Communications Monitoring Report 2018

Link directly to Retail Fixed Internet Sector and Broadband Availability Retail Fixed Internet Sector and Broadband Availability

In 2017, the retail fixed Internet sector (hereafter, Internet sector) was the fastest-growing sector of all telecommunications sectors. Revenues grew by 7.7%, while subscriptions grew by 3.9% compared to the previous year. From 2013 to 2017, revenues for this sector grew at an average annual rate of 9.2%. The vast majority of revenues (80% or $8.8 billion) were generated by residential access services followed by business access and transport services (14% or $1.5 billion). Other types of revenues, such as revenues derived from equipment sales and rentals, as well as Internet applications, represented the balance and comprised only a minor part of telecommunications service providers’ (TSPs) revenues.

More Canadians are subscribing to Internet services, 86% of households subscribed to these services in 2017. Canadians continued to use more data; subscribe to faster, larger packages; and allocate more money to Internet services.

Subscriptions increasing by 4.2%.

Business Internet revenues grew at a much slower pace than residential revenues. However, more Canadian businesses subscribed to Internet service packages, with subscriptions increasing by 4.2%.

Fibre deployment continued, improving the availability of fibre–to-the-home (FTTH) Internet services from 29% to 35%. These deployments were mainly in large urban areas.

Internet service providers (ISPs) are categorized in three main groups: incumbent TSPs^{Footnote 1}, cable-based carriers^{Footnote 2} and other service providers^{Footnote 3}.

Link directly to i. Residential market i. Residential market

As mentioned above, more Canadians are subscribing to Internet access services. Canadians households continue to use more data, subscribe to faster, larger packages, and allocate more money to Internet services.

Link directly to Subscriptions Subscriptions

Residential subscriptions reached 12.8 million (…) which is approximately 3 times higher than the population growth rate

In 2017, a variety of ISPs provided Internet services, including traditional telephone and cable companies, fixed wireless service providers, and resellers. Residential subscriptions reached 12.8 million (86% of the 14.8M households), a 3.9% increase from 2016, which is approximately 3 times higher than the population growth rate^{Footnote 4}. Cable-based carriers and incumbent TSPs had the vast majority of the market by subscriptions (87%). Other entities continued to increase their subscribers, reaching 13%, up from 10% in 2013.

Growth rates in Internet subscriptions were approximately 2 to 4 times higher than population growth rates from 2013 to 2017.

Canadians continued to subscribe to Internet services with higher monthly data allowances. In 2017, while most subscriptions (51%) included monthly data transfer limits of 300 gigabytes (GB) or more, approximately 2% of subscribers had transfer limits below 20 GB. Average data use continued to increase concurrently, with a nearly 30% increase in average combined download/upload data usage since 2016, reaching 166.2 GB per month in 2017.

Canadian households paid approximately $94 million in Internet overage charges

Less than 5% of subscribers went over their limit in 2017

In 2017, Canadian households paid approximately $94 million in Internet overage charges, which represents roughly 1% of total residential Internet service revenues, and is down from the approximately $100 million paid in 2016. Less than 5% of subscribers went over their limit in 2017, consistent with 2016 levels.

All information in this section regarding gigabytes per month usage, and subscriptions by advertised speed and advertised download capacity, is from the larger ISPs. They reported approximately 87% of total residential high-speed Internet service subscriptions in 2017.

In 2017, 36% of high-speed residential Internet service subscriptions were for a service that met the Commission’s target speeds of 50 Mbps download and 10 Mbps upload, compared to 23% in 2016. Subscriptions to services meeting these speeds (50/10 Mbps), and with unlimited monthly data transfer, more than doubled to 28% of total high-speed subscriptions from 14% in 2016.

While the majority of Canadians did not subscribe to an Internet service meeting the Commission’s target speeds of 50 Mbps downstream and 10 Mbps upstream (50/10 Mbps service), they were increasingly subscribing to faster Internet services. Services with speeds slower than 16 Mbps went from 69% of high-speed residential Internet service subscriptions in 2013 to 36% in 2017, while subscriptions to services including speeds of 50 Mbps and higher went from just 5% of high-speed residential Internet service subscriptions in 2013 to 39% in 2017.

In 2017, gigabit services, defined as those with speeds of 940 Mbps and faster, accounted for 1.2% of subscriptions, and generated an average of $83.65 in one-month revenue, which is far below the reported lowest price for most packages for this type of service. This may be due to promotional activity and bundling. All of the reported gigabit services had unlimited data allowances.

