Skip to main content
SEO

What to Do When AI Overviews Show Your Bad Reviews

AI Overviews and chatbots now summarise your reputation. Here's what to do when AI surfaces your bad reviews — and how to shape what it says.

By NetTrackers

There's a new layer to online reputation that didn't exist a couple of years ago. When someone asks Google's AI Overviews, ChatGPT, or Perplexity about your business, the AI doesn't just list results — it summarises them. And that summary can pull your worst reviews to the surface, phrased as fact, before the user has clicked anything. "While [Business] is generally well-reviewed, some customers report slow service and billing issues" is now something an AI will happily tell a prospective customer. This article covers what to do about it.

Why this is a different problem

Traditional reputation management dealt with a list of blue links you could influence and reorder. AI-generated summaries are different in three ways.

First, they're synthesised, not listed — the AI reads across many sources and produces a verdict, so a single strongly negative theme can dominate the summary even if it's a minority of your reviews. Second, they carry authority — users tend to accept an AI summary as a neutral consensus rather than one opinion among many. Third, they're less transparent — you often can't see exactly which sources shaped the answer, which makes them harder to correct.

There's a fourth difference worth naming: AI summaries are increasingly the first and sometimes only thing a prospect reads. In the old model, a searcher scanned ten links and formed their own impression. Now a growing share of people read the AI's paragraph, feel they have their answer, and never scroll to the reviews themselves. That concentrates enormous influence in a few sentences you didn't write and can't directly edit — which is precisely why the underlying signals matter more than ever.

How AI decides what to say about you

AI systems draw on the same broad signals that shape search: your reviews and their content across platforms, mentions of your business in articles and forums, your own website and profiles, and the overall sentiment of everything written about you. If your reviews frequently mention a specific complaint, that theme becomes prominent in the AI's mental model — and in its summary.

The important implication: you shape AI summaries the same way you shape search, by shaping the underlying corpus of information. There's no "edit" button, but there is a body of evidence the AI reads, and you can influence its balance.

A worked example of how a theme forms

Picture a Birmingham accountancy firm. Over 18 months it accumulates 40 reviews. Thirty-two are warm and four or five stars. But six of them, clustered around a chaotic tax-season, mention "slow to reply" and "hard to get hold of". When a prospect asks an AI "is [Firm] any good for small business accounts?", the AI reads all 40 — and because the responsiveness complaint appears repeatedly and in similar language, it surfaces as a named theme: "Reviews are largely positive, though several clients mention slow communication." The firm is genuinely good, and the complaint is a minority view, but the repetition and consistency of that phrase gave it weight the algorithm couldn't ignore. The fix isn't to argue with the AI; it's to resolve the communication issue operationally and then build a run of newer reviews that describe fast, clear responses — giving the AI a fresher, stronger counter-signal.

What to do

Fix the underlying theme, not just the reviews. If AI keeps mentioning "slow service" or "hidden costs", that's because enough real reviews say it. The durable fix is operational — solve the actual problem — so that newer reviews stop reinforcing the theme. AI weights recent, consistent signals; change the reality and the summary follows, over time.

Build volume and recency of positive signals. A steady stream of genuine, recent, specific positive reviews shifts the balance the AI reads. Recency matters: AI summaries lean on current sentiment, so an active review habit reshapes the picture faster than an old stockpile of five-stars. This is the single most effective lever.

Get the specifics into your reviews. Because AI reads review content, encourage genuine reviews that mention the things you do well and want associated with your name. If your strength is "always on time and transparent about pricing", reviews saying exactly that give the AI positive specifics to surface — and counterbalance old complaints.

Strengthen authoritative third-party mentions. AI trusts credible sources. Genuine press coverage, legitimate directory profiles, case studies and authoritative content about your business give the AI trustworthy positive material to draw on, diluting the weight of a few loud negative reviews.

Make your own content answer the concern. If a recurring criticism is unfair or outdated, address it factually on your own site — an honest FAQ, a transparent pricing page, a clear explanation of your process. AI systems read your owned content, and a clear factual rebuttal gives them an alternative to the complaint.

Optimise for GEO. Getting cited accurately by AI is an emerging discipline — generative engine optimisation — covered in GEO Explained: Getting Cited by ChatGPT and AI. The same practices that get you cited favourably in AI answers also help ensure the AI's summary of your reputation is fair and current.

