Last Updated | November 18, 2024
Ookla’s Speedtest® Methodology
The definitive source for network performance, quality, and coverage intelligence
Introduction
When Ookla® was founded in 2006, high-speed home internet was just becoming available in many locations, and consumers needed a reliable way to test their network speed. Speedtest®, the company’s flagship product, is the most accurate tool for measuring a network’s capabilities. Today, modern connected experiences require reliable, consistently high performance across many additional metrics beyond download speed, including low latency for gaming, high-quality video streaming, good upload speeds for teleconferencing, and fast page load times while browsing on a mobile device.
In this evolving connectivity landscape, Ookla remains fiercely committed to measuring worldwide internet performance — as well as availability, quality of service (QoS), and quality of experience (QoE) of fixed broadband and mobile networks. For over 17 years, Ookla has been perfecting the way we test connectivity services and transparently reporting on their performance. By capturing real-world metrics and delivering insights back to network operators and policymakers, Ookla serves as the bridge between consumers and the telecommunications industry.
Every day, millions of people press the “Go” button on Speedtest to better understand the performance and quality of their internet connections. This consumer-initiated testing reflects how people actually experience the internet, when and where it actually matters to them. Because Ookla has the world’s largest network of testing servers located across every country and major population center, Speedtest does not introduce any extra travel time as the data passes between the device and server — which provides the most accurate measure of real-world network speed.
In addition to consumer-initiated testing, Ookla collects billions of mobile network samples daily, which measure radio signal levels, network coverage and availability, and QoE metrics for a number of connected experiences, such as streaming video, video conferencing, gaming, web browsing, and CDN and cloud provider performance.
We then apply rigorous aggregation and analysis to this robust global dataset, ensuring that our measurements provide the most accurate reflection of real-world consumer network experience — which is why Ookla is the preferred data provider for telecommunications operators, regulatory bodies, trade bodies, analysts, journalists, nonprofits, and other connected industries worldwide. Ookla’s enterprise clients use these robust global datasets and solutions to improve networks and create better connected experiences for consumers.
This guide shares the details of Ookla’s methodology, including how our methods differ from other network testing providers, why Ookla is uniquely capable of measuring super-fast fiber and 5G connections, and definitions for all the metrics we cover.
How Ookla collects data
Every time a user presses “Go” on Speedtest, Ookla captures robust data about that fixed broadband or mobile network connection. Additionally, we gather billions of daily mobile network samples measuring network quality of service (QoS) and quality of experience (QoE).
|
Consumer-initiated testing measures a fixed broadband or mobile network’s QoS: full throughput capacity, latency, ping, and jitter |
Mobile network samples measure a radio network’s QoS, including signal, coverage, and availability, as well as quality of connected experiences (including Wi-Fi) |
Consumer QoE measures quality of video streaming, video conferencing, gaming, CDN and cloud provider performance, and web browsing on mobile networks |
Speedtest measures the full throughput capacity of an internet connection
Consumer-initiated testing is the most reliable way to understand the throughput capacity (maximum experienced speeds) of a network connection. When a consumer decides to take a test using Speedtest, it is run as a dedicated foreground service in the application (i.e., a user is intentionally interacting with the app). Foreground testing provides the most accurate measure of internet performance capability because it is able to use enough data to flood the internet connection and measure the full capability of both the network connection and the device. The connection must be fully utilized to accurately assess network performance and QoS metrics such as download speed, upload speed, latency, packet loss, jitter, and other indicators of network conditions.
A test taken with Speedtest measures the characteristics of the communication network between a device and multiple servers using our global server network. The fewer links between a device and a server, the more relevant the measurement is to quantifying and understanding the networking capability of a particular device. To that end, Ookla operates a vast testing infrastructure of over 15,000 servers worldwide which allows us to ensure an accurate, meaningful view of network performance.
When aggregated, these measurements describe the network’s real-world performance, including information about locations, times, service providers, and devices.
For the quality of service metrics Ookla collects, please see the Metrics and Definitions section of this guide.
Ookla measures a mobile network’s radio signal, coverage, and quality of connected experiences
Ookla collects mobile network samples from over one billion consumer devices worldwide. These samples provide real-time insights into radio network coverage, signal conditions, QoS, spectrum usage, and network equipment — at a fine level of geographic detail. In addition to device radio measurements, these samples also provide insights into quality of experience while video streaming, video conferencing, gaming, and connecting to content delivery networks (CDNs) and cloud providers.