While Canadians subscribed to faster services, they also consumed more data.

While Canadians subscribed to faster services, they also consumed more data. The average amount of data downloaded by residential Internet service subscribers increased by 31% between 2016 and 2017 to 153GB per month, and by an average of 36% annually from 2013 to 2017. Average upload amounts also increased by 15% in 2017, reaching an average of 13GB per month. These increases from 2013 to 2017, resulted in average uploaded data per subscriber doubling while average downloaded data tripled.

While 85% of households subscribed to broadband Internet service^{Footnote 6}, regional differences exist. British Columbia and Alberta led with around 90% of households subscribing to broadband Internet services, while Saskatchewan, Nova Scotia, and the North^{Footnote 7} trailed at around 77%.

Link directly to Revenues Revenues

The Internet service revenue share of the top five companies (Bell, Rogers, Shaw, TELUS and Videotron) declined slightly, from 74.8% in 2016 to 74.3% in 2017.

In the meantime, independent ISPs slowly increased their revenue share, growing from 8% in 2013 to 12% in 2017. This reflects their limited share of the residential Internet service market.

Independent ISPs and incumbent TSPs increased their share of residential Internet service revenues at the expense of the cable-based carriers’ revenues from 2013 to 2017. The independent ISPs were the only type of provider to increase their share of revenues from 2016 to 2017. Cable-based carriers continued to have the highest percentage of residential Internet service revenues.

Resellers consistently reported the lowest average revenue per user (ARPU), while the highest ARPU was attributed to the other facilities-based TSP category, which consists mainly of fixed wireless and satellite-based service providers.

The industry-wide ARPU increased by 5.5% from 2016 to 2017, with an average growth rate of 7.1% from 2013 to 2017.

ARPU may vary from Figure 5.9 below, which uses data from only the larger providers, who hold 89% of all high-speed residential Internet service subscriptions. ARPU is calculated based whole-year revenues and on year-end subscription data, not data from a particular month. This table also contains data from dial-up services.

Per subscriber revenues increased rapidly for the increasingly popular higher-speed Internet service packages. For instance, one-month average reported revenues for Internet service packages with download speeds of 50 Mbps and faster increased by 12% from 2016. This is especially impactful in the marketplace, as these packages represent 39% of the market, as shown in Figure 5.8. Meanwhile, packages with download speeds under 16 Mbps (36% of the market share) have seen one-month average revenues increase by 3.6%. The average revenues generated per subscriber with services offering speeds between 16 and 50 Mbps increased by 6.8% from 2016 to 2017. These services represented 26% of the Internet service market in 2017.

fibre-based Internet service subscriptions (…) went from 3% in 2013 to 12% in 2017

As part of their efforts to compete with cable-based carriers, incumbent TSPs continued to build fibre to the home (FTTH) networks and promote fibre-based Internet services. This resulted in a significant increase in their share of fibre-based Internet service subscriptions, which went from 3% in 2013 to 12% in 2017.

Link directly to ii. Business market ii. Business market

Link directly to Revenues Revenues

Independent ISPs, which are not affiliated with Canadian incumbent TSPs or cable-based carriers, had 24% of the access service revenues while having 13% of the subscribers in 2017. This may be due to a large proportion of independent ISP revenues coming from business information technology services providers that supply high-capacity connections to large enterprises.

$2.3 million in Internet overage charges

Canadian businesses paid around $2.3 million in Internet overage charges, with only around 1% of subscriptions going over their monthly limits in an average month.

Link directly to Subscriptions Subscriptions

From 2013 to 2017, the growth in the number of subscriptions to business access services was strong, and exceeded residential subscription growth rates (sometimes by a large margin), except for 2016, when changes in company reporting resulted in a change in results.

Link directly to Technology Technology

Cable-modem-based Internet access service consistently showed increases in market share over the period from 2013 to 2017. These increases made serving businesses over existing DOCSIS [Data Over Cable Service Interface Specification] cable networks an increasingly important part of the business of cable-based carriers. Other technologies, such as fixed wireless and satellite, had only a minimal share of the business market.

Link directly to iii. Broadband service availability iii. Broadband service availability

Broadband deployment improved since 2016 in some ways, but availability in sparsely populated areas did not improve as much as more dense areas, except for Long-Term Evolution (LTE)-Advanced coverage. Also, among other factors, newer, more accurate information about the extent of deployment reduced coverage in some areas.

