload speed of 62.40 Mbps and 39.66 Mbps respectively, in Q2 2023. ENet also recorded the lowest median multi-server latency across mobile and fixed networks.
Haiti: Digicel was the fastest mobile operator in Haiti with a median mobile download speed of 10.53 Mbps and median upload speed of 6.99 Mbps. SpaceX Starlink had the fastest median fixed download speed at 60.24 Mbps, while Natcom had the fastest median fixed upload speeds (17.76 Mbps) and lowest median fixed multi-server latency at 32 ms.
Jamaica: Flow was the fastest mobile operator in Jamaica with a median download speed of 35.56 Mbps. Flow also had the lowest mobile median multi-server latency at 36 ms. SpaceX Starlink had the fastest median fixed speeds at 84.93 Mbps.
Mexico: Telcel had the fastest median download speed over mobile at 48.76 Mbps, and for 5G at 223.93 Mbps. Telcel also had the lowest mobile median multi-server latency at 64 ms. Totalplay was fastest for fixed broadband (87.03 Mbps) and had the lowest median multi-server latency at 24 ms. Among Mexico’s most populous cities, Guadalajara recorded the fastest median mobile download speed of 39.13 Mbps, and Monterrey the fastest median fixed download speed of 78.30 Mbps.
Peru: Claro was the fastest mobile operator with a median download speed of 22.67 Mbps, and had the highest mobile network Consistency in the market with 80.4%. Apple devices had the fastest median download speed among top device manufacturers at 29.68 Mbps.
Trinidad and Tobago: Digicel had the fastest median download speed over mobile at 37.34 Mbps, and highest Consistency of 87.7%. Digicel+ had the fastest median fixed broadband download and upload speed at 99.11 Mbps and 98.32 Mbps respectively, and the lowest median multi-server latency at 7 ms.
United States: T-Mobile was the fastest mobile operator with a median download speed of 164.76 Mbps. T-Mobile also had the fastest median 5G download speed at 220.00 Mbps, and lowest 5G multi-server latency of 51 ms. Spectrum edged out Cox as the fastest fixed broadband provider with a median download speed of 243.02 Mbps. Verizon had the lowest median multi-server latency on fixed broadband at 15 ms.
Venezuela: Digitel was the fastest mobile operator with a median download speed of 9.53 Mbps, and had the highest mobile network Consistency in the market with 58.1%. Airtek Solutions had the fastest fixed median download speed of 73.44 Mbps, and lowest median multi-server latency at 8 ms.

Asia

Afghanistan: The fastest mobile operator in Afghanistan was Afghan Wireless with a median download speed of 7.17 Mbps. It also had the lowest median multi-server latency at 78 ms, and highest Consistency of 58.1% in Q2 2023.
Bangladesh: Banglalink was the fastest mobile operator in Bangladesh with a median download speed of 23.47 Mbps in Q2 2023. DOT Internet was the fastest fixed broadband provider with a median download speed of 90.88 Mbps and had the lowest median multi-server latency at 4 ms.
Bhutan: There was no fastest mobile operator in Bhutan during Q2 2023, but TashiCell had the lowest median multi-server latency at 42 ms, and offered the highest Consistency in the market with 83.8%.
Brunei: There was no statistical winner for fastest mobile download performance during Q2 2023 in Brunei, but Apple devices had the fastest median download speed at 143.97 Mbps.
Cambodia: Cellcard recorded the fastest median mobile download speeds at 31.60 Mbps during Q2 2023. SINET had the fastest median fixed download speed at 42.26 Mbps.
China: China Mobile was the fastest mobile operator with a median download speed of 132.81 Mbps. China Mobile also had the fastest median mobile 5G download speed at 279.14 Mbps. China Unicom was fastest for fixed broadband at 222.22 Mbps.
Georgia: There was no statistical winner for fastest mobile download performance during Q2 2023 in Georgia. Geocell recorded the lowest median mobile multi-server latency at 39 ms, while Magti recorded the highest mobile Consistency with 90.0%. MagtiCom had the fastest median fixed speed at 27.81 Mbps. MagtiCom also had the lowest median multi-server latency at 11 ms.
Indonesia: Telkomsel was the fastest Indonesian mobile operator with a median download speed of 28.71 Mbps. Telkomsel also had the lowest median mobile multi-server latency at 46 ms.
Japan: There was no statistical winner for fastest mobile download performance during Q2 2023 in Japan, however Rakuten recorded the fastest mobile upload speed at 19.90 Mbps. So-net had the fastest fixed download and upload speeds, at 276.58 Mbps and 179.51 Mbps respectively, and the lowest median multi-server latency at 9 ms.
Malaysia: TIME was the fastest fixed provider in Malaysia with a median download speed of 108.38 Mbps, and had the lowest multi-server latency at 9 ms.
Pakistan: Transworld had the fastest median fixed broadband download speed in Pakistan at 17.10 Mbps, and the highest Consistency, at 36.6%.
Philippines: Smart delivered the fastest median mobile download speed in the Philippines at 35.39 Mbps.
South Korea: SK Telecom recorded the fastest median mobile download and upload speeds at 161.16 Mbps and 16.37 Mbps respectively. LG U+ had the lowest median multi-server latency in the market at 63 ms. KT delivered the fastest median fixed download speed at 131.09 Mbps.
Sri Lanka: SLT-Mobitel delivered the fastest mobile and fixed broadband speeds in Sri Lanka at 20.71 Mbps and 38.97 Mbps, respectively in Q2 2023. Dialog had the lowest median mobile multi-server latency at 35 ms, and the highest Consistency, at 81.8%.
United Arab Emirates: etisalat by e& recorded the fastest median download speeds across both mobile and fixed, at 216.65 Mbps and 261.98 Mbps respectively in Q2 2023. etisalat by e& also had the fastest median 5G download speed at 680.88 Mbps and lowest median mobile multi-server latency at 35 ms. du recorded the lowest fixed multi-server latency, at 12 ms.
Vietnam: Vinaphone had the fastest median mobile download speed in Q2 2023, at 52.58 Mbps. It also had the lowest median mobile multi-server latency at 34 ms, and highest Consistency at 94.8%. Viettel was the fastest fixed provider with a median download speed of 105.72 Mbps.

