Why the feature table tells you nothing
Every row in a feature table is a claim the competitor made about themselves, collected from the surface they control most tightly. Two products can both tick inventory management and differ so completely in what that means that no customer would consider them substitutes. The table records the word, not the capability, and the gap between those two is where every wrong strategic decision comes from.
The second failure is subtler. A feature table frames the question as what should we build to match them, which is the question that produces a worse copy of an existing product. The question that produces a business is which customers do they lose, and why.
The question worth asking
Every product serves a centre and abandons an edge, and the abandonment is deliberate. A team with finite engineers decides that the customers who need a different data model, a different price point, a different language or a different compliance regime are not this quarter's problem — and then makes that decision again every quarter. Competitor analysis is the work of finding out where their edge is.
Three questions locate it, and all three are answered by their users rather than by them: which requests do they close without acting on, which complaints recur across years rather than releases, and which segment writes great product, but in a public review.
Where the answers are written down
Four sources, in descending order of honesty. Public issue trackers are the most honest, because a closed-as-wontfix ticket is the company stating in writing what it will not do. App-store reviews come next: they are written by people who paid, at the moment they were angry, and the one-star tail is where the abandoned edge complains. Support forums show the workarounds users maintain by hand, which is a precise inventory of what the product does not do. Community threads are the least structured and the best for phrasing — they show how the segment describes its own problem, which is the language your page has to speak.
Sort what you find by the reason given, never by the product. A reason that appears against every serious product in the category is a property of the category — that is a market. A reason that appears against one product only is that product's bug list, and a company built on a competitor's bug list has a six-month shelf life, because bugs get fixed.
An empty market is a warning, not an opening
Finding no competitor feels like finding an unclaimed field, and it almost never is. Someone has looked before you. The three reasons a market stays empty are all worth more of your time than the celebration: it has no money in it, it cannot be reached at a cost that leaves a margin, or it has already defeated everyone who tried and those companies are no longer visible precisely because they failed.
The healthy shape is not zero competitors. It is several competitors whose users are visibly unhappy in the same way. That combination proves two things at once: money moves in this market, and the people taking it are not satisfying the people paying.
How this site scores competition
Competition is one of the axes behind every Opportunity Score here, and it is computed from evidence rather than from a count of logos. Each market hub also carries a table of the products people in that market are reporting problems with, aggregated by how many distinct people reported them — which is the supply side of exactly the analysis this guide describes, already done.
The pairing that matters is demand against competition, not either alone. High demand with high competition is a market you are late to; high demand with low competition is the shape worth a year of your life, and it is what the matrix view exists to show.