| August 2, 2022

Find Out Which 50 Top European Vacation Destinations Have the Fastest Mobile Internet Speeds

Vacation season is upon us with millions of tourists already flocking to Europe’s top summer destinations despite the heat. Whether you’re a European looking for the best “staycation” or an American taking advantage of the exchange rates, our analysis of mobile internet performance in 50 popular destinations across the continent should help you see how well you will be able to connect (or disconnect, should you so choose) on vacation this summer.

We divided these towns and cities into four categories that might appeal to different types of travelers: big cities, beach life, cultural havens, and forest getaways. Then we compared Speedtest Intelligence® data for median download speeds and multi-server latency across each category. Of course many of the best vacation destinations belong in multiple categories, so be sure to check each list to find your favorite locale, and you’ll want to remember that your experience may vary based on your carrier’s roaming agreement.

“Big cities” with internet performance as exciting as the lifestyle

Oslo, Norway topped our list of 15 European vacation destinations full of big city excitement with the fastest mobile internet with a median download speed of 156.99 Mbps during Q2 2022. Perfect if you’re planning to do live updates from the Øya Festival. Two other Nordic cities rounded out the top three with Stockholm, Sweden showing a median download speed of 136.74 Mbps and Copenhagen, Denmark coming in at 133.16 Mbps. With speeds like these you should be able to host as many “see how glorious my vacation is?” video calls as you want.

There was a large gap between these three cities and the next fastest group, led by Lisbon, Portugal at 86.49 Mbps. Istanbul, Turkey and Rome, Italy had the lowest median download speeds on this list. These are still adequate mobile speeds, especially if you want to lean into the vacation experience and report back on it later (if ever).

Nicosia, Cyprus had the lowest median multi-server latency on our “big city” list at 17 ms during Q2 2022. This metric measures how quickly your device gets a response after you’ve sent out a request during three stages so you can understand if you’ll run into unnecessary lags in your connection. We’ve previously reported how latency in Europe increased while roaming, and higher latency could mean that you encounter glitches in video calls or even have to shut down apps to get your phone to respond in a timely fashion. Rome had the highest latency on this list.

Where internet speeds are fit for “beach life”

Nothing says summer like relaxing on the beach. If your beach vacation needs fast internet access, surf the web in Biarritz, France and Korčula, Croatia. These two locales had the fastest median download speeds on our beach life list during Q2 2022 at 155.65 Mbps and 133.67 Mbps, respectively, though this is not a statistically significant difference. Varna, Bulgaria and Faro, Portugal rounded out the top four at 104.00 Mbps and 101.74 Mbps, respectively.

Amalfi, Italy and Cefalù, Italy are good beach destinations if you’re looking for encouragement to log off. With median download speeds of 31.40 Mbps and 38.18 Mbps, respectively, during Q2 2022, your internet connection should be adequate (though less inspiring than the Tyrrhenian Sea).

Piran, Slovenia had the lowest median multi-server latency (20 ms) on this list while Cefalù had the highest (56 ms).

“Cultural havens” with research-ready internet speeds

Travelers looking for maximum culture and fast internet speeds can do no better than Tromsø, Norway, which had a median download speed on mobile of 251.52 Mbps during Q2 2022, almost 2.3x faster than runner-up Amsterdam, Netherlands. Whether you’ve traveled to the northernmost bit of Norway for the wooden houses, one of the many summer festivals, or just to escape the heat, you’ll have fast enough internet to thoroughly research all the places you can go if you extend your vacation.

At the other end of the spectrum, if you’re traveling to Yerevan, Armenia or Florence, Italy to enjoy the museums and other cultural highlights of either, you might want to take pictures instead. With median download speeds of 24.11 Mbps and 29.84 Mbps, respectively, these cities were the slowest on this list. All the more reason to immerse yourself in the sights and wait to connect to Wi-Fi before uploading your photos.

Helsingør, Denmark had the lowest median multi-server latency, while Edinburgh, Scotland, United Kingdom had the highest. High latency can impede the performance of your phone even when you’re in an area with great speeds — something to keep in mind if you’re relying on your phone to locate the next Fringe venue.

Navigate internet speeds in “forest getaways”

Some people swear by the forest for the best vacations, especially in the heat of summer. The Alpine cities of Innsbruck, Austria and Digne-les-Bains, France had the fastest median download speeds over mobile on our list of forest getaways during Q2 2022 at 87.34 Mbps and 67.30 Mbps, respectively, though there was no statistical winner between the two. There was also no statistically significant difference between the median download speeds in Zakopane, Poland (43.52 Mbps); Killarney, Ireland (35.90 Mbps); and Sighișoara, Romania (30.48 Mbps). While you may prefer to be offline in the woods, having a strong and fast internet connection available can help you navigate with GPS, check out the names of all the flora you’re encountering, or reach help in case of emergency.

Sighișoara and Killarney had the lowest median multi-server latency on this list at 24 ms and 25 ms, respectively. Digne had the highest latency at 36 ms.

Regardless of where you travel, mobile internet performance will affect your trip. We hope this list prepares you for the getaway you want. If we missed your favorite European vacation destination, take a Speedtest® on Android or iOS to show off your speeds and tweet us your suggestions for next year’s list or share your ideas via this Google form.

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.

| August 1, 2023

European Vacation: Your Guide to Roam Like at Home Performance this Summer

Summer is here and with it the August vacation season that most of Europe has been waiting for. We’re here with fresh data from Speedtest Intelligence® to help you know in advance if your phone will be a help or a hindrance on your travels. We’ve examined results from Android devices in the European Union (EU) and the United Kingdom (U.K.) during Q2 2023 and included data on 5G performance to see if “Roam Like at Home” is living up to its promise.

