AI Reputation Management: What LLMs Say About Our Brand, and What We Do About It

By Sean Chaudhary · Published Oct 13, 2026

What AI reputation management actually means

When a buyer asks ChatGPT, Perplexity or Google’s AI Overviews about your company, the answer is a summary of everything the model can find: your site, your reviews, directories, news, forums and competitors’ pages. It decides what you’re good at, who you’re for, what you cost and whether people like working with you.

AI reputation management is the work of watching those answers, checking every claim against reality, and fixing the sources that feed them. It’s entity work, review work, PR and honest positioning, all measured every single day.

Our numbers, last 30 days

We track AlchemyLeads across ChatGPT, Google AI Overviews and Perplexity in Searchable. From September 13 to October 13, 2026:

• Sentiment score: 80 out of 100 in ChatGPT’s brand perception of AlchemyLeads, with 61% positive, 36% neutral and 2% negative mentions.
• Across all three engines: 76.7 out of 100 over the past week, up 3.1 points.
• Volume: 2,551 AI mentions of AlchemyLeads in the last 30 days.
• Brand facts: 77 of 77 statements AI engines made about us checked out as correct.
• Branded prompts: we appeared in 99.4% of 312 answers when someone asked about AlchemyLeads by name.

What AI gets right about us

A lot, honestly. Because we publish our pricing, our ideal client and our results, the engines repeat them back accurately:

The engines also get our positioning right. One answer described AlchemyLeads as “a highly specialized, boutique consultancy” with transparent pricing from $6,850 a month, next to much larger agencies with custom, less visible pricing. That’s accurate: we stay small on purpose, so senior people do the work.

When that same answer compared us with two much bigger agencies, it ended with the exact line we’d want a buyer to read:

That line comes from years of saying the same thing, the same way, everywhere: on our site, in our entity map, in our schema, in our reviews and in every listing we control. We keep a single source of truth for every claim and number we publish, so the engines never get two versions of the story.

What AI flags as our weaknesses

Here’s the part agencies rarely publish. Across engines, the negatives come down to three themes:

• Premium pricing. ChatGPT says pricing from $6,850 a month “creates a high bar for success” and makes us a poor fit for many small businesses.
• Self-published proof. Our strongest results come from our own case studies, and the engines note they aren’t independently audited.
• A “small” independent review base. The engines count 11 Clutch reviews and 14 on Trustpilot, and point out there are no new Trustpilot reviews in the past year. That picture is incomplete. They tend to miss our 22 five-star Google reviews, our largest review source, and our client video reviews. In total we have about 58 five-star reviews and testimonials.

The one we accept: price and fit

The pricing note is correct, and we’re comfortable with it. We published our prices and our ideal client on purpose. If you’re under roughly $2M in revenue, we’re usually the wrong partner, and an agency built for high-volume small business work will serve you better. We say so on our AI SEO cost guide and in our fit guide.

That transparency does real work. A buyer doing research in ChatGPT gets an honest answer before they ever book a call. The good-fit companies show up already knowing our pricing, and the poor-fit ones self-select out. AI reputation management includes deciding which “weaknesses” are simply your positioning.

Our review count, in context

All in, we have about 58 five-star reviews and video testimonials: 22 on Google, 11 on Clutch, 14 on Trustpilot and 11+ client video reviews.

That’s 58 five-star reviews and testimonials for an agency that rarely has more than 50 clients at a time. Over the past few years we’ve averaged about 35 to 40 clients on the roster. Larger agencies serve hundreds or thousands of clients, so their review counts naturally run higher. For a roster like ours, it means a massive share of our clients have gone on the record.

The quiet Trustpilot page, explained

Here’s something the AI summaries miss. Every one of our 14 Trustpilot reviews is five stars. We have zero four-star reviews there. Yet our TrustScore shows 4.5.

That’s because a TrustScore is a weighted score. Trustpilot blends baseline reviews into the calculation and gives recent reviews far more weight, so a profile with no new reviews drifts below its true average. We don’t get new Trustpilot reviews because its tools for inviting customers and showcasing reviews are mostly part of its paid plans, and we haven’t paid for them. So our review requests go elsewhere.

We’ve put our effort into review content we own. When a client wants to vouch for us, we point them to Google or ask them to record a short video review we can share with future buyers. We now have more than 11 client video reviews, and our most recent Google review came in about a month ago. We share a selection of them on our reviews page.

So yes, our Trustpilot profile looks quiet, and the 4.5 reflects that quiet period. Every review on it is still five stars. If we turn on Trustpilot’s paid tools later, we expect that score to catch up quickly.

