Being a SaaS (Small Analytics company) in the Earth Observation market

The title of this blogpost should have referred to the conference we attended two months ago. However, I decided to extend the topic to the general talk about the market of Earth Observation analytics. At this moment the community people are discussing whether or not the SaaS model is applicable to the Earth Observation, dispirited by the sad final of Descartes Labs — the analytics company with the large invested capital and powerful products as it’s told in the requiem by their former CEO.

Actually, Earth Observation (EO) is an umbrella term for the very large number of remote sensing technologies and the analytics is supposed to reveal the value of the EO data for the numerous domain applications. Let’s represent the market value-added chain, like this:

This scheme can be addressed to the much more sophisticated and well-designed market classification by Aravin Ravichandran (the author of TerraWatch Space podcast, very recommended).

Back to so called analytics layer. If you build some SaaS product to automate the analysis you have to be cautious about your market positioning as long as the Data providers rule the game. Whenever it comes to SaaS / Analytics — it’s somewhere between or across the Distribution platforms and Applications, but far behind the sales Distribution (“data resellers” according to Aravin) where the distributors negotiate with every client — actually providing the consultancy services as well.

The thing why I started with this, having said that I’m not a great market designer — is to show you that SaaS highly depends on the Data as a service model and both contribute significantly to market development by increasing the data consumption and engaging more analysts to start using EO data.

The commercial model

In his most cited article, Joe Morrison criticized each part of this commercial Distribution model on which we are based: Distribution, Licensing, and Pricing — claiming Data providers are those who limit the market growth and build up the long creepy procedure of emails— sales people — FTP download with every single purchase of satellite image.

I would step into the shoes of Data providers to mark what has changed since the good old “DVD” days.
Of course, we are not in the market of Netflix where movie streaming services replaced good old DVDs… but things are changing even here.

• Distribution:
  Providers like MAXAR moved their most recent archives to the cloud. (Maxar guys told me that in some years they were named as the largest customer of AWS, breaking down their distribution model while their cloud costs were skyrocketing). There are petabytes of older archive data that are ready to be ingested into the cloud on demand — still, it can take an unpredictable amount of time in our experience — but a lot of data analysts (and developers) can connect online to the most recent and high-resolution imagery captured by the satellites.
• Licensing:
  Data licenses have always been the stumbling block in EO market. The Data providers sell a lot of premium satellite imagery to the Internet giants. And the companies like ESRI or Mapbox have their own license terms for the data. Mapbox and Bing were the first who allowed this data to be freely used for digitizing and editing maps for Openstreetmap. Thereafter MAXAR started supporting OSM directly.
  To summarize, there are two main options regarding the licensing / usage of EO data (not to mention the smaller alternatives like “open data program”):
  - Buy data for the full price, go for the partnership with some Data provider
  - Use the data within the confines of and under the specific terms of some products (like OSM, ArcGIS, etc.)
• Pricing:
  When Morrison blames satellite companies for being greedy about their archive data — you shouldn’t forget to take into account the operational costs related to the Distribution and the cloud costs if it comes to the large data storage. The options here are the same as what we have for the Licensing: place a single order and pay the full price to the commercial imagery distributor, otherwise, get this data as a service and pay for the license to the software vendor (up to $100K / annually for some ESRI’s products).
  Remember video and music subscriptions on demand? If you can stream movies you don’t need to download’em…

The streaming services that can be integrated into third-party workflows and applications are definitely the new and trending business model. However, the reality is that the major part of this data is still distributed as files. Go catch the new fascinating opportunities… but change the customer’s mindset first.

To be or not to be the SaaS company

There are much more fishes in the EO analytics market than you can see in the picture, and the ocean doesn’t seem very blue… 😬
To name a very few of those who we might consider our competitors (which means we strive to do our best to develop the market together): Eofactory, Picterra, Luxcarta (BrightEarth)... What proves them to be SaaS companies — one can start playing around with the product as soon as he has an account registered (not so true of Luxcarta, you still need to contact them first). On the other side, this is also the significant number of open-source tools and models available. Now there are also Microsoft and Facebook covering the whole World with open datasets and making your professional margin miserably low.

Is there room for the new SaaS product to grow up in such a market environment? What it is to make revenue with SaaS between the open data and large tailored enterprise solutions? Look, when Mark from Descartes Labs is regretting the way they wouldn’t but could be the “modern AI consulting” company — he is talking about large customized enterprise solutions, isn’t he?

And why do we believe that this model is applicable to EO market and strive to prove it on our own? We can freely speak it out as all our growth based on a very little capital (mostly bootstrapping), being a small analytics and software development company. Yet we define our business model as a trade-off between survival and a strategic product development. To be ore specific, Mapflow has two revenue streams: Subscriptions and Tailored Projects (services). The first category comprises all customers who purchase Mapflow’s subscriptions, the second category constitutes the customers who order custom solutions powered by Mapflow requiring its extension, model customization or integration with external products. If we would want to maximize our revenue we would go for the large enterprise solutions (modern AI consulting) but we are supposed to fall into the Project’s category. Our optimistic forecast is based on the assumption that the contribution of Mapflow Subscriptions (SaaS) will rise with the increase in the number of clients and their average bill, while the contribution of the tailored project will steadily diminish as Mapflow’s functionality evolves and starts covering more use cases.
The following gives us this optimism:

• The market. There is still poor adoption or consumption of Satellite / UAV data (❗️Earth Observation is about 2% of all Downstream Space applications vs 14% Upstream, see this report). That means that the Data proposal and the growing number of sensors are far exceeding the demand. Developing this market through the services is a lengthy road, so the innovators should be prepared for the long-term commitment to create an impact and turn the ideas into commercial reality. Building a SaaS / B2B product looks like a viable and reasonable strategy here.
  If it doesn’t challenge you, it doesn’t change you (c) — as fitness coaches are used to repeating.
• The technologies. The commercial potential of some disruptive technology is a tricky question. When it comes to the market, things are changing rapidly, and “getting outside the building” means stop getting funds from the research institute or your regular work and start talking to your first customers.
  In general, there is a demanding and rapidly growing market for EO / Deep learning technologies but this market is becoming very competitive and the AI-powered applications become IT as usual with its fierce competition for customer loyalty.
  Providing a cloud-based, no-code imagery analysis platform featuring AI-models and commercial imagery integrations is the most scalable approach we can figure out.
• The community. What drives us the most, is the opportunity to cooperate with data analysts and mappers across industries who explore the World in 2D, 3D, time-series, etc., including Data providers and Software vendors, UAV enthusiasts, and cadastre engineers. The community has great potential to evolve.
  What used to drive the market at the early age of Geospatial technologies — was the mass adoption of Navigation Services and Digital Maps. Over the past several years the market has been disrupted again by the increasing number of affordable Earth Observation sensors and imagery processing (AI) methods making it more realistic to update maps in quasi-real time.
  As they used to say in Openstreetmap — “The map is never complete”. Not only because of the blanks or incompleteness of the map features. There are changes that happen on all over the World. And having them timely detected and interpreted in the way of location intelligence is the Map no one can complete but everyone can contribute to.