Windward is an Israeli startup that provides maritime-risk-analytics-as-a-service (MRAaaS?). I’ve known about it for a few years and I’m intrigued by how the company has built a product using readily available data sources and yet has limited competition—at least from an outsider’s perspective.
This post covers the company’s basic details, how its product creates value for its customers and how the business derives value from customers. I also look at the competitive dynamics of Windward’s industry and conclude that maritime analytics is an area where experience present a strong barrier to entrance for would-be competitors.
About the business
Windward tracks commercial shipping vessels using a range of data sources including ship transponder data and satellite imagery. To this data, it applies machine learning models to detect and predict risky or abnormal patterns of behaviour.
Detail from Windward’s homepage
Analysts at the Rand Corporation point out that Windward’s product is effectively military-style signals intelligence (SIGINT) for commercial use. It has three primary use cases based on its analytics: intelligence analysis, sanctions compliance, and fleet and vessel risk assessment. Customers include the United Nations Security Council, various coastguards, hedge funds and maritime insurers such as Gard AS.
Windward was started in 2010 by former Israeli navy intelligence officers Ami Daniel and Matan Peled and now has more than 80 employees based in offices in Tel Aviv, London and Washington DC. It has raised around $40m from investors including insurance fund XL Innovate, Aleph and Horizon Ventures.
How does Windward create value?
The primary value Windward provides to its customers are the insights gleaned from analysing historical patterns of behaviour for different ships and understanding behaviours in relation to specific maritime contexts.
For example, when an oil tanker’s transponder signal is lost—known as “going dark”—the tanker’s location, historical activity and patterns of potentially illicit activity can all play a role in predicting what has occurred notes Ron Crean, the company’s commercial VP.
The tanker’s transponder may have poor reception or it may have been turned off deliberately so the ship can maneuverer without scrutiny. Where the ship is when it goes dark plays an important role in whether a ship is flagged as engaging in potential illicit behaviour; some areas are higher risk than others.
“Going dark in the eastern Mediterranean, Gulf or off the coast of Venezuela is certainly different than doing so in the North Sea” says Crean.
So while knowing where a ship is at any given time is potentially useful, it is predicting where a ship may have been and what it might have been doing and therefore what it may be likely to do in the future that is Windward’s real value. I have written about how the value of data products is linked to prediction.
This predictive ability underpins the company’s three core uses cases—intelligence analysis, sanctions compliance and insurance risk assessment. Sanctions compliance seems to address the most commercially viable and acutely felt need.
A March 2020 article in Wired notes that sanction surveillance is Windward’s most lucrative product line and is proving to be an attractive offering to state clients. Increasing tensions between Iran and the US and Europe throughout 2019 and the imposition of fresh sanctions against Iran in early 2020 will have helped drive demand for Windward’s newly launched sanctions compliance product among state and private customers.
How does Windward capture the value it creates?
While limited information is available on how Windward delivers its three core use cases to its customers, its website suggests that each use case is served by its subscription-based web platform. Information is also provided to subscribers via configurable email reports and an API. Like many business-to-business software providers, Windward’s pricing model is opaque.
Given the scope of the data that Windward provides and its real-time nature, it seems likely that subscriptions would be priced in hundreds of thousands of dollars. Windward may also charge usage fees for its API and additional subscription fees for access to particular sets of behavioural insights—for example, fishing or oil sanctions compliance.
Industry dynamics
Given that the satellite and transponder data that Windward uses to generate its analytics can be bought from numerous commercial providers, supplier power seems to be relatively low. This would also seem to suggest that the industry has a low barrier to entry.
However, there seems to be very few companies that offer maritime analytics for sanctions and intelligence work. US firm Hawkeye 360 specialises in radio frequency data analytics and sells a maritime radio frequency solution that geolocates frequencies beyond the normal Automatic Identification System (AIS). But Hawkeye 360 is listed on Windward’s website as a partner—presumably Hawkeye supplies some of the data the Windward’s models require.
Detail from Hawkeye 360’s homepage
Other companies commonly cited as competitors seem to really be telemetry or analytics for asset management. These companies may be substitutions for Windward’s fleet and vessel risk management use case but are clearly inadequate with respect to sanctions evasion and intelligence work.
So why are there no direct competitors? It looks like the maritime risk analytics industry may have experience as a barrier to entry. Michael Porter notes in the canonical Competitive Strategy, “Experience is a more ethereal entry barrier than scale, because the mere presence of an experience curve does not ensure an entry barrier.”
Porter goes on to suggest that for a barrier to exist, the experience must be proprietary and not available to would-be entrants through copying, hiring competitors’ employees or by buying the relevant knowledge in.
Windward relies tacit knowledge of how commercial ships operate, particularly when their crews are seeking to obfuscate illegal activities. In theory, this combination of tacit knowledge and capability could create competitors to Windward in places that former navy personnel (and perhaps sanction evaders?) and data scientists can mingle. Though in practice, this seems to be the style of business that Israel is particularly good at creating.
After 10 years of operation the company will have a vast database of behavioural patterns that are more than the sum of the constituent data sources. These are derived from Windward’s data science capability and acquired knowledge of maritime signals intelligence and operating sea-going vessels.
Given these two factors, Windward’s specific tactic knowledge and its database of behavioural data, it does meet Porter’s requirements for creating an experience-based barrier to entry. Hiring former naval intelligence officers or Windward employees might be a good starting point but without the additional data you would likely be playing catch up for a long time.
Windward is an interesting example of where tactic knowledge, readily available data and earned insights have created a robust barrier for would-be followers.