Broadband service availability is calculated using information provided by ISPs. For 2013 to 2015, locations were considered to be serviced if their dissemination block representative point fell within an area of broadband service coverage. As of 2016, ISED pseudohouseholds^{Footnote 9} are used, along with 2016 census demography. Broadband service availability data may not take into account capacity issues or issues regarding line of sight.^{Footnote 10}

Unless otherwise noted, broadband service availability figures exclude wireless mobile technology. “Satellite access services” in this section refer to direct-to-home (DTH) satellite, and not to the technology used to connect communities to the Internet (e.g. satellite link transport).

Fibre-based Internet service availability (…) [went] from 28% in 2016 to 35% in 2017

Fibre-based Internet service availability continued to increase, going from 28% in 2016 to 35% in 2017. Incumbent TSPs used their fibre infrastructure to make gigabit service available to over 3.9 million households, while cable-based carriers used mainly DOCSIS 3.1 technology to make gigabit service available to over 6.6 million households. However, in general, fibre-based gigabit services have far faster upload speeds than their DOCSIS-based counterparts.

Incumbent TSPs and other non-traditional television providers continued to increase the availability of IP technology-based television service (IPTV), proving a source of competition to traditional cable-based systems, while leveraging their broadband infrastructure to provide services outside of Internet and legacy phone service. These FTTH deployments occurred mainly in large urban areas.

the availability varied greatly between urban (97%) and rural areas (37%)

Services at speeds meeting or exceeding the Commission’s target of 50 Mbps download and 10 Mbps upload with an option for unlimited monthly data transfer were available to 84% of Canadian households. However, the availability varied greatly between urban and rural areas, with only 37% of rural households having access to this kind of service, versus 97% in urban areas. Subscriptions to a 50/10 Mbps service with unlimited monthly data transfer more than doubled to 24% of Canadians households, compared to 11% in 2016.

The total footprint for all areas with access to broadband service speeds of 30 Mbps and above and 50 Mbps and above did not increase in 2017, but slightly declined due to more precise information filed with the Commission (e.g. the use of precise coverage instead of modeling).

British Columbia, Quebec and Ontario led the way in terms of availability of 50 Mbps service.

Availability continued to vary by province when it comes to higher-speed broadband services. Saskatchewan and the North had less coverage at the 50 Mbps level. British Columbia, Quebec and Ontario led the way in terms of availability of 50 Mbps service.

The vast majority of areas within the provinces and territories that had broadband service had it at 5 Mbps or faster, with the exception of Nunavut, which had virtually complete coverage at 1.5 Mbps speeds, but far less at 5 Mbps or faster. In fact, only 29.9% of households in Nunavut had access to speeds of at least 5Mbps and none had access to 10Mbps or higher speeds.

only 39% of rural households [have] access to services with download speeds of 50 Mbps or faster, and small population centres [have] 88% availability

Rural and small centre populations continued to trail in the availability of advanced broadband services, with only 39% of rural households having access to services with download speeds of 50 Mbps or faster, and small population centres having 88% availability. This is in contrast to near-ubiquitous availability of such services in medium and large centres.

Advancements in the deployment of rural broadband were mainly in the 16+ Mbps and 25+ Mbps categories, going from 61% to 69% for 16 Mbps or higher availability, and from 60% to 66% for 25 Mbps for higher. This is due mainly to continued deployments of LTE-based fixed wireless technology. Deployment in lower-speed categories did not increase as appreciably, due to deployment being centred mainly on already built areas with slower service speeds. Due to the filing of more precise data replacing previously used coverage models, the availability of some service speeds decreased slightly in 2017.

Link directly to iv. Broadband applications iv. Broadband applications

Link directly to Data usage Data usage

Streaming services

Streaming services over the Internet consume data at different bitrates. The bitrates are measured in megabits per second (Mbps) or gigabytes per hour (GB/hr). Generally, video streaming services consume data at higher bitrates than audio streaming services because the video stream contains more information.

Usually, a higher bitrate can also enable a higher-quality audio or video stream. However, depending on the end-user device, a higher-quality stream may be indistinguishable from a lower-quality stream. For example, on a smartphone, an ultra-high-definition (UHD) video stream may not show a perceivable difference when compared to a lower-resolution video stream due to the relatively small screen size.

Figure 5.20 and Figure 5.21 illustrate the average and the range of the amount of data some streaming services can use on an hourly basis over wireline (broadband) and wireless (LTE) Internet connections, respectively.

The rates at which streaming services consume data can vary significantly.

While some services allow the user to manually control their quality and how much data they use, other services are set automatically. See methodology and observations below for more details.

