Why the future of marketing is offline

Every quarter it costs a little more to reach people who believe a little less of what they see. You do not need a study to know it. It is in your cost per lead.

I wanted to understand why, so I went looking for the numbers. Then I went outside.

Start with supply. Graphite checked 55,400 newly published articles with three different detectors, and one in two had been written by AI. When something costs nothing to produce, there is no limit to how much of it gets produced. Your customer's attention did not double to make room for it.

Then look at what that does to demand. Pew Research followed 69,000 real Google searches. When an AI summary sat at the top of the page, 8% of searches ended in a click. Without one, 15%. You can publish the best page in your category and still never be visited.

So everyone ends up bidding for the same shrinking click. Google's average cost per click passed five dollars in 2025, and roughly a third of internet users now block ads altogether. If you run a B2C or D2C brand, this is the line in your P&L that keeps drifting the wrong way, and nothing on a screen is going to reverse it.

In Spain, spending on door drops fell 12% last year. Brands did not leave because paper stopped working. They left because nobody could prove that it did.

That was a marketing director describing her last campaign to me. And yet paper still gets something a screen has lost. JICMAIL measures mail in the UK the way television is measured, and this is what one piece of paper gets once it is inside a home.

Those are the highest readings in five years, in the same years that digital attention collapsed. Put the two together and the picture is simple. Attention has moved to the quietest channel. The only thing missing was data.

Carto runs two services today, hand to hand distribution and door drops, and our first seven campaigns were a mix of both: a festival in Cádiz, a museum just off the Ramblas, a studio in Poblenou. What makes them different is what comes back.

Every crew is tracked by GPS. Every zone has photo proof. Every flyer carries an offer and a code, so we know which street answered and at what hour. Then we match each route against Spain's cadastral record, building by building, so we can tell you how many mailboxes were actually reached. Not estimated. Counted. I have not seen anyone else do this.

The trail does not stop at the door either. A client installs one line on their website, the same way they would install Google Analytics, and our pixel follows each response through to the booking, the sign-up or the sale. No cookies, no personal data. For the first time, an offline campaign is trackable from the hand or the letterbox all the way to the purchase.

Across those seven campaigns, measured response has been between 3.3% and 4.8%, and one of them was still receiving responses weeks after the last flyer went out. That is data no model can generate. Someone has to carry it back from a real street.

They did their job. Routes, municipal permits, crews, printing, measurement: every part of the operation now runs on the platform, and it runs the same way in any city. We have mapped the ten largest cities in Spain.

Spain is the start. The technology is built and ready to scale, and we will be taking it to the rest of the EU, the UK and the US very soon.

This is also why AI makes Carto stronger instead of weaker. AI will keep making digital content cheaper. A flyer in a real hand cannot be generated. It has to be taken there, and that cost is exactly what makes it worth something.

And the flyer is only the first thing we carry. What we are really building is a measured way to put something in front of a real person in a real place. Sidewalk robots have already made more than ten million deliveries, and humanoids are next. We are designing Carto for hybrid crews: robots carrying volume along planned routes, people doing what people do better, reading a street and handing something over with a smile. In a few years the thing being delivered might be digital, or something none of us has thought of yet. The platform that tracks our crews today is the one those robots will plug into.