Competitive Analysis Template: A Rival Profile You Can Score

The folder called Competitors

Every strategy team has one. A shared folder with a slide per rival: logo, revenue, headcount, a product grid with green and red cells, a few screenshots of the pricing page. Some slides are three years old. Nobody can say which numbers came from a filing and which from a sales rep's memory of a conference conversation.

Then a real decision arrives. We are about to bundle a new module at no extra charge; what will the two biggest rivals do? The folder has nothing to say. It describes the rivals. It never committed to what they would do next.

A competitive analysis template is a fixed set of fields you fill in for each rival that matters to a decision. The one on this page has two features most templates lack. Every line carries its source and a grade: filed fact, the rival's own claim, or our inference. And the file ends in a prediction log, a dated, falsifiable guess about the rival's next move that you score six months later.

The method behind it, including how to draw the field from the customer's side and where the legal line runs, is on the competitor analysis page. This page is the document you fill in. Its point of view: a profile without sources is gossip, and a profile without a prediction is a scrapbook.

Every line needs a source. Every profile needs a bet.

The template, part one: the rival profile

Fill one profile per serious rival. Three to five rivals is typical for a single decision. Each row has three columns: what you found, where you found it, and its grade (F for a filed or observed fact, C for the rival's own claim, I for our inference).

Two rules hold the profile together. A line without a source does not go in. And beliefs go in quotation marks, in the rival's words, never paraphrased into yours. Paraphrase is where analysts quietly replace a rival's theory of the market with their own.

The template, part two: the prediction log

The profile exists to feed this table. One row per rival per decision.

The probability column matters more than it looks. “They might cut price” can never be wrong. “Seven in ten they discount renewals within two quarters” can be, which is the only way the file gets better. The scale on this page is our convention, not a research standard; any fixed scale the team uses consistently will do.

Where the evidence is

For listed rivals, the best sources are the ones they are legally required to write. In the US, the annual report on Form 10-K has an Item 1 business description, an Item 1A on the most significant risks, and an Item 7 in which management gives its perspective on the past year's results ¹. The SEC's rules for the business description say it may include “trends in market demand and competitive conditions” ². Smaller reporting companies are asked to describe, where material, their “competitive business conditions,” their competitive position in the industry and their methods of competition ².

Field 2 has the most underused source. Most companies put executive pay in a separate proxy statement ¹, and its compensation discussion must explain what the program is designed to reward and may describe “what specific items of corporate performance are taken into account” in pay decisions ³. If a rival's chief executive is paid on revenue growth and adjusted margin, you know which of the two it will defend when you force a choice.

• Search across years. The SEC's full-text search covers electronic filings since 2001 ⁴. Search a rival's filings for your company's name, a product category or a phrase like “pricing pressure” and read how its language changed.
• Read beliefs in sequence. One shareholder letter is a statement. Five in a row show what management keeps promising and what it quietly stopped mentioning.
• Private rivals. Use what they publish (pricing pages, release notes, job postings), what customers and distributors tell your sales team, and your own win-loss reviews. Grade most of it C or I and say so.
• Keep the legal and ethical line. Public filings, products bought and taken apart, customer conversations and published job ads are fair. Posing as a customer and taking a rival's confidential documents from a new hire are not. The competitor analysis page covers the definition in law.

A filled-in example

Everything below is invented, including the filings. Notice the grades, and notice that the prediction can be wrong.

Suppose September comes and Rival A announced a time-tracking partnership in June. The prediction was wrong, and the outcome column says which line misled us: field 3. Rival A's shareholder letter said buyers choose on price; its behavior said it did not fully believe that. Next time, weight its actions over its words, and grade that belief lower.

Filling it in without fooling yourself

1. Start from the decisionWrite the decision and the date rivals will see it before opening a single filing. It tells you which rivals and which fields matter.
2. Fill fields 2 and 3 firstWhat a rival is paid on and what it believes predict most moves. They are also the fields teams most often leave blank.
3. Grade every lineF, C or I, with the source. When a profile is mostly I, say so at the top.
4. Write the prediction before the meetingEach analyst writes a prediction and probability alone, then the team compares. Writing first stops the most senior person's guess from becoming everyone's.
5. Assign the signalsEach early signal gets an owner and a check frequency, or nobody will notice it.
6. Score on the dateMark each prediction right, wrong or partly, and name the profile line that misled you. Update that line.

Daniel Kahneman, Dan Lovallo and Olivier Sibony list competitor neglect among the biases that make a base case too optimistic, and suggest war games as one remedy ⁵. The prediction log is the cheap, written version of a war game: it forces someone to play each rival before the decision is final.

The case against prediction, and running the file over time

The strongest objection: rivals are not that predictable. A new chief executive, a board fight or an acquisition can change a company's behavior overnight, and a confident log invites false precision. Better, the argument goes, to watch closely and respond fast.

Watching and responding is exactly what the log organizes. The prediction decides which signals are worth watching and what you will do when one appears. And the scoring is the point. A team that has scored 20 predictions knows which rivals it reads well and which it does not, and which kind of source (their words or their actions) has been the better guide. A team that only watches learns none of that, because it never wrote down what it expected.

Three problems appear when a strategy team runs these files for years.

• Stale profiles. Put a review date on each profile and re-grade it after every earnings season. A profile older than a year should say so on the first line.
• Predictions nobody scores. Put the score dates in a shared calendar with an owner. If the team stops scoring, it is writing opinions with numbers on them.
• The comfortable field. Teams update capabilities and signals, which are easy, and leave beliefs and incentives untouched. Rotate which analyst owns fields 2 and 3 for each rival so a fresh reader checks them.

Monday: one rival, one prediction

Pick the next decision a rival will see. Open that rival's latest proxy statement and annual report, and fill fields 2 and 3 with quotes and page references. Then write one prediction with a probability and a score date, and put the date in the calendar.

The folder called Competitors will still be full of logos and grids. The difference is one page that says what a rival will do, why, and how sure you are, and that will tell you in six months whether you understood it. A competitive analysis that cannot be wrong has never been tested.

Questions

What should a competitor analysis template include?

For each rival that matters to a decision: the unit that competes for your customers, what management is measured and paid on, what it believes about the market in its own words, its commitments and capabilities, dated recent signals and the risks it admits. Then a prediction log: the rival's likely response, a probability, the reasons, early signals with owners, your counter-move and a date to score it.

Where do I find information about a public competitor?

Start with what it must file. In the US, the 10-K's business description, risk factors and management's discussion, plus the proxy statement's compensation discussion, which explains what executive pay rewards. The SEC's full-text search covers filings since 2001. Add earnings calls, shareholder letters, pricing pages, job postings and your own win-loss reviews.

Why grade each line as fact, claim or inference?

Because the three deserve different weight. A filed number is a fact, a statement in a shareholder letter is the rival's claim about itself, and a conclusion drawn from job postings is your inference. Grading each line shows how much of a profile rests on guesswork and which lines to check first when a prediction misses.

How do you score competitor predictions?

Write each prediction with a time window and a probability, and set a score date. On that date, mark it right, wrong or partly right, and name the profile line that misled you. Over a year, the scores show which rivals your team reads well and whether a rival's words or its actions have been the better guide.

Where to go next

• Competitor Analysis: A Practitioner's Method for Predicting Moves
• Competitive Benchmarking Without Fooling Yourself
• Competitive Intelligence
• Competitor Myopia
• Signaling Theory