For the quality of service metrics Ookla collects, please see the Metrics and Definitions section of this guide.
Measuring 5G, fiber, and other super-fast connections
Other testing providers often report much slower speeds than Speedtest. That’s because their tests either run in the background of other apps without consumers’ knowledge or simply download a very small file from a CDN or server. These quick tests cannot scale up to assess the full capacity of a connection or accurately determine upload speed — and the results do not indicate how a connection might actually perform while using large amounts of data over a 5G or fiber connection.
Accurately detecting and measuring a 5G network connection is likewise impossible when passing a very small file to a CDN. Because a significant amount of demand on the network must be generated for 5G carrier aggregation to even take effect, shorter-duration tests are incapable of ensuring the network assigns the maximum number of component carriers to the user equipment (UE) to trigger a 5G connection. So that consumers can reliably see when they’re testing a 5G connection, Ookla has directly partnered with device manufacturers worldwide to implement accurate in-app 5G detection in Speedtest.
The ability to measure a network’s full capacity is especially critical with modern network speeds increasing rapidly across connection types like 5G and fiber broadband. Speedtest uses a client and server testing engine that dynamically scales the number of connections to multiple servers, in order to saturate and accurately measure client-side connections up to 10 Gbps. This allows Ookla to measure the full extent of real-world performance and overcome the effects of network bottlenecks such as TCP slow start (a transmission control protocol that avoids sending more data than the network is capable of forwarding).
Measuring latency and network responsiveness
Latency (referred to as “Ping” in Speedtest results) is the bidirectional round-trip time for a packet of data to travel between two endpoints, measured in milliseconds (ms). This delay impacts all connected experiences, and certain connected experiences suffer more from a lagging network, such as streaming live video or playing a massively multiplayer online game. However, for modern use cases where connectivity is a safety issue — such as autonomous vehicles and augmented reality in manufacturing — understanding the factors impacting latency is critical.
Latency is not a singular metric; accurately measuring latency depends on the network environment and the round-trip destination. To quantify this variability, Ookla tests to multiple servers and uses three primary real-world scenarios for latency measurements: base latency to a single server, multi-server latency measured from multiple nearby servers, and latency under load (that is, when the network connection’s throughput is maximized).
For the different latency metrics captured by Ookla, including gaming latency, CDN latency, and video conferencing latency, please see the Metrics and Definitions section of this guide.
Measuring video experience
Video testing in Speedtest utilizes a real video player in the Android and iOS apps, and captures metrics from adaptive bitrate playback on both fixed and mobile networks. Video testing in Speedtest provides users with insights about the quality of their video streaming experience in real-world conditions. With this information, a consumer can see if their network is delivering the expected video playback quality.
For the video metrics captured by Ookla, please see the Metrics and Definitions section of this guide.
Measuring mobile gaming experience
Based on data from billions of consumer-initiated performance tests and tests to popular gaming servers, Ookla’s gaming metrics are representative of the real-world network performance that customers experience when playing video games on their mobile devices.
For the gaming metrics captured by Ookla, please see the Metrics and Definitions section of this guide.
The Ookla difference: quantity + quality
|
Foreground testing Unlike other solutions that primarily test in the background, Ookla collects data as a foreground service in Speedtest and associated mobile applications. Speedtest is uniquely able to saturate a connection to measure throughput, and Ookla data is collected in the places consumers are using their devices, at times they care about connectivity. |
Measuring real-world user experience Unlike testing providers who only test to a single CDN (which introduces variability and doesn’t reflect real-world network experience), Ookla provides real-world measurements to show how gaming and video streaming are impacted by network conditions, CDN, cloud, or device performance. |
Providing real-time insights This constant flow of immense amounts of data allows us to precisely track how networks respond to events like large crowds, the capabilities of new devices, the impact of network upgrades, and the rollout of new technologies like 5G. Ookla’s data and analysis are cited in tens of thousands of articles per year, and governments and regulatory bodies use Ookla data to inform policy and spending decisions. |
Different testing approaches help diagnose different root causes of network issues
Ookla provides multiple means of testing a network connection to help internet service providers, mobile network operators, telecommunications regulators, and other enterprises diagnose network problems and improve connectivity.
