The quickest test to see if Google and AI understand what you're selling
- Redazione
- 2 days ago
- 9 min read
Many companies don't just have an online visibility problem.
They have a deeper problem: Google, generative AI, and customers don't always understand what they're selling in the same way.
When this happens, traffic becomes more generic, campaigns generate less consistent leads, and positioning gets trapped in overly vague categories.
The problem is that internally the company knows exactly what it's doing, but the market doesn't read intentions. Google doesn't read intentions. Generative AIs don't read intentions.
They interpret signals: words, pages, categories, content, links, consistency between what you say and what someone is searching for. Then they put you in a box.
Sometimes that box is right. Sometimes it's comfortable, reassuring, and wrong.
It happened to us with a seemingly innocuous word: web agency.
For years it has been a useful definition, because in everyday language many companies use it to indicate any digital partner: website, advertising, SEO, campaigns, content, online marketing.
Except that for Google and for AI, “web agency” tends to mean one thing above all: an agency that makes websites .
And we are not primarily this.
The website can be part of the job. Landing pages can be part of the job. But our real focus is elsewhere: lead generation, customer acquisition, advertising, tracking, data, and market research .
This difference, which was obvious to us, was not necessarily obvious to the algorithms.
Summary
The test: ask AI as if you were a potential customer
We ran a simple test. We asked an AI model:
“Find me a web agency in Bergamo.”
Hangler was not among the first proposals.
Then we asked why .
The answer, artificial politeness aside, was quite clear: for that request the model was looking for agencies more closely associated with website development , e-commerce , SEO and generalist digital communications.
It wasn't an AI error. It was information.
The model was doing what many users do: it took a word, compared it to what it found online, and built a list consistent with the most likely meaning of that word.
At that point we asked the right question:
“If someone were looking for lead generation agencies in Bergamo, who would you suggest?”
In that case Hangler appeared.
The lesson wasn't "we should call ourselves a web agency."
In fact, it was almost the opposite: if in order to be found we have to use a category that doesn't well represent what we sell, we risk attracting the wrong intention .
The point isn't to please AI. It's to understand what AI has understood.
You should not use ChatGPT, Gemini, or Perplexity as oracles.
They're not positioning tribunals. They can make mistakes, invent, mix sources, and reason through questionable associations.
But they are useful as a mirror.
If you ask an AI “ who does X in my area? ” and your company never appears, or appears in the wrong category, you don’t have the absolute truth.
You have a signal.
And that signal needs to be verified with Google, Search Console, Google Ads, Analytics, CRM, and sales conversations.
In our case, the signal was consistent with something we were already seeing in the data: when the communication was closer to the “web agency” category, the traffic was more generic.
As the message became more direct about lead generation , customer acquisition, and market research, intent improved.
Here's the thing: AI didn't tell us what we were supposed to be. It showed us how we could be interpreted.
And in marketing, sometimes, this is the most useful diagnosis.
What we changed
We didn't "redo marketing because AI told us to." We did something more normal and more useful: we removed ambiguity .
We stopped treating “web agency” as the main container and started making more explicit what we really do.
Less generic category, more words that adhere to the real offer: lead generation, contact acquisition campaigns, tracking, conversions, market research, customer analysis, data useful for decision-making.
This doesn't mean a web agency can't handle marketing, advertising, or digital growth . It just means that, in the language of users and algorithms, certain words convey very specific expectations.
When someone searches for “web agency Bergamo,” they often expect someone to build a website.
If you search for “lead generation Bergamo,” “B2B lead generation agency,” “customer acquisition campaigns,” or “market research for companies,” the intention is different.
And that's where we need to be clear.
Why clarity of positioning also changes leads
Clarity isn't just about "telling yourself better." It's about bringing in more coherent people.
When a page uses too many generic words, Google may show it for broad searches, AI may file it in a vague category, and the user may click expecting something different.
The result is not only less qualified traffic, but weaker leads : off-target inquiries, contacts looking for something else, advertising budget spent explaining what should have been clear before the click.
This is a huge issue in lead generation, because the problem isn't generating leads at all. The problem is generating leads with an intention that's compatible with what you're selling.
A page that says “web agency” attracts a different demand than a page that says “B2B lead generation,” “customer acquisition campaigns,” or “market research for companies.”
It's not a linguistic nuance. It's a commercial difference.
If the positioning is vague, the lead tends to be vague too. If the promise is precise,
Instead, the contact already starts from a cleaner ground: he has understood what you do, why you could be useful to him and what kind of problem you can help him solve.
The data: what happened in the countryside
After focusing , Google Ads campaigns began working towards a more aligned intent. The interesting fact isn't that we spent more, because the budget remained essentially the same. The point is that, for the same investment, conversions increased significantly.
Metrics | Before | After | Variation |
Click | 786 | 1,619 | +106.0% |
Conversions | 11 | 41 | +272.7% |
Average CPC | €0.47 | €0.23 | -51.1% |
Total cost | €368.03 | €367.55 | -0.1% |
With virtually the same spend, conversions went from 11 to 41.
This is the most important data, because it shows that the problem was not “advertising more”, but better aligning search intent , ad promise and page content.
When a user searches for something, clicks an ad that says that, and lands on a page that confirms that fact, the system works better. Google understands better. The user understands better. The campaign wastes less .
This doesn't just apply to us.
