The mistake: reading one report and deciding it worked, or didn't.
Why it hurts: AI answers vary from run to run, even for the same prompt. One run is a snapshot.
The fix: track the same prompts in a Scheduled Report and read the trend on Trends over several runs.
The mistake: optimizing keywords and links and assuming AI answers follow.
Why it hurts: a top-ranked page doesn't guarantee a mention. AI answers lean on the pages they cite and on what they can state plainly.
The fix: track prompts, not only keywords, and work on the pages AI cites. See ChatGPT SEO vs Google SEO.
The mistake: one plan for ChatGPT, Claude, Gemini, Grok, Perplexity and Google AI Mode (AI Overviews).
Why it hurts: each reads different pages and runs different searches. Strong on one often means weak on another.
The fix: check Visibility by platform on the Overview, then filter Sources and Fan-outs to the weak one.
The mistake: rewriting your pages and waiting.
Why it hurts: many answers cite other people's pages: reviews, comparisons, forums, press.
The fix: find the pages AI cites for your prompts in Sources, then get onto them with Backlink outreach and Community.
The mistake: "the leading, innovative, all-in-one platform".
Why it hurts: there's nothing to repeat. AI platforms repeat facts.
The fix: say what you do, for whom, where, at what price and compared with whom. "Mailchimp is an email marketing platform for small businesses, with a free plan" gives an answer something to use.
The mistake: looking only at your own score.
Why it hurts: the brands named instead of you show exactly what the answers want.
The fix: on Competitors, check share of voice; on the Prompts page, open the answers that miss you and see who is named and which pages are cited.
The mistake: counting on feed posts and follower growth.
Why it hurts: posts behind a login rarely reach an answer.
The fix: focus on public discussions AI does cite. Sources lists the Reddit threads, YouTube videos, LinkedIn posts and forums in your answers. Answer the right threads with Community.
The mistake: different names, categories, prices or locations on different sites.
Why it hurts: AI platforms pick one version, or get it wrong.
The fix: run Brand Audit to see how AI platforms describe you and where they're wrong, then correct the source pages and your listings. Add your other names as aliases in Brand settings, so answers that use them count.
The mistake: a strong page for your main product and nothing for the questions around it.
Why it hurts: prompts in the uncovered categories name someone else.
The fix: on the Prompts page, open the Categories tab or filter by category and look for the prompts that miss you. Content writes articles aimed at those prompts.
The mistake: asking ChatGPT now and then and going by feel.
Why it hurts: you can't see progress, drops or which fix worked.
The fix: a baseline report, a weekly Scheduled Report, and a monthly look at the score, share of voice and citations.
Run a report and open How to fix your results
Fix your facts (8) and your own pages (5, 9)
Get onto the pages AI cites (4, 7)
Measure weekly and read the trend (1, 10)
Email [email protected] and I'll help.
