Why hasn’t the Earth observation industry taken off?

The Earth observation industry seems to be on a slow burn – but why? Brace yourself for a ‘rollercoaster ride’ through Earth observation as this article unravels the mysteries behind the muted growth: from startup struggles to market nuances.

I first considered the question of why the Earth observation (EO) industry hasn’t taken off around five years ago. Back then, the EO sector was starting to receive a lot of attention; governments and the private sector were investing more and more, and new operators and service providers were emerging with new types of sensors and ideas for innovative applications. However, commercial growth (for EO data and services) remained slow. Whilst investment and the supply side looked positive, traction for actual uptake of solutions was sluggish.

Fast forward to today, and we could still ask the same question. On the face of it, not too much has changed. The short and simple answer to the question is that the data supply still isn’t there to meet various industry requirements and thus expand the market. The longer answer is somewhat more complex.

The current situation

On the face of it, the commercial EO industry certainly still hasn’t taken off. The same two companies are responsible for over 50% of data/imagery sales, and the value-added services industry has slow growth and remains fragmented. Established companies such as Airbus and Maxar have pointed out that growth from EO has been challenging. Maxar was not helped by the failure of WorldView-4 and delays to Legion. For someone who has spent nearly 25 years in EO, and most of that in the commercial aspects of the business, that could sound a little depressing.

Much is made of new companies emerging, with new constellations of satellites with various sensors, but the simple truth is that the majority of these companies are not up to full satellite capacity. Without a full complement of satellites, they cannot deliver on the goals they set out to achieve, which are aligned to the requirements of various end-user customers. Downstream, new EO services companies building analytics and delivering information also need this data supply to feed their algorithms to build solutions. Therefore, they cannot achieve all they set out to do either.

This means we’re still ‘as we were’. There are excellent commercial data sources out there from the aforementioned Maxar and Airbus, plus e-GEOS, MDA and so on, but they are not focused on building low-cost, high-revisit and/or coverage constellations which much of the anticipated market expansion will look to utilize. Indeed, most of the market for commercial EO procurement still remains for defence purposes (particularly US defence). It is only relatively recently that companies such as Planet and Iceye have started coming up to full capacity. 

The challenges to face

For companies to achieve their goals, more financing is needed in many cases. At the same time, we’re being told that the financial environment for raising funds is more challenging. But without more funding, companies cannot get to where they need to be. And without the data these companies collect, new service areas cannot emerge and revenue generation remains stunted. Add to this a more crowded field of even more companies – both operators and services companies – emerging and competing for funds, and the likely outcome is only in one direction: a consolidation of the sector, either through acquisition or simply due to companies running out of funds.

How to develop the EO market

What was true five years ago, and still remains the case, is that Earth observation demand by vertical markets should be considered in two different, but connected camps: new ways of meeting the requirements of existing vertical markets (i.e. the experienced or ‘established’ end user), and the potential to open up into new areas. There is a key difference requiring the need for some separation here: the end users with longer-term usage of EO solutions have a better understanding of its capabilities; on the other hand, the ‘new’ markets have not been exposed to EO and thus do not really understand (or care) about its capabilities.

Whatever the differences, end-user requirements would still centre around receiving cost-effective information based on high-frequency change detection, irrespective of the application. Delivering this in a regular fashion to multiple end users and, in doing so, opening up EO to a mass market would shift it from a business-to-government (B2G) to a business-to-business (B2B) solution. This shift has been slow to occur, however, and this is unlikely to change significantly in the short to medium term.

From B2G to B2B

The difficulties in developing B2B is reflected in the business plans of new EO companies. ‘Established’ customers in existing vertical markets (in defence, mining, oil & gas and, to a degree, civil government, agriculture, etc.) understand data/imagery and the EO industry better, so selling into these industries is seen as lower-hanging fruit. Customers in oil & gas and agriculture have worked with EO data since the onset of Landsat, and SAR/AIS has been used to track ships for decades.

Potential new ‘big ticket’ markets for EO, such as in finance, carbon trading, environmental social governance (ESG) and the like, may appear attractive, but it is hard to demonstrate the path to revenue generation. To some degree, this has been one of the issues hindering expansion of the commercial base for EO; in order to generate revenue as soon as possible, existing markets are the first targets for sales. If new EO companies can get off the ground, then they can look to new areas to exploit. These new end-user markets care less about EO imagery and more about information. But this also means that if the information cannot be delivered when it is needed, then it will not take off as an application. These end users do not understand (and will be turned off by being told): “You can’t have your information today because the satellite didn’t pass”.

The myth of oversupply

Building these markets is a challenge without the desired satellite infrastructure being in place. The simple fact of the matter is few companies are there yet. Not many EO companies have significant revenue generation and/or are up to full capacity.

I often read of EO ‘oversupply’, but that is simply not true. There may be a lot of new companies, but most do not have many satellites. The ‘oversupply’ claim comes from looking to a future (unlikely) scenario whereby all companies will be up to full capacity. Perhaps a case could be made for 1m multispectral data being available from multiple sources. But lower-cost SAR is only just getting there, hyperspectral and emissions monitoring companies are just starting, one or two thermal infrared satellites have gone up, a couple more companies are getting into 10cm data, there is still no high (~5m) short-wave or Lidar… so where exactly is the oversupply?

This creates a kind of ‘chicken and egg’ conundrum. The market is there to be exploited, new (and existing) EO companies know their potential future customer needs well, now they just need to get there. And to do that, more investment is going to be needed, and some companies will fail. I can think of over 20 companies in EO alone that will go for a Series A round this year. It simply won’t work out for all of them.