5G boosts European roaming speeds but latency varies 

Speed and latency are two important indicators of online performance. Speed will have a greater effect on streaming content while latency will affect gaming and video chatting. Both come into play while browsing online maps for that hot new gelato place your friends told you about. 

While the EU extended “Roam Like at Home” regulations until 2032, our previous analyses demonstrated that mobile speeds are usually slower when roaming than when exploring your home country. The actual speeds vary dramatically based on roaming agreements between mobile operators, something you can’t really control unless you’re prepared to do the research then shop for a whole new plan before your trip. What you can control is whether you spring for 5G before your big trip. The data below represents the local speeds and multi-server latency for each country as well as those experienced by residents of the named country while roaming in the EU or the U.K. during Q2 2023.

Roaming speeds can’t live up to local performance

Median Mobile Roaming Performance in Europe
Speedtest Intelligence | Q2 2023
Country Local Download (Mbps) Roaming Download (Mbps) Local 5G Download (Mbps) Roaming 5G Download (Mbps)
Austria 59.28 42.77 161.00 80.51
Belgium 48.00 59.86 157.23 124.64
Bulgaria 78.27 48.97 252.75 93.24
Croatia 73.65 47.86 184.29 124.38
Cyprus 56.93 40.83 197.22 64.24
Czechia 46.50 31.96 104.36 72.26
Denmark 118.48 51.74 206.95 115.82
Estonia 64.99 77.51 * *
Finland 78.00 53.43 217.13 106.54
France 61.64 43.46 187.42 95.56
Germany 46.30 39.54 114.07 84.27
Greece 57.41 66.17 143.26 153.28
Hungary 40.44 48.28 85.12 132.98
Ireland 34.49 36.10 98.70 85.38
Italy 36.97 37.68 130.85 99.38
Latvia 59.73 65.13 220.51 146.15
Lithuania 64.70 53.04 * *
Luxembourg 74.86 46.88 * *
Malta 44.56 48.11 109.97 110.80
Netherlands 96.77 37.61 128.04 87.84
Poland 42.14 34.23 78.82 99.51
Portugal 57.86 39.91 212.95 79.88
Romania 43.34 59.59 153.35 99.84
Slovakia 43.16 41.53 139.53 85.16
Slovenia 52.73 49.86 145.83 95.50
Spain 32.33 35.77 85.75 116.48
Sweden 84.00 51.24 160.93 129.43
United Kingdom 40.74 48.06 108.74 98.92

Although EU residents can roam call, text, and surf without additional charges, a drop in speed is an issue while roaming according to Q2 2023 data. Residents from 17 countries showed faster local download speeds than roaming speeds when considering results over all technologies during Q2 2023. The 11 exceptions were Romania (which was much faster roaming), Belgium, Hungary, Estonia, the U.K, Greece, and Spain, which all showed faster roaming speeds, and Latvia, Malta, Ireland, and Italy which showed only slightly improved speeds.  

While roaming speeds varied widely based on country of origin, every country on this list saw much higher median download speeds when roaming on 5G than roaming on all technologies. Despite this uplift, 20 countries showed faster local 5G download speeds than roaming 5G speeds during Q2 2023. The exceptions were Hungary, Spain, and Poland, which all saw faster 5G roaming than locally, while Greece had marginally better 5G roaming than at home, and people from Malta didn’t really see any major difference. Estonia, Lithuania, and Luxembourg did not have enough 5G samples to qualify for analysis.

Latency suffers awfully when roaming

Median Mobile Roaming Latency in Europe
Speedtest Intelligence | Q2 2023
Country Local Multi-server Latency (ms) Roaming Multi-server Latency (ms) Local 5G Multi-server Latency (ms) 5G Multi-server Latency (ms)
Austria 32.19 91.24 30.96 111.08
Belgium 36.51 84.96 31.89 67.61
Bulgaria 31.72 125.60 26.41 124.63
Croatia 39.03 85.25 35.96 80.37
Cyprus 27.90 200.18 23.52 184.68
Czechia 34.72 80.18 30.77 78.72
Denmark 29.47 94.77 28.08 81.85
Estonia 33.10 87.00 * *
Finland 33.67 102.84 32.16 99.14
France 46.97 92.61 44.39 87.49
Germany 41.78 87.19 39.07 77.57
Greece 38.16 137.29 34.63 132.91
Hungary 36.27 85.39 35.06 77.26
Ireland 36.21 116.62 31.70 118.47
Italy 50.87 100.20 49.96 93.87
Latvia 29.88 101.91 26.28 94.70
Lithuania 34.03 107.79 * *
Luxembourg 33.45 67.70 * *
Malta 47.44 146.57 38.69 135.43
Netherlands 33.24 83.25 31.17 79.81
Poland 42.41 108.69 40.25 100.36
Portugal 35.98 123.05 31.62 126.54
Romania 40.65 122.91 34.22 121.21
Slovakia 30.47 80.28 26.39 70.96
Slovenia 30.61 76.68 27.70 75.27
Spain 51.53 112.81 47.11 102.94
Sweden 37.52 114.69 32.97 99.48
United Kingdom 50.45 111.01 46.04 107.59

Residents of every country surveyed had a much higher multi-server latency when roaming during Q2 2023. This is because roaming signals are routed through a user’s home network, making latency a significant issue for Europeans when traveling. The difference was smallest in Italy and France where locals had about a 97% higher latency when roaming compared to at home. On the other end of the scale, residents of Cyprus saw over a 615% higher latency while roaming compared to at home. A high latency means a long response time for any request you send over the internet, perhaps adding more frustration to your vacation than you would want.

The results were similar over 5G with Italy showing the smallest difference between latency while roaming and at home (about 88% greater) and Cyprus having the largest difference (about 685%). Surprisingly, 5G did not reliably improve multi-server latency. 

