Competitive Intelligence Playbook

Competitive intelligence is a routine, not a report. Write down the few questions it must answer and who acts on each, watch a short list of public sources on a fixed schedule, and turn every change that matters into a battlecard, a talk track or a claim you can prove. Review it every quarter with named owners, and track the status quo as closely as any rival: the spreadsheet and doing nothing.

Most startups do competitive analysis once, for a pitch deck, and then the market moves. A rival changes its price, ships the feature your homepage says it lacks, or starts showing up in deals you expected to win. In my view, the analysis is only worth what it changes next quarter, so it needs owners, a calendar and a place where sales can use it.

The one-time version, from listing every alternative to the investor competition slide, is in Competitive Analysis for Startups, and the Competitor Research Grid holds it. This playbook is what happens after: the questions, the collection, the analysis, the cards, the claims and the review that keep it current.

Start from your deals, not from the rivals you hear about most. Pull every opportunity from the last two quarters that reached a demo, and write down the one alternative the buyer would have chosen without you. If the notes don’t say, ask the buyer; it is a one-line question.

Then rank the alternatives by share and match the depth of tracking to the rank (Figure 1). My rule of thumb, not a benchmark: an alternative named in 20% or more of deals gets weekly signals and a battlecard, 5% to 20% gets a monthly check, and under 5% gets a quarterly glance. That keeps the hours on what touches revenue, not on the rival that posts the most.

Doing nothing is a real competitor. In a series of decision-making experiments, and in faculty members’ choices of health plans and retirement programs, Samuelson and Zeckhauser (1988) found that people disproportionately stick with the status quo, which they define as doing nothing or keeping a current or previous decision^[1]. Doing nothing and the spreadsheet have no changelog to watch. What you track for them is the cost of the problem and the reasons buyers give for staying put, which come from your own calls and from interviews run the way the Win-Loss Analysis Playbook describes.

Without questions, competitive intelligence turns into a channel of screenshots nobody reads. Write a short list of questions, and tie each to a decision someone will make with the answer: the CEO on pricing and the roadmap, sales on objections, marketing on messages and claims, product on gaps.

For each question, name the decision it feeds, the owner who keeps the answer current, and how often it is checked. If a question feeds no decision, drop it. My rule of thumb, not a benchmark: no more than six questions at seed, and one owner for each alternative you track weekly.

Put the questions on one page and have the people who will use the answers sign it (Figure 2). Signing sounds formal, but it does two useful things: it records who asked for what, and it makes the time budget visible before anyone starts collecting. The Market Research Playbook uses the same habit for the questions a pitch must answer.

Almost everything a startup needs is public or already in its own calls. Pricing pages show packaging and price moves; changelogs and help centers show what shipped and how it works; review sites show what buyers praise and complain about after they bought; job posts show where a rival is investing; and win-loss interviews show why buyers chose.

If a rival, or a rival’s parent company, is public, read its filings. The SEC’s EDGAR system gives free access to company filings, with full-text search across more than 20 years of them^[2]. For an annual report, the SEC’s Regulation S-K asks for a description of the business, which may cover trends in market demand and competitive conditions, and for a discussion of the material risk factors^[3],[4]. Anyone who makes a materially false or misleading statement in a filed report can be liable to investors who relied on it, so the language is careful and worth reading closely^[5].

Put every source on a cadence that matches how fast it changes and the tier from Step 1, with an owner and a minute budget (Figure 3). Save a dated copy of any page you might quote, because pricing pages change without notice. An entry with no source and no date is a rumour, and it stays off every card.

The lines I don’t cross

In my view, intelligence is only worth having if you could explain to the competitor how you got it. That rules out pretexting: posing as a buyer, a job candidate, a student or a researcher to get a demo, a price sheet or a conversation. It also rules out asking a rival’s staff, former staff or customers for anything confidential, and signing up for a trial whose terms forbid use by competitors (Figure 4).

These lines match the profession’s own code and the law. The code of ethics of SCIP (Strategic and Competitive Intelligence Professionals) asks members to disclose all relevant information, including their identity and organization, before all interviews^[6]. US trade-secret law lists misrepresentation, and breach or inducement of a breach of a duty to maintain secrecy, among the improper means of acquiring a trade secret^[7]. Some software terms forbid this use outright: Salesforce’s Main Services Agreement, which also governs its free trials, bars Salesforce’s direct competitors from the services and forbids access for “benchmarking or competitive purposes”^[8].

More of the legal side is covered in Competitive Analysis for Startups. My lines are stricter than the law on purpose: a team that never has to ask “is this allowed?” collects faster. This is orientation, not legal advice.