For example, a user could have a poor video streaming experience on a particular network. Testing to the mobile network operator’s servers, the content delivery networks (CDNs) serving the video content, and the individual streaming service provider helps identify the point of failure. A multi-pronged testing approach allows us to show whether the issue is caused by a local cell site, a peering problem with a particular CDN, a widespread degradation in performance to multiple CDNs, or a larger issue upstream with the network provider.
|
Consumer-initiated testing with Speedtest |
Ookla’s client-to-server testing helps diagnose network performance issues, down to the “last mile” of an internet connection — the final leg of a telecommunications network that reaches an end user’s device, such as fiber or cell towers. -> |
Speedtest maximizes an individual user’s connection to provide the most accurate measure of throughput capacity. Poor performance in the last mile of a connection can help rule out upstream issues with the network provider or content providers. |
|
Dedicated Speedtest Server Network™ |
Distant servers can introduce latency and additional network bottlenecks. With multiple servers in every country and major population center, Speedtest ensures an accurate and meaningful view of networking performance by minimizing the number of data links between device and server. -> |
The fewer links between a device and a server, the more relevant the measurement is to qualifying and understanding the networking capability of a particular device. Because Ookla tests to multiple nearby servers concurrently, we can accurately measure latency. |
|
Testing to CDNs and cloud providers |
While Speedtest measures the full throughput capacity of a network connection, file-transfer testing to multiple CDNs and cloud hosting servers helps diagnose the root cause of degraded network experiences across multiple services. -> |
Testing to CDNs and cloud providers helps identify performance and peering issues between a network provider and the services that are delivering content on the internet. |
|
Testing to individual services |
Testing to providers of video streaming, online gaming, video conferencing, individual CDNs, and other web services helps identify performance and quality issues specific to those services. -> |
Testing individual services helps providers understand their users’ quality of experience (QoE) and diagnose problems that impact specific connected experiences, such as routing issues, peering issues, or issues with the individual service. |
Ookla provides a reliable, consistent test experience across devices
Speedtest provides an accurate, consistent test experience across devices. We have mobile applications for both Android and iOS that are optimized to measure the specific intricacies of a mobile connection. These apps are available in the App Store, Google Play, Samsung Galaxy Store, Huawei App Gallery, and Amazon Appstore.
There are also dedicated Speedtest apps for computers (Mac and Windows) and an AppleTV app. For developers who want to use the trusted technology and global server network behind Speedtest to power their own solutions, we also offer a command-line interface tool (Speedtest CLI) and Speedtest Powered™ integrations via web and mobile SDKs, an HTML5-based web test, or embedded solutions for networking hardware and CPE devices. Our original web test at Speedtest.net covers most other platforms, and we continue to explore new platforms for Speedtest.
How Ookla aggregates and analyzes data
We use a rigorous statistical sampling methodology to combat sampling bias and ensure data accuracy. Through consumer-initiated testing, Speedtest gives every user a voice to describe what the internet connection is like on their device at the times and places that matter most to them. Whether a user takes a Speedtest once per month or once per hour, our sampling methodology makes sure that each user’s voice is heard and is not drowned out by high-volume testers.
When we aggregate data from all of the tests we’ve received, each unique Speedtest user’s results are averaged to create a single sample that summarizes their internet experience for that time period and geographic area. We then evaluate each service provider based equally on the samples provided by each of its users. This removes the potential for results to be skewed by individual outliers or short-term fluctuations in service or user behaviors.
Certain types of test results are excluded from our published results and Speedtest Award™ verification. Network engineers frequently use Speedtest while developing next-generation technologies. To ensure that our results represent the true commercially available user experience, results from our CLI tool are excluded from our published results. This is because these tests are often performed in controlled environments that do not necessarily reflect the real-world conditions consumers see. For various reasons, results from our Speedtest Powered solutions for integrated network testing are excluded from our aggregates.
Finally, Ookla’s data science team employs a wide variety of automated checks and filters. This additional processing helps to ensure that each result represents true user behavior, that no anomalous testing behavior has been detected on that network, and that no bias is introduced to our aggregates, either intentionally or unintentionally.