What this data tells you is that it’s worth upgrading to 5G before a trip if speed is your main concern. Whether 5G will help you with your latency will depend heavily on which country you are traveling from. Regardless, you should expect a high latency while roaming. Plan ahead accordingly.

Outbound roaming speeds vary dramatically

As discussed above, roaming speeds depend heavily on where you are traveling from, where you are traveling to, and what agreements your home mobile operator has with operators in the country you’re visiting. That said, it’s fun to look at what’s possible, so we looked at the very best and very worst roaming speeds in Europe during Q2 2023.

Fastest roaming speeds in Europe

Customers from Italy roaming in Portugal saw one of the fastest median roaming download speeds over 5G in all of Europe during Q2 2023 at 317.94 Mbps. Estonians roaming in Finland saw a 278.69 Mbps median 5G download speed, and Finns in Sweden experienced a median 5G download of 252.57 Mbps. 

Looking at the highest results from all technologies combined, Latvians roaming in Finland experienced a median download speed of 137.49 Mbps, while Greeks saw median download speeds of 129.75 while roaming in Bulgaria and 125.83 Mbps while roaming in Austria.

Slowest roaming speeds in Europe

On the other end of the scale, Slovakian visitors to Poland had one of the slowest median download speeds over all technology while roaming during the same time period at 10.82 Mbps. Polish roamers saw a median download speed of 13.27 Mbps in the U.K. and 14.78 Mbps in Romania. Ten of the 20 slowest roaming speeds on our country-by-country list for all technologies were for Europeans roaming in the U.K. This means that travelers going to the U.K are not only seeing poorer performance, they could also be paying more for the privilege as they do not benefit from “Roam Like at Home” regulations.

Germans in Italy had the slowest median 5G download speed on our list during Q2 2023 at 33.64 Mbps. Austrians in the U.K. saw a median download speed over 5G of 41.57 Mbps, and Portuguese roamers in Spain experienced a median download speed over 5G of 44.18 Mbps.

Whether you choose to roam, pick up a local SIM or eSIM, or simply spend less time online, we hope this information helps you make the most of your vacation. If you have an extra minute between activities, take a Speedtest® to see how your performance compares. 

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.

| February 16, 2023

European Commission Uses Ookla® Data to Demonstrate Territorial Disparities and the Digital Divide in the EU

The characteristics of places might not define the identity of people living there, but they affect the daily life of every citizen and can prevent them from reaching their full potential and the best life. Persisting territorial disparities, related to phenomena such as depopulation or limited access to services, can become even more visible by observing places rather than individuals. In fact, growing territorial inequality, further exacerbated by the COVID-19 pandemic, limits on people’s lifestyle and life choices, and creates a feeling of being left behind.

The Joint Research Centre of the European Commission recently published a report focused on understanding and identifying territorial disparities and challenges in the EU across a variety of sectors, including access to broadband and the performance of the digital connection, using the Speedtest® by Ookla® Global Fixed and Mobile Network Performance data for Q4 2020. A full comprehension of these challenges and their interlinkages is essential to produce evidence to improve policies, especially for the benefit of those living in vulnerable conditions, in line with the principle of leaving no place and no one behind.

Key findings

High-speed broadband increasingly represents an essential infrastructure to drive the economic and social development of territories. Therefore, the lack of broadband network or poor access to a high-speed connection might cause significant disparities among places and citizens and leave some areas behind in terms of access to services and opportunities, diminishing the quality of life of residents. The connectivity gap, often recognized as an urban-rural digital divide, represents an important challenge to face for some countries and policy makers and may require a joint effort by both public and private initiatives.

Results show that significant differences exist in network speed across the EU-27 countries. Territorial disparities are even more relevant, with most Member States in urban areas enjoying easy access to the highest fixed broadband available (> 100 Mbps), whereas in rural areas a significant percentage of residents have access to an average speed below the minimum standard of 30 Mbps. Only very few countries show access to over 100 Mbps broadband for rural populations (Denmark, Sweden, the Netherlands, Luxembourg).

Regarding mobile broadband, the average speed is generally lower than fixed broadband in all countries, with only a few areas having more than 100 Mbps average speed. Interestingly, the Alpine region across France, Italy and Austria appears to have better connectivity with mobile broadband (over 30 Mbps) than with fixed broadband (under 30 Mbps). The same pattern can be observed in Central Italy and Sardinia, Croatia, and partially in Greece.

To strengthen the evidence of such urban-rural digital divide, the spatial patterns of access to broadband were combined with the population density and the classification of degree of urbanization of municipalities in cities, towns and rural areas (see Figure 01). Results confirm that urban areas present the highest speed in broadband connection, revealing how the areas already most connected in terms of physical networks (i.e., with roads and railways) are also the most connected from the digital point of view. Access to a good broadband connection is most problematic in remote municipalities (with 45 minutes or more from the nearest city by drive), where the average speed is significantly lower than the national averages, especially in countries such as Belgium, Spain, France, and Portugal.

To explore the urban-rural digital divide further, the analysis employed Machine Learning to identify patterns of similarities in terms of vulnerabilities across all areas, regardless of their degree of urbanization, considering the performance of the broadband infrastructure, the population distribution and the remoteness classification of all areas. The areas (belonging to group 0 in Figure 2) are characterized by low speed, high latency, low population density, remoteness, and are identified as the most vulnerable places. However, this analysis also showed that the bad performance of broadband networks is not limited to rural areas. This means that the digital divide is not merely a matter of urban-rural, but mostly a matter of cities versus non-cities: places that would not be considered as disconnected from the physical perspective (not remote areas, but towns close to cities) can still be disconnected from the broadband perspective.

Digitalization can be an opportunity only if its rollout is quick enough to enable rural businesses to remain competitive. However, it is useless without a parallel development of digital literacy and skills for residents in rural areas. Furthermore, digital connectivity is only one player in the game, and it cannot alone overturn the depopulation trends and the other vulnerabilities affecting rural areas.