Turning your website into a source AI trusts

One under-used lever deserves expanding. AI systems lean heavily on clear, structured, factual content — and much of that can live on your own site, which you fully control. In practice: publish a genuinely useful FAQ that answers the real questions prospects ask, including any recurring concern, in plain factual language; keep a transparent, specific pricing or process page rather than a vague "contact us for a quote"; make sure your key facts (what you do, where, for whom, your credentials) are stated plainly and consistently everywhere they appear; and use clear headings and straightforward structure so machines can parse the meaning. If a recurring criticism is that you're "expensive with hidden costs", a candid page explaining exactly how your pricing works gives the AI a factual, first-party source to weigh against the complaint. You're not tricking the model; you're giving it better, clearer evidence.

What not to do

Don't try to game it with fake reviews or manipulated content — AI systems increasingly detect and discount manipulation, and the platform-level risks (suspension, consumer-law exposure) are unchanged. Don't ignore it either; AI summaries are becoming the first impression for a growing share of prospects, and an unmanaged one can quietly cost you enquiries you never see.

Monitor what AI says about you

Periodically ask the major AI tools about your business, as a customer would: "Is [Business] any good?", "What do people say about [Business]?", "Best [service] in [town]?" See what they surface. This is now a core reputation-monitoring task, and most businesses don't do it at all — which means they have no idea what a growing share of their prospects are being told.

Make it a proper routine rather than a one-off. Once a month, run the same handful of questions across Google's AI Overviews, ChatGPT and Perplexity, and keep a dated note of what each says. Watch for named themes ("some customers mention…"), factual errors (wrong opening hours, a service you no longer offer, an out-of-date location), and whether the overall verdict is trending warmer or cooler. Because the models draw on shifting sources, the answer changes over time — so a summary that was fair in spring may harden around an old complaint by autumn if you've stopped generating fresh positive signals. Treating this as a recurring check, like glancing at your reviews, is what keeps you from being blindsided.

Frequently asked questions

Can I directly edit or remove what an AI says about my business?

No. There's no edit button and no single source to correct. AI summaries are synthesised from many signals — your reviews, mentions, your website and overall sentiment. You influence them indirectly by improving the underlying corpus: resolving real issues, building genuine recent reviews, strengthening authoritative mentions and publishing clear factual content on your own site.

Why does the AI focus on one complaint when most of my reviews are positive?

Because AI weights consistency and repetition. If several reviews use similar language about the same issue, that theme reads as a pattern the model feels obliged to mention, even if it's a minority. The counter is fresher, specific positive reviews and operational fixes that stop new reviews reinforcing the theme.

How quickly will the AI summary change if I improve things?

There's no fixed timeline, but it's gradual rather than instant. AI leans on recent sentiment, so a steady flow of new, genuine, specific positive reviews tends to shift the picture faster than an old stockpile ever could. Expect a matter of weeks to months as fresh signals accumulate, not overnight.

Does the same reputation work help me appear in AI answers at all?

Yes. The practices that make your reputation summary fairer — genuine reviews, authoritative third-party mentions, clear structured content on your own site — are largely the same ones that help AI cite you accurately and favourably. This overlap is the emerging field of generative engine optimisation (GEO Explained: Getting Cited by ChatGPT and AI).

Should I worry about AI or focus on traditional Google reviews?

Both, because they draw on the same foundations. Genuine, recent, specific reviews and a strong, accurate first-party web presence improve your standing in ordinary search and in AI summaries. You don't need a separate "AI strategy" so much as a solid reputation strategy that's aware AI is now reading it.

How do I check what AI is saying without special tools?

Just ask, as a customer would. Once a month, put the same few questions to Google's AI Overviews, ChatGPT and Perplexity — "Is [Business] any good?", "What do people say about [Business]?" — and keep dated notes. Watch for named complaint themes, factual errors, and whether the tone is warming or cooling over time.

Where NetTrackers fits

We build every campaign to be found in traditional search and cited fairly by AI — combining reputation management with AI SEO. If you don't know what ChatGPT or Google's AI is saying about your business, that's exactly the blind spot we help close. Month-to-month, no contracts. Book a free strategy call.