Your Customers Are Searching for You in AI. You Probably Have No Idea.

A couple of days ago, a friend of mine who works in consumer goods complained to me.

He said, "Liu Run, I noticed something really strange. Our old customers keep asking us lately — did you change ownership? Did you drop your prices?"

No. Nothing had changed.

So what was going on?

He eventually got someone to look into it. Here's what he found: across several AI search engines, whenever a user asks, "What do you think of brand XX?" the AI's answer was getting his company's description wrong. Some answers quoted prices that were too high. Others listed a competitor's product as his flagship.

He was stunned.

I was taken aback for a moment too. Then I told him: you need to take this seriously.

A New Channel You Can't See

What exactly is AI search?

Think about how you look things up these days. You used to open a search engine and sift through ten results yourself. Now? You just ask an AI, and it hands you a direct answer — a tidy paragraph with citations.

Your customers are using this exact method to look up your brand. Your products. To figure out who's better — you or your competitors.

Here's the question: how often do you show up in that answer? And when you do, is the AI saying good things or bad things? Is it recommending you, or your rival?

You have no idea.

Your entire playbook used to revolve around traditional search rankings. Which keywords you rank for, where your traffic comes from, what your click-through rate is — all crystal clear. But now? Customers aren't clicking links anymore. The AI serves up the answer right to them. Whether you got cited or got ignored — you have no way to find out.

This is a new, massive, and completely invisible channel.

What Makes It So Hard?

Measurement.

Traditional search has mature tooling. It tells you traffic, rankings, clicks. You buy the tools, read the data, adjust your strategy — that loop has been running smoothly for over a decade.

But AI search is different. It's a black box. Why did the AI mention your competitor in its answer but not you? You can't figure it out. You want to peek behind the scenes and check "where do I rank in this particular AI?" — sorry, that dashboard doesn't exist.

Part of your marketing budget is flowing into a space you can't measure. And that's deeply uncomfortable.

Even worse — you can't even see what your competitors are up to. In traditional search, you could install a plugin and check competitor rankings. In AI search? You ask the same question ten times, and the AI might give you ten different answers. How do you even track that?

So, naturally, a whole crop of tools has sprung up specifically to tackle this problem.

What Do These Tools Actually Do for You?

I've sorted the common tools on the market into categories. Regardless of what they call themselves, the core comes down to three things.

First: counting how often you get mentioned.

This means tracking, across major AI search engines, how frequently your brand gets mentioned or recommended. Think of it as the "rank tracking" of traditional SEO — except instead of tracking your position on a search results page, you're tracking the probability that you appear in the AI's answer.

Some tools go more granular, telling you which categories of questions you get mentioned in most, and which topics you're completely absent from. That's far more useful than a single total number. Because once you know where you're missing, you know where to focus your efforts next.

Second: seeing what the AI says about you.

Getting mentioned isn't enough. Whether the AI is praising you or trashing you, whether it's putting you in the recommendation slot or making you the one getting thrown under the bus in a comparison — the difference is enormous.

There's a category of tools dedicated entirely to this. They analyze the tone of the AI's answer when it mentions you: positive, negative, or neutral. The really aggressive ones will even tell you, point by point, which specific information sources caused the AI to form that impression of you.

What does that amount to? It's like hiring an intelligence operative who combs through AI answers every day and delivers you a daily sentiment briefing.

Third: digging into why the AI answers the way it does.

This is the most technically demanding piece.

Every answer the AI generates cites certain information sources. It might be an industry website, a forum, a product review. What your brand looks like in the AI's eyes depends heavily on what those sources have written about you.

There are tools designed specifically to dig up these sources and tell you: these are the main websites shaping your brand's AI image. Then you can strategically cultivate coverage on those sites, and fill in any descriptions that work against you.

Put bluntly, the AI doesn't form opinions out of thin air. Its opinions are fed by outside content. What you need to do is influence the content that feeds it.

So, What Should You Look for When Choosing a Tool?

Prices vary wildly. From a few dozen dollars a month on the low end, to enterprise-level deals that require separate negotiation — the spread is staggering.

I think when you're choosing, you don't need to get dazzled by fancy features. Just focus on three questions.

What's your scale?

If you're a small team just testing the waters, find something under a few hundred dollars a month that shows you basic visibility data and sentiment analysis. That's enough. Don't jump straight to an enterprise solution — you won't use half the features, and the money's spent for nothing.

If you're a mid-to-large company where brand equity is your lifeblood, then it's worth investing in tools that offer deep analysis, cross-language tracking, and cross-region monitoring. Because your customers are spread across regions and languages, and the way AI describes you in each one could be completely different. Small tools can't cover those blind spots.

Where's your pain point?

Some companies' problem is "we're simply not getting mentioned." What you need then is visibility monitoring plus topic gap analysis — figuring out which topics you're invisible on.

Other companies' problem is "we're getting mentioned, but the information is wrong." What you need then is sentiment analysis plus source attribution — figuring out which content is skewing the narrative, then correcting it precisely.

Does your team have someone who can act on the data?

This is the most easily overlooked point.

A tool gives you dashboards, reports — but if no one on your team is regularly reviewing them, interpreting them, and turning them into action, that money is wasted. No matter how good the tool is, it only gives you a pair of eyes. It can't replace the thinking head on your shoulders.

One Last Thing

My friend eventually spent some money and bought one of these tools. After using it for a month, he got a report.

The report stated clearly: on a certain topic, his brand's probability of being mentioned by AI was only a third of his competitor's. On another topic, he was mentioned frequently — but the AI's tone leaned negative, mainly influenced by a few negative review posts.

He held the report and told me, "Liu Run, I can finally see it clearly."

I said, "Now that you can see it, things get a lot easier."

An invisible channel doesn't mean it doesn't exist. If you don't measure it, it'll sit there quietly, handing your customers off to your competitors — one by one.

In this era, whoever opens the black box first gains the upper hand.