The paperwork is ready. The buyer is missing

Think of an estate agency and a notary's office in any Spanish city. The agency uses AI to write adverts, prepare property details and respond to enquiries. The Spanish notary, a legal professional exercising public authority and charging regulated fees, uses it to organise documents, check charges on properties and prepare draft deeds. The professionals review and sign. The responsibility remains theirs.

Month by month, both offices work faster and better. Preparing a transaction costs less. Errors are spotted sooner. The improvement is real.

But once the paperwork is ready, there is no extra family waiting with a mortgage approved.

A fishing rod, a lake and some fish

A simple image helps explain it. Imagine several people fishing around a lake. The fish are the customers: in our case, people buying or selling a home. The lake is the market, with the fish it already holds. The fishing rod is the tool each person uses.

AI is a better fishing rod. It lets you fish with less effort and in less time. But a better fishing rod does not put more fish in the lake. And if everyone gets the same new rod, the lake still has the same fish and, at the end of the day, each person takes home more or less their usual catch. The difference is that catching it took less time.

We have already explained why greater productivity does not mean higher revenue. Here we take it a step further: what happens when the estate agency opposite and the notary's office in the next street improve too?

The agent does not stock the lake

A sale needs an available home, an affordable price and a bank willing to lend. Interest rates determine the repayments. A flawless advert does not provide the deposit the buyer lacks. An instant draft does not get their loan approved.

The agency has a better fishing rod. The fish depend on what happens outside its office. It can handle more transactions, but it can only complete those that exist.

Everyone gets a new rod

The first agency to adopt a good tool can win customers from its competitors. For a while. Then they copy it. In a sector where AI costs little, an improvement in marketing lasts only as long as it takes the agency next door to buy the same software.

And when a sale moves from one agency to another, the sector still completes the same sale. One gains what another loses. Nobody has caught more fish.

In 2003, economists Chang-Tai Hsieh and Enrico Moretti described something similar among estate agents in the United States: high, rigid commissions attracted more agents, and each sold fewer homes per hour worked. More effort to share the same market. Their study is not about AI, but it illustrates the mechanism: competing for existing buyers does not increase the number of property sales.

Working faster does not mean charging less

The agency charges a percentage of the sale price. Writing the advert in two minutes does not change that percentage. Market custom sets it, not the time spent.

The notary's office makes this even clearer. In Spain, it charges according to a regulated fee schedule under Royal Decree 1426/1989. The rules provide for reductions and allow a discount of up to 10 %, but none of those figures depends on the hours AI saves. A notary who works twice as fast charges the same for the same deed.

In Norway, a notary is not involved in an ordinary property sale. The estate agent, or eiendomsmegler, or a lawyer handles the oppgjør: the settlement process coordinating payment, registration and the discharge of charges. With agreed commissions or fees, the logic is the same.

So the client does not pay less. The buyer does not receive the saving.

The saving changes hands

Where does it go, then? First, to work previously done outside the business. The agency used to commission a marketing firm to produce its adverts, copy and campaigns. Now it does that work in-house with AI. It used to have property details translated for foreign buyers. Now it translates them itself. It entrusted some of the paperwork to an administrative services firm. Now it handles that in the office.

The agency cuts a bill. The supplier loses that income. Not a single extra home has been sold. What has changed is who does the work.

For the agency, that means a higher margin. For the economy, it is not growth: it is money moving from one pocket to another, or disappearing from the marketing firm's revenue.

The vacancy that never opens

Next, inside the business. If the team can prepare the same files with less effort, the next hire is no longer needed. The agency does not dismiss anyone: it simply does not look for the sales agent it had planned to recruit. The notary's office handles, with the same staff, a workload that previously required another clerk.

If the catch does not grow, neither does hiring. Employment does not fall overnight. It stops growing. And that makes no headlines, because a vacancy that is never advertised makes no noise.

The lake can grow a little

There is a counterargument, and it comes from the same sector. A faster service can save a transaction facing tight deadlines. Helping a foreign buyer in their own language brings supply and demand closer together. Better prepared documents prevent delays and sales falling through.

All of that counts. But the tool does not determine the outcome. There still needs to be a suitable home, a willing seller and money to pay for it.

A better rod, the same fish

The agency and the notary's office reflect thousands of small and medium-sized businesses. They work better with AI. They see it in their accounts: fewer supplier bills, fewer new hires, more time. They do not see it in their customer numbers.

Across an entire sector, the result is a more efficient economy that grows no faster. That may be part of the answer to why AI is visible in every office but not in the growth figures.

The productivity gains are real. The extra activity still has to be found.

Sources

Text by Roberto Ortega, prepared with AI editorial assistance. The Norwegian and English editions are editorial adaptations of the Spanish original.