Limited services and data overage fees

When a customer with an Internet data plan with a data cap exceeds their monthly usage limit, overage fees are applied.

Generally, usage limits are significantly lower and overage fees higher for wireless data plans compared with wireline broadband Internet plans. Currently, the average overage fee per GB is $2.38 for wireline broadband Internet services and $55.00 for mobile wireless services.

To address bill shock associated with mobile wireless overage fees, the Wireless Code requires service providers to suspend data overage fees once they reach $50 within a single monthly billing cycle, unless the account holder or authorized user expressly consents to pay additional charges.

Figure 5.22 and Figure 5.23 below illustrate how overage fees (average and range) can translate to the usage cost of some services ($/hr) over wireline and wireless Internet plans, respectively.

When reading the charts below, it is important to note that customers incur overage fees only if they exceed their plan limits. Customers can avoid paying any overage fees if they use less than the maximum amount of data included in their plan each month. Service providers offer plans with various limits and, in some cases, it may be in the consumer’s best interest to change to a plan with a higher limit if they continue to exceed their usage limit each month.

Since many wireline plans include unlimited data usage, Figure 5.22 shows the lower range as zero.

The “Wireless code data overage cap” line in Figure 5.23 illustrates how much of each service a customer could use, after they hit their plan limit, before hitting the overage cap.

Link directly to Unlimited services and fair/acceptable-use policies Unlimited services and fair/acceptable-use policies

the Wireless Code prohibits a service provider from applying overage charges and from putting any limit on the use of such a service purchased on an unlimited basis

For unlimited wireless plans, the Wireless Code prohibits a service provider from applying overage charges and from putting any limit on the use of such a service purchased on an unlimited basis, unless these limits are clearly explained in the fair-use policy.

For such plans, the fair-use policy may reduce the data connection bitrate once a specified limit has been reached until the monthly billing cycle renews. Bitrate reduction to the end-user usually results in a service degradation so that basic services (e.g. email, web browsing) continue to work, but other services requiring faster connections (e.g. streaming services) are disrupted.

Figure 5.24 compares the average bitrate requirements of some services to some fair-use-policy bitrates for unlimited wireless plans. Generally, if a service’s bitrate requirement exceeds the fair-use-policy threshold, the service will be degraded or disrupted, or become unusable.

When reading the chart below, it is important to note that customers can avoid triggering bitrate reductions by not exceeding the usage limits specified in their plans.

Link directly to Methodology and Observations of broadband measurement Methodology and Observations of broadband measurement

Methodology

To collect data for this sub-section, the CRTC used a test environment that aims to replicate how a typical consumer would utilize online streaming and real-time communications services. The services were accessed by a typical wireline residential broadband service, and a national LTE cellular data network, using mainstream off-the-shelf consumer electronics: Android- and iOS-based tablets and phones, smart TVs, Windows-based laptop and desktop computers, and various set-top streaming devices. A web browser was used to access the streaming services on the PCs, and official applications (apps) were used on the other devices.

To measure the data consumed by these services on the wireline connection, a specially-configured Windows-based computer was inserted between the Internet connection and the local network. Using data traffic measurement tools, all data flowing between the test device and the Internet was captured for analysis. For the LTE connection, readily available applications as well as integrated capabilities within the phones operating systems were used to measure the data traffic.

The maximum bitrates of the wireline and LTE Internet connections were tested, and found to be significantly higher than the maximum observed bitrates of the streaming services tested; in other words, the Internet connections did not limit the bitrate of the streams in any way. To ensure accuracy, multiple measurements were taken for each service and quality level (where this setting was available), and background data usage (i.e. the usage of background apps and services, other than the one being tested) was minimized.

Observations

The streaming services deliver their data in different patterns, some with single large bursts with gaps in between, and others with more continuous data. The measurements were conducted over sufficient periods of time such that a representative average could be obtained. It should also be noted that most services can modify their burst size dynamically thus making the average rate an important factor in determining the data rate requirement for most services.

In a typical consumer scenario, the available bandwidth at any given moment can vary due to numerous reasons, including resource sharing between multiple devices on a home network or moving around to/from different coverage areas on a mobile network. Although the end-user’s internet connection is one factor in determining the quality and stability of a stream, other factors can include network congestion, server load, network/server latency, and end-user device capability. In addition, many services can dynamically and automatically adjust their quality (and therefore the amount of data consumed) based on several factors that contribute to a stable audio and/or video stream to the end-user.

Due to the limited number of samples and the diversity of network configuration and equipment, the reported values in this section should be viewed as average-case estimates, not worst-case limits.