TerraWatch Space

For more analysis and insights on EO, become a Pro subscriber of the @terrawatchspace newsletter. newsletter.terrawatchspace.com/earth-observat…

Some big M&A news in Earth observation: Canadian space industry leader MDA Space is acquiring approximately 70% of French EO analytics firm CLS for €567M in cash - making it one of the bigger transactions in the downstream EO market. I think this deal says two things about the business of EO today: 1️⃣ EO analytics is still mostly a services business, not a scaling SaaS business. The projected 2026 revenues of CLS is €286M, suggesting a 3-4x multiple. Software companies are valued several times more, because their margins are in the 60-80% range and they hold as they grow. But, most of EO businesses are generally services firms, where growth means hiring more people, with margins in the 20-30% range. 2️⃣ AI for EO is still an unanswered question. Nearly every downstream EO company now calls itself AI-driven, on the assumption that AI turns analytics into software, and software into software margins. Yet CLS - profitable, growing 22% a year since 2023, with a customer base very few EO analytics firms can match - was still valued as a services company. So far, AI has changed what EO products can do and how quickly we can get to a result. Whether it changes what they cost to deliver is the question the sector has not answered.

For more EO insights and analysis, subscribe to the @terrawatchspace newsletter! newsletter.terrawatchspace.com

I first created the 5Ms Framework a couple years back to make it easy for those outside the Earth observation and geospatial bubble to be able to understand the various applications of satellite data. But since then, it has found new purposes in our advisory work with startups, investors and enterprise users - some examples: 1. AI and EO foundation model fit AI fits best where the task can be learned from data. Mapping and Modelling are the strong fits as pattern recognition and prediction are what these models do best. Measuring, Monitoring and Managing are weaker: Measuring use cases need continuous validation against real-world reference data. Monitoring use cases need context as models can flag change but not judge whether it matters. Managing needs human judgement and coordination. 2. Edge computing fit Highest fit where latency matters and the output is compact. Monitoring and Managing are the natural areas for onboard processing. Mapping, Measuring and Modelling are the weakest fits: Mapping use cases are archive-driven, the Measuring ones require heavy calibration and the Modelling ones are data and fusion-heavy. 3. Business model fit The business model shifts left to right - from selling data to selling outcomes. Mapping and Measuring are essentially subscription models for a data layer. Monitoring is a managed service, priced per asset or area. Modelling produces a prediction of specific variables, monetised through API access. Managing is scoped to one event, generally through premium, project-based models.

July is here! Are you curious to know the Earth observation, geospatial and remote sensing events happening globally in the second half of 2026? Check out our consolidated list of conferences, workshops, symposiums and more: terrawatchspace.com/eo-conferences/ If there is a relevant one missing, please submit your entry through the form on the page.

Big news in the climate risk space (with some implications for Earth observation): Financial data provider @MSCI_Inc is acquiring First Street, a climate risk modelling provider, for $120M. Great deal for MSCI and a good outcome for First Street, which started as a non-profit and is now becoming part of a major provider. Given that a lot of the underlying data behind First Street is geospatial/EO-derived - flood, wildfire, heat, storm, drought, sea-level rise, property, infrastructure, land use etc. - a couple of thoughts on this: 1. Climate risk is becoming an increasingly decisive input into what an asset is worth and whether it can be insured or financed (homes, infrastructure, loans and entire investment portfolios). Increasingly, as this deal suggests, those risk scores will come from private companies. The concern is that climate risk assessment could end up becoming a closed layer (a black box). Methodologies may be published or peer-reviewed, but not the same as being able to audit the full model. First Street was actually one of the more transparent players here with peer-reviewed, openly published methodology. Worth watching whether that "openness" survives absorption into a commercial owner. 2. Physical climate risk models are useful, and the market clearly needs them. But the problem is there are plenty of private climate risk providers, each validates its own model and may produce different numbers - which means competition gives you more black boxes that don't reconcile. There is no trusted public benchmark for climate risk scores, so the market increasingly prices risk using numbers that no independent authority has really validated. So the market ends up pricing off private numbers that no independent authority has ever validated. Forcing every vendor to open its model is not (and should not) be the answer. A more useful step might be an independent body that reviews and certifies these models against common standards. As climate risk consolidates into a handful of private providers, that may be the institution the market is missing.

Become a Pro subscriber to receive exclusive EO insights! newsletter.terrawatchspace.com/earth-observat…

In the new edition of TerraWatch Pro newsletter, we covered the implications of a) NASA adding 8 new commercial EO providers to the Commercial Satellite Data Acquisition (CSDA) program (now totalling 23 companies) and b) the increased budget for CSDA - a floor of $55.3M and a ceiling of $97.5M to be split across the 23 vendors. As always, the weekly briefing includes curated news, analysis and insights on EO and its applications.

