Competitive Intelligence: A Practical Evidence Checklist

Start with the decision

SCIP describes competitive intelligence as gathering and analyzing information about competitors and their industries to help an organization make decisions. For a small team, a useful starting point is one choice that actually needs to be made. In our example: should we change our price for the customers we serve? The companies, prices, and customer accounts in this article are fictional.

That question determines what to investigate. You need to understand which customers can use the competing offer, what they receive, the commitment required, and whether price is affecting their choices. Your own costs and ability to serve those customers matter too. A folder full of competitor announcements will not answer all of those questions.

Write down the decision before collecting more material. Then ask what evidence could change it. If no plausible finding would make you reconsider, the research may be serving a decision already made. You can work through the argument below or go directly to the practical evidence checklist and use it with a decision of your own.

Why Sun Tzu invests in knowledge

The final chapter of The Art of War opens with the expense of mobilizing an army: 百姓之费，公家之奉，日费千金. Households and the state bear the cost. The chapter then describes prolonged confrontation in pursuit of a single day's victory. Intelligence appears inside this discussion of expensive, consequential action.

Sun Tzu condemns the commander who begrudges rewards while remaining ignorant of the enemy's situation: 而爱爵禄百金，不知敌之情者，不仁之至也. The moral force of the passage comes from the people and resources exposed by that ignorance. Saving on information can leave a much larger commitment poorly informed.

The chapter names the advantage of knowing beforehand as 先知. The next passage rejects divination and deduction from comparisons or calculations as substitutes for knowledge obtained from people who know the enemy's circumstances. Its concern is access to the actual situation before action.

The ancient context is military espionage. The chapter describes five kinds of spies, including people used to transmit false information, and places their work under close command. Those methods do not become a modern business research program merely by changing their names. What we carry into this article is the question they make urgent: what knowledge does this costly decision depend on, and how will we obtain it?

The final chapter returns to the first

The opening chapter supplies a framework for assessment: 故经之以五事，校之以计，而索其情. It asks the reader to examine five matters, compare conditions, and seek the actual situation. Its seven comparisons then ask which side has the better leadership, conditions, organization, strength, training, and clarity of rewards and penalties.

The final chapter supplies a vital part of the knowledge that such a comparison requires: 必取于人，知敌之情者也—knowledge must come from people who know the enemy's situation. Read together, the first and last chapters connect the questions of assessment with the work of learning about the other side.

Conditions within your own organization can be assessed internally. The other side's circumstances require information about that side. The text does not say that every answer to the five matters and seven comparisons comes from a spy. The useful connection is that a comparison needs a basis in what is actually happening.

For a modern team, we can organize that connection as a loop: define the decision, collect relevant evidence, compare the conditions, choose an action, and update the assessment as new information arrives. This is our synthesis of the chapter relationship. It gives the evidence checklist below a purpose: each entry should help answer a question that matters to the decision.

Separate facts, inferences, and unknowns

A fact in your research record is a specific observation with a traceable source. For example: on the date you checked, the competitor's public page displayed $19 per user per month. Save the wording, the page address, the plan name, and the conditions shown. If you did not observe the earlier page, call this the current advertised price; you have not yet established that the price was reduced.

An inference is a conclusion you draw from observations. ‘This could make the product more attractive to some of our customers’ is an inference worth investigating. ‘Our customers will leave unless we cut our price’ requires evidence about those customers and their alternatives. Recording it as a hypothesis lets other people examine how you reached it.

An unknown is a missing answer that could affect the choice. Do both plans include the same service? Is the advertised rate available month to month? Do the customers you serve qualify? Unknowns make the next research step easier to identify. Leaving them visible is more useful than quietly filling them with assumptions.

These labels describe the status of a particular statement. A company's claim is evidence of what it claims. A review is evidence of an account someone published. Establishing that the underlying product works as described may require a different check. Keep the observation separate from the wider conclusion it is being used to support.

Choose sources for the question

The U.S. Small Business Administration distinguishes research using existing material from direct research with consumers. Existing sources can save time, while direct research can explore questions specific to your customers and business. Its market research and competitive analysis guidance provides a useful modern starting point. Choose a source according to the answer you need.

For our pricing decision, the public terms help establish the advertised commitment. Customer conversations can reveal how people considered their options. Internal sales records may show whether similar customers have raised price objections. Those sources contribute different pieces; one does not automatically replace the others.

A source also has a perspective. A pricing page is produced to sell an offer. A reviewer reports an experience from a particular time and plan. A customer interview contains the participant's recollection, shaped in part by the questions asked. Record those circumstances so that a later reader can judge whether the material fits the claim.

Check whether apparently separate sources are independent. Three articles repeating the same announcement still lead back to one announcement. Several reviews may be informative, but their number alone does not establish that they represent your target customers or the current product.

Modern interviews need a clear purpose and an honest introduction. SCIP's code of ethics for competitive intelligence calls for disclosure of identity and organization before interviews and for honest, realistic conclusions. The example here uses public material and voluntary customer conversations; customers contribute knowledge of their own choices and experiences.

