What questions should a campaign test?
Start with the questions a buyer will ask before, during, and after the campaign, not with the campaign slogan. Group prompts by problem, audience, comparison, use case, objection, and next action. Include branded questions, unbranded category questions, and prompts that mention the campaign offer only indirectly.
Write each prompt as a buyer might type or speak it. Avoid turning every variation into a separate question when the underlying intent is the same. Record the intended answer, the page that should support it, and the competitor most likely to appear instead. Keep campaign prompts separate from evergreen tracking prompts so a short-term launch does not distort the wider visibility picture.
A campaign can look successful in branded searches while remaining absent from the category questions that create demand. The question set should therefore include both recognition and discovery. For a broader planning framework, the product launch process can help connect campaign timing with the pages and answers that need to be ready.
Set the pre-launch visibility baseline
Record the campaign baseline before publishing creative, landing pages, or public announcements. Check whether ChatGPT, Perplexity, Gemini, Claude, Grok, Google AI Overviews, and Google AI Mode mention the brand, cite its pages, name competitors, or provide no useful brand answer for each priority prompt.
Capture the exact answer, cited URLs, position or prominence where the engine exposes it, and the date of the check. The answer text matters because a brand mention without a supporting citation may not send a reader to the campaign page. Also note whether the cited source is owned, earned, partner-controlled, or unrelated to the business.
Use the same prompt wording for the baseline and the first comparison after launch. Otherwise, a changed prompt can look like a visibility improvement. Cituna tracks these seven engines every day and shows which competitors and pages they cite instead. Its tracking scope does not include Microsoft Copilot, so teams that report across engines should label that gap rather than imply complete coverage.
Choose the campaign visibility decision rule
Decide in advance what evidence will trigger a change to the campaign, rather than judging visibility from a single answer. A useful rule separates three outcomes: the brand is mentioned and supported by the intended page, the brand is mentioned but supported elsewhere, or the brand is missing while a competitor is cited.
Treat the second outcome as a content and distribution problem, not as a complete win. The campaign may have recognition, but the wrong page is carrying the explanation. Treat the third outcome as a prioritization signal. If several high-value prompts produce the same competitor and cited source, investigate that source before rewriting every campaign asset.
Set a review window that fits the campaign cycle and define who can change the landing page, supporting content, internal links, or outreach. Do not react to one volatile answer without checking the same prompt again and comparing the cited sources. The goal is a repeatable decision, not a screenshot that supports a preferred conclusion.
Audit the landing page against buyer prompts
Make the campaign landing page answer the priority buyer prompts directly, with clear evidence for each important claim. A visitor and an answer engine should be able to identify what the offer is, who it suits, which problem it addresses, how it differs from alternatives, and what action follows.
Map every priority prompt to a section or supporting page. Remove gaps where the campaign uses a headline or product name but leaves the practical question unanswered. Replace vague claims with specific explanations, qualifying conditions, examples, and links to relevant evidence. Keep the campaign promise consistent across the ad, landing page, product information, and supporting materials.
Do not force every prompt onto one page. If a comparison, implementation detail, or eligibility condition needs more space, link to a focused page and make the relationship clear. The page intended to earn a citation should be stable, crawlable, and easy to interpret without campaign jargon. Review the page before launch, because post-launch edits can make it harder to identify which change affected visibility.
Check citation paths beyond the landing page
Trace the pages that answer engines already cite and decide whether the campaign can improve those citation paths. A campaign landing page is not automatically the strongest source for every question. Product documentation, comparison pages, customer education, partner pages, and authoritative third-party coverage may carry different parts of the buyer journey.
For each missing or weak prompt, ask whether the intended page has enough context to stand alone. Check that internal links connect the campaign page to supporting explanations and that important claims are not confined to images, scripts, or forms. Where a third party is cited instead, compare its wording with the campaign page and identify the specific information it supplies more clearly.
Avoid changing every cited page at once. First separate a source problem from a message problem. If the campaign page is clear but engines consistently cite an older or external explanation, improve the relationship between the pages and make the preferred source easier to find. Campaign teams working across countries can use the new market launch checklist to keep regional citation paths distinct from the main launch.
Test competitors and substitution prompts
Test the prompts where a buyer could choose a competitor instead of the campaign offer. Include direct comparisons, best-for questions, migration questions, alternatives, and prompts that describe the problem without naming any vendor. These prompts reveal whether the campaign is visible only after a buyer already knows the brand.
Record which competitor appears, what evidence supports its inclusion, and whether the answer gives a reason to choose it. Then compare that reason with the campaign page. A competitor may win because it explains a use case better, has a more suitable page, is associated with a clearer category term, or is cited by a source the brand does not control.
Fix the mismatch that matters most to the campaign objective. If the competitor wins on an unanswered use case, improve the relevant page. If the competitor wins because the campaign page makes a broad claim without proof, add precise supporting information. Do not copy competitor wording blindly. The test is whether a buyer can understand the campaign's fit and limitations from a credible source.
Launch with a change log and ownership
Launch only after every visibility change has an owner, a date, and a stated reason. Create a simple record for prompt edits, landing-page revisions, internal-link changes, supporting content, outreach, and campaign activation. Note which prompts each change is expected to affect and whether the change is intended to improve mentions, citations, or both.
Keep campaign changes separate from unrelated SEO work. Without that separation, a later improvement cannot be attributed confidently. Preserve the previous page copy and prompt set so the team can compare the pre-launch and post-launch states. Record whether a change was published, indexed, cited, or merely planned.
Assign one person to review answer evidence and another to approve material campaign changes when the business requires separation of duties. Cituna joins AI answers to Google Search Console data and provides SEO, AEO, and GEO fixes, which can help connect answer visibility with search performance instead of treating either signal as the whole result. The operating rule remains simple: every change should have a hypothesis that the next review can test.
Review post-launch evidence before scaling spend
Review post-launch answers against the pre-launch baseline before increasing campaign reach or budget. Re-run the same priority prompts, compare mentions and citations, inspect which pages engines cite, and check whether competitor substitution changed. Then separate a durable improvement from a temporary answer fluctuation by looking for the same pattern across repeated checks and related prompts.
A useful review has four conclusions: keep the campaign as planned, fix the page before scaling, change the message, or pause the campaign while the evidence is resolved. A mention without a citation may justify improving the source path, while a citation to the wrong page may justify clearer internal linking or a more focused supporting page. A missing brand across discovery prompts points to a larger category or authority problem than a missing mention in branded prompts.
Compare answer visibility with Google Search Console impressions, clicks, and query patterns, but do not treat search traffic as proof that an answer engine cited the campaign. Keep the launch record updated with the observed result and next action. That creates a reusable decision history for the next campaign rather than another isolated snapshot.
Related reading
- AI Visibility Audit Checklist: 12 Checks Before You Publish
- How often should I check my AI visibility
- AI Visibility Review Meeting Agenda
Sources consulted
- Google Search Central (developers.google.com)
- Google Search Console Help (support.google.com)
- OpenAI Platform Documentation (platform.openai.com)
- Perplexity API Documentation (docs.perplexity.ai)
Drafted with AI assistance from our own research and Search Console data, and reviewed by Rahul A before publishing. Rules and prices change; check the linked official source before you act.