nPerf Coverage & Speed Maps (2G/3G/4G/5G)

It is the answer to 'how good is my mobile network really' - collected by the people standing in it. nPerf operates a global crowdsourced measurement platform whose iOS, Android and web apps run download, upload and latency tests from ordinary handsets in ordinary places, then aggregate those geolocated results into interactive maps: a combined 2G/3G/4G/5G signal-coverage layer, bitrate speed maps, and a dedicated 5G view.

The structure underneath is what makes it usable rather than merely pretty. Each market resolves to its own country page - India's signal map being the one verified end-to-end this pass - and within it each major carrier holds a dedicated layer: Jio Mobile, Airtel Mobile, Vi Mobile and BSNL Mobile for India, with rosters varying by country. City pages drill below the national view for Mumbai, Delhi, Bengaluru, Kolkata, Chennai, Ahmedabad, Hyderabad and Pune among other metros. Behind the display sits a discipline most coverage maps skip: GPS precision must be 50 m or better for a point to count toward coverage, 200 m or better for a download-bitrate point, and the whole test pool rolls on a two-year window purged monthly, so yesterday's dead zone does not haunt today's map.

Within the Datadory catalog - 1,744 datasets averaging 7.81 - the integrated telecommunication services slice holds 8 cataloged datasets averaging 7.12, and this record scores 6/10: unusual breadth across operators, technologies and metros, offset by inferred rather than publisher-documented field definitions. Get a sample of this dataset and we return rows shaped exactly like the dictionary below, cut to the markets you name.

Records arrive keyed to a layer, not to a picture:

# one map layer, keyed the way a delivery arrives
country_code     : IN
city             : Mumbai
operator         : Jio Mobile
technology_layer : 5G
measurement_type : bitrate

# same market, rival carrier, same shape
city             : Mumbai
operator         : Airtel Mobile
technology_layer : 4G+
measurement_type : bitrate

# rural-style row where only signal exists
country_code     : IN
operator         : BSNL Mobile
technology_layer : 3G
measurement_type : signal

# latency measurement rides the same key
city             : Delhi
operator         : Vi Mobile
technology_layer : 4G
measurement_type : latency

Read the anatomy, not the values above - those are illustrative shapes, and the live rows ship with your sample. One record fixes whose network (Jio Mobile, not 'the Indian market'), which radio (5G, distinct from 4G+, because marketing merges what engineering separates), what was measured (signal presence versus download bitrate versus latency - a bar with five bars of signal and no throughput is exactly the case these fields exist to expose) and where (Mumbai at city grain, or the country view without it). Because every carrier wears the same key structure, a Jio-versus-Airtel comparison in one city is a group-by rather than a research project.

Five fields carry every record: operator names the carrier behind the layer; technology_layer sets the radio filter across no-coverage, 2G, 3G, 4G, 4G+ and 5G; measurement_type splits signal, bitrate, upload and latency measurements apart; country_code gives the ISO market code; and city is the optional drill-down target.

Additional fields on request: the geolocated test-level measures that feed the tiles - coordinates, GPS precision, timestamp and the numeric results per measurement type - plus operator rosters beyond India's four verified carriers. Their exact shapes are pinned against live records when your sample is cut, which is also where column naming locks for your pipeline.

Three chips summarize the footprint:

• Geography: a global map index resolving to dedicated country pages - India, France and the US among dozens - with city-level drill-downs for Mumbai, Delhi, Bengaluru, Kolkata, Chennai, Ahmedabad, Hyderabad, Pune and other metros, and per-operator layers whose rosters vary by market (four verified for India).
• Time frame: a rolling two-year window of app-test observations, purged monthly. The maps describe current conditions rather than preserve history, which is precisely why they stay honest about today's network - and why long-run trend work needs a complementary statistical series rather than this one.
• Granularity: individual geolocated tests aggregated into map tiles, splittable per operator, per technology layer and per measurement type. Quality gates apply before aggregation: 50 m GPS precision or better for coverage points, 200 m or better for download-bitrate points.