Without policy support, a lack of or poor access to high-speed broadband might leave some areas behind. Access to broadband and data might help to foster new business and economic activities, especially in vulnerable areas such as rural regions. Exploiting the potential that connectivity and digitalization represent for education and training, cooperation and networking, access to services and markets, can make these areas more attractive to people and businesses.

Cohesion Policy will support the implementation of the EU’s digital agenda. In particular, European Regional and Development Fund investments will focus on digitalization of services for businesses and citizens and rollout of the high-speed broadband. The support will go where it is most needed, i.e. areas where there is a weak take-up of digital technologies or no, or very slow, or very expensive, broadband access or where there is not enough commercial potential to attract private investors.

To read more about this work, the full report is available here.

We are grateful to Patrizia Sulis for her work with our data within the European Commission’s report and guidance in composing this article. For more information, please contact Patrizia Sulis – Scientific Officer – Joint Research Centre, European Commission.

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.

| November 30, 2022

A Gigabit Society Requires a Data-Driven Regulatory Approach: Key Takeaways From Ookla's Southeastern Europe Regulatory Summit

Ookla®, in partnership with Ibis Solutions, hosted the inaugural Southeast Europe Regulatory Summit on November 4, 2022 in Belgrade, Serbia with participants from across Europe and a guest speaker from Saudi Arabia. The summit examined how crowdsourced data can support regulators and drive effective policymaking while reflecting on the regional telecommunication landscape, its challenges, opportunities, and how to ensure consumers have good quality of experience (QoE).

Key Takeaways

  • Spectrum is key when it comes to 5G. CST shared how data-driven, evidence-based policy decisions enabled a 5G rollout.
  • The European Commission’s goal of gigabit connectivity is one of the driving forces behind the rollout of fiber networks. So far, the progress is mixed across the region in terms of fiber network rollout and adoption.
  • Good quality data and sharing best practices is essential for effective policymaking. Crowdsourced data helps measure progress and assess real service performance experienced by end users.

A data-driven approach to 5G rollouts will bear fruit

The majority of countries across southeast Europe have launched 5G, including Croatia, Cyprus, Greece, Montenegro, North Macedonia, Romania, and Slovenia. However, Serbia, Albania, and Bosnia and Herzegovina are still yet to auction dedicated 5G spectrum.

Chart of 5G performance and availability across southern Europe

Within Southeast Europe, 5G Availability (the proportion of users of 5G capable devices who spend the majority of their time on 5G networks) reached almost 50% in Cyprus. Yet, four out of seven countries had 5G Availability below 10% in Q3 2022. 5G performance depends heavily on the 5G spectrum holding. As a rule of thumb, the larger the share of the C-Band spectrum, the faster the 5G download speed, while access to low-band spectrum has an impact on 5G Availability. 

Therefore, having a spectrum roadmap is key to ensuring effective spectrum assignment. Salem Askar, a Senior Specialist in the  QoS Department at CST (Communication, Space and Technology Commission), shared how data-driven decision-making informed the development of its spectrum strategy. This resulted in Saudi Arabia improving its mobile performance, and its ranking in the Ookla’s Global Index – from 105th place in September 2017 to 10th place in October 2022.

CST committed to a new, evidence-based approach to spectrum assignment drawing from technology trials and consultations with network operators, device and chipset manufacturers, infrastructure providers, end users, and network data providers, including Ookla. Most notably, CST needed data to expand its analysis of spectrum usage, the performance of various bands, and existing network infrastructure to see where investments had been made within certain bands (e.g., extra capacity in specific bands in urban areas and coverage of rural areas using adequate bands). If you would like to find out more, read this case study. Looking ahead, CST properties revolve around the improvement of citizens’ lives by giving them access to minimum speeds of 100 Mbps, releasing more spectrum for 5G and beyond. 

However, spectrum is just one part of the 5G puzzle. Konstantinos Masselos, President of Hellenic Telecommunications and Post Commission and BEREC Vice Chair 2022, shared that BEREC’s focus extends beyond 5G spectrum allocation. That includes following up on developments for a number of relevant issues covering the whole 5G ecosystem (i.e., end user, interoperability, new business models and value chains, privacy, Quality of Service (QoS), roaming, roll out, security, state aid) and identifying regulatory challenges. A key focus area is sustainability and how to make networks not only faster but also more energy efficient and environmentally friendly. BEREC’s work on 5G aims to contribute to the maximum benefit for end users in terms of availability, pricing, quality, security, and other choices.

Regulatory incentives propel the move to fiber

Chart of fixed performance and FTTH/B subscription percentage of households across southern Europe

The demand for gigabit-capable broadband access increased over the past few years, mainly due to the increasing number of people working from home and the growth in service availability. According to Speedtest Intelligence®, Romania had the fastest 5G speeds in Q3 2022, with a median fixed download speed of 131.96 Mbps and 94.29 Mbps upload speed. These speeds are not a surprise given that Romania has one of the highest Fiber to the Home/ Building (FTTP/B) penetration rates across the region — nearly 54% of households — which is partially driven by government-backed fixed infrastructure projects such as RoNet, and the special attention given to rural and disadvantaged areas. Slovenia is also progressing well in terms of access to fiber, thanks to the Slovenian government’s introduction of a next-generation broadband network development plan, which is in line with the EU’s ‘gigabit society’ plan. It is aiming to provide internet access to all households in rural or urban areas with at least 100 Mbps that can be upgradable to gigabit speed. Serbia’s growth in FTTP/B connections is gaining pace as well, with Yettel’s entry into the fixed segment in November 2021 resulting in increased market competition. Pavle Mijuskovic, Deputy Executive Director, the Department for Electronic Communication Networks and Services (EKIP) of Montenegro shared that (FTTH/B) networks covered around 70% of households, while the real subscriptions are much lower — only about 45% of households. The need to close the gap in rural areas is something that Montenegro focuses on now. Mr. Mijuskovic also highlighted that widespread fiber rollout necessitated a change in the way infrastructure is deployed: putting cables underground at lower power grades and reusing the post or street lights to deliver fiber to households. 

