Your customer is trying to buy. Stop making them ask.

Somewhere this week, a contractor needed a price on something you sell.

He did not call. He typed the product into his phone between two job sites, looked for a number, did not find one, and bought from whoever had one posted. He was not comparison shopping. He was trying to give somebody money and could not work out how much to send.

It could as easily be a designer pricing a specification at eleven at night, or a homeowner halfway through a project. The behavior is the same.

That is happening to European manufacturers and to American dealers at the same time, for opposite-looking reasons that turn out to be the same reason.

Two ways of making a customer ask

A European manufacturer builds a good product, sets up in the United States, and puts up a website that says contact us to discuss your requirements. In Madrid or Milan or Antwerp that is correct and even courteous. The relationship comes first, a person answers, a conversation begins, and everyone knows how this works.

In America it reads as a closed door. The buyer wanted a specification sheet and an indicative price. He got a form. He will not fill in the form, because filling in the form means a salesperson calls him for a week, and he has eleven other things to price before Thursday.

An American dealer has the opposite problem and the same outcome. He has the inventory, he knows the price, he has been doing this for forty years. None of it is visible. His website has a phone number, an address, and a photograph of the yard taken in 2011. His customer, who genuinely likes him, still checks somewhere else first, because somewhere else loads in two seconds at eight o'clock at night.

Neither of these is a marketing problem. Both are the first place a commercial operation leaks, and neither shows up in a sales report. Nobody logs the order they did not place.

Two markets, two ways of buying

The gap underneath both is cultural, and it is not that one side does it better. In Europe, commerce runs on the relationship. The buyer calls, meets, emails, builds trust between people first, and lets the information follow. In America, the buyer serves himself first and builds the relationship second. Neither instinct is wrong. They are different, and each is valuable in its own market. A company selling into the United States has to sell the American way while it is there, not the way that works at home. The American dealer's problem is not culture, it is visibility, but it ends in the same place: a buyer who serves himself and cannot find what you already know.

And into that gap, everyone is selling AI

Both of these companies are being pitched artificial intelligence, hard. Agents, copilots, forecasting, dynamic pricing, automated quoting, intelligent search. Some of it is real. Very little of it is finished.

Look at what happened to Starbucks, which is not a company short of money or engineers. It rolled out an AI inventory counting system across North America and retired it about nine months later after it could not count reliably at scale. That is not a story about AI failing. It is a story about that application of it not being ready, at a company with more resources than anyone reading this.

Now put the same decision in front of a dealer doing sixty million dollars a year, or a manufacturer three years into an American entry that is not producing. Spend the budget on an AI initiative whose return nobody can yet demonstrate, or spend a fraction of it making your inventory and your prices visible to the person already trying to buy from you.

One of those has an unproven return. The other has a customer standing at the door with a wallet.

The catalog comes before the copilot.

The order matters more than the technology

There is a sequence to this and most people are running it backwards.

The catalog comes first. Not a beautiful one. A working, interactive one: what you sell, in the colors, sizes, and accessories it actually comes in, what it costs or at least what it lists at, and whether you have it. A manufacturer nervous about publishing prices can publish a suggested retail price and a note that project pricing is quoted. That gives nothing away. It lets a distributor or dealer see their margin at a glance, and it tells a buyer he is in the right place, which is the only job that page has. For the American dealer the work is the same page from the other direction: the forty years of stock and pricing he already carries, finally made visible, so the customer who likes him can see it before he checks somewhere else.

Then the specification. Downloadable, current, in U.S. sizes and the certifications American specification actually requires, and findable without an account. Designers and specifiers cannot put your product into a drawing they cannot download at eleven at night.

Then availability, and the logistics behind it. For stock items that increasingly means days, not weeks. For project and specification work it can be far longer, and that is fine, as long as you say so. Silence on lead time reads as cannot deliver.

And get the language right, in both directions. A Spanish site and an English site is the baseline, but a machine translation is not a U.S. website. Have a native English speaker review it, not only for the words but for context and tone, because a phrase that sells in one market can quietly undercut you in another.

Only then is there any point talking about intelligence layered on top, because a system that reasons over a catalog you do not have is reasoning over nothing.

The person still matters more than the machine

None of this replaces the person. The U.S. market runs on trust and on someone who shows up, across distribution, specification, and the big-box and co-op channels. What it changes is what that person does with their day.

The one who works best knows the product cold, knows the website just as well, and uses the tools on it to serve the customer in the moment, pulling the spec, the availability, the quote while the conversation is live. On top of a site that already answers the basics, that person spends their time where people win: the relationship, the negotiation, and the deals the site surfaced but cannot close. Staffing only for how business is done at home, and not for how America buys, is the quiet miss. It is a different commercial culture, not a worse one, and the United States needs someone who knows its version.

Where AI does earn its place, right now

This is not an argument against the technology. It is an argument about what you build first.

Two applications are ready and both are worth doing once the fundamentals are in place. A chatbot that answers real questions from your own documentation, so a buyer at nine at night gets an answer instead of a form, or a search assistant that pulls the right specification and can prepare a bid off a project price rather than the listed one. And intake: capturing what someone wants, routing it to the right person, and making sure nothing sits in an inbox over a weekend. Both do the same thing. They stop the customer having to ask twice.

What is not ready is anything where a mistake is expensive and nobody is checking. Pricing that moves on its own. Inventory that reports itself. Forecasting that goes straight into a purchase order. The technology will get there, and for this industry it is worth letting it get there first. Building the fundamentals now and letting the technology catch up is not falling behind. It is spending in the order that actually returns.

The uncomfortable part

Most owners already know this. What stops them is that a digital catalog is unglamorous work. It is a spreadsheet, an argument about SKUs, someone photographing products against a wall, and three months of nobody being impressed.

An AI initiative is a press release. It is a slide at the board meeting. It sounds like the future rather than like admin.

But the contractor between two job sites does not care which one you chose. He cares whether he found a price. And the manufacturer whose American numbers are not what the plan said, or the dealer watching his customer buy somewhere else at night, will not find the answer in a technology budget. It is usually further back, in the ordinary question of whether a stranger can work out how to buy from you without speaking to anybody.

Fix that first. The rest gets easier, including the AI.