Glossary
Here at Ookla, we are continuously researching new and better metrics to describe real-world internet performance and coverage. Our metrics help consumers to understand their network performance and service providers to understand their customer experience, gain competitive insights, and identify how to improve their networks.
Operational definitions
It’s important to define three of the key phrases we use in our articles and reports.
|
Median |
The median is the middle value in a set of data when ordered from least to greatest. When measuring the performance of a network, the median value captures the typical user’s experience. In statistical terms, median is less likely to be influenced by outliers (such as a limited number of users on extremely high-speed connections) than mean (the average found by adding all numbers in the data set and then dividing by the number of values in the set). |
|
Top providers |
To be designated as a top provider, an operator or ISPs must account for 3% or more of total samples in the market for the period. We use this designation to ensure that most consumers in the area actually have access to the provider to qualify it as the fastest, most consistent or having the best coverage. |
|
Modern chipsets |
Ookla maintains a list of device models built on chipsets capable of utilizing the latest network technologies. When calculating Speed Score and determining the winners of Speedtest Awards, we use results from devices built on modern chipsets so that an operator’s score is not negatively impacted if a portion of its subscriber base continues to use older technology. |
Quality of service (QoS) measurements
Performance metrics
|
Download speed |
Download speed measures how quickly you can pull data from a server on the internet to your device. Most connections are designed to download much faster than they upload. This is because the majority of online activity, like loading web pages or streaming videos, is driven by downloading content. Download speed is measured in megabits per second (Mbps). |
|
Upload speed |
Upload speed measures how quickly you send data from your device to the internet. A fast upload speed is helpful when sending large files via email, or in using video-chat to talk to someone else online (since you have to send your video feed to them). Upload speed is measured in megabits per second (Mbps). |
|
Latency / Ping |
Also called latency, ping is the reaction time of your connection — that is, how quickly your device gets a response after you've sent out a request. A fast ping means a more responsive connection, especially in applications where timing is everything (like video games). Latency is measured in milliseconds (ms). Ookla measures several types of latency; see the Latency section of this guide for more information. |
|
Jitter |
Also called Packet Delay Variation (PDV), jitter is a measure of the variability in ping over time. Jitter is not usually noticeable when reading text, but when streaming and gaming, a high jitter can result in buffering and other interruptions. This metric shows the average difference between consecutive latency measurements. Minimum Jitter: Minimum Jitter shows the average difference between consecutive latency measurements. Minimum Jitter measures the jitter associated with the lowest-latency testing server prior to the download phase of a Speedtest. Primary Server Jitter: Primary Server Jitter is a measure of the variability in Primary Server Latency, which is an interquartile mean (IQM or midmean, a statistical measure of central tendency based on the truncated mean of the interquartile range) of the idle latency tests taken prior to the download portion of a Speedtest. Multi-Server Jitter: Multi-Server Jitter is a measure of the variability in latency that is calculated over all servers used during server selection conducted prior to the download phase of Speedtest. It is designed to represent the jitter a user should expect to encounter when their network is not under heavy load. |
|
Packet loss |
Packet loss occurs when a packet of data being sent over the internet is not received or is incomplete. This is described in the percentage of packets lost compared to packets sent. Packet loss can be a result of network congestion, poor signal, or poor line quality. Even small amounts of packet loss can severely degrade the quality of experience for real-time applications such as video conferencing and gaming. |
|
Fixed network connection type |
Device Connection Types
Fixed Technology Types:
|
|
Mobile network connection type |
Network Generations
Mobile Device Technologies
|
|
Speed Score™ |
Speed Score™ is a performance measure that incorporates each provider’s download and upload speeds to rank network speed performance. Beginning with data collected in 2022, download speed contributes 80% and upload 20%; this reflects the growing importance of upload capability in the day-to-day experience of consumers. Both the upload and download speed components are measured with a modified trimean which combines speeds from the 10th percentile, 50th percentile (also known as the median), and 90th percentile in a weighted average using a 1:8:1 ratio, respectively. We place the most emphasis on the median speeds, as those represent what most network providers’ customers will experience on a day-to-day basis. |
Radio signal metrics
|
Operator |
Information about the subscriber network and the network to which the device is connected. |
|
Signal strength and quality |
