The AI Investor Answer Test: Four Questions Every Small-Cap IR Team Should Run Monthly

Ask three mainstream AI assistants what your company’s biggest risk is, what could change its valuation, and whether management has given guidance. If the answers are stale, vague, or confidently wrong, the problem is not the chatbot—it is the evidence trail your IR program has left behind.

AI systems increasingly turn public-company filings, releases, earnings materials, and transcripts into investor-facing summaries. Research tools now market AI analysis across public-company documents, while current AI coverage continues to focus on whether spending is producing revenue, profit, and cash flow. For a small-cap issuer, that means discoverability is no longer just a search-engine problem. It is an evidence problem.

The test is not “Does AI mention us?”

A company can appear in an answer and still be poorly represented. The useful question is whether an investor receives a defensible, current explanation of the business without having to reconstruct it from scattered documents.

Run the test monthly using the same four questions:

• What is the company’s biggest risk? The answer should reflect the risks management has actually identified, not a generic industry concern or an outdated event.

• What could change the company’s valuation? The answer should connect business drivers to measurable milestones, while avoiding invented certainty about market performance.

• Has management given guidance? The answer should distinguish formal guidance, operating targets, historical results, and management commentary.

• What changed in the latest update? The answer should identify the new information rather than repeat the company description.

Score the answer, not the wording

Use a simple four-part score for each response:

• Accuracy: Are the facts correct?

• Currency: Does the answer reflect the latest material disclosure?

• Attribution: Can the answer be traced to a public company source?

• Context: Does it include the relevant limitation, risk, or qualification?

Score each category from zero to two. A zero means missing or wrong; one means incomplete; two means clear and supported. The total is less important than the failure pattern. A low attribution score points to weak source architecture. A low context score suggests that the company has published claims without enough explanation around them. A low currency score often means the latest update is not easy to connect to the prior record.

Fix the evidence before optimizing the language

The tempting response is to rewrite the website, add more keywords, or publish another summary. That treats the symptom as the problem.

Start with the public record:

• Check that the business description is consistent across the website, filings, releases, and presentations.

• State which metrics matter and define them the same way each time.

• Separate historical performance from targets and expectations.

• Explain material changes instead of assuming investors will infer them.

• Retire obsolete language that still appears in evergreen pages or downloadable decks.

• Make the latest update easy to find and easy to relate to the previous one.

Only after those corrections should wording and discoverability become the focus. Better phrasing cannot repair an unsupported claim, an omitted risk, or a stale investor presentation.

Turn the audit into a monthly IR habit

Keep a record of the four prompts, the date tested, the answers received, and the source trail behind each answer. Track recurring errors rather than chasing every variation in output. AI systems can phrase the same conclusion differently; the underlying factual failure is what matters.

FiskLabs’ discoverability or AI-monitoring capability can help repeat this baseline audit across reporting cycles. The value is not a flattering score. It is a short list of public evidence that needs correction before investors—or the systems summarizing the company for them—fill the gaps.

The goal is not to make an AI assistant say more about the company. It is to make the company’s public record clear enough that an answer can be accurate, current, and properly qualified.