Europe

Albania: Digicom was the fastest fixed broadband provider in Albania in Q2 2023, recording a median download speed of 93.40 Mbps. It also recorded the highest Consistency in the market, at 86.0%. There was no winner for fastest mobile operator in the market.
Belgium: Proximus recorded the fastest median mobile download speed during Q2 2023, at 78.01 Mbps. It also recorded the highest Consistency in the market, at 90.5%. Telenet had the fastest median fixed download speed at 143.42 Mbps. Among Belgium’s most populous cities, Ghent recorded the fastest median mobile download speed of 187.90 Mbps, and Antwerp the fastest median fixed download speed of 87.72 Mbps.
Denmark: YouSee was the fastest mobile operator in Denmark with a median download speed of 140.59 Mbps. Hiper was fastest for fixed broadband at 268.02 Mbps.
Estonia: The fastest mobile operator in Estonia was Telia with a median download speed of 101.32 Mbps. Telia also had the lowest median multi-server latency on mobile at 31 ms. Elisa was the fastest fixed broadband provider, with a median download speed of 94.70 Mbps.
Finland: DNA had the fastest median mobile download speed at 99.07 Mbps. Lounea was fastest for fixed broadband at 105.84 Mbps and had the lowest median multi-server latency at 11 ms.
Germany: Telekom was the fastest mobile operator in Germany with a median download speed of 93.39 Mbps, and a median download speed with 5G at 187.25 Mbps. Vodafone recorded the fastest fixed broadband performance, with a median download speed at 121.76 Mbps. It also recorded the highest Consistency in the market, at 83.8%.
Latvia: BITĖ was the fastest mobile operator in Latvia during Q2 2023, with a median download speed of 114.51 Mbps. LMT recorded the lowest mobile multi-server latency, at 26 ms. Balticom was fastest for fixed broadband with a median download speed of 243.92 Mbps. Balticom also had the lowest median fixed broadband multi-server latency at 4 ms.
Lithuania: The mobile operator with the fastest median download speed was Telia at 117.68 Mbps in Q2 2023. It also recorded the highest Consistency in the market, at 95.0%. Cgates was fastest for fixed broadband with a median download speed at 161.67 Mbps.
Poland: UPC was the fastest provider for fixed broadband with a median download speed of 223.32 Mbps in Q2 2023. There was no statistical winner for fastest mobile operator during Q2 2023, however Plus recorded the fastest median 5G download performance, at 153.19 Mbps.
Switzerland: Salt blazed ahead for the fastest fixed broadband in Switzerland, with a median download speed of 358.73 Mbps. Salt also had the lowest median multi-server latency over fixed broadband at 8 ms, and highest Consistency in the market, at 94.1%.
Turkey: Turkcell was the fastest mobile operator in Turkey with a median download speed of 58.52 Mbps. Türk Telekom had the lowest median mobile multi-server latency at 39 ms. TurkNet was fastest for fixed broadband, with a median download speed of 62.80 Mbps. It recorded the lowest median fixed multi-server latency, at 13 ms, and highest Consistency, at 80.5%. Among Turkey’s most populous cities, Istanbul recorded the fastest median download speeds across mobile and fixed, of 39.89 Mbps, and 40.27 Mbps respectively.

Oceania

New Zealand: Speedtest Intelligence data showed no winner for fastest mobile operator in New Zealand during Q2 2023. 2degrees had the lowest median mobile multi-server latency at 40 ms, and the highest Consistency, at 91.6%.

The Speedtest Global Index is your resource to understand how internet connectivity compares around the world and how it’s changing. Check back next month for updated data on country and city rankings, and look for updated Ookla Market Reports with Q3 2023 data in October.

About the Author

Mark Giles

| More Articles

Ookla Research™ Articles

Isla McKetta | June 27, 2017

Which Airport Has the Fastest Internet in Asia?

Travelers jetting off to Asia this summer will probably want to know whether you can connect to the internet upon landing and whether that internet is fast enough to help you nail down any final travel details before hitting the hotel and sleeping off the jet lag.

Using Speedtest data for March-May 2017, we analyzed the speeds of free airport Wi-Fi and local cellular signals at the busiest airports in Asia to see what your best options are and where you’re flat out of luck.

Fastest airport Wi-Fi

Dubai reigns when it comes to free airport Wi-Fi. In fact, this airport has the fastest Wi-Fi we’ve seen at any airport in Asia, Europe or Africa. And their average upload speed is even faster than their download. Travelers to second-place Seoul are also in excellent shape if they need to connect to the internet while in transit.

Tokyo, Delhi and Singapore have decently fast download speeds over airport Wi-Fi while Bangkok’s and Hong Kong’s are merely okay. Sadly, the rest of the airports offer painfully slow free Wi-Fi.

You might think airport Wi-Fi is similar to the average mobile Wi-Fi speeds of the country, but instead some of the fastest countries — Singapore (111.59 Mbps), Hong Kong (63.70 Mbps) and China (47.64 Mbps) — have poor to average airport Wi-Fi speeds. Though sitting near the top of the airport Wi-Fi pack, South Korea’s 66.67 Mbps, Japan’s 42.00 Mbps and Thailand’s 30.48 Mbps country averages show the Wi-Fi at their premier airports could be a lot faster. India’s average download speed (12.39 Mbps) is right in line with the Wi-Fi at Indira Gandhi International Airport.

On the other end of the spectrum, the United Arab Emirates has clearly prioritized airport Wi-Fi because the Wi-Fi download speed at Dubai International is nearly double the country average of 22.12 Mbps.

Fastest airport cell

In countries including China and India, you can’t connect to the free airport Wi-Fi without an in-country mobile number, so we checked Speedtest results for users on cell networks as well.

The average mobile download speed at Singapore’s Changi Airport is nearly as fast as the country’s average of 46.12 Mbps. Considering Singapore ranks second fastest in the world for mobile downloads, that’s a hard speed to beat. Dubai also has wonderfully fast speeds, and they beat the country average of 29.81 Mbps.

East Asia’s airports form a strong middle of the pack with cellular download speeds ranging from 18.18 Mbps in Bangkok to 27.12 in Guangzhou. These are comparable to the country average mobile download speeds of 33.63 Mbps for China, 19.70 Mbps for Hong Kong, 18.48 for Japan, and 14.58 for Thailand. Flyers at Delhi’s Indira Gandhi Airport, though, will be sorely disappointed with the 5.85 Mbps on offer. But that’s only slightly slower than India’s 7.62 Mbps mobile download average.

Wi-Fi or cell?

If you’ve already nailed down your international SIM card options, you’re going to have a lot better luck in many parts of Asia on a cellular signal than you would using the free airport Wi-Fi.

Of course you’re in good shape either way in Dubai, Seoul’s Wi-Fi download speed is slightly faster than that on cell, and in India you’ll definitely want to use the airport Wi-Fi if you can access it. Everywhere else the local cell performance puts airport Wi-Fi to shame.

Regional trends

Southeast Asia

If you’re jet-setting through Southeast Asia, count on any time spent in the Singapore Airport for your internet needs and plan to enjoy a more disconnected experience in India and Thailand, especially if you don’t have an Indian phone number and have to rely on cell service at the Delhi airport.

