iana.ca

Learn

How to read the government availability map

The federal broadband map is the official word on who can get what. It's useful, but it's built from provider-reported service areas, statistical households and a 75% rule, and in our view it tends to make coverage look better than it is.

Innovation, Science and Economic Development Canada (ISED) publishes the National Broadband Data, collected with the CRTC and documented on its Universal Broadband Fund pages. It’s open data, and this site uses it on the home map and every community page.

Hexagons and roads

Hexagons of roughly 25 square kilometres report only 50/10 and 5/1; one marked 5/1 may have faster service. Road segments about 250 metres long get the best tier reported along them: 50/10, 25/5, 10/2, 5/1 or none. Our community pages use the wired roads layer (cable, DSL and fibre); fixed wireless, cellular and satellite aren’t counted. A tier counts only when both download and upload are met.

Hexagons: about 25 km² each 50/105/1none tiers: 50/10, 5/1, none scope: unspecified in the data dictionary Roads: about 250 m segments 50/1025/510/25/1none best tier reported along each scope: wired only (cable, DSL, fibre) none: no wired service reported wireless, cellular, satellite: not counted

Top: three hexagons, each about 25 square kilometres, shaded by the only tiers the hexagon layer reports: 50/10, 5/1, none. The hexagon layer's scope is unspecified in the data dictionary. The hexagon layer only reports 50/10 and 5/1; a hex marked 5/1 may have faster service below 50/10. Bottom: a road chopped into segments of about 250 metres, each shaded by the best tier reported along it: 50/10, 25/5, 10/2, 5/1, none. The roads layer used here is wired only (cable, DSL and fibre); none means no wired service reported, and fixed wireless, cellular and satellite are not counted.

Two shapes of the same data. Tiers are download/upload in Mbps: 50/10 means at least 50 down and 10 up. Hexagons are about 25 km² each. The hexagon layer only reports 50/10 and 5/1; a hex marked 5/1 may have faster service below 50/10. Road segments of about 250 m each get the best tier reported along them; the roads layer this site uses counts wired service only, so none means no wired service reported, not no internet at all. Illustration Source: National Broadband Data, Open Government Portal (ISED)

The 75% rule

A hexagon gets a tier when more than 75% of households in it can get that speed, so one shaded 50/10 can have up to a quarter without it. ISED says it’s possible not every household has access to the service indicated. In a rural hexagon, in our view, that quarter is likely the farms and acreages at the end of the line.

The 75% rule 50/10 4 without 50/10 16 of 20 households can get 50/10: 80%, more than 75%, so it shows 50/10 In our view: the 4 left out are likely farms and acreages at the end of the line

A hexagon with 20 households along its roads. 16 of them, along the main road and a short lane, can get 50/10; that is 80%, more than 75%, so the whole hexagon is shaded 50/10. The other 4, down a lane near the edge of the hexagon, can't. A note marked as our view says those are likely the farms and acreages at the end of the line.

The 75% rule. A hexagon is shaded 50/10 when more than 75% of its households can get 50/10. Here 16 of 20 can (80%), so the whole hexagon shows 50/10, and 4 homes can't. In our view, in a rural hexagon, those are likely the farms and acreages at the end of the line. Illustration

Modelled homes, reported footprints

Instead of real addresses, the data uses pseudo-households: statistical locations from Census counts, placed along roads; ISED says it’s accurate to within 250 metres in most cases.

Modelled homes, reported footprint provider-reported footprint modelled household road segment: 50/10 your house: can't order it each road segment: about 250 m ISED: accurate to within 250 m in most cases

A provider's reported service footprint covers part of a road. A modelled household sits on the road inside the footprint, so the road segment of about 250 metres next to it shows 50/10. A real house down a lane that branches from that served segment, outside the footprint, can't order the service. A note says ISED describes the data as accurate to within 250 metres in most cases.

Modelled homes, not real addresses. The data places statistical households along Statistics Canada's road network, refined with provincial address data where available, and gives each road segment a tier from them. A road can show as served because a modelled household falls inside a provider's reported footprint, even if your house, down a lane from that served segment, can't order the service. The footprints come from facilities-based providers (the companies that own the networks). Illustration Source: ISED, National Broadband Data information

Service areas are those publicly advertised and available, as reported by facilities-based providers. ISED doesn’t run speed tests for it. So a footprint can include homes needing a new drop, or where equipment is full; tiers are plan speeds, and on DSL real speed falls with the length of the copper line. ISED says providers correct their data directly through the CRTC. We haven’t found a process on ISED’s page for a household to challenge its own square.

In our view, it overstates availability

In our view, the errors mostly run one way: the 75% rule rounds up, modelled households smear coverage along roads, footprints are generous and plan speeds are best cases. None is dishonest alone; together they make the map look better than reality, especially outside cities. In Sturgeon County, Cybera found mapped 50/10 where speed tests showed under 5/1. In rural southern Alberta, RMA speed tests in 2022 typically came in around 19/2, against the 50/10 target.

Using it well

Check the retrieval date on each map: new builds take time to appear, and old claims linger. Ask whether fibre goes all the way to the building (coax vs DSL vs fibre). The About page explains how to send a correction.

Sources