RSRP: Reference Signals Received Power is a measure of the power of the cellular signal when a mobile test was taken (by Speedtest Android users). RSRP is measured in decibel-milliwatts (dBm). RSRQ: Reference Signals Received Quality is a measure of the quality of the cellular signal when a mobile test was taken (by Speedtest Android users). RSRQ is measured in decibels (dB). RSSI: Received Signal Strength Indicator is a measure of the strength of the cellular signal when a mobile test was taken (by Speedtest Android users). RSSI is measured in decibel-milliwatts (dBm). RSSNR: Reference Signal Signal to Noise ratio is a measure of the power of reference signals relative to the amount of noise present in the signal. RSSNR is measured in decibels (dB). |
|
Information about frequency band |
ARFCN: An absolute radio-frequency channel number (ARFCN) is a code that specifies a pair of physical radio carriers used for transmission and reception in a mobile radio system, one for the uplink signal and one for the downlink signal. EARFCN: EARFCN stands for E-UTRA Absolute Radio Frequency Channel Number. In LTE, the carrier frequency in the uplink and downlink is designated by EARFCN, which ranges between 0-65535. EARFCN uniquely identifies the LTE band and carrier frequency. EARFCN is independent of channel bandwidth. UARFCN: The UTRA Absolute Radio Frequency Channel Number (UARFCN) is a unique number given to each radio channel within the frequency bands used by the UMTS UTRA. The UARFCN can be used to calculate the carrier frequency. |
|
Data connection type (e.g., LTE, Wi-Fi) |
Data connection type:
Cellular connection type:
|
|
Radio access network |
|
Quality of experience (QoE) measurements
Video experience
|
Adaptive Average Bitrate |
The average media bitrate recorded during the adaptive bitrate stage of the video test. Expressed in megabits per second (Mbps). |
|
Adaptive Rebuffering Ratio |
The ratio of time spent stalling to the time spent playing during the adaptive bitrate stage of the video test. |
|
Fixed Stage Rebuffering Ratio |
The ratio of time spent stalling to the time spent playing during the fixed resolution stage of the video test. |
|
Adaptive Start Time |
The time spent waiting for the video to start playing in the adaptive bitrate stage of the test. Measured in seconds. |
|
Fixed Stage Start Time |
The time spent waiting for the video to start playing in the fixed resolution stage of the test. Measured in seconds. |
|
Adaptive Stall Severity |
Adaptive Stall Severity is the ratio of time spent rebuffering to time spent playing in the adaptive bitrate stage of the video test, given that a stall event has occurred. |
|
Adaptive Stall Events |
During the adaptive bitrate stage of the video test, if a user experiences a stall (pause in playback for rebuffering) it is considered a stall event. The “Adaptive Stall Events” measure is the percentage of video samples where at least one stall event has occurred. |
|
Adaptive Primary Resolution |
"Adaptive primary resolution" shows the percentage of samples by primary resolution category (e.g., 720p, 1080p). During the adaptive bitrate stage of the video test, the video resolution fluctuates to optimize the video streaming experience. When a video is played, the amount of time spent in each resolution (e.g., 720p, 1080p) is recorded; the resolution used most often is recorded as the "primary resolution" for that playback. |
|
Overall Highest Successful Resolution |
For each sample, the highest resolution achieved during the video test is recorded. "Overall Highest Successful Resolution" shows the percentage of samples that achieved a given resolution (e.g., 720p, 1080p) as their highest successful resolution. |
Gaming experience
|
Game Latency |
Game Latency is a measure of latency to popular gaming server locations. A server selection process is performed which selects the best (lowest latency) gaming servers for the user, per game, for a set of popular online games (such as Fortnite and Free Fire). Latency tests performed to the selected gaming servers and results are recorded as Game Latency. Game Latency can be used to understand the typical responsiveness of your network when playing popular online games such as Fortnite or Free Fire. |
|
Game Jitter |
Game Jitter is a measure of the variation in latency to popular gaming servers. See Game Latency for more information. |
|
Download Speed |
Download Speed is a measure of how quickly data can be pulled from a server on the internet to the user’s device. |
|
Upload Speed |
Upload Speed is a measure of how quickly data can be transferred from the user’s device to a server on the internet. |
|
Game Score™ |
Game Score™, available for 5G and All Network Generations, is based on Ookla’s consumer initiated Speedtest results as well as Consumer QoE™ latency and jitter measurements taken to real-world game servers. Game Scores are composed of eight components, each measuring a different aspect of consumer gaming experience. Each of these components is evaluated and scored on a scale of 0-100 for each eligible provider. Finally, scored components are combined in a weighted average to produce a Game Score for each eligible provider. |
Latency
|
Latency/Ping |
Latency (or Ping) is the reaction time of your connection-how quickly your device gets a response after you've sent out a request. A low latency (fast ping) means a more responsive connection, especially in applications where timing is everything (like video games). Latency is measured in milliseconds (ms). |
|
Minimum Latency |
Minimum Latency (formerly, Latency) measures the best case latency for the user at the time they decide to take a Speedtest. Across one or more pings made prior to the download speed test, the lowest ping value is determined – this represents the Minimum Latency. |
