What has to be true for AI to recommend a brand?

Translated from the original Spanish article.

Two things are changing in how customers discover and choose a brand. And if you understand what has to be true for AI to recommend one company over another, you can genuinely influence that recommendation.

I'll explain with real examples. First, the two things that have changed.

One is how specific queries have become. This is easy to see if you think in categories.

When someone used to look for a supplier on Google, they searched for the service in general terms and, from there, had to sift through all the options.

They would add one or two filtering criteria to the search. A very common one is geographic; another, a specific technical requirement.

“industrial security door manufacturer in Spain” / “security door manufacturer for airports”.

The customer had to assess, among the results in that general category, which suppliers best fit their particular case.

The second thing that has changed, hand in hand with the first, is how far along in the buying decision a customer is by the time they finally visit a company’s website.

Searches today are not only very specific; they are consultations.

We’ve gone from searching for a “financial consultancy in Madrid” to “I have a mid-sized company in the professional services sector in Madrid. Our revenue is roughly 20-25M a year. Over the last three years, we’ve had an EBITDA of 5M. I’m looking for financing for an internationalization project, because we’ve been exploring two new markets and, now that we’ve validated their potential, we need muscle to grow in them. I want to understand which financing options might be most attractive, and what kind of financial consultancy could best help me”.

By the time they reach the supplier’s website, they have a much clearer idea of the solution they need for their problem.

So, back to the beginning: what has to be true for AI to recommend a brand?

Well, the first thing, as obvious as it is forgotten, is that the AI knows the brand offers a specific service for the case at hand. If our brand offers industrial doors for airports, with a particular locking mechanism, but doesn’t explain that on its website, the AI will have a very hard time knowing it.

This means that what we call transactional pages (the service or product pages) are essential for discovery. And it pays to expand their structure around more specific, particular cases.

The second thing that has to be true is that the AI perceives the supplier as an authority, an entity worth recommending.

That rests on the experience it can demonstrate in similar cases, its knowledge of the subject, external recognition, and the profile of its professionals.

Both things that have to be true, and the factors behind them, are in our hands, if we know how to work on them.