During the panel session, Mr. Masselos shared his view that “development of broadband to the home is a very important topic,” with regulatory actions in place to achieve a target of all homes across the EU to have ‘gigabit connectivity’ by 2030. He brought attention to the fact that in Greece there is a significant gap between fiber rollout and adoption. Despite improvements in Fiber to the Premise (FTTP) coverage, which stood at roughly 20% of households in 2021 (an increase from 7% in 2019) the actual take-up is in the low single-digit partially because of the higher cost of fiber subscription. Greece is also looking to promote open wholesale networks to increase coverage and prevent overbuilding of fiber networks. With regards to FTTH deployment and adoption, Mr. Masselos emphasized the need to formulate policies that will encourage investment and will also support the demand side so that fiber infrastructure is accessible and economically viable in the long term. In addition to being available, services over FTTH/FTTP must also be competitively priced.

The power of data in understanding a digital divide 

The vast majority of operators in the region plan to retire copper networks and migrate their DSL users over to FTTH/P. As such, fiber investment will remain a priority for telcos and will often be supported by subsidies, EU funding, and co-investment. At the European level, the policy goal is for every European household to have access to high-speed internet coverage by 2025 and gigabit connectivity by 2030. Regulators, alongside BEREC and the European Commission, work to promote connectivity and enable access to high-speed networks. Therefore, it’s vital to understand the real service performance as experienced by end users. Bryan Darr, Vice President, Smart Communities at Ookla shared how crowdsourced data can be used by governments to help them measure progress.Map of Fixed broadband Speedtest results in central Serbia

Darr shared a snapshot of fixed network performance in Central Serbia by layering the speeds with the fastest speeds on top, providing evidence of where broadband speeds are being met and where they are not (see map above). Whereas fiber penetration is increasing across Serbia, there are still areas where users do not experience speeds beyond 25 Mbps (red dots on the map). What our data reveals is that good performance tends to be centered in urban and metropolitan areas. Similarly, the OECD report “Bridging Digital Divides in G20 Countries” found that in rural areas of the EU and G-20 countries, fixed broadband download speeds are close to 50% slower than speeds in cities and urban hubs. 

Map of mobile performance, coverage, and signal measurement near Novi Sad, Serbia – Ookla Cell Analytics

Similarly, using Ookla Cell Analytics data we can identify poor coverage areas and assist with network planning to optimize mobile network connectivity and performance. It all starts with understanding service availability, including signal strength (RSRP), signal quality (RSRQ), and areas of high demand. Customer throughput depends on capacity; even though there might be a high signal strength and good network quality, the actual speed is determined by network capacity which depends on the number of devices within this location. 

Good quality data and sharing best practices is the best tool in the regulator’s arsenal

Mr. Masselos shared that good quality data is a necessity for efficient policymaking and EETT has done work in this space already. For example, by looking at broadband prices versus performance and QoS across Greece and other countries, “we see these comparisons cannot be one dimensional, but two dimensional. I feel we need to do more of these types of analyses and develop relevant methodologies.” Mr. Askar of CST echoed this sentiment by saying that to achieve goals regulators ought to bring knowledge from different parts of the world. Pavle Mijušković of EKIP stressed that to overcome the digital divide there is a need to use a combination of all available technologies to provide services. Furthermore, he suggested the improvement of capacities and capabilities will help drive digital transformation and strengthen the digital awareness of society.

Ookla’s Co-Founder and CEO Doug Suttles summed up the event by saying that a hidden benefit Ookla can provide is to “figure out the best practices, as relationships with operators in 150 countries help a lot on that. Having regulators exchange knowledge on how to create better connectivity for cost-effectiveness, is vital to our industry.” Our event partner, Ibis Solutions, a software products vendor in the southeast Europe region, also underlined the potential of Ookla metrics for regulatory agencies such as public portals with new features. Vladimir Rakić, Software Development Director at Ibis Solutions summarized the summit by saying “attending regulatory agencies recognized how crowd metrics could help address massive outages or network performance issues that further result in a huge amount of user complaints.”

Ookla’s mission is to measure, understand, and help improve connected experiences. Events such as this regulatory summit help with the cross-pollination of ideas and best practices. Telecom regulators who attend Ookla’s future summits will also be able to facilitate discourse on related topics, challenges, ideas on elevating the subscriber experience in their respective regions, and how to improve their country’s rank on the Speedtest Global Index™. If you are a regulator looking to make more data-driven, evidence-based policy decisions, inquire here

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.

| February 20, 2019

An Expansive Analysis of European Mobile Roaming Speeds and Behaviors

Last year we took a look at how free roaming was working out for EU citizens in terms of speeds and latency. This year we’ve expanded our analysis to all European countries and included data on Wi-Fi roaming behavior. We’ve also added a little insight into how roaming might affect download speeds for visitors to Mobile World Congress (MWC) in Barcelona from February 25- 28, 2019.

Except where noted, this analysis is based on Speedtest data from Android devices on 4G LTE cellular connections during Q3-Q4 2018. We included data on any country with greater than 30 samples in all categories.

Most Europeans experience slower downloads while roaming

Roaming traffic is subject to deals struck between individual mobile operators on how that traffic will be prioritized, so roaming speeds can depend not only on the country of origin but also the country of destination and the plan a subscriber has selected.