This applies to any company that uses words that are too big, too vague, or too convenient. "Innovative solutions," "digital partner," "professional services," "360-degree consulting," "quality and expertise" are phrases that often seem safe, but they don't help Google or AI understand why they should recommend you.
They are fog with the logo on them.
How to Know if Your Business Is Google-Friendly and AI-Friendly
The method is simple and can be done in half an hour.
First, ask the AI a general question, like a real customer would. Don't ask "find my company," because that would be cheating. Instead, ask something like:
“Find me companies in [area] that do [generic category].”
Then look at who's leaving and, above all, ask why. The useful part isn't just the list of names, but understanding the criteria used to build it:
“What criteria did you use to choose these companies?”
“What keywords or concepts did you associate with my request?”
“What kind of companies did you exclude?”
This is where the problem often arises.
You may think you're selling “digital consulting,” but AI associates that word with training.
You may call yourself a “creative studio,” but the market is looking for “social media agency.”
Maybe you write “technological systems”, but the customer searches for “photovoltaic for warehouses”.
You may talk about “efficiency solutions,” but no one understands whether you sell software, consulting, machinery, or energy.
Then do the second test : use the word closest to your true value.
In our case, not “web agency” but “lead generation agency” or “market research agency”.
If you don't get noticed with the generic category, but you do with the specific one, you might not have a visibility problem. You might have a category problem. And that's good news, because it can be fixed.
How to Check Your Ranking with Google, Search Console, and Google Ads
AI testing alone isn't enough. It needs to be cross-referenced with Google.
Search for the same words anonymously. Look at the SERPs, ads, and competitors that appear, and check if your page truly speaks the same language as the search.
Then go into Search Console and see which queries you're getting impressions for. Not just clicks: impressions . Impressions tell you where Google is trying to place you.
If queries appear that aren't relevant to your offering, Google may have misunderstood you. If relevant queries aren't appearing, you may not have given them enough signals.
Finally look at Google Ads .
Campaigns are a cruel lens because they combine intent, message, and conversion. If you're getting clicks but few conversions, you might be attracting the wrong people.
If you have few impressions, perhaps the demand is too low or the keyword is too narrow.
If you have conversions but few leads, perhaps the promise is strong but not qualifying enough.
The point isn't to stop at "the campaign works" or "the campaign doesn't work." We need to ask ourselves if Google, AI, and users are understanding the same thing .
The three questions to ask yourself
The first question is:
If an ideal client were to search for us without knowing our name, what words would they use?
Not the ones we use in meetings. Not the more elegant ones. The ones a person with a problem to solve would actually use.
The second is:
Are those words clearly present on the website, service pages, ads, and content?
It's not enough to simply name them once at the bottom of the page. They must be part of the structure: titles, H1s, texts, FAQs, case studies, examples, internal links, and dedicated pages.
The third is:
Is our category the right one or is it just the most convenient?
This is the most annoying question, because many companies define themselves with broad categories so as not to give up anything. The problem is that the more you broaden the definition, the more interchangeable you become.
AI, Google, and customers don't have time to interpret all your complexity. They assign you a position. If you don't give it to them, they'll invent one.
What to learn for SEO, AI visibility, and lead generation
Teaching is not “write for AI.”
The lesson is: write so that a person, Google, and an AI model come to the same conclusion about what you do, who you do it for, and why they should choose you.
This is the new minimum positioning hygiene.
For years, we've been primarily concerned with being found on Google. Now we must also worry about being understood by the systems that summarize, recommend, and compare . But the foundation remains the same: clarity, specificity, consistency.
If you sell lead generation, don't hide behind "360-degree digital marketing." If you do market research, don't just say "strategic consulting."
If you are strong on a specific issue, name that issue.
The market doesn't always reward those who do more things. It often rewards those who can make themselves understood most quickly.
A practical test
Take your homepage and ask yourself: if you removed the logo, would a person understand in ten seconds what you're selling?
Then ask an AI. Then Google it. Then look at the data.
If the four answers don't match, you don't just have a copywriting problem. You have a problem with the alignment between positioning, demand, and distribution.
And it's one of the most costly problems, because it's not immediately visible. It manifests itself in overpriced campaigns, off-target leads, pages that don't convert, content that drives useless traffic, and AI that recommends someone else.
We ran the test starting with a very simple question: "Find a web agency in Bergamo."
The answer wasn't what we expected. That's exactly why it was useful .
FAQ
How do I know if AI has understood what my company is doing?
Ask an AI model to search for companies in your category, without mentioning your brand. Then ask what criteria, keywords, and concepts it used to build its answer. If you don't appear, or appear in the wrong category, you have a signal to verify with Google, Search Console, and real data.
Why can a category that is too general reduce the quality of leads?
Because it attracts different intentions. A broad definition can drive traffic, but not necessarily traffic consistent with what you're selling. In lead generation, accuracy matters more than quantity: if the user searches for one thing and the page promises something else, the lead will be weaker.
Is optimizing for AI different from SEO?
Partly, yes, but the foundations remain similar: clear content, direct experience, solid technical structure, readable text information, internal links, and consistency across the website, ads, and company profiles. The difference is that AI tends to synthesize and categorize, so ambiguity weighs more heavily.
What is the fastest test to do?
Ask the AI for a general search in your category, then a more specific search for your actual service. Compare the results with Google, Search Console, and Google Ads. If the answers don't match, you probably have a ranking or language issue.
Comments