The ones we’re working on: proof and reviews

The AI engines want to see more third-party proof, and that’s fair:

• Fresh reviews in more places. We keep collecting Google and video reviews, and we’re adding more to Clutch, so the engines see recent, independent voices.
• Proof that lives off our site. Mentions in trade publications, industry lists and interviews carry more weight with AI than anything we say about ourselves. That’s the digital PR work we do for clients, pointed at ourselves.
• Clearer evidence on our own pages. Every case study names the metric, the timeframe and the source, so a model can quote it accurately.

It changes every day

AI sentiment moves daily. Over one recent week our score swung by several points from one day to the next, with no change on our end. Engines re-crawl, re-rank sources and phrase things differently each time someone asks.

So we treat it as ongoing work. We check sentiment, brand facts and competitor comparisons every day, and we look for patterns over weeks before we change anything. One bad day is noise. A new weakness that sticks for two weeks is a signal.

Showing up next to much bigger agencies

AI answers often place us alongside well-known names in search, including Omniscient Digital, Siege Media, WebFX, First Page Sage, iPullRank and Directive. Many of them are much larger teams than ours, and they’ve earned their spots. We’re a senior boutique team, and it’s good company to be in.

Across the 1,227 brands that showed up in the answers we track over the last 30 days, AlchemyLeads ranks seventh for visibility, and when we’re mentioned we’re usually mentioned first. Bigger agencies still win more of the generic “best agency” prompts, and that’s where our AI SEO work is focused now.

For a smaller team, the advantage comes from knowing exactly what each engine gets right and wrong about you, and fixing the sources quickly.

Why you should track branded prompts

Most AI visibility tracking focuses on category prompts like “best SEO agency for manufacturers.” Those matter, and your reputation shows up most clearly in the branded prompts: the questions buyers ask once they already know your name.

Our branded prompts are where we see the full picture: we show up in 99.4% of those answers, and they’re what surface the sentiment, the brand facts and the head-to-head comparisons. Branded prompts show you what AI tells a buyer who is about to call you.

A good starter set of branded prompts:

• “Is [brand] legit?” and “[brand] reviews”
• “How much does [brand] cost?”
• “[brand] vs [competitor]” for your top three competitors
• “Who is [brand] best for?”
• “How do I hire [brand]?”

How to run your own AI reputation audit

1. Track branded and category prompts across ChatGPT, Perplexity and Google AI Overviews, daily. (Here’s the stack we use.)
2. Read the actual answers. Scores are a summary. The sentences tell you which source the model is leaning on.
3. Verify every brand fact against one source of truth for your pricing, ICP, results and services. Fix the page that’s wrong, then watch the answers update.
4. Sort weaknesses into “positioning” and “problems.” Accept the first, fix the second.
5. Build proof off your site: reviews, earned media, lists and partner pages.
6. Benchmark against competitors weekly on visibility, sentiment and position.
7. Connect it to revenue. Watch for AI conversation queries in Search Console and track AI referral traffic, so reputation work ties back to pipeline.

Frequently Asked Questions

What is AI reputation management?

AI reputation management is monitoring what AI assistants like ChatGPT, Perplexity and Google AI Overviews say about your brand, checking those statements for accuracy, and improving the sources they rely on, such as your website, reviews, listings and earned media.

How do I check what ChatGPT says about my brand?

Ask it the questions your buyers ask, including branded ones like “Is [brand] legit?” and “[brand] vs [competitor].” To track it reliably, use an AI visibility tool that runs the same prompts daily across ChatGPT, Perplexity and Google AI Overviews and scores sentiment and accuracy.

Why does AI sentiment about my brand change every day?

AI engines pull from live sources, re-rank them and phrase answers differently each time. Small daily swings are normal. Look for patterns that hold for a week or more before acting.

Can you fix negative AI sentiment?

Often, yes. Inaccurate facts can be fixed by correcting the source pages and keeping your claims consistent everywhere. Fair criticisms take real change, like earning more independent reviews. Some negatives, like premium pricing, may simply reflect your positioning.

How many reviews does AlchemyLeads have?

AlchemyLeads has about 58 five-star reviews and testimonials: 22 Google reviews (5.0), 11 Clutch reviews (5.0), 14 Trustpilot reviews (all five stars) and 11+ client video reviews. Google is its largest review source.

Why is AlchemyLeads’ Trustpilot score 4.5 if every review is five stars?

All 14 of AlchemyLeads’ Trustpilot reviews are five stars. Trustpilot’s TrustScore is a weighted score that favors recent reviews, and AlchemyLeads doesn’t use Trustpilot’s paid invitation tools, so with no new reviews the displayed score sits at 4.5.

Which prompts should I track for AI reputation?

Track branded prompts (reviews, pricing, comparisons, who you’re best for) to see sentiment and facts, and category prompts (best provider for your niche) to see visibility against competitors.