Set against the wider Datadory catalog - 1,744 datasets averaging 7.81 - this industry slice holds 8 cataloged datasets averaging 7.12, and this one scores 6/10, carried by breadth across operators, technologies and metros rather than by documented schema depth.

API, files, or your warehouse. Daily, weekly, or hourly.

The maps move fast upstream; your delivery cadence is set separately, to match the decision you are feeding - and changing it is a settings conversation, not a re-integration project. Pick API for live lookups, files sized for overnight loads, or a direct pipe into Snowflake, BigQuery or Redshift.

Every delivery ships the field dictionary unchanged and sample rows for validation, flattened so one record equals one operator-technology-measurement combination for a place - which makes a carrier-versus-carrier comparison in a single city a group-by rather than a scraping-and-parsing project. Name the markets, operators and cadence you care about and the sample comes back shaped to them before any commitment.

Ranked by how directly one record settles their day job:

1. Market researchers and consultants. Benchmark operator 2G-5G coverage and bitrate per country and city inside telecom market studies, with a crowdsourced evidence base behind every ranking.
2. Competitive intelligence teams. Track rivals' coverage and speed positions market by market and counter a competitor's coverage claim before it calcifies into customer perception.
3. Data scientists and ML engineers. Model network-quality geography from geolocated tiles, stating the two-year rolling window as a constraint rather than discovering it later.
4. Journalists and academics. Cite per-operator crowdsourced maps in service-quality reporting - independent measurement, not the operator's own press kit.
5. Investors and quant researchers. Read network-quality gaps between carriers as an alternative signal beside the subscriber counts everyone already has.
6. Sales and growth teams. Qualify territories by the networks customers actually get before routing field teams into them.

For contrast inside the same industry: World Bank WDI - Mobile cellular subscriptions counts how many SIMs exist per 100 people but says nothing about how well they connect; ITU World Telecommunication/ICT Indicators supplies official country time series; and Wikipedia - List of Mobile Network Operators names who operates where. None of them measures whether the network under a user's thumb actually performs - which is the question this record exists to answer.

Market researchers and consultants score 3 out of 3 in the persona pack here: operator-by-city benchmarking is the core use case, and this is the measurement layer beneath it. Competitive intelligence teams also hold relevance 3 - coverage claims are competitive currency in telecom, and this is the counter-evidence source. Data scientists and ML engineers (relevance 2) get geolocated test tiles for quality-surface models. Journalists, academics and students (2) get citable per-operator maps for service-quality stories. Investors and quant researchers (1) and sales and growth teams (1) round out the pack - alternative-signal reads and territory qualification respectively. Persona-by-persona detail lives on the industry hub at /industries/integrated-telecommunication-services.

Inside this slice the datasets answer different questions, and mixing them up is how bad analysis gets written. World Bank WDI - Mobile cellular subscriptions scores 10/10 and answers how many - subscriptions per 100 people across decades of country-years. ITU World Telecommunication/ICT Indicators answers what the administrations reported, harmonised globally. This record scores 6/10 and answers something none of the statistical series can: how the network performs where a person is actually standing. The trade-off is explicit - less documentation depth and a shorter memory (two years, rolling) against ground-truthed, operator-split, street-level measurement. Adoption statistics and quality measurement complement rather than compete, which is why serious telecom work runs both.

Four things, stated up front.

First, this is current-state measurement, not an archive. The test pool rolls two years and purges monthly, so the maps describe now. If your question spans years, pair this record with a statistical series such as World Bank WDI or ITU indicators rather than substituting one for the other.

Second, the verified dictionary covers five fields; anything deeper - the geolocated test-level measures behind the tiles, or operator rosters beyond India's four confirmed carriers - is pinned down against live records when we cut your sample, which is also where column naming locks for your pipeline.

Third, operator coverage varies by market. Four Indian carriers are verified end-to-end this pass; other countries' rosters were not enumerated, so tell us which markets you need and the sample confirms what exists there rather than promising blind.

Fourth, the illustrative rows above show shapes, not values - live records arrive with your sample so validation happens against the real thing. Get a sample of this dataset scoped to your markets, operators and cadence - daily, weekly, or hourly - before anything is scheduled.