Price Benchmarking of Satellite Broadband

Authors: Christopher Baugh | Carlos Placido (independent adviser)

Satellite broadband is the one application whose mere existence hinges on pushing the limits of frequency reuse and spectral efficiency. With the proliferation of scalable HTS networks globally, broadband pricing is increasingly becoming –for better or worse– a barometer of regional and local satellite bandwidth pricing trends. Generational leaps in HTS technology modulate market price expectations, not just for broadband but also –often unfairly– for other SATCOM data applications, whether fixed or mobile. 

Unlike wholesale FSS pricing based conventionally on monthly cost per allocated MHz or guaranteed Mbps, satellite broadband plans are openly offered at retail levels to both residential and business users, but without disclosing underlying capacity considerations. Therefore, key questions arise as to how to systematically benchmark satellite broadband offerings, and how to infer effective bandwidth and cost metrics. 



 

In NSR’s Satellite Capacity Pricing Index, 7th Edition (2021) report (SCPI7), NSR analyzes bandwidth pricing per application, frequency band and region. In addition to wholesale-level pricing, NSR collected retail broadband pricing data from 248 HTS residential and business broadband service plans, offered across 44 countries by 9 major satellite broadband providers, namely Eutelsat Konnect, HughesNet, Kacific, Orbith, SES Astra Connect, Thaicom IPStar, ViaSat, Xplornet Canada and YahClick. Such database feeds an extensible Excel dashboard that allows a quantitative analysis of broadband service offerings, based on selectable filters that include service provider/s, region/s, specific countries and data rates. 

Based on filter selections, five inter-related principles drive the pricing analysis and dashboard visualizations: 

• Advertised Data Rate (Mbps): The pricing database includes offered downstream and upstream data rates, as advertised in the respective service plans offered by satellite broadband providers. These are the maximum allowed data rates, which are generally not guaranteed all the time. 

• Traffic Allowance (GB): As in cellphone subscription plans, allowance or (daytime) traffic caps denominated in Gigabytes (GB) are behind satellite broadband service plans, whether explicitly disclosed or not. Traffic volume allowance is key to determine the marginal cost per GB transported through the system.

• Effective Data Rate (Kbps): The effective data rate is calculated in the dashboard by dividing the monthly traffic allowance by the peak time hours across 30 days. NSR assumes that such peak-time window is 12 hours (editable).

• Average Cost per Gigabyte ($/GB): A simple but key metric obtained by dividing the monthly service cost (after temporary discounts) by the traffic allowance in GB. Note in the map the paradox that lowest $/GB figures are observed in high-ARPU countries like Australia and Canada. 

• Overbooking: The overbooking or contention ratio is not a straightforward metric, but implicit overbooking rates can be inferred by dividing the advertised downstream data rates by the effective data rate. 

 With the above-described analysis conducted on all collected service plans, one can drive interesting conclusions per region: 

• SA/SEA: South Asia and South East Asia rank first in terms of effective data rates and average cost per GB (lowest cost). This is driven by competitive offerings by IPStar and Kacific. Judged by these metrics SA/SEA is currently world’s most competitive satellite broadband region. 

• WEU: While Western Europe ranks highest based on advertised data rates, the region ranks third on the effective data rates given that traffic caps are lower relative to the advertised data rates.

• NAM: North America remains world’s largest and most mature market for satellite broadband but –interestingly– ranks second in terms of the effective data rate and third –after SA/SEA and WEU– in terms of average price per GB. 

• MEA: HTS service plans surveyed for Middle East and Africa have the highest cost per GB and lowest advertised speed. Note, however, that lower advertised data rates also place the region as the least contended. In other words, judged by the analysis, users pay more per GB but may expect a service that more consistently performs near the offered data rates. 

Since the SCPI7 dashboard has user-selectable filters, this type of analysis can be done more granularly in order to spot differences between service providers, countries, and service plans; or any combination. Commercially available service plans drive the analysis, so it does not include Starlink, which offers residential broadband in specific territories but on a trial basis without commercial plans yet announced. Nevertheless, the dashboard analysis is extensible to Starlink or other players (whether GEO or LEO) as additional service plans become available. 

Impact on Pricing by LEOs and VHTS 

Disregarding architectural differences between GEOs and LEOs, Starlink could soon become an interesting leading case for high-speed broadband. Residential users’ excitement about Starlink’s “better than nothing” trial while paying $99/month signals that in high ARPU countries, there is a market for high-speed sat-broadband, possibly at higher price points provided the advertised speeds are delivered more consistently and with low contention. 

The other major development to alter future satellite broadband speed and pricing remains in GEO. GEO-VHTS satellites planned by ViaSat and Hughes aim to achieve large economies of scale, minimizing the capital cost per production Mbps. Provided production capacity meets projected demand, these programs will be able to push the limits of bandwidth pricing, supported by software-defined (SDN) capabilities allowing operators to hedge bets as to where demand growth will be. This changes the paradigm with the net effect of further reducing the lifecycle cost per bit for satellites that live 15+ years.

Bottom Line 

NSR’s SCPI7 Satellite Broadband Dashboard provides a framework to benchmark broadband services, both current and future. Longer term, LEO constellations and VHTS programs will further commoditize satellite bandwidth, becoming more evident towards 2022-2023 when such programs gain traction and a subset of the LEO constellations complete their first shells and work out commercial distribution and landing rights. It is possible that we will see a gradual departure from region/country-level pricing, as ultra-local supply-demand dynamics and pricing power could modulate pricing. Software-defined capabilities are now focused on network-centric issues (beam steering, orchestration, interference avoidance) but for smart platforms it is only a matter of time before we start seeing “uberization” and AI come into the picture for satellite bandwidth pricing as well.