A practical evidence checklist

Use these seven checks for one decision. They are an editorial research aid, not a translation of Sun Tzu's seven comparisons. The aim is a record that another person can follow from the original material to the proposed action.

1. Name the decision and its scope

   State what you may do, for whom, and by when. ‘Should we change the monthly plan for teams with fewer than ten users?’ is easier to investigate than ‘What is our competitor doing?’ Name who will make the decision and what finding could change it.

2. Record the observation and its source

   Write what you actually saw or heard. Save the original address or document, the relevant wording, and the date checked. Keep a participant's report identifiable as a report. If a summary omits the qualifications, return to the original material before relying on it.

3. Keep the source's perspective visible

   Note who produced the information, what they were in a position to know, and the purpose of the material. A customer can describe their own purchase; that does not give them knowledge of the competitor's finances. A sales page can describe an offer without explaining why customers accept it.

4. Check date, audience, and conditions

   Record the plan, market, customer type, version, and period the information concerns. Ask whether those conditions match this decision. An old review of an enterprise plan may have little bearing on today's monthly offer for a small team. Update the record when the relevant conditions change.

5. Write the conclusion beside the evidence

   State the narrow conclusion the material supports, then label the additional inference you want to test. Ask what other explanation could fit the observation and whether an independent source can help distinguish them. Keep repeated versions of the same claim traceable to their common source.

6. List the unknowns that could change the choice

   Focus on unanswered questions with consequences. In the price example, whether our customers qualify for the competing plan matters more than a complete history of the competitor's website. Say what remains unknown rather than assigning a made-up confidence percentage to the whole decision.

7. Choose the next check and a review trigger

   Specify the missing answer, how to investigate it, and who will do so. Record when to reconsider: for example, if the public terms change or comparable customers begin choosing the other offer because of price. A useful next check should have a plausible effect on the decision.

Follow the price beyond the headline

Return to the fictional $19 offer. Reading the terms now reveals annual billing and a minimum of ten paid seats, even if the team is smaller. Your $29 plan is billed monthly and allows smaller teams. The two headline numbers describe different commitments. The lower advertised per-user rate still exists, but its relevance depends on which customers are comparing the plans.

For a ten-person team willing to commit annually, that offer might be an important alternative. For a three-person team seeking monthly flexibility, the minimum and commitment could change the comparison. You still need to check features, service, and the customer's priorities. The terms alone do not tell you which product anyone will choose.

Suppose a recent prospect says that both products made their shortlist and that monthly flexibility influenced the choice. Record this as one person's account. It supports investigating flexibility as a factor for similar prospects; it does not establish what all small teams value. Compare it with other relevant conversations and your own recorded sales outcomes before generalizing.

The team now has a better next step than an immediate price cut: investigate customers who genuinely overlap with the competing offer. If their choices are being influenced by the lower price, consider a response in that segment. If the comparison turns on commitment or service, a different response may be appropriate. The same observation can lead to different actions as the missing context becomes clearer.

Decide what is worth checking next

Information takes time and resources to obtain. The final chapter's criticism of skimping on intelligence makes sense alongside the cost of acting without it. It does not mean every unanswered question deserves an investigation of unlimited depth.

Prioritize an unknown by asking whether a different answer could change the choice. Reading the terms addresses a basic ambiguity in the price comparison. Learning how relevant customers choose between the offers addresses a different ambiguity. Collecting more announcements may add little if neither issue is clarified.

Match the next step to the stakes and the opportunity to change course. A small, reversible experiment can provide useful observations before a wider commitment. A difficult-to-reverse price change may justify more careful investigation of customer overlap and your own economics. Specify what the experiment would teach you; a trial without a learning question can create activity while leaving the original uncertainty intact.

Proceed when the evidence supports a bounded action, keep the unresolved assumptions visible, and set a reason to review. You may also choose to wait or redirect the effort. This checklist helps expose the basis of a judgment; it cannot promise that a source is truthful, a market stays still, or a decision succeeds.

Return knowledge to the next decision

The final chapter ends by calling intelligence 三军之所恃而动也—something the army depends on in order to act. The knowledge has a destination. It must reach the assessment and the people making the choice.

Chapter 3 adds the choice of whether to engage at all. Our reading of how Sun Tzu explains when to fight develops that connection: knowing conditions can support proceeding, waiting, or declining. For the pricing team, understanding a competing offer is useful when it changes what the team is prepared to do.

The connection between the opening assessment and the closing intelligence chapter also appears in our thirteen-chapter reading map. The circle returns to its starting point. In a continuing research practice, that return carries updated evidence into the next assessment.

Keep the decision question at the top of the record. Put the source next to the observation, the inference next to its supporting evidence, and the next check next to the uncertainty it addresses. When new information arrives, revisit the judgment that depends on it. That is how competitor research becomes part of an ongoing decision process.