Competitor maps that carry their source

4 min read

A competitor slide gets one question in the room: says who?

Take The Trade Desk. Crunchbase, the company database many investors use, keeps a list of competitors for each company. On 29 October its first five for The Trade Desk were Integral Ad Science, MNTN, Topsort, Viant and Moloco. Specialists in the same category, ad-tech companies that look like it. The field talks about someone else. “Amazon is generally considered the apex predator for The Trade Desk’s budgets right now,” AdExchanger wrote on 11 September 2026, citing an analyst at New Street, and the same article names Google’s Buyer Direct as the longer-term threat (AdExchanger).

Both lists are right. They answer different questions. The database files The Trade Desk with the companies it resembles. The article tells you who fights it for the same budget.

How do you build a competitor map with sources?

You keep the source with the name. The two lists answer different questions. We checked this on 27 companies: of the real competitors that articles named for them, 7 in 10 were not on the company’s Crunchbase list. That does not make the articles the better list. Crunchbase’s first five were real competitors more often than the names the articles gave. You want both, as two columns, not one answer.

And a pair named in an article is a lead, not a fact. We took 150 pairs of companies that articles named as competitors, and checked each against what both companies sell on their own websites: about 2 in 3 really compete. The misses look like this: Alphabet and Nike (different markets), AMD and Microsoft (partners).

In Ins7ghts, our research platform (a knowledge graph of the data and AI news we read every day), every pair comes with the article that named it. Many articles that put two companies side by side are thin, for example paid market-report releases listing twenty “key players”. So we open the article before a pair goes on a map.

Get the AI & data signal, daily.

41k+ read Ins7ghts, 25k+ read AI Beat. Ins7ghts and AI Beat: two free emails every morning on AI and data. Unsubscribe anytime.

The check, step by step

  1. Write down which question you are answering: who looks like it, who it is filed with, or who it fights.
  2. For “who it fights”, collect the pairs that articles name as competitors, one article per pair.
  3. Keep the article with the pair, always.
  4. Open the article and check: same market, and not a partner, customer or owner.
  5. Put the database’s list next to it as a second column, not as the answer.
Looks like it a look-alike search Filed with it a company database Named against it the articles Open the article: compete, other market, partner, customer or owner? Compete: on the map The rest: off the map

The test behind the three questions, on 29 companies, gets its own article on 19 November. And before any of this, make sure each name is the right company: Stop your AI confusing Alphabet with alphabet.

The kit has the prompts, the map template and the checklist. It comes by mail once you confirm your email.

Get the kit for this check

Two prompts, a map template and a checklist. Confirm your email and it arrives by mail. You also get Ins7ghts: one free email every morning on AI and data. Unsubscribe anytime.

Small print. Pairs named as competitors in articles of late August to mid October 2026: 98 of 150 really compete, going by what both companies sell on their own websites. Two companies named in the same articles but not as competitors compete far less often, 30 times in 150. The 7 in 10 is 52 of 73 real competitors, over 27 companies, against Crunchbase lists read on 29 October 2026; the extra names the articles give lean to big platforms that compete on one product line. AI models did the checking, against the companies’ own websites; no person checked them. The Trade Desk article was opened and read on 7 November 2026.

Yves Mulkers

Yves Mulkers is the founder of 7wData and a widely followed voice in the data and AI community. He curates the 7wData and AI Beat newsletters, reaching hundreds of thousands of data and AI professionals, and writes on data strategy, analytics, AI, and the evolving data ecosystem.