|
Unloaded Latency |
Unloaded latency is our classic measure of latency, looking at the time it takes for a bidirectional round trip on the last mile of a connection under no load. This is a best case measure of latency. |
|
Primary Server Latency |
Primary-Server Latency is the interquartile mean (IQM or midmean, a statistical measure of central tendency based on the truncated mean of the interquartile range) of the idle latency tests taken prior to the download portion of a Speedtest. Primary-Server Latency can be used to understand the typical responsiveness for the user at the time they decide to take a Speedtest. |
|
Multi-Server Latency |
Multi-server latency measures latency to multiple connections and then takes an average. This is a better representation of a customer’s experience as their connection pulls media and other information from a variety of locations around the internet. Poor multi-server latency can show where improvement is needed in peering arrangements and transit. |
|
Loaded Latency |
Loaded latency measures the round-trip time delay of a connection, as measured when the connection’s bandwidth is fully saturated by another flow. We capture loaded latency for both download and upload, which gives an approximation of how a network performs under poor conditions (such as poor buffer management or a background app fully utilizing the connection’s bandwidth). |
|
Game Latency |
Game Latency is a measure of latency to popular gaming server locations. A server selection process is performed which selects the best (lowest latency) gaming servers for the user, per game, for a set of popular online games (such as Fortnite and Free Fire). Latency tests performed to the selected gaming servers and results are recorded as Game Latency. Game Latency can be used to understand the typical responsiveness of your network when playing popular online games such as Fortnite or Free Fire. |
|
CDN Latency |
CDN Latency is a measure of latency to popular Content Delivery Networks. CDN Latency can be used to understand the typical responsiveness of your network when accessing resources hosted via Content Delivery Networks (e.g. websites, images). This can be used to measure the responsiveness to major content and cloud providers used by many applications and services (e.g. Amazon Web Services, Google Cloud). |
|
Video Conferencing Latency |
Video Conferencing Latency is a measure of latency to popular Video Conferencing server locations. Video Conferencing Latency can be used to understand the typical responsiveness of your network when using video conferencing apps when working from home or connecting with friends and family. |
Jitter
|
Game Jitter |
Game Jitter is a measure of the variation in latency to popular gaming servers. See Game Latency for more information. |
|
CDN Jitter |
CDN Jitter is a measure of the variation in latency to popular CDN servers. See CDN Latency for more information. |
|
Video Conferencing Jitter |
Video Conferencing Jitter is a measure of the variation in latency to popular video conferencing servers. See Video Conferencing Latency for more information. |
Consumer Sentiment
|
Ratings |
The five-star ratings system gauges a consumer’s overall satisfaction with your brand. At the end of a test, users may receive a prompt, “Rate your provider,” and are presented with a range of one to five stars. The name of the provider associated with the corresponding test result is also presented inline. The Rating is the mean of all the user rating samples. |
|
Net Promoter Score™ (NPS) |
The Net Promoter Score (NPS) is a key performance indicator of customer experience. At the end of a test, users may be asked to answer "How likely is it that you would recommend [provider] to a friend or colleague?" on a scale from 0-10. Users are categorized into Detractors (score 0-6), Passives (score 7-8), and Promoters (score 9-10). NPS is the percent of Promoters minus the percent of Detractors and is displayed in the range from -100 to 100. Providers are ranked in descending order of NPS. |
Ookla metrics and scores
|
Speed Score™ |
Speed Score fully accounts for the range of speeds a provider offers in a single metric. This makes it easier to compare mobile operators and ISPs on several measures of speed using one metric. Speed Score™ is a performance measure that incorporates each provider’s download and upload speeds to rank network speed performance. Beginning with data collected in 2022, download speed contributes 80% and upload 20%; this reflects the growing importance of upload capability in the day-to-day experience of consumers. Both the upload and download speed components are measured with a modified trimean which combines speeds from the 10th percentile, 50th percentile (also known as the median), and 90th percentile in a weighted average using a 1:8:1 ratio, respectively. We place the most emphasis on the median speeds, as those represent what most network providers’ customers will experience on a day-to-day basis. |
|
Coverage Score™ |
Using billions of measurements from hundreds of millions of mobile coverage scans collected by Speedtest® and Consumer QoE™ users each day, Ookla® provides a robust and comprehensive view of worldwide mobile coverage. Our unbiased, data-driven approach controls for extraneous variables and ensures an equitable accounting for all coverage scans. Because coverage is a spatially focused metric, only scans with precise and legitimate location information are included as we build samples that normalize data by user, operator, location and timeframe.