Thanks for covering it and the nice write-up @Via_Satellite! satellitetoday.com/imagery-and-se…

Reflections from EO Summit 2026: As the main organiser of the conference, I did not (sadly) get to experience it like a regular attendee, but here are my 3 main takeaways from the conference: 1️⃣ AI: The magic wand for EO adoption, or the illusion of it? AI is now the new magic wand for EO adoption - it does make building something with EO easier than ever, with faster demos and pilots. But getting started was not the hard part though, going operational is. Things like trust, liability and integration into workflows still stand in the way of AI. The risk is mistaking faster pilots for real adoption. 2️⃣ EO is horizontal, but adoption is vertical EO is one horizontal technology, but every use case it serves in every industry is its own world. Crop intelligence, nature monitoring and methane detection may derive from the same data, but they have different processes with different customers who have completely different problems and different willingness to pay. This leaves EO companies with a choice: go deep on a single use case and expand into adjacent ones, or build an abstraction layer that works across many. The latter is the path and the promise embeddings seem to be going for. 3️⃣ The invisibility thesis is going mainstream The mental model of adoption is shifting from "get users to buy EO" to "get EO into the tools they already use." Success stops looking like someone consciously adopting EO and starts looking like EO vanishing into a process they already have. While I am glad the invisibility thesis is getting wider traction, there might be a risk in EO disappearing: the further EO goes into the background, the less the people depending on it understand what they are actually depending on.

🛰️Next week, #SatCen will contribute to two major Earth Observation and innovation events: the EO Summit 2026 in London on 22 June and the Φnnovation Summit in Frascati on 23 June. Sergio Albani, SatCen Head of the Research, Technology Development and Innovation Unit, will share perspectives on the evolution of the Earth Observation ecosystem, innovation, and the role of Earth Intelligence in supporting operational decision-making. We look forward to engaging with the wider Earth Observation, space and innovation communities. #EarthObservation #Innovation #GEOINT #SatCen #RTDI #EuropesEyesintheSky #Space

⏳ Counting down to EO Summit 2026: 7 days to go!! As we look forward to welcoming 300+ attendees in London next Monday, we would like to take a moment to give a shoutout to all our amazing sponsors who made this event possible. 💎 Platinum sponsor: @PixxelSpace 🟡 Gold sponsors: @Esri | @UP42_ ⚪ Silver sponsors: @EarthDailyA | @iceye_global | @AbsolutSensing | Orbital Eye | @KuvaSpace | @ChlorisGeo 🛡️ Defence track sponsor: @satlantis_ | @EUSPACEIMAGING | @Argotec_Space 🌪️ Disaster track sponsors: @Open_Cosmos | TelePIX 🟤 Bronze sponsors: Attis | @AistechSpace | Another Earth | @NV5Geospatial ☕ Coffee sponsors: @graniotES | @RssHydro --- Last few tickets remaining on eosummit.com

1 week to go for @EOSummitConf 2026! We just got the first set of badges printed for 300+ attendees gathering in London on June 22 to discuss commercial, civilian, environmental and defence applications of Earth observation. We will have a great mix of participants from several sectors including insurance, finance, agriculture, forestry, energy, utilities, nature, climate and sustainability as well as EO companies, geospatial solution providers, investors, non-profits and government representatives. I am proud that EO Summit has become the only forum that brings together the entire EO ecosystem - from satellite companies and space agencies to end-users and NGOs. 🎟️ Last few tickets remaining on eosummit.com

Wrote a thing - a long read on some thoughts on the future of Earth observation, especially the open data policy which we all seem to be taking for granted. At a time when 60+ countries seem to be launching national EO satellite constellations, the question is not whether we will have enough capability - we do. The harder question is whether we will have the will to come together to create a global infrastructure for EO. Sixty countries launching their own EO satellites could mean 60 contributions to a global Earth observation system or just 60 isolated national EO programs that do not add up.

Announcing the Parametric Insurance panel at the EO Summit in London on June 22 which will focus on the use cases of satellite data for designing new parametric solutions. Our sincere thanks to @EarthDailyA for sponsoring this session. Join us to learn about the applications of geospatial and remote sensing in the insurance sector!

Join us at EO Summit on June 22 in London to learn about Google's EarthAI models and use cases of AlphaEarth Embeddings. Less than a week remaining for booking your tickets for @EOSummitConf!

Join us for EO Summit in 2 weeks - we have an amazing line-up of speakers from some incredible organisations across several sectors. Really proud of our logo wall! I can't wait for June 22, see you in London! @EOSummitConf

Switzerland has decided it would be opting out of participating in the Copernicus programme for the next cycle (2028-2034). This follows their previous decision in 2024 to not participate in the current cycle. Why? Copernicus has an open-data policy and so Switzerland can access almost all of the data. But, it won't be able to take advantage of membership benefits like receiving real-time data, activating Copernicus services (in case of emergencies) and influencing the direction of the programme. What's worth watching is the bigger question this raises: the financial sustainability of open data. Copernicus is only getting more expensive from here: new Sentinel missions, more data to process and more investment in storage and dissemination. All of it will be funded by the countries in Copernicus. So if other countries reach the same conclusion that Switzerland has - that open access makes paying in optional - there will be questions about the willingness of contributors to keep funding something others can use for free, or to continue today's open data policies for non-members. Important to remember: The open-data policy has always been a political choice, not a law of nature. And the more it looks like it can be enjoyed without contributing, the more pressure builds to revisit the terms for those who don't. In short: Please do not take open satellite data as a given!