What competitive intelligence actually is, and what it is not

Competitive intelligence is collecting public information about your competitors and turning it into decisions.

That is the whole definition. The industry around it is large and the idea is simple, and confusing the two is how companies end up buying a platform when they needed an hour a month.

The four types

They answer different questions and need different work.

Strategic. Where is the market going, who is entering, who is consolidating. Sources: filings, funding announcements, hiring patterns, industry press. Consumers: executives, planning cycles.

Tactical. What are they doing right now. New pages, new ads, price changes, campaigns, launches. Sources: their own website, the ad libraries, search results. Consumers: marketing, weekly.

Sales. How do we win against them in a live deal. Sources: win-loss interviews, sales calls, their pricing. Consumers: sellers, in the moment.

Product. What are they building and what do customers want that neither of us has. Sources: their changelog, their job listings, their reviews, their documentation. Consumers: product and roadmap.

Most teams need the tactical and sales layers. Most software in this category is sold on the strategic layer, which is a mismatch worth noticing before you buy anything.

Where the line is

Straightforward, and worth stating because people get anxious about it.

Fine. Reading their website, their pricing page, their sitemap, their job listings, their filings, their reviews. Using the public ad libraries. Signing up for their newsletter. Asking your own customers why they chose you. Buying their product under your own company name.

Not fine. Pretending to be a prospect to get a sales demo under false pretences. Pretending to be a journalist or a job applicant. Asking a candidate from a competitor to disclose confidential information. Anything that induces someone to break a confidentiality agreement.

The practical test: would you be comfortable if the competitor learned exactly how you obtained it. Almost all of the value sits comfortably on the right side of that line, which is the point most people miss when they imagine this work is somehow shady.

What good looks like at a small company

You do not need a programme. You need three things.

A short list of real competitors. Three to five, taken from sales calls, search results and AI answers, not from who you talk about internally.

A one page profile each, kept current. Their positioning sentence, their price, who they beat you with, who you beat them with, their loudest claims, their most common complaints, what they did this period, and what you are doing about it.

A monthly hour. Sitemap diff, ad libraries, pricing screenshot, new reviews, and twenty buyer questions asked in the AI assistants.

That is competitive intelligence. It fits on three pages and it produces decisions, which is more than most programmes do.

What good looks like at a larger company

Two things get added, and they are about distribution rather than collection.

Battlecards in the CRM. The intelligence has to reach the seller at the moment a competitor's name appears on a deal, or it does not exist. This is what the enterprise platforms are actually for.

Win-loss as a habit. A single field on every closed lost deal recording who they chose, populated consistently, is worth more than most software. Add ten interviews twice a year and you have better intelligence than any vendor can sell you.

The failure modes

Collecting without deciding. The most common by far. A thorough document, a nodding meeting, no changed behaviour. If a piece of research does not end in three dated actions, it did not happen.

Obsession. Reacting to every competitor move turns your roadmap into theirs. Most of what a competitor does is not aimed at you and does not deserve a response.

Stale documents. A battlecard nobody updated in eight months is worse than none, because sellers quote it. Date everything.

Measuring the wrong things. Follower counts, estimated revenue, modelled ad spend. Comfortable numbers that change no decision.

Getting it to happen every month

The honest obstacle is not knowing how. It is that a recurring task with no deadline loses to everything with one.

Two fixes that work. Put it in one person's job description with a named monthly date. Or automate the collection so only the deciding is left, which is the part that needs a person anyway. Our own Figo does the collection weekly across nine channels for $49 a month and writes the summary; a page monitor at $12 a month covers a narrower version.

The decision part does not automate, and anyone selling you that is selling something else.