|
|
Consistency |
While fast speeds are paramount, a consistent experience is also a worthy measure of a network’s quality. Consistency Score™ is the metric we use to identify fixed broadband or mobile networks that provide a consistent quality of service. It reflects the percentage of a provider’s data samples that meet minimum thresholds for download and upload speeds, depending on the type of network. Consistency Score for fixed broadband uses thresholds of 25 Mbps minimum download speed (the recommended speed for streaming in 4K) and 3 Mbps minimum upload speed. Consistency Score for mobile uses thresholds of 5 Mbps minimum download speed (the recommended speed for streaming in HD) and 1 Mbps minimum upload speed. The higher a provider’s Consistency Score, the more likely a consumer will enjoy acceptable internet performance and quality. |
|
Service |
Service measures indicate how likely, on average, a user is to have service available in the places they go. To determine this, the area of interest (such as a selected country) is broken into small spatial polygons that tile the area. For the time period and provider of interest, we determine the set of small spatial tiles where at least two users/devices have been at least one time; this is referred to as the user-footprint. From there, we compute the percentage of the user-footprint for which service was available. Three different types of service are available in separate views:
The percentage of the user-footprint where access to 4G LTE service is available (including roaming)
The percentage of an operator's 5G-active tiles where 5G service is available (including roaming); a device is considered “5G-active” if it is known to have reported 5G service in the past 6 months
The percentage of the user-footprint where access to any kind of service is available (including roaming)
The percentage of the user-footprint where access to the network to which it is subscribed is available (not roaming) |
|
Availability |
Whether or not service is available is closely tied to coverage. We divide our data on availability into three separate views: General Availability, 4G Availability, and 5G Availability. These metrics indicate how likely a user, on average, is to have service available in the places they go. They are individually defined as:
The percentage of an operator’s known locations where a device has access to any kind of service (including roaming).
The percentage of an operator’s known locations where a device has access to 4G LTE or greater service (including roaming).
The percentage of an operator’s known locations where a device has access to 5G service. |
|
Video Score™ |
Video Score™ is based on the adaptive bitrate stage of Ookla’s consumer-initiated video experience test. Video Score is a weighted sum of five components, each measuring a different aspect of consumer video experience. These components are evaluated and then scored on a scale of 0-100 for each eligible provider. |
|
Game Score™ |
Game Score™, available for 5G and All Network Generations, is based on Ookla’s consumer initiated Speedtest results as well as Consumer QoE™ latency and jitter measurements taken to real-world game servers. Game Scores are composed of eight components, each measuring a different aspect of consumer gaming experience. Each of these components is evaluated and scored on a scale of 0-100 for each eligible provider. Finally, scored components are combined in a weighted average to produce a Game Score for each eligible provider. |
Ookla retains ownership of this content, including all of the intellectual property rights, data, images, graphs, and analysis. This content may not be quoted, reproduced, distributed, or published for any commercial purpose without prior consent.
The definitive source for network performance, quality, and coverage intelligence