East Asia

As mentioned above, all the airports we analyzed in China, Japan and Korea had strong cellular speeds. Tokyo’s Haneda Airport and Seoul’s Incheon Airport also had good download speeds over their free airport Wi-Fi networks. China’s free airport Wi-Fi, on the other hand, is varying degrees of slow.

China

Within China, Hong Kong has the fastest free airport Wi-Fi, but you will be much better off with cellular networks at any of the airports we surveyed in China. Guangzhou has the fastest average download on cell while Hong Kong is the slowest, but with averages between 22 Mbps and 28 Mbps, you should be just fine.

Watch this space for upcoming articles comparing Wi-Fi and cellular speeds at airports across the globe. Until then, if you think your local airport is over- (or under-) rated, take a Speedtest on Android or iOS and show us what you’re experiencing.

About the Author

Isla McKetta

| More Articles

Ookla Research™ Articles

Isla McKetta | January 24, 2018

GOOOOAL: Which World Cup Finalist Scored the Fastest Internet in their Capital City?

Whether you call it soccer or football, everyone calls the World Cup fun. We couldn’t wait for the actual match-ups in June, so we decided to pit the qualifying countries against one another to see who has the fastest internet speeds in their capital cities. The results might surprise you.

Get ready to watch Russia best Brazil and Portugal defeat Iran; meanwhile, Argentina and Nigeria and Belgium and England are preparing for penalty shoot-outs.

Using data from Speedtest Intelligence for Q3-Q4 2017, we’ve calculated which capital cities of World Cup-qualifying countries have the fastest mobile and fixed broadband speeds. We also took a peek at the fastest carriers and internet service providers (ISPs) in each capital using Speed Score, a comprehensive metric that combines measures of internet performance at all levels.

Mobile winners

Iceland’s sixth place ranking for mobile download speed in the Speedtest Global Index^TM virtually assured that Reykjavík would come out at the top of the list of fastest World Cup contenders. Canberra represents Australia well with a second place finish for mobile download speeds among World Cup capitals. And Brussels, Belgium barely surpasses Bern, Switzerland for a third place finish.

**Mobile Internet Speeds**
Capitals of World Cup Qualifying Countries | Q3 – Q4 2017
Country	Capital City	Average Download (Mbps)	Average Upload (Mbps)
Iceland	Reykjavík	55.49	21.53
Australia	Canberra	44.24	12.60
Belgium	Brussels	42.52	16.74
Switzerland	Bern	42.02	17.52
South Korea	Seoul	41.85	14.15
Denmark	Copenhagen	41.78	18.29
Croatia	Zagreb	41.16	16.40
Sweden	Stockholm	40.12	12.63
Spain	Madrid	38.30	14.02
Portugal	Lisbon	30.60	11.39
Serbia	Belgrade	30.33	12.49
France	Paris	29.03	9.26
Poland	Warsaw	26.94	9.84
Germany	Berlin	25.83	9.51
England	London	25.09	11.49
Russia	Moscow	21.89	8.49
Japan	Tokyo	19.89	7.10
Uruguay	Montevideo	19.82	11.49
Mexico	Mexico City	19.11	11.51
Peru	Lima	18.33	12.90
Tunisia	Tunis	18.27	8.07
Brazil	Brasília	18.00	8.64
Morocco	Rabat	17.32	9.76
Colombia	Bogotá	16.87	9.50
Nigeria	Abuja	16.17	6.76
Iran	Tehran	15.05	7.04
Argentina	Buenos Aires	13.77	7.70
Egypt	Cairo	13.15	6.33
Panama	Panama City	12.89	8.45
Saudi Arabia	Riyadh	12.28	8.88
Senegal	Dakar	8.85	3.81
Costa Rica	San José	5.97	3.33

Looking at the group draw, Group A fares the worst with 16th place Moscow, Russia being the capital city with the fastest mobile downloads in the group. In Group B, Spain comes out on top. Australia wins Group C, Iceland takes Group D, Switzerland leads Group E and South Korea has the fastest mobile download speed in Group F. Belgium finishes first in Group G and Poland prevails in Group H, despite a 13th place finish overall.

From a regional perspective, European capitals top the rankings with all 14 European World Cup capitals sitting in the top half of the list. Latin American, Middle Eastern and African cities fare worst. Asia’s two contenders are split with Seoul boasting the fifth fastest mobile download speed among World Cup capitals and Tokyo, Japan coming in 17th.

The fastest World Cup capital in Latin America (Montevideo, Uruguay) shows a 64.3% slower mobile download speed than Reykjavík. First place among African World Cup capitals, Rabat, Morocco is 68.8% slower than Reykjavík for mobile downloads. And Tehran, Iran, the fastest World Cup capital in the Middle East, is 72.9% slower than Reykjavík.

Fastest carriers

We also looked into which carriers were fastest in each of the 32 World Cup capital cities.

With Speed Scores ranging from 8.89 in Dakar, Senegal to 46.57 in Brussels, mobile carrier Orange was fastest in four cities and tied for fastest in one. Vodafone was fastest in both Lisbon, Portugal and Madrid, Spain with comparable Speed Scores in the two locations. The rest of the cities show the diversity of fastest carriers that you might expect from a worldwide competition.

**Fastest Carriers Speeds**
Capitals of World Cup Qualifying Countries | Q3 – Q4 2017
Country	Capital City	Fastest Carrier	Speed Score
Argentina	Buenos Aires	Personal	16.15
Australia	Canberra	Telstra	50.21
Belgium	Brussels	Orange	46.57
Brazil	Brasília	Claro	24.72
Colombia	Bogotá	Avantel	20.93
Costa Rica	San José	ICE	8.30
Croatia	Zagreb	Hrvatski Telekom	49.35
Denmark	Copenhagen	TDC / Telia	45.34 / 45.09
Egypt	Cairo	Orange	16.50
England	London	EE	36.83
France	Paris	Orange	33.15
Germany	Berlin	Telekom	53.54
Iceland	Reykjavík	Nova	64.61
Iran	Tehran	MTN IranCell	15.89
Japan	Tokyo	SoftBank	27.26
Mexico	Mexico City	AT&T	20.26
Morocco	Rabat	inwi	20.51
Nigeria	Abuja	MTN	29.23
Panama	Panama City	Cable & Wireless Panama / Movistar	14.85 / 14.80
Peru	Lima	Entel Peru	20.73
Poland	Warsaw	T-Mobile	36.07
Portugal	Lisbon	Vodafone	42.44
Russia	Moscow	MegaFon	37.06
Saudi Arabia	Riyadh	Zain	13.20
Senegal	Dakar	Orange	8.89
Serbia	Belgrade	Vip mobile	45.56
South Korea	Seoul	LG U+	50.03
Spain	Madrid	Vodafone	40.17
Sweden	Stockholm	Telia	54.49
Switzerland	Bern	Sunrise / Swisscom	42.14 / 41.91
Tunisia	Tunis	Ooredoo / Orange	19.90 / 19.89
Uruguay	Montevideo	Antel	20.35

Fixed broadband winners

Given that Iceland ranks second in the world for fixed broadband download speed on the Speedtest Global Index and has the world’s highest gigabit user penetration (GUP), we’re not surprised to see Reykjavík shut out the competition by coming out on top of World Cup contenders for fixed broadband speed, too. Seoul, South Korea comes in second for fixed broadband download speed among World Cup capitals and Paris, France takes third.

