You already paid to get them. Then you let them walk away.

01Why this is the cheapest money in the account

Think about what you actually paid for on a cold click. You paid to stop a stranger. You paid to make them curious. You paid for the click. You paid to land them on your page. That is the hard, expensive part, and you have already done it.

Then most of them leave. And most businesses simply let them go, like the whole thing never happened.

Reaching that same person again costs far less, because you are not buying discovery any more. You are buying a reminder for somebody who already knows your name. The maths is not even close.

The expensive part is already paid for. Then most owners throw it away.

02The maths that makes it obvious

Round numbers. Say you spend 2,000 and it brings 1,000 people to your page at 2 a click. 3 percent ask about you. That is 30 enquiries, so about 67 each.

Now look at the other 970. You paid for every single one of them. They came, they looked, they left. For most businesses that is the end of the story.

Reaching all 970 again costs much less. Say half a rupee or half a dollar a click, so about 485 to reach the lot. If just 1 percent of them come back and ask about you, that is roughly 10 more enquiries for 485.

Cheaper per enquiry, from an audience you had already bought and were about to waste. And I used a deliberately gloomy 1 percent. In real life the gap is usually wider, because a person who has already read your page is not a stranger any more.

Put your own numbers in. Only 2 matter. Your cold cost per click, and the share of visitors who do nothing. That second number is almost always above 90 percent, which is exactly why this works and exactly why so few people bother.

03The shadow funnel

Here is the version people do not run. Do not retarget on 1 platform. Put the pixel for every platform you might ever advertise on, on your pages. A pixel is just a small piece of code that lets a platform remember who visited you. Then run retargeting across all of them at once, for anyone who came.

Somebody reads your page on a Tuesday. Then they see you in their Instagram feed. Then before a video they were watching. Then on a news site they read. Nothing you did was expensive, and the effect on a person who was already interested is out of all proportion to the spend.

This is also where 2026 changed the game a little. People no longer check you only on search. They ask a chatbot who is good for this job, they read your page, then they go quiet for weeks. I wrote about that in how buyers now ask an AI who to hire. The shadow funnel is what keeps you in front of that person while they are still deciding.

04What to actually show them

Not the same ad again. They saw it. They clicked it. It did its job. Showing it again is just repeating yourself louder.

A visitor who did not act had a reason. Usually 1 of 3. They were not sure it works. They were not sure it works for somebody like them. Or the timing was wrong. Retargeting is where you answer those, 1 ad each.

That last one matters more in high ticket service work than people expect. A stranger reading about Meta ads at midnight is not booking a big engagement that week. Being useful now and still being there later is the whole play.

This is where the way we work now helps a lot. The AiOS command center and the Telegram assistants mean the useful thing you hand a cold visitor, the guide, the article, the answer, gets made and sent without you sitting down to build it. Same for STORM, the steady stream of posts and articles. The warm pool keeps seeing something new from you instead of the same ad for the 11th time.

05The other cheap one: your own name

While we are on cheap and boring, run search ads on your own brand terms. On your own name. It is one of the most reliably profitable things available, and it gets skipped because it feels silly to pay for traffic you would have got free.

Sometimes you would have got it free. Sometimes a competitor is bidding on your name and taking a person who was specifically looking for you. It is cheap because nobody else scores well on your own name, and losing a buyer who typed your name is a genuinely stupid way to lose one.

06The 2 mistakes, and where this fits

First mistake. Retargeting everybody the same way. Somebody who read 1 paragraph and somebody who read your pricing page twice are not the same person. The second one deserves a different ad.

Second mistake. Running it forever. There is a point where you become the thing following somebody around the internet, and it stops persuading and starts annoying. Cap the window. If they have not moved in a month, they are not moving because of an 11th showing.

And be honest about what retargeting is. It is not a growth plan on its own. It only reaches people your cold campaigns already found, so it grows exactly as far as the top of your funnel does and no further.

What it is, is the cheapest win available, and the first thing to switch on before you start spending harder on the cold side.

07Questions, answered