Consider the following table where an Albanian experiences a mean download speed of 54.56 Mbps at home and then 38.47 Mbps while roaming elsewhere in Europe. This is expected as a roamer does not usually have a direct relationship with the mobile operator handling their data and calls abroad.

Mean Mobile Download Speeds in Europe
Speedtest Data | Q3-Q4 2018
Country Local Speed (Mbps) Roaming Speed (Mbps) % Difference
Austria 38.23 35.78 -6.4%
Belarus 16.15 19.49 20.7%
Belgium 52.58 35.42 -32.6%
Bulgaria 47.28 32.35 -31.6%
Croatia 43.83 43.60 -0.5%
Cyprus 37.13 20.58 -44.6%
Czech Republic 44.91 13.43 -70.1%
Denmark 48.83 34.89 -28.6%
Estonia 36.43 38.98 7.0%
Finland 39.33 38.20 -2.9%
France 39.94 34.97 -12.5%
Germany 33.77 28.57 -15.4%
Greece 41.35 38.08 -7.9%
Hungary 49.57 28.19 -43.1%
Iceland 69.27 35.58 -48.6%
Ireland 28.23 31.49 11.6%
Italy 32.18 43.12 34.0%
Kazakhstan 22.93 13.80 -39.8%
Latvia 30.88 33.79 9.4%
Liechtenstein 56.48 36.66 -35.1%
Lithuania 41.49 33.43 -19.4%
Luxembourg 50.91 25.08 -50.7%
Malta 56.34 34.27 -39.2%
Montenegro 45.45 49.97 10.0%
Netherlands 56.06 33.86 -39.6%
Norway 68.49 38.69 -43.5%
Poland 28.74 29.71 3.4%
Portugal 32.06 36.95 15.3%
Romania 36.64 30.74 -16.1%
Russia 20.91 20.47 -2.1%
Serbia 43.41 21.64 -50.1%
Slovakia 33.47 31.80 -5.0%
Slovenia 35.51 36.41 2.5%
Spain 36.07 22.37 -38.0%
Sweden 44.87 34.59 -22.9%
Switzerland 47.59 30.36 -36.2%
Turkey 38.19 28.77 -24.7%
Ukraine 26.07 25.48 -2.3%
United Kingdom 30.84 38.76 25.7%

Residents of the Czech Republic will face massive speed disappointment when roaming through the rest of Europe. Other countries with much better speeds at home than abroad include Luxembourg, Serbia, Iceland and Cyprus.

In ten European countries, citizens experience faster mobile downloads while roaming than they do at home. These include: Italy, the United Kingdom, Belarus, Portugal, Ireland, Montenegro, Latvia, Estonia, Poland and Slovenia. Most of these are among the slowest countries on this list, so it would make sense that their citizens would experience better speeds elsewhere in Europe than they do at home.

Europeans are connected to Wi-Fi most of the time

Customers sometimes try to get better speeds and avoid roaming fees (for those outside the E.U.) and data overages by connecting to Wi-Fi. The following table compares the percentage of time spent on Wi-Fi by a resident of a country with that of a visitor to the country.

Percentage of Time Spent on Wi-Fi in Europe
Speedtest Data | Q3-Q4 2018
Country Local Customers Visitors % Difference
Albania 61.9% 59.8% -3.4%
Andorra 73.6% 69.0% -6.2%
Armenia 61.5% 68.1% 10.7%
Austria 65.6% 36.7% -44.1%
Azerbaijan 67.0% 70.7% 5.5%
Belarus 63.9% 58.3% -8.9%
Belgium 71.1% 32.9% -53.7%
Bosnia and Herzegovina 71.9% 66.0% -8.2%
Bulgaria 65.2% 46.8% -28.2%
Croatia 66.1% 41.0% -38.0%
Cyprus 70.9% 58.4% -17.7%
Czech Republic 75.1% 38.9% -48.1%
Denmark 70.4% 52.9% -24.9%
Estonia 61.2% 45.6% -25.4%
Finland 56.6% 47.1% -16.7%
France 60.2% 43.1% -28.5%
Georgia 61.4% 62.2% 1.3%
Germany 72.2% 42.4% -41.3%
Greece 73.5% 52.2% -28.9%
Hungary 71.9% 35.0% -51.3%
Iceland 65.7% 58.2% -11.4%
Ireland 67.2% 52.4% -22.0%
Italy 64.1% 48.9% -23.7%
Kazakhstan 57.7% 64.7% 12.3%
Latvia 60.9% 43.8% -28.1%
Liechtenstein 71.7% 58.2% -18.9%
Lithuania 66.8% 43.3% -35.1%
Luxembourg 63.6% 26.0% -59.1%
Macedonia 65.3% 52.9% -18.9%
Malta 74.3% 58.0% -22.0%
Moldova 67.2% 67.3% 0.2%
Montenegro 63.6% 65.2% 2.5%
Netherlands 73.0% 42.5% -41.7%
Norway 74.7% 59.7% -20.1%
Poland 62.5% 48.1% -23.1%
Portugal 69.1% 54.7% -20.9%
Romania 62.4% 48.7% -21.9%
Russia 58.8% 65.9% 12.2%
San Marino 66.8% 39.7% -40.6%
Serbia 68.6% 61.2% -10.7%
Slovakia 69.6% 35.1% -49.6%
Slovenia 63.3% 26.4% -58.3%
Spain 70.8% 53.0% -25.2%
Sweden 71.8% 44.9% -37.4%
Switzerland 62.4% 47.2% -24.4%
Turkey 61.7% 73.0% 18.4%
Ukraine 61.5% 62.4% 1.5%
United Kingdom 71.3% 54.0% -24.3%

Finland showed the lowest time spent on Wi-Fi by residents at 56.6%. Kazakhstan was second at 57.7% followed by Russia (58.8%), France (60.2%) and Latvia (60.9%). The Czech Republic showed the highest time spent on Wi-Fi by residents at 75.1%. Norway was second at 74.7% followed by Malta (74.3%), Andorra (73.6%) and Greece (73.5%).

