Every few weeks someone asks me how I get customers. The honest answer is that I have almost never marketed a product directly. I have marketed the gap between two products that already have customers.

That distinction is the whole thing, so let me be concrete about it.

The cheapest distribution is someone else's audience

Before any of the SEO mechanics below, there is a better move, and it is the one I would actually lead with.

Find someone who already owns the audience you want, and build the product with them.

I do this with course creators. A course creator has students, and they charge a one time fee. That is their whole problem: the revenue stops the moment the course is sold. So I go to them and propose we build a SaaS together, aimed at the thing their course teaches. I build it. They put it in front of their students.

A meaningful share of those students convert. In my experience it lands in the double digits, somewhere around twenty to thirty percent. And I spent nothing to acquire any of them.

FastLien came from exactly this. A course creator with a course and a book on US tax lien investing. I had no idea what a tax lien was. He had the audience and the domain knowledge, I had the ability to build the thing his students needed after the course ended. It has been making good money for both of us since.

Four things make this work better than it has any right to:

It creates new money rather than splitting existing money. He was not earning recurring revenue before. I did not have an audience before. Neither of us gave anything up. That is what makes the conversation easy.

Competition is close to zero. You plant a flag in someone's audience. It is unlikely that a second person walks up to the same creator and makes the same offer. Compare that to competing for a keyword.

The power dynamic is balanced. They depend on you for the product, you depend on them for distribution. Neither side can squeeze the other, so the deal stays fair without anyone policing it.

You pay nothing for traffic. US ad costs right now are brutal. In this model there is no ad account, because the creator is sending you the customers.

One hard constraint: do this B2B. B2C is harder, you need far more volume and the price points have to be lower. I would go B2B every time.

Ask what your market uses before and after you

The second question I ask, once a product exists, is not "who else is in my category". It is what does my target market use immediately before and immediately after using my product.

That question points straight at your integrations, and your integrations point straight at your distribution.

REreferrals is for US real estate agents. Agents use a CRM after they get a referral, so we integrate with the CRMs they already have: Real Geeks, Brivity, Follow Up Boss. We are not competing with those tools, we are feeding them.

The same logic scales up past individual tools. Associations and system integrators serve entire groups of the buyers you want. I have a product that integrates with a company serving franchises. If they send you one franchise, they send you all of them. One sale, many customers through the door.

That is the highest leverage version of the seam idea. Instead of standing between two products, you stand next to the one organisation that already aggregates your entire market.

Why I skip the marketplaces

The naive version of this is marketplaces. You built an e-commerce tool, so you list in the Shopify app store. Their customers are your customers, and they want more features for their users. It works, and I do not use it.

Two reasons. It is easy for everyone to get listed, so you are immediately standing in a crowd. And the house makes the rules. You are building a business on ground you do not own, where the terms can change and you have no say.

So instead of asking which platform has my customers, I ask a different question: who influences my target market?

Usually it is a person, not a platform. A podcaster, a YouTuber, someone who runs a course, someone with a newsletter or a community. Their audience is already qualified in a way that a marketplace listing never is, because it arrived through trust.

And here is the part that makes it work rather than just being networking: you are not asking them for a favour. You are unlocking recurring revenue for them. A one-off affiliate payout is a transaction. A recurring share of every customer they send, for as long as that customer stays, is an income stream. That changes the conversation completely, because now they have a reason to keep mentioning you next month.

The honest numbers on this

I have run both versions, so let me give you the results rather than the theory.

Looking at signup attribution on one of my products:

• A large media and education partner in the space sent over 1,100 signups. Two of them became paying customers.
• A course creator sent about 200 signups, and ten converted. Roughly five percent.
• Two community and Facebook group partnerships sent about 350 signups between them, converting three.
• Plain organic search sent 290 signups, and twenty three converted. Nearly eight percent, the best of the lot.

Read that carefully, because it is not the result I expected either. The audience partnerships produced by far the most volume. Search produced the most customers.

The difference is intent. Someone who clicks a link in a newsletter is curious. Someone who typed the problem into Google is stuck. Curiosity signs up and disappears. Stuck pays.

That does not mean audience partnerships are worthless. The course creator converted at five percent, which is a genuinely good channel, and the pattern there is instructive: the tighter the audience and the more specific their reason for being there, the better it converts. Big and broad performed worst. Small and specific performed best.

