Over the last few years, we spoke with dozens of publishers — care services, communities, commerce apps, content platforms. The businesses looked different. The moment they got stuck looked almost identical.
Whenever we met a publisher about to start selling ads, we kept seeing the same scene. An ad inquiry lands in the inbox. It should be good news — someone wants to pay to advertise on your service. But the person stares at the reply and hesitates. "What price do I quote? Which placement can I sell? Where do I collect the creative, how do I block off the schedule? Who handles the report and the settlement when it's over?"
Traffic, loyal users, even advertisers showing up on their own — and still, starting an ad business was the hard part. This is the story of how our team kept watching that, and slowly defined the real problem.
At first, we thought it was "they can't find advertisers"
Honestly, we were confident about that assumption. If publishers weren't making money from ads, the obvious explanation was that they couldn't find advertisers. So we spent months thinking about distribution