When it comes to time spent on Wi-Fi by visitors, Luxembourg had the lowest percentage at 26.0%. Slovenia was second at 26.4% followed by Belgium (32.9%), Hungary (35.0%) and Slovakia (35.1%). Turkey showed the highest time spent on Wi-Fi by visitors at 73.0%. Azerbaijan was second at 70.7% followed by Andorra (69.0%), Armenia (68.1%) and Moldova (67.3%).

Luxembourg saw the largest difference in time spent on Wi-Fi between residents and visitors with visitors using Wi-Fi 59.1% less than residents. Slovenia was close behind at 58.3%, followed by Belgium (53.7%) and Hungary (51.3%). On the other end of the spectrum, visitors to Turkey were on Wi-Fi 18.4% longer than residents followed by Kazakhstan (12.3%) and Russia (12.2%).

Roaming dramatically increases latency in Europe

Because roaming signals are routed through a user’s home network, latency is always an issue in roaming. Speedtest data shows that latency while roaming is a much larger issue for residents of some countries than it is for others.

Comparing European Latency In-Country and Abroad
Speedtest Data | Q3-Q4 2018
Country Local Latency (ms) Roaming Latency (ms) % Difference
Austria 24 83 245.8%
Belarus 32 75 134.4%
Belgium 24 81 237.5%
Bulgaria 24 126 425.0%
Croatia 32 86 168.8%
Cyprus 20 194 870.0%
Czech Republic 24 86 258.3%
Denmark 24 98 308.3%
Estonia 23 78 239.1%
Finland 26 104 300.0%
France 42 87 107.1%
Germany 33 87 163.6%
Greece 27 137 407.4%
Hungary 22 94 327.3%
Iceland 18 163 805.6%
Ireland 34 114 235.3%
Italy 52 116 123.1%
Kazakhstan 35 164 368.6%
Latvia 24 94 291.7%
Liechtenstein 40 90 125.0%
Lithuania 26 108 315.4%
Luxembourg 23 73 217.4%
Malta 18 141 683.3%
Montenegro 18 42 133.3%
Netherlands 26 81 211.5%
Norway 36 109 202.8%
Poland 33 104 215.2%
Portugal 27 102 277.8%
Romania 26 131 403.8%
Russia 42 161 283.3%
Serbia 22 76 245.5%
Slovakia 30 69 130.0%
Slovenia 21 69 228.6%
Spain 45 118 162.2%
Sweden 30 118 293.3%
Switzerland 26 75 188.5%
Turkey 26 115 342.3%
Ukraine 35 116 231.4%
United Kingdom 37 107 189.2%

Residents of Cyprus saw an average latency of 194 ms while roaming in Europe. The country with the second highest latency for residents roaming abroad was Kazakhstan at 164 ms, followed by Iceland (163 ms). Russia (161 ms) and Malta (141 ms). In contrast, Montenegro had a lower latency for residents roaming abroad than Italy did for residents using their mobile phones locally.

Roaming performance at MWC

MWC, the largest mobile conference in the world, has Barcelona teeming with visitors from across the globe all trying to connect to their home networks. During February 2018 we saw an average download speed of 34.31 Mbps while roaming in Barcelona and a mean latency of 201 ms. Visitors from the U.K. saw an average download speed of 49.00 Mbps and a mean latency of 111 ms, while those from Italy averaged a download of 32.88 Mbps and a latency of 128 ms.

What will the performance look like at this year’s conference? Schedule a meeting or come see us in Hall 2 at Booth 2i25 to learn more about our roaming data.

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.

| November 30, 2021

5G in Europe: EU Targets Require a Rethink

Monitoring progress of the EU’s Digital Decade

We recently attended the European 5G Observatory’s workshop, “5G in the Digital Decade,” which provided an overview of progress towards the EU’s 5G Action Plan targets and also looked ahead to the new targets as set out in the EU Digital Decade. The European Commission presented its vision of the region’s Digital Decade in March 2021, providing targets (the Digital Compass) for the digital transformation of Member States by 2030. Progress towards these targets is measured through the existing Digital Economy and Society Index (DESI), a key component of which focuses on connectivity. The relevant 2030 goals here are to provide Gigabit access to all EU households and 5G coverage in all populated areas.

The connectivity component of DESI is made up of a mix of demand and supply-side factors, including fixed and mobile broadband coverage and take-up, 5G spectrum availability and broadband pricing. The latest iteration was published in November using 2020 data, with Denmark, the Netherlands and Spain securing the top three spots for connectivity, and Greece and Bulgaria the bottom two.

ookla_desi_eu_markets_1121-01

As 5G evolves, so should the Commission’s 5G targets

The Commission’s DESI has evolved from its inception in 2014 to now track the Digital Compass’ targets. However, 5G is a rapidly evolving technology, and it was clear from the discussion during the Observatory’s workshop that there will be a need to review the actual targets and by implication the measurement criteria. The key concern lies in the “5G everywhere” target, both from a coverage and a performance perspective. As the end goal is the digital transformation of Member States, 5G coverage should be extended beyond the current target of all populated areas, to provide greater geographical coverage — in particular to support emerging vertical industry needs, for example in agriculture or logistics.

