A few weeks ago, the LoRa Alliance announced that it had achieved coverage in 100 countries, thanks to its pretty extensive ecosystem. At the time, we wrote that despite Sigfox’s recent struggles, one could never rule the company out of the running. A couple of weeks after that, Sigfox announced a deal with Groupe PSA and IBM, somewhat affirming our position, and after that article, Sigfox’s CEO Ludovic Le Moan reached out, to talk about annual results and the plan for Eutelsat to launch a constellation of satellites to expand the network footprint.
We began by asking for details of the Groupe PSA deal, which apparently constitutes 250,000 devices. Le Moan said that Sigfox was working with IBM on other similar deals, some of which will be announced soon, involving parcel tracking and pallet tracking for German automakers. Airbus and Michelin were specified.
Le Moan also drew attention to the manufacturing deal that Sigfox struck with Alps Electric, which sees Alps use Sigfox as its prime LPWAN connectivity technology in its customer designs. To this end, Alps is making Sigfox trackers, and DHL is apparently a big customer for Alps, and is set to be one of the soon-to-be announced deals.
It is Le Moan’s view that asset tracking should be the main focus of the IoT for the next few years, as it is the easiest way to demonstrate a return-on-investment to skeptics, and to that end, we agree. Asset tracking is a good foundation on which to add more functions, such as environmental sensing, and begin tying this operational data in with established business processes in the back-end, where nifty analytics tools can be used to identify opportunities or costs that could be cut. Starting small and working up the value chain is the best way to ensure leadership buy-in.
As for the satellite project, Le Moan said that given his experience over the past 15 years, it is clear that full global coverage is needed. To this end, a terrestrial-only approach will have obvious gaps where water and inhospitable terrain are encountered, by a satellite option would alleviate this problem.
This is where Eutelsat comes in. Le Moan said that the power usage on the end device would be in the 15-20mW range, connecting to the satellites in a low-earth orbit (LEO) constellation of 30-50 satellites, around 600-700km above them. This allows these devices to send Sigfox messages, with use cases including livestock monitoring in countries like Australia and Argentina – where huge ranches are quite challenging to cover.
Le Moan said that the vision for Sigfox was still the one-stop shop approach, with his particular obsession being to keep the prices low, focused on the small messages. With Eutelsat on board, as a full Sigfox Network Operator (SNO), Sigfox should be able to sell a sweeter service to customers – an SLA-backed guarantee that a device will always be able to send a message.
Of course, Sigfox has contractual agreements with the SNOs, to the extent that there are mandated footprint requirements, such that they can advertise a country as ‘covered.’ In return, the SNOs get the bulk of the revenue derived from the connected device. Around 40% of the revenue goes to Sigfox itself, which also makes money by selling the base station equipment to the SNOs – as Sigfox controls that IP, although it did just open up the end-device radio protocols.
As for the Sigfox results, there are 6.2mn active devices on its network, sending 13mn daily messages. The target for 2019 is to hit 10mn, but Le Moan believes it could reach 12mn. In 2020, it hopes to double the 2019 figure, but the official press release mentions its ‘ambitious 5-year strategy.’ Called 1B23, the plan is to connect 1bn devices by 2023, which would require truly exponential growth.
Last year, Sigfox copped to having just 2.5mn active devices, and €50mn in revenue, up from around 1.5mn in 2016 according to our calculation and missing the 2017 revenue target by €10mn. Riot covered that climb-down at the time, noting that this was far from the expected number of devices.
Since that announcement, the company has taken a kicking in the press, as we outlined in the recent LoRa discussion. When asked about the high leadership staff turnover, Le Moan said that he had hired people that were good for the early stages of company growth, but that Sigfox now needs different people with different skills to scale. He said that it was, of course, difficult to let the early employees go, but that the team now looks properly aligned, in terms of vision. He noted that it has taken some time to get there, and that building a company from the ground up is exhausting – but that he thinks Sigfox has pulled it off.
Another criticism that we have leveled at the company in the past was that Sigfox had never managed to find a SNO partner for the USA and Germany – two of the largest markets for its services, but two that apparently couldn’t find anyone wanting to run the Sigfox network in the territory.
We put this to Le Moan, who said that the issue was that Sigfox couldn’t find partners for these two critical territories early enough, and so was forced to run them itself in order to get the ball rolling. Le Moan said that it couldn’t let them go, so had to set them up, but that now there is a potential SNO for Germany, which would leave Sigfox running only the USA and its domestic French market.
We asked if Sigfox had any appetite for moving to higher bandwidth communications, to which Le Moan said no – that Sigfox had technology that was perfect for small and slow messages, and that adaptions would require fundamental changes. It is focused on improving its power consumption, but not in increasing bandwidth.
Similarly, the goal of the “0G” network positioning is that it provides customers to always connect to the global network, which opens up opportunities for Sigfox to serve as an ‘out-of-band’ tool for redundancy and backup connections, or even as a function to provision a device once it is sent to a buyer.
To this end, the service has to remain low-cost, if it is going to be adopted at scale. As such, Le Moan explained that keeping the exclusivity deals with the SNOs is key, as having two rival SNOs competing would increase the price of a connection, as the cost of twice as much capex and opex would be felt by the end-customer.