How to research competitors when your market publishes almost nothing
A practical method for competitor research in defence, deep-tech and other low-visibility markets, using tenders, patents, hiring signals and event lists.
By Prathmesh Soni. Published .
Key takeaways
- In low-visibility markets, competitor research means collecting many weak public signals and cross-checking them, because no single source tells the whole story.
- The most useful signals are usually government tenders, patents, hiring activity, event exhibitor lists and funding announcements.
- Mark every finding with how confident you are in it. A competitor map with honest confidence levels is more useful than a confident map that is partly wrong.
Why is competitor research so hard in markets like defence?
In consumer or SaaS markets, competitors publish pricing pages, customer logos and product docs. In defence and other deep-tech markets, most of that is never made public. Contracts are confidential, customers rarely talk about suppliers, and many products are only described in vague terms.
That leaves founders guessing about who they are really up against. The usual desk research (search the category, read a few websites) produces a list of names but very little about capability, customers or momentum.
Where do you find signals when competitors don’t publish?
Look for places where companies have to leave a trace, even when they would rather not. In India, for example, these sources tend to be the most useful:
- Government procurement portals such as GeM and the Central Public Procurement Portal, which show who bids on and wins public tenders.
- Defence innovation programmes such as iDEX, whose challenge winners and partners are announced publicly.
- Patent databases like Google Patents, which show what a company is actually building and how actively it is filing.
- Hiring activity on LinkedIn and job boards. The roles a company hires for say a lot about what it is about to build or sell.
- Exhibitor lists from industry events such as DefExpo and Aero India.
- Startup databases such as Tracxn and Crunchbase, plus funding announcements in the business press.
How do you turn scattered signals into a usable competitor map?
- Build a long list of every company that could plausibly compete, including adjacent players and large incumbents.
- For each one, record what they build, who they sell to, how far along they are, and the source for each fact.
- Group them by capability and by customer, not just by category label, because two “drone companies” can serve completely different buyers.
- Place them on a simple positioning grid, for example capability depth against customer focus, and look for the empty spaces.
- Write down what the map means for your own positioning in two or three plain sentences.
How do you know the research is good enough to act on?
Treat every claim as provisional until two independent sources agree on it. Tag each finding as confirmed, likely or unverified, and say so in the final document. Founders can then decide how much weight to put on each point instead of treating the whole map as fact.
It also helps to have someone outside the company review the conclusions. Founders are close to their own product, and a second opinion catches blind spots on both sides: competitors you underestimated and strengths you didn’t realise you had.
What this looked like for a defence-tech startup
We used this approach with BORTEX, a Bengaluru defence-tech startup building AI systems that combine air, land and sea sensor data for border monitoring. The goal was a clear picture of the competitive landscape before the founder committed to his next moves.
“The analysis you gave me actually helped me understand my competitor market much better. Doing a competitor analysis in a deep-tech sector, especially for defence, is quite tough, and I definitely benefited from this. Apart from that, this also served as a word of confidence—a valuable second opinion.”
Shubham Datta, Founder, BORTEX
Frequently asked questions
How long does competitor research take in a low-visibility market?
It depends on how many players there are and how much is public, but expect it to take longer than in a typical SaaS market, because most of the time goes into finding and cross-checking weak signals.
Is it legal to use tender and patent data for competitor research?
Yes. Public procurement records, patents, job postings and press releases are published openly. The method relies only on public information.
What should the final output include?
A list of competitors with sources, a positioning grid, the gaps you could own, and a confidence label on every important claim.