A change becomes intelligence only when someone writes down what it means. I log every change in three lines: what changed, what it means for buyers and for you, and what you will do. The change log in the Competitor Research Grid has a column for each, plus an owner and a date.

Each quarter, and after any move by a weekly-tier alternative, plot the alternatives on the two criteria buyers weigh most. Use the columns from your competition slide, which The Market and Competition Slides in Your Investor Pitch shows how to choose. Where the criteria can be measured, plot numbers, not impressions; in my view, a map drawn from impressions flatters the company that drew it.

The gap is the space buyers value that no alternative fills. After each move, check whether it is still empty. If a rival is moving toward it, sharpen your message around what is still different; if it has closed, rework your positioning with Step 2 of the Foundations Playbook.

A battlecard is the intelligence in the form a rep can use with a buyer on the line: where the rival wins, where you win and the proof, dated facts, the top objections with answers, and when to walk away. The Battlecard Builder asks for each of these and flags undated facts and words such as “always” and “best”.

Start with where the rival wins. Buyers already know it, and in my view a card that admits nothing reads as spin and costs the rep their credibility at the first pushback. Saying the strength first, then when it matters less, is also the honest order.

Write each talk track in three moves: acknowledge the objection, reframe it with a question, and prove the difference with something the buyer can see. Build them from the objections reps actually hear, collected in Step 3, not from what marketing expects them to hear.

Make one card for each weekly-tier alternative, including a “why change” card for doing nothing and the spreadsheet, which argues against the status quo rather than a rival. Keep the cards where reps work, show the last-checked date on each, and retire any fact older than 90 days, which is the Battlecard Builder’s own rule.

Comparative claims are where competitive intelligence becomes a risk. The FTC’s policy on comparative advertising encourages naming or referring to competitors but requires clarity and, if necessary, disclosure to avoid deceiving buyers. It evaluates comparative advertising in the same way as all other advertising, and holds that disparaging advertising is permissible so long as it is truthful and not deceptive^[9].

The FTC also expects advertisers to have a reasonable basis for objective claims, express or implied, before they are disseminated. A claim that says “tests prove” or “studies show” needs at least the level of support it advertises^[10]. A rival can act too: section 43(a) of the Lanham Act lets anyone who believes they are or are likely to be damaged sue over commercial advertising that misrepresents the nature, characteristics or qualities of their own or another person’s goods or services^[11].

If you sell in the EU or the UK, the conditions are written as a list. Article 4 of the EU directive on misleading and comparative advertising permits a comparison only if, among other conditions, it is not misleading, objectively compares one or more material, relevant, verifiable and representative features, and does not discredit or denigrate a competitor’s marks, goods or services^[12]. Regulation 4 of the UK’s Business Protection from Misleading Marketing Regulations 2008 sets the same tests^[13].

In my view, battlecards, sales decks and sales emails deserve the same standard as ads, because buyers forward them. The law points the same way: the FTC Act declares unlawful deceptive acts or practices in or affecting commerce, not only deceptive ads^[14], and the FTC’s policy statement on deception lists false oral or written representations among the practices found deceptive^[15]. The Lanham Act clause above also reaches “commercial advertising or promotion”, not ads alone^[11]. This is orientation, not legal advice.

In practice: compare on facts you checked against the rival’s own current page, state the basis and the date, drop “only” and “best” unless you can prove them, and recheck every comparison when the rival ships. The Claims & Proof Checker flags comparative claims in a draft before anyone sees it.

Hold a 60-minute review each quarter. Each owner brings the head-to-head record for their alternative from the CRM, what changed, which facts are older than 90 days, and which questions now have answers. End with written decisions, not a slide.

Read head-to-head win rates next to their counts. At startup volumes a single deal swings the rate, so look for a direction that holds for two or three quarters before you change a card or a price; the Win-Loss Analysis Playbook covers reading the range.

Prune and promote. Alternatives move between tiers as the deals show, and questions that fed no decision are dropped. Between reviews, the owner updates a card the same week a weekly-tier rival changes its price or ships into the gap, as Competitor B did.

What changes by stage?

What mistakes should I avoid?

• Tracking only the rivals you can name. Buyers disproportionately stick with the status quo, so doing nothing and the spreadsheet need tracking too (Step 1).
• Collecting everything because it is there. If no decision needs it, it is noise (Step 2).
• Posing as a buyer to get a demo or a price. It is pretexting, and it taints everything you learn (Step 3).
• Reporting moves without saying what they mean. A screenshot is not intelligence (Step 4).
• A battlecard that admits no rival strengths. Buyers know them, and reps lose credibility (Step 5).
• A comparison that was true last quarter. Recheck every claim when the rival ships (Step 6).
• No owner and no review. The research goes stale within a quarter (Step 7).