The market potential

Various market figures illustrate EO potential. Euroconsult states that the total market for EO data and services in 2022 is US$4.64 billion (and a five-year compound annual growth rate [CAGR] of 5%). Of this, the market for EO commercial data only was US$1.78 billion. It is driven by submetric resolution usage. In total, all optical data finer than 1m ground resolution accounted for US$1.2 billion in 2022, or nearly 65% of total revenue. This also indicates the importance of the defence sector to the EO industry. However, despite the emergence of new companies over the last five years, only a small number of EO operators generate over US$100 million in revenue.

The remaining US$2.86 billion is related to the delivery of value-added solutions (including products, services and analytics/information.) More moderate-resolution datasets tend to be utilized more to build solutions rather than represent significant revenues as imagery alone, particularly for applications in which greater coverage and spectral capacity are key requirements (such as in natural resources monitoring, agriculture and environmental applications.)

The total addressable market (TAM) for EO is challenging to establish for ‘the whole of EO’ across all verticals. It is also hard to do at a company level, but at least there are some numbers out there. The three EO companies which went public through a special purpose acquisition company (SPAC) released company-specific TAMs in investor reports prior to the public offering. These numbers stated TAMs at between US$40 billion and US$140 billion. Obviously, the TAMs will differ depending on which markets each company wants to serve, but just taking these as is, it means a market penetration today of only 3%-10%. In other words, there is still plenty of margin for growth.   

The light at the end of the tunnel

That’s where the investment community comes in. We’re being told that it is a challenging investor environment for EO, and indeed the space industry in general. According to analysis from early-stage investor Space Capital, space industry investment plummeted from a US$47 billion peak in 2021 to US$20 billion in 2022. The stock performance of SPAC’d companies since their public offering hasn’t helped matters. However, investors need to invest to make money, so they invest in perceived ‘safer’ industries and technology. The investments are still there to be made in EO, but the technology needs to look like a safer bet. This perhaps means greater scrutiny of business cases compared with a few years back.

Nevertheless, on a positive note, several companies (Blacksky, Capella, Pixxel, Wyvern Satellite Vu, Kuva Space, etc.) all managed significant raises in the last year. Indeed, the EO sector has raised US$5.5 billion over the last decade, and most of that was raised in the last five years, albeit with a recent slowdown.

At Euroconsult, we have taken part in many EO due diligence processes over the years. This has been mostly for private equity, banks and debt deals, but we are starting to more interest from venture capital too. I now also work with startups leading into the due diligence process. As a broad generalization, most investors I’ve encountered tend to believe the company can build the technology that they are setting out to do, but they are less convinced about the path from this technology to revenue generation. This includes what the company is selling (just imagery, imagery plus analytics, just analytics…) and how these products are developed. There still seems to be a gap in language between the EO community and investors. To become a safer bet, work is required on explaining how data from the satellites can be converted into products that customers want.

One way to get around this is for the EO company to focus on what it knows, or the new bit of technology it is developing, whether a new type of sensor, a new way of processing data, etc. This also streamlines capital expenditure (CapEx), reducing risk. Why build a new analytics platform or your own satellite bus if it’s not your core business or part of a wider business strategy? There are plenty of solutions already out there. This is also giving rise to new space/satellite-as-a-service companies, such as Loft Orbital or Muon, which offer quicker ways to space. And for EO companies, that can mean quicker ways to generate revenues.

I also suspect new EO operator companies will eventually fall into one of two camps: general data suppliers of whatever types of data they are collecting, or vertically integrated companies focused on data delivery and service generation in a couple of markets. Trying to target numerous markets for data and services from your dataset may have seemed like a good way of generating big addressable markets to look attractive to venture capitalists in the past, but the reality is that this is hard to achieve and perhaps even unfeasible. There will be some compromise in the data – the data needed for mapping could be used for agriculture, but it is unlikely to be the ideal imagery – and it also means the need to hire specialists and sales teams in each of the vertical markets you are aiming to address, all of which increases the costs. It is more likely that the EO operator may acquire specialist service providers in the future if it wants to double-down on a specific vertical. We have already seen a little of this activity, e.g. a couple of Planet acquisitions in the agriculture space in the last couple of years.

Looking ahead

A paradigm shift is unlikely to occur over the coming year. In fact, on a negative note, I think some EO operators and even some analytics companies will restructure/consolidate or simply cease operations. However, others will continue to emerge, resulting in more and varied satellites and sensors.

Realistically, the industry probably still needs a couple more years of development and investment before it starts to reach its full potential. During 2024, low-cost SAR operators will start to reach full capacity. Companies such as Iceye are already making strides in the market, and Umbra’s open-data policy and transparent pricing is encouraging. Wyvern, Pixxel and OSK will continue to roll out hyperspectral satellites, and we may see 5m ground resolution SWIR, but none of these operators will become close to full capacity within the year. Meanwhile, Satellite Vu will launch further units for thermal imaging, and we may even see the launch of NUVIEW’s first commercial Lidar satellite. This in itself would all be great to see. It will allow companies to test what can be done with the different datasets, but it won’t lead to an immediate shift in market expansion. These companies need to come to full fruition.

If we consider EO’s current position in the growth cycle, apart from a couple of exceptions – such as optical datasets in support of defence, or SAR in maritime domain awareness – I think it is fair to say that the sector is still in the take-up phase. A little perspective is needed. In the satellite communications business, for example, the operators were government entities for years prior to being fully commercialized. But now, no one thinks of communication/TV/broadband as being ‘government-orientated’. Perhaps what is needed for the EO sector to further develop is a little less hype, and a little more patience.