In terms of 5G performance, the ITU’s IMT-2020 requirements are for a user-experienced data rate of 100 Mbps download and 50 Mbps upload. However, not all 5G networks are created equally. The increasing use of dynamic spectrum sharing (DSS) means that operators can boost 5G coverage and capacity by using existing spectrum formerly dedicated to other generations of mobile technology. As a result, network performance will vary widely based on the bands used. There are also differences in 5G rollout, with initial 5G radio equipment tending to be deployed on existing tower infrastructure, and the coverage of higher capacity C-band spectrum will be limited relative to that of lower frequency bands as result.

ookla_median-download-speeds_europe_1121-01

Divergence between the DESI and end-user internet speed

What’s also interesting are the outliers and how their performance differs from their position within the DESI. Bulgaria (which came second to last in the connectivity portion of the DESI), does lag behind the European average on fixed broadband speeds, but is significantly faster on mobile speeds. Bulgaria’s median mobile download speed of 83.71 Mbps makes it the seventh fastest market globally. The reverse is true for Spain, which came third in the DESI, and where median fixed download speeds of 118.25 Mbps are well above the European average, putting it in 11th place globally, but where mobile speeds of 34.63 Mbps lag well behind the European average, and place it 51st globally. Other examples of divergence between the DESI and actual internet speeds on the ground include markets such as Belgium, Cyprus, Estonia and Lithuania.

To support Member States, the Commission has proposed a “Path to the Digital Decade”, a governance framework which in addition to the development of roadmaps at a Member State level and regular reporting on progress, will also include a framework to address progress shortfalls and support for multi-country projects. Member States have been encouraged to make use of the Recovery and Resilience Facility, an EU-wide fund of EUR 723.8 billion designed to mitigate the impact of the COVID-19 pandemic, from which they are required to devote at least 20% of their allocation to digital projects. As of October 2021, digital investment plans totalled 26% of the fund, of which 11% is dedicated to connectivity. With such significant public funds directed at improving connectivity, and the digital services that it underpins, the Commission must ensure that its targets and measurement methodology keep pace with the evolution of network technologies.

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.

| October 8, 2018

Just How Much Does Roaming Affect Mobile Performance in Southeast Europe?

This week the Wholesale Agreements and Solutions Group (WAS) will be meeting in Rovinj, Croatia as part of a GSMA working group to discuss the big deals that will affect worldwide roaming speeds over the next year. That made us curious — what kinds of roaming speeds might these telecom execs experience while they’re deciding our roaming fate?

We looked into Speedtest Intelligence® data from over 2.7 million consumer-initiated tests in Southeast Europe during Q1-Q2 2018 on devices using LTE networks to see how median speeds and latency were affected as mobile users roamed around the region. Roaming speeds include Android results only.

Though we expected to see a general decline in speeds and increase in latency (the time it takes for data to travel from your phone to the nearest server and back) while roaming, we were surprised by how much a person’s home country affected their mobile roaming experience.

Albania, Bosnia and Herzegovina, Macedonia and Montenegro are excluded from this analysis due to insufficient sample size during the examined time period. Data for Serbia includes the disputed region of Kosovo.

There’s no correlation between local download speed and decrease in roaming speed

Median Local vs. Roaming Download Speeds
Southeast Europe
Q1-Q2 2018 | 4G LTE
Country Local Download (Mbps) Roaming Download (Mbps) % Difference
Bulgaria 37.97 18.21 -52.0%
Croatia 34.50 31.67 -8.2%
Greece 33.61 27.87 -17.1%
Romania 28.47 20.89 -26.6%
Serbia 34.19 30.32 -11.3%
Slovenia 26.26 28.26 7.6%
Turkey 29.93 21.75 -27.3%

Slovenians are the only Southeast Europeans who see better mobile speeds while traveling than they do at home. Considering Slovenia’s local mobile download speed is the lowest on this list, we can presume that residents are encountering better infrastructure abroad and seeing better speeds as a result.

That does not that mean that countries with the fastest local download speeds see the largest declines in speed when roaming. For example, Bulgaria shows the fastest median mobile download speed in Southeast Europe and sees the largest declines in download speed when roaming. But Croatia, which has the second fastest mobile download speed in Southeast Europe, shows the second smallest decline in download speed when roaming.

On the other hand, Romania, which ranks sixth on our list for median local download speed shows the third largest decline when roaming. While Turkey, which ranks fifth in the region for local download speed, shows the second largest decline when roaming.

In fact, roaming speed is affected by in-country infrastructure and also by how operators prioritize out of country traffic. These decisions are made between each individual operator in each individual country, often at meetings like those in Rovinj this week where operators negotiate who has access to what networks and how much that access should cost them. This is why some operators offer extensive roaming packages and why those often come at a cost.

Roaming dramatically increases latency

Latency is measured in milliseconds, and a mobile user usually won’t notice a lag between taking an action on their phone and seeing an expected result when latency is low. Even in Croatia, the country with the highest local latency on our list, latency is very close to what the rest of the countries experience.

Median Local vs. Roaming Latency
Southeast Europe
Q1-Q2 2018 | 4G LTE
Country Local Latency (ms) Roaming Latency (ms) % Difference
Bulgaria 23 135 487.0%
Croatia 26 84 223.1%
Greece 23 144 526.1%
Romania 24 118 391.7%
Serbia 19 31 63.2%
Slovenia 19 67 252.6%
Turkey 24 130 441.7%

However, latency can skyrocket when roaming as some operators route a user’s traffic back through their home country, which leads to a much different user experience. This is definitely what we found in our analysis of Southeast Europe. As we saw with download speed, the percentage change between local and roaming latency was not correlated to local latency.

For example, Slovenia and Serbia show the same median local latencies (19 ms), but Slovenia shows a large increase in latency while Serbia’s increase is the smallest on our list.

This is because latency, like roaming speed, is a lot more complex than simply what infrastructure consumers are on. There’s also the matter of network agreements… the very deals that telecom execs are hammering out right now in Rovinj.

Are you at WAS? Stop by booth 23. We’d love to share a sneak peek of our roaming data with you. We have the insights and data for use in decisioning, sales, negotiation and quality steering related to wholesale and roaming solutions.

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.