So the refinement I would make to my own advice is this. Partner with people who own audiences, yes. But qualify the audience the same way you would qualify a keyword. A hundred people with the exact problem beat a thousand people who find the topic interesting.

The market is not the platform, it is the seam

FatturaExpress exists because Stripe does not do Italian electronic invoicing, and Fatture in Cloud does not do Stripe. Both of those companies are much bigger than me. Both have users who are stuck in exactly that gap. Neither of them wants to build the wire between them, because for each of them it is a niche edge case owned by the other one.

So I built the wire.

I did not have to explain what electronic invoicing is, or convince anyone they needed it, or compete with Fatture in Cloud. The demand already existed, fully formed, sitting between two products that people were already paying for. My job was to be findable at the exact moment someone realised the two things they use do not talk to each other.

This is what I mean by integration marketing. Not "we have an integrations page". The integration IS the product, and the same asset that delivers the value is the asset that gets found.

Where the traffic actually comes from

I looked at my own Search Console before writing this, because I did not want to give you theory.

The pattern that shows up across products is that people search for the seam. Not for a category, not for a solution, for the specific pair of tools they are already holding. Queries hitting my sites right now include things like "callrail follow up boss integrations", "realscout follow up boss integration", "brivity integrations", "airtable hubspot integration".

Nobody searching that is browsing. They have two tools open in two tabs and a problem between them.

Almost nobody writes those pages, because each one individually looks too small to bother with. That is precisely why they work. The competition is not other companies, it is the fact that the page does not exist.

Free tools outperform blog posts at the top of the funnel

The single best acquisition asset I have built recently is not a piece of content. It is a lookup tool.

On FatturaExpress there is a small utility that finds a company's SDI code from its VAT number. It takes five seconds and it is free. It currently ranks around position 3 to 4 for the searches people actually type, and it pulls thousands of impressions a quarter.

Here is why it beats a blog post. A blog post about electronic invoicing attracts people researching electronic invoicing. The lookup tool attracts people who are, at that exact second, trying to send an invoice and blocked. That is not a reader. That is a customer with a problem in their hands.

If you can find the ten second job that sits immediately before the thing you sell, build that job as a free tool. It will outperform anything you write about the topic.

Programmatic pages, where the seam repeats

Once you find a seam that repeats along some dimension, generate the pages.

FastLien has a page per US state, because tax lien rules are set at state level and every investor searches by their state. Those state pages are consistently the highest-click pages on the site. FastLand has comparison pages for pairs of homebuilders, because that is how buyers actually think, one builder against another. Several of those rank on the first page.

The rule I use is simple: the page must be genuinely useful on its own, and the dimension has to be something people really search along. Geography and versus-comparisons are the two that keep working for me. Generating ten thousand pages along a dimension nobody searches gets you nothing, slowly.

Now the part that is usually left out

All of the above is about getting found. None of it is about getting paid, and the gap between those two is where most of this advice quietly fails.

Two examples from my own data, both embarrassing.

Treendly's single highest-traffic page has around forty thousand impressions a quarter. It ranks for an internet slang term that has nothing whatsoever to do with trend forecasting. It converts at a click-through rate of well under one percent, and the people who do click are not looking for a product. That page is my biggest source of impressions and my smallest source of customers.

LeadHall has an integrations page sitting at position 1.9 in Google. Second result in the world. It gets zero clicks.

Ranking is not the goal. Being found by someone who is not your buyer is the same as not being found, except it feels like progress, which makes it worse. I wrote a while back about giving growth experiments enough volume to be evaluable, and this is the mirror image of that: volume that cannot convert is not an experiment, it is a decoration.

So the test I apply now, before building any of this, is one question. Is the person who searches this the same person who pays? If the honest answer is no, the page can rank first in the world and it will not matter.

What I would tell someone starting today

Do not look for an untapped market. There is almost certainly not one, and if there is, you probably cannot afford to create demand for it.

Look instead at two tools that a lot of people already pay for, that do not talk to each other properly. Or find the one person whose audience is small enough to be specific and who would rather have an income stream than a thank you. Build the wire. Make the wire findable by naming both tools plainly. Then build the free five second version of the job that happens right before your product is needed.

You are not creating demand. You are standing in a place where demand already passes through, and where nobody has bothered to put up a sign.

The longer version

I gave a talk at PugliaTechs in Bari about the broader version of this: building products without permission, without a team, and without money. It is fourteen minutes and it covers where the products come from before any of the above applies.

If the player does not load, it is here on YouTube.