Most competitive research fails for a structural reason rather than an analytical one. Someone is asked to “look into” a competitor, they gather a large amount of material, and the work ends as a document nobody uses. The gathering was fine. What was missing was a cycle.

Competitive intelligence has a standard four-phase shape, borrowed from national intelligence practice and adapted for business. It is worth learning properly, because each phase has a specific output, and skipping any one of them produces a recognisable failure mode.

Phase 1: Planning and direction

The cycle starts with a question, not with data. Specifically, it starts with a decision that someone is going to make, and the question that decision hangs on.

This distinction matters more than any other single thing in this article. “Tell me about Competitor X” is not a question; it has no answer, no stopping point, and no way to be wrong. “Should we match Competitor X's move into mid-market, and if so what would it cost us?” is a question. It implies what evidence would count, and it tells you when you are finished.

Good direction-setting produces a short list of what practitioners call intelligence requirements. Each one should:

  • Name the decision it feeds
  • Be answerable from evidence you can plausibly obtain
  • Have a deadline tied to when the decision gets made

If a requirement fails any of those three, it is a curiosity, and curiosity is what turns a two-week project into a two-month one.

Phase 2: Collection

Collection is gathering evidence against those requirements. The discipline here is coverage: you want to hit several independent source types, because any single type has a systematic bias.

Broadly, sources divide into two families:

  • Primary — evidence you generate. Win/loss interviews with your own prospects, conversations at trade shows, direct product trials, customer advisory boards. Slow, expensive, and much more valuable than most teams assume.
  • Secondary — evidence that already exists. Company websites, pricing pages, regulatory and financial filings, job postings, patent records, review sites, engineering blogs, press coverage, community forums.

A rule that saves a lot of pain: record the source with the fact, at the moment you capture it. Not afterwards. An unattributed claim discovered three weeks later is worthless, because you cannot tell whether it came from a filing or from someone's guess in a forum, and you will not be able to reconstruct which.

We cover where to look in detail in the public record and what it will tell you.

Phase 3: Analysis

This is the phase teams skip, usually without noticing. Collection feels productive — the folder fills up, the document gets longer. Analysis feels like staring at a wall.

The difference between data and intelligence lives entirely in this phase. Data is “Competitor X posted eleven enterprise sales roles last quarter.” Intelligence is “Competitor X is staffing for an enterprise motion they have not announced; on their historic hiring-to-launch lag that lands in roughly two quarters, which is inside our renewal window for the six accounts most likely to be targeted.”

The move from one to the other requires three things:

  1. Structure. A framework that forces comparison rather than description — SWOT done properly, a positioning map, or a capability matrix.
  2. Inference, labelled as inference. You will have to go beyond what is directly stated. That is the job. What is not optional is marking where you did it, so a reader can weigh it.
  3. A so-what. Every finding needs to end in an implication for the decision from Phase 1.

A useful test: read each finding and ask whether a reasonable person could disagree with it. If nothing in your analysis is disagreeable, you have written a description, not an analysis.

Phase 4: Dissemination

Intelligence that does not reach the decision-maker in a usable form has not been produced, it has been filed. Dissemination is a real phase with real design decisions.

The main one is format, and it follows from audience:

  • An executive making a strategic call wants the conclusion first, the reasoning available, and the evidence in an appendix.
  • A salesperson in a live deal wants one page they can read between calls — see what belongs on a pocket card.
  • A product team wants the capability comparison and the gaps, in enough detail to argue about.

The same underlying analysis serves all three. The packaging is different, and getting it wrong wastes the whole cycle.

Why it is a cycle and not a line

Two reasons.

First, dissemination generates new requirements. You present findings, someone asks a question you cannot answer, and that question is the first requirement of the next iteration. A cycle that never produces new questions is usually a sign the analysis was too safe.

Second, intelligence decays. A competitive picture has a shelf life measured in months — pricing changes, people leave, positioning shifts. A one-off analysis is accurate on the day it is delivered and progressively less so afterwards, which is why a watchlist matters more than a single deep study.

Where the cycle usually breaks

The failures cluster in three places:

  • No Phase 1. Research begins without a decision attached. Everything downstream is then unfalsifiable, because there is no standard for what counts as enough.
  • Phase 2 eats Phase 3. Collection expands to fill the available time and the analysis gets whatever is left, usually the last afternoon.
  • Phase 4 as an afterthought. A thorough analysis delivered as a sixty-slide deck to someone with fifteen minutes is a failure of dissemination, regardless of how good the analysis was.

If you fix only one of these, fix the first. The other two are usually symptoms of it.

Running the cycle without a dedicated team

The cycle is not scale-dependent — a founder can run all four phases in a week. What it is, is discipline-dependent, and the discipline is hardest to sustain when competitive research is the fourth priority of someone whose first three are operational.

That is the practical case for structured, repeatable competitive reports: not that the cycle is beyond a small team, but that a small team rarely reaches Phase 3 before something more urgent arrives. Our Competitive Intelligence report runs the full cycle against a company you name, with sources cited and estimates labelled as estimates.