**Fixed Broadband Internet Speeds**
Capitals of World Cup Qualifying Countries | Q3 – Q4 2017
Country	Capital City	Average Download (Mbps)	Average Upload (Mbps)
Iceland	Reykjavík	142.89	154.28
South Korea	Seoul	130.75	131.96
France	Paris	112.58	55.86
Sweden	Stockholm	98.77	66.68
Spain	Madrid	86.59	73.43
Japan	Tokyo	75.88	70.46
Denmark	Copenhagen	72.74	52.13
Switzerland	Bern	68.82	54.44
Poland	Warsaw	62.57	16.19
Portugal	Lisbon	55.80	30.97
England	London	52.53	16.12
Germany	Berlin	46.84	9.52
Russia	Moscow	45.25	42.96
Belgium	Brussels	43.25	9.63
Panama	Panama City	29.11	5.93
Australia	Canberra	28.85	12.46
Serbia	Belgrade	26.45	5.59
Croatia	Zagreb	26.20	11.40
Mexico	Mexico City	24.11	10.14
Uruguay	Montevideo	23.02	5.82
Argentina	Buenos Aires	22.03	4.26
Brazil	Brasília	21.57	5.29
Saudi Arabia	Riyadh	20.93	9.05
Peru	Lima	18.15	3.51
Colombia	Bogotá	13.43	6.48
Morocco	Rabat	11.83	2.51
Iran	Tehran	9.33	4.18
Costa Rica	San José	8.79	4.29
Nigeria	Abuja	8.07	5.27
Tunisia	Tunis	7.82	4.49
Senegal	Dakar	7.42	3.11
Egypt	Cairo	5.61	1.92

Group A again suffers on the fixed side with leader Russia coming in 13th based on Moscow’s fixed broadband download speed. Spain’s still the front-runner of Group B. France takes Group C, Iceland wins Group D, Switzerland tops Group E, South Korea reigns over Group F, England heads up Group G and Japan starts Group H based on average download speeds over fixed broadband in their respective capitals.

European capitals again fare well, with 12 of the 14 placing in the top half of fastest World Cup capitals for fixed broadband download speed. Belgrade, Serbia and Zagreb, Croatia rank 17th and 18th, respectively. Tokyo ranks much better for fixed broadband download speed than for mobile, which puts both Asian World Cup capitals in the top six.

With the exception of Panama City, Panama, which ranks 15th, all Latin American World Cup capitals are in the bottom half of the list for download speed over fixed broadband. As are all Middle Eastern and African capital cities.

Panama City’s fixed broadband download speed is 79.6% slower than Reykjavík’s. Riyadh, Saudia Arabia boasts the title of fastest World Cup capital in the Middle East, but is still 85.4% slower for fixed broadband downloads than Reykjavík. The fastest World Cup capital in Africa — Rabat, Morocco — is 91.7% slower than Reykjavík.

Fastest providers

Comparing Speed Scores for fixed broadband across World Cup capitals, Vodafone had wins in Berlin, Germany and Lisbon and Orange took Paris and tied for first in Madrid. The rest of the fastest ISPs vary by location as listed below:

**Fastest ISPs Speeds**
Capitals of World Cup Qualifying Countries | Q3 – Q4 2017
Country	Capital City	Fastest ISP	Speed Score
Argentina	Buenos Aires	Cablevisión Fibertel	21.72
Australia	Canberra	iiNet	33.23
Belgium	Brussels	Telenet	66.95
Brazil	Brasília	NET Virtua	27.30
Colombia	Bogotá	ETB	19.17
Costa Rica	San José	Cabletica	8.28
Croatia	Zagreb	vip	30.23
Denmark	Copenhagen	Fiberby	103.26
Egypt	Cairo	TE Data	4.84
England	London	Hyperoptic	117.40
France	Paris	Orange	107.20
Germany	Berlin	Vodafone	55.46
Iceland	Reykjavík	Nova	278.06
Iran	Tehran	Mobin Net	11.74
Japan	Tokyo	So-net	118.05
Mexico	Mexico City	Axtel	45.83
Morocco	Rabat	Maroc Telecom	9.25
Nigeria	Abuja	MTN	10.73
Panama	Panama City	Cable Onda	25.08
Peru	Lima	Movistar	16.64
Poland	Warsaw	UPC	82.72
Portugal	Lisbon	Vodafone	61.80
Russia	Moscow	MGTS	62.00
Saudi Arabia	Riyadh	STC	16.46
Senegal	Dakar	Tigo	6.42
Serbia	Belgrade	SBB	34.60
South Korea	Seoul	KT	162.45
Spain	Madrid	Masmovil / Orange	101.52 / 101.34
Sweden	Stockholm	Ownit	158.78
Switzerland	Bern	Fiber7	241.93
Tunisia	Tunis	TOPNET	7.61
Uruguay	Montevideo	Antel	22.01

Did your team not come out as expected? Or are you defending a tight match? Take a Speedtest on Android, iOS or on the web and we’ll check back in on scores closer to the main event.

About the Author

Isla McKetta

| More Articles

Ookla Research™ Articles

Isla McKetta | May 2, 2018

The American Globetrotter's Guide to Roaming Speeds

Mobile roaming has come a long way from the days when I spent most of my tour of China touring hotel lobbies desperately hoping to connect my U.S. flip phone to the Wi-Fi. Not only can you actually get a signal in most countries these days, some carriers offer special packages for the jet set so you don’t have to pay extra for roaming calls and data.

But how are the speeds?

Using Q1 2018 Speedtest® data, we’re here to report on mobile roaming speeds for U.S. consumers in 15 popular destinations, including which carriers are fastest where. For overall speeds we look at data from all devices and when we analyze carriers we look only at data for modern (LTE-capable) devices.

Where roaming speeds will (and will not) let you down

Get thee to Canada! Our analysis of roaming Speedtest results found that U.S. customers in Canada saw a mean download speed of 42.03 Mbps during Q1 2018. That’s not quite as fast as the 45.28 Mbps Canadians receive on their home mobile networks, but it beats the 27.08 Mbps average in the U.S.

**Roaming Speeds for U.S. Customers Abroad**
Q1 2018
Country	Download (Mbps)	Upload (Mbps)
Canada	42.03	13.50
South Korea	21.81	8.60
Mexico	18.02	10.18
Spain	13.23	7.09
Italy	12.70	6.38
France	12.48	5.45
Australia	11.84	6.96
Japan	10.91	4.79
United Kingdom	10.40	5.68
Germany	9.02	4.03
Costa Rica	7.72	4.11
China	7.05	3.91
Dominican Republic	5.75	3.58
India	2.96	1.96
The Bahamas	1.70	2.99

Second place South Korea showed roaming speeds for U.S. travelers about half as fast as those in Canada. Mexico was third fastest. The middle tier of the roaming speed ranking is taken up mostly by western European countries (with Japan and Australia to break up the pack).

At the bottom of the spectrum, Bahamian roaming speeds are painfully slow. They aren’t much better in India or the Dominican Republic.

A lot of factors go into the roaming speeds you’ll experience abroad, including how carriers prioritize out of country traffic, something that’s decided between each individual carrier in each individual country.

How does your carrier stack up?

Your roaming experience on your next trip is going to depend a lot on which carrier you have, so we broke our roaming speed analysis of Speedtest results on modern devices down to the carrier level.

**US Carrier Speeds While Roaming Abroad**
Q1 2018 | Mean Download (Mbps)
Country	AT&T	Sprint	T-Mobile	Verizon Wireless
Australia	21.24	N/A	2.14	22.14
Canada	26.53	27.65	53.56	43.22
China	17.23	4.77	1.15	13.15
Costa Rica	13.67	N/A	0.70	14.86
Dominican Republic	11.00	N/A	0.57	7.68
France	22.72	N/A	1.96	26.30
Germany	20.55	N/A	1.86	20.58
India	4.92	1.70	0.79	7.13
Italy	24.05	N/A	1.99	25.19
Japan	18.22	24.79	1.40	11.46
Mexico	19.95	9.66	17.22	22.35
South Korea	27.97	17.49	21.67	N/A
Spain	29.27	N/A	1.18	24.82
The Bahamas	1.79	N/A	0.25	3.53
United Kingdom	19.87	9.07	1.74	16.61

From the above, it looks like there’s no one right answer for the fastest roaming carrier. And there are other things to consider when roaming, too, like does your carrier offer a special plan that includes free roaming or are you paying through the nose.

It’s important to remember that roaming comes at a cost to carriers, which means that if your carrier includes free or low-cost roaming on almost all types of plans, the trade-off might be that you get slower speeds than you would with another carrier.

So if speed is your primary criterion, there are two standouts on this list. Verizon wins eight of the 15 countries we analyzed and AT&T wins six. T-Mobile and Sprint each win one country. We excluded Sprint from the running in eight countries because of a low number of test results.

Are you roaming (for business or pleasure) this summer? Take a Speedtest on Android or iOS to show us how fast (or slow) your connection is.

About the Author

Isla McKetta

| More Articles

Ookla Research™ Articles

Isla McKetta | December 10, 2018

The World’s Internet in 2018: Faster, Modernizing and Always On

When it comes to the internet, the news is mostly good for 2018. Download and upload speeds are increasing across the globe on both mobile and fixed broadband. 5G is on the horizon and gigabit service is expanding.

We looked at data from Speedtest Intelligence^TM, Ookla’s flagship data platform, from December 2017 through November 2018, to analyze everything from global internet speeds to the world’s fastest countries to when people are online. We also investigated which parts of the world are seeing the most benefit from high speed LTE on mobile and gigabit speeds on fixed broadband.

Mobile speeds increased more than 15% in 2018

Graphic-Tables-Final-v2_mobile-average-speed-3

The world’s average mobile download speed of 22.82 Mbps increased 15.2% over the past year, while mobile upload speed increased 11.6% to reach 9.19 Mbps.

Graphic-Tables-Final-v2_android-vs-ios-1

With a mean download speed of 27.84 Mbps and a mean upload of 10.61 Mbps, worldwide speeds on iOS devices were faster than those on Android (21.35 Mbps download, 8.73 Mbps upload) in 2018. This is likely due to market factors as Android devices are more popular in emerging markets where internet speeds tend to be slower.

However, download speed on Android devices increased 19.0% and upload speed increased 15.1%, more than those on iOS (18.0% increase for download and 11.1% increase for upload), which is good news for those emerging markets.

Fixed broadband speeds increased more than 26% in 2018

Graphic-Tables-Final-v2_fixed-average

On a global level, fixed broadband speeds were nearly twice as fast as those on mobile in 2018. The world’s average download speed on fixed broadband was 46.12 Mbps, 26.4% faster than last year. Upload speed increased 26.5% to 22.44 Mbps.

All said, though, both mobile and fixed broadband speeds increased at a slower rate in 2018 than they did in 2017.

Countries with the fastest internet in 2018

Graphic-Tables-Final-v2_top-5-mobile

The countries with the fastest mean download speeds over mobile in the past 12 months were: Norway (63.19 Mbps), Iceland (58.68 Mbps), Qatar (55.17 Mbps), Singapore (54.71 Mbps) and the Netherlands (53.42 Mbps).

Graphic-Tables-Final-v2_top-5-fixed

Singapore showed the world’s fastest mean download speeds over fixed broadband during the past twelve months at 175.13 Mbps. Next fastest were Iceland (153.03 Mbps), Hong Kong (138.31 Mbps), South Korea (114.67 Mbps) and Romania (109.90 Mbps).

To keep up with month-to-month internet speeds at a global level, visit the Speedtest Global Index.

Most improved countries for 2018 internet speeds

Graphic-Tables-Final-v2_most-improved-mobile

Costa Rica saw the largest increase in mobile download speed over the past year at 194.6%. Myanmar was second with 121.8%, Saudi Arabia third (113.2%), Iraq fourth (92.3%) and the Ukraine fifth (82.1%).

The countries with the largest improvements in upload speeds were Bangladesh (179.2%), the Ukraine (172.5%), Costa Rica (163.4%), Myanmar (146.9%) and Iraq (126.7%).

Graphic-Tables-Final-v2_most-improved-fixed

Paraguay saw the biggest increase in mean download speed over fixed broadband in the world over the past year at 268.6%. Guyana was second with 113.5%, Libya third (108.0%), Malaysia fourth (89.5%) and Laos fifth (76.2%).

Libya showed the most improvement in mean upload speed over fixed broadband during the past twelve months at 176.4%. Guyana was second with 116.1%, Malaysia third (95.2%), Belize fourth (88.9%) and Iraq fifth (76.8%).

4G is increasing mobile speeds

2017-2018-LTE-growth

When we looked specifically at Speedtest results on 4G, we saw that mean download speeds increased in most countries. Costa Rica was most improved for 4G download speed as was the case with overall download speed on mobile increasing 184.3% year over year. Saudi Arabia was second at 110.2% and Myanmar third at 78.0%.

Most encouragingly, we saw the number of Speedtest results over 4G increase in all but 15 countries. This could indicate that 4G availability is expanding. Tanzania saw the greatest increase with 355.0% more tests over 4G in 2018 than in 2017. Malta was second at 267.2% and Algeria third at 143.7%. We can see this expansion on the map above as 4G results fill in areas of the globe that were previously blank.

Gigabit coverage is expanding globally

gigabit-fade-1

Gigabit is in the news as ISPs across the globe expand their high-speed networks. We looked at Speedtest results on fixed broadband in excess of 750 Mbps to see which cities are benefitting most. Comparing locations with 100 or more gigabit-speed results in 2017 with those in 2018, that expansion becomes obvious. In 2017, 60 countries met our gigabit test threshold. In 2018, 16 additional countries joined our gigabit list. We’re also seeing that more cities around the world now have access to gigabit speeds.

Also exciting is that cities which already had gigabit in 2017 saw increases in the number of gigabit-speed results in 2018 as ISPs continue to build out infrastructure across cities. Many cities saw their first real gigabit expansion in 2018. For example, we saw the number of gigabit tests in New Delhi increase from 119 gigabit speed results in 2017 to 20,239 in 2018, that’s a mind-boggling 16,908% increase. Chennai, India saw a 7,481% increase (from 763 to 57,840) and Cormeilles-en-Parisis, France jumped 6,480%. Huge leaps in the number of gigabit-speed results were also seen in Gdańsk, Poland (6,338%); Rome, Italy (4,909%); Lancashire, United Kingdom (3,962%); Ota, Japan (3,240%); São Paulo, Brazil (2,947%); Hangzhou, China (2,669%) and Turda, Romania (2,636%).

When people are online

point-3-sec-final-1

The internet is always on, but we were surprised to see how consistently and steeply usage fell off on both mobile and fixed broadband after 9 pm local time. Normalizing the time of day for Speedtest results from around the globe, we found that usage bottoms out at 4 am and then climbs steeply again until 10 am. From there, the internet gets gradually busier until that night time drop off.

Internet in the world’s largest countries

The world’s five most populous countries are notably absent from the lists of fastest and most improved countries on mobile and fixed broadband. China, India, the U.S., Indonesia and Brazil represent about 46% of the world’s population, which makes their internet speeds worth noting nonetheless.

**Internet Speeds in the World’s Largest Countries**
Speedtest Data | December 2017-November 2018
Country	Mean Mobile Download Speed (Mbps)	% Improvement in Mobile Speed	Mean Fixed Download Speed (Mbps)	% Improvement in Fixed Speed
China	30.96	-5.8%	76.03	42.5%
India	9.11	15.2%	23.00	50.4%
United States	28.50	22.3%	92.77	37.3%
Indonesia	10.39	5.3%	14.89	18.3%
Brazil	18.65	29.3%	22.95	39.4%

Mobile internet speeds in the world’s largest countries

China had the fastest average mobile download speed among the world’s most populous countries in the past twelve months. However, China’s mobile download speed decreased during that time. The United States is fast catching up with China on mobile download speed.

Brazil occupied a middle ground for mobile download speed among the world’s most populous countries and showed the fastest rate of increase. India and Indonesia were at the bottom of this list. While the two nations show similar mobile download speeds to each other, India’s mobile download speed is improving much more quickly than Indonesia’s.

Fixed broadband speeds in the largest countries in the world

The United States showed the fastest fixed broadband download speed among the world’s most populous countries over the past year. China was second, India and Brazil nearly tie for third and Indonesia follows.

India showed the largest improvement in mean download speed over fixed broadband of the world’s five largest countries. China was second, Brazil third, the U.S. fourth and Indonesia fifth.

We’ll be back throughout 2019 to report on the state of the world’s internet as it evolves. Until then, take a Speedtest to find out how your network compares.

About the Author

Isla McKetta

| More Articles

Ookla Research™ Articles

Isla McKetta | September 4, 2019

In-Depth Analysis of Changes in World Internet Performance Using the Speedtest Global Index

A lot has changed in the two years since 2017 when we first began ranking mobile and fixed broadband speeds of countries around the world with the Speedtest Global Index^TM. 5G is being deployed around the world and fiber continues to make gigabit speeds a reality in more and more countries. We’ve been tracking it all and are here to report on how much speeds have increased, which countries are leading internet performance and which are falling behind, and what trends we see across continents.

World mobile speed increased 21.4% with fixed broadband up 37.4%

World-Download-Speeds-2019-OG2

Looking just at the last year, the world’s mean download speed over mobile increased 21.4% from 22.81 Mbps in July 2018 to 27.69 Mbps in July 2019. Mean upload speed over mobile increased 18.1% from 9.13 Mbps to 10.78 Mbps. The world average for download speed over fixed broadband increased 37.4% from 46.48 Mbps in July 2018 to 63.85 Mbps in July 2019. Mean upload speed over fixed broadband increased 48.9% from 22.52 Mbps to 33.53 Mbps.

Shake-ups in the country rankings for internet performance

Fastest-Countries-Mobile-2018-2019

Mobile speeds in the fastest countries have skyrocketed in the past year which has dramatically shifted the rankings. South Korea, which was not even in the top ten a year ago, saw a 165.9% increase in mean download speed over mobile during the past 12 months, in large part due to 5G. Switzerland’s mean download speed increased 23.5%. Canada’s was up 22.2%, Australia 21.2%, the Netherlands 17.3%, UAE 11.1%, Malta 10.3% and Norway 5.8%. Qatar remained in the top ten, although the country’s mean download speed over mobile actually dropped 1.4% from July 2018 to July 2019.

Individual mobile operators can make a huge difference in a country’s speeds. In 2017 we were excited to see Telenor uncap their mobile speeds, which drove Norway to the top of the Speedtest Global Index. A big part of South Korea’s mobile success in the past year is the way KT, LG U+ and SK Telecom banded together to release 5G at the same time. Switzerland has also benefited from 5G and Sunrise leads the country with 262 5G deployments across the country while Swisscomm has 52.

Fastest-Countries-Fixed-2018-2019

Fixed broadband rankings on the Speedtest Global Index have not changed as dramatically during the past 12 months as those on mobile. Singapore remains the fastest country with an increase in mean download speed over fixed broadband of 5.6%. Taiwan had the largest jump in speeds among the top 10 with a 166.5% improvement in fixed download speed between July 2018 and July 2019. Mean download speed over fixed broadband increased 52.4% in South Korea, 26.4% in Macau, 21.7% in Romania, 21.0% in Switzerland, 19.3% in the United States and 3.5% in Hong Kong.

Monaco and Andorra did not have enough tests to qualify for the Speedtest Global Index one year ago, but massive fixed broadband improvements in both countries inspired us to lower our test count threshold for inclusion and also share these smaller countries’ success stories.

Technologies paving the way: 5G and gigabit

The presence of 5G is not enough to change a market

As discussed above, 5G has the potential to rocket a country to the top of the mobile rankings on the Speedtest Global Index. In practice, we’ve seen 5G speeds that were over 1000% faster than those on LTE.

Mobile-Download-Speeds-by-Country

In reality, though, unless 5G is commercially available widely across a country and from all mobile operators (as was the case in South Korea), the change in speeds at the country level is not that significant. Though commercial 5G was launched widely across Switzerland by Sunrise and Swisscom in April 2019, the country’s mean download speed only increased 2.8% in the three months since. The average mobile download speed in the U.S. has actually declined slightly since 5G was initially deployed. This is because 5G is still only available in a very limited number of markets to consumers with 5G-capable devices.

Visit the Ookla 5G Map for the latest on 5G deployments across the globe.

Gigabit is a game-changer, if you can get it

Unlike 5G, fiber connections have been rolling out since 2007, opening up the possibility of gigabit-speed fixed broadband. That said, it’s costly and time-intensive to lay miles and miles of fiber so progress has varied widely across the globe.

Gigabit-Test---Performance_Singapore-1

Geographically small countries like Singapore have the advantage when it comes to fiber, because It’s easier and cheaper to lay fiber optic cable across the country’s small footprint. Singaporean internet service providers (ISPs) have used this advantage to go beyond mere gigabit and offer connections as fast as 10 Gbps. This is reflected both in Singapore’s dominance of the fixed rankings on the Speedtest Global Index and in the fact that 2.87% of their total Speedtest results over fixed broadband are gigabit-speed (800 Mbps or higher).

Gigabit-Test---Performance_Brazil-1

Brazil offers a good contrast for how difficult it can be for gigabit to reach the masses. While the first Brazilian ISP to offer fiber-to-the-premises (FTTP) initially did so in 2007, easy access to gigabit speeds was slow to follow. However, that may be starting to change. Between June and July 2019 we saw a large uptick in gigabit-speed results in Brazil, where the proportion of gigabit speed tests increased from 0.02% of total fixed broadband tests to 0.17%. This corresponded with a large increase in mean download speed at the country level.

Comparing world mobile and fixed broadband at a glance

We were curious to see just how different internet performance experiences were around the world, so we plotted average mobile download speed against average download speed on fixed broadband. All of the graphs below use a percentage difference from the global average, a number that changed between 2018 and 2019.

fade-Performance-vs-Global---Quadrant-All

Speed Leaders

There was not much change in the list of countries that showed above-average download speeds on both mobile and fixed broadband between July 2018 and July 2019, the “Speed Leaders.” What did change was that fixed broadband speeds increased significantly enough among the group to bring the whole pack closer to Singapore and Hong Kong. On the mobile axis, South Korea’s major increase in download speed made that country more of an outlier, pushing the boundaries of what great performance can look like.

Fixed-Focused countries

Between July 2018 and July 2019 we saw the number of countries considered to be “Fixed-Focused” (having faster download speeds over fixed broadband than the world average while their average mobile download speeds were slower than average) increase. Ireland was the only country that solidly fit this category in 2018. Thailand and Chile started near the midline for fixed speeds and below-average for mobile speeds in 2018. 2019 found both countries squarely in the Fixed-Focused category. Israel also edged into this category as their mobile download speed fell between July 2018 and 2019.

Mobile-Focused countries

The “Mobile-Focused” category saw the most movement between July 2018 and July 2019 as some countries (the UAE and Qatar) increased their fixed speeds sufficiently to join the Speed Leaders. Meanwhile, Bosnia and Herzegovina’s mobile download speed increased year-over-year to move them into the Mobile-Focused quadrant. Georgia’s mobile download speed decreased enough to move them from Mobile-Focused to Speed Laggers.

It will be interesting to see how many of these Mobile-Focused countries double down on their mobile investments and explore 5G alternatives to fixed broadband.

Speed Laggers

No country wants to be in the position of having slower than average mobile and fixed broadband speeds. We saw 57 countries in this “Speed Laggers” quadrant in July 2018 and 78 in July 2019. This increase is mostly due to our expansion of the number of countries we consider for the Speedtest Global Index based on test count. There were enough countries in this category that we’ve considered them separately by continent below.

Regional views of mobile and fixed broadband performance

Mobile-and-Fixed-Broadband-Improvement-by-Continent-02

We aggregated Speedtest results by continent to analyze mobile and fixed broadband performance by continent.

Mobile-and-Fixed-Performance-by-Continent-01

Asia had the highest percentage increase in mobile download speed followed by North America, Oceania, South America, Africa and Europe. Oceania had the fastest mean download speed in July 2019. North America placed second, Europe third, Asia fourth, South America fifth and Africa sixth.

On the fixed broadband side, South America saw the highest percentage increase in download speed. Asia came in second, Europe third, Africa fourth, North America fifth and Oceania sixth. North America had the fastest mean download speed in July 2019. Europe was second, Asia third, and Oceania fourth. As we saw with mobile, South America and Africa again ranked fifth and sixth, respectively.

A zoomed-in view of the speed quadrants separated by continent offers a more detailed view of each country’s role in these rankings.

Africa mostly lags in internet speeds

2019-Performance-vs-Global---Africa

In July 2019, all but two African countries in the Speedtest Global Index fell into the Speed Laggers category, having mobile and fixed broadband speeds that were below global averages. The exceptions were South Africa and Guinea, which both had fast enough mobile speeds to place them in the mobile-focused quadrant.

Asian markets show a wide breadth of internet performance

2019-Performance-vs-Global---Asia

Asia was the most diverse continent we examined in terms of internet performance. We saw a plurality of countries in each of the four quadrants in July 2019. Most of the Speed Leaders were in East Asia: China, Hong Kong (SAR), Japan, Macau (SAR), South Korea, and Taiwan. If we include Singapore, another Speed Leader, these are among the wealthiest nations in Asia (using GDP per capita). Two of Asia’s Fixed-Focused countries are in Southeast Asia (Malaysia and Thailand) and one is in the Middle East (Israel).

The Speed Laggers category contained countries from South Asia (including Afghanistan, India and Pakistan), Southeast Asia (Brunei, Cambodia, the Philippines and Vietnam) and the Middle East (Jordan). Mobile-Focused countries in Asia were mostly Middle Eastern, including Kuwait, Lebanon, Oman, Saudi Arabia and Turkey.

European mobile performance is mostly strong, fixed varies

2019-Performance-vs-Global---Europe

With the exception of Ireland, the European countries on the Speedtest Global Index fell into the Speed Leaders, Mobile-Focused, and Speed Laggers categories. All of the Speed Laggers (Belarus, Kazakhstan, Russia and the Ukraine) were from Eastern Europe. Countries from Southeast Europe (including Albania, Bosnia and Herzegovina, Croatia, Greece, Montenegro, Moldova, Serbia and Slovenia) and Central Europe (Austria and the Czech Republic) made up the bulk of the Mobile-Focused category.

Speed Leaders included countries from the Baltics (Estonia, Latvia and Lithuania), the Nordics (Denmark, Norway and Sweden), Central Europe (Poland and Romania), and Western Europe (including Belgium, France, Germany, the Netherlands, and Spain).

North American internet performance is sharply divided

2019-Performance-vs-Global---North-America

Canada and the U.S. are the only two North American countries in the Speed Leaders category. Panama is the only North American country under Fixed-Focused. Mexico and all of the Central American countries fall into the Speed Laggers category. There are no North American countries that are Mobile-Focused.

Each country in Oceania has a very different internet story

2019-Performance-vs-Global---Ocean

Oceania is represented in three of the four quadrants: Speed Leaders (New Zealand), Mobile-Focused (Australia) and Speed Laggers (Papua New Guinea) with Fiji straddling the divide between Speed Laggers and Mobile-Focused.

South America mostly lags in mobile and fixed internet speeds

2019-Performance-vs-Global---South-America

Most of the South American countries represented on the Speedtest Global Index are in the Speed Laggers quadrant (Argentina, Bolivia, Brazil, Colombia, Ecuador, Paraguay, Peru, Suriname and Venezuela). Chile is an exception, being part of the Fixed-Focused group, as is Uruguay which sits in Mobile-Focused.

Global internet speeds are improving on average and 5G and gigabit are compounding those advances where available. However, not all countries are benefitting equally. We’ll be interested to see how 5G continues to push mobile speeds in the next year and also whether 5G Wi-Fi becomes a game changer for fixed broadband. Remember to check the Speedtest Global Index on a monthly basis for updated country rankings. And take a Speedtest to make sure your experience is represented in your country’s averages.

Editor’s Note: This article was edited on September 10, 2019 to correct an error in the labeling on the first image. The colors in a later image were updated for consistency.

About the Author

Isla McKetta

| More Articles

Ookla Research™ Articles

Ookla | March 13, 2020

Tracking COVID-19’s Impact on Global Internet Performance (Updated July 20)

We are no longer updating this article as internet speeds in most countries have stabilized to pre-pandemic levels. For ongoing information about internet speeds in specific countries, visit the Speedtest Global Index^TM or contact our press team.

Ookla^® closely monitored the impact of COVID-19 on the performance and quality of global mobile and broadband internet networks in the early days of the pandemic. We shared regular information based on Ookla data to assist in the understanding of this unprecedented situation. You can still download the July 20, 2020 CSV here which contains all the public data we tracked in this article. If you are looking for information on internet or online service outages, please check Downdetector^®.

About the Author

Ookla

| More Articles

Ookla Research™ Articles

Isla McKetta | January 4, 2021

Slack Starts the New Year Late

2021 started with a sigh today as workers across the globe rushed back to their desks only to find that Slack was down. Users flocked to Downdetector® to report problems with Slack across the globe just after 7:00 a.m. Pacific. Issues were reported in Asia (Japan), Australia, Europe (Finland, France, Germany, Italy, the Netherlands, Poland, Russia, Spain, Sweden and the United Kingdom), North America (United States and Canada) and South America (Brazil). The most frequently reported issue was the inability to connect at all and the majority of reports came from the U.S.

Downdetector_Slack_Outage_1020

Although the spike has passed and the outage seems to be resolving itself, the service is not fully operational yet. Now might be a good time to take a breath and catch up on all that post-holiday email. The little red notification will be back on your desktop soon enough.

Downdetector data can help your team resolve service issues faster and improve customer experience when an outage occurs, which becomes all the more important during periods of high usage. Contact us here to learn how your network operations center can get faster outage detection.

About the Author

Isla McKetta

| More Articles

How Boingo Leverages Ookla's Comprehensive Network Intelligence

Ookla Joins Ericsson's Leading Telecommunications Open RApp System

New Year, Great Data: The Best Ookla Open Data Projects We’ve Seen So Far

Get the latest Ookla Research™

Share

Highlighting the digital divide in the U.S.

Building a data affluence map

Internet speeds are beautiful

Use Ookla Open Datasets to make your own maps

Administrative unit data

Adding data from Ookla Open Datasets

Exploratory data analysis

Examining test counts

Data distribution

Mapping average speed

Rural and urban analysis

Data distribution overall and over time

Spotlighting regional variances

What else you can do with this data

About the Author

Katie Jolly

Ditch the Lag: Cities with Great Gaming Culture and Low Ping

Get the latest Ookla Research™

Share

The top contenders

The rest of the pack

About the Author

Isla McKetta

48 New Ookla Market Reports Available for Q2 2023

Get the latest Ookla Research™

Share

Africa | Americas | Asia | Europe | Oceania

About the Author

Mark Giles

Which Airport Has the Fastest Internet in Asia?

Get the latest Ookla Research™

Share

Fastest airport Wi-Fi

Fastest airport cell

Wi-Fi or cell?

Regional trends

Southeast Asia

East Asia

China

About the Author

Isla McKetta

GOOOOAL: Which World Cup Finalist Scored the Fastest Internet in their Capital City?

Get the latest Ookla Research™

Share

Mobile winners

Fastest carriers

Fixed broadband winners

Fastest providers

About the Author

Isla McKetta

The American Globetrotter's Guide to Roaming Speeds

Get the latest Ookla Research™

Share

Where roaming speeds will (and will not) let you down

How does your carrier stack up?

About the Author

Isla McKetta

The World’s Internet in 2018: Faster, Modernizing and Always On

Get the latest Ookla Research™

Share

Mobile speeds increased more than 15% in 2018

Fixed broadband speeds increased more than 26% in 2018

Countries with the fastest internet in 2018

Most improved countries for 2018 internet speeds

4G is increasing mobile speeds

Gigabit coverage is expanding globally

When people are online

Internet in the world’s largest countries

Mobile internet speeds in the world’s largest countries

Fixed broadband speeds in the largest countries in the world

About the Author

Isla McKetta

In-Depth Analysis of Changes in World Internet Performance Using the Speedtest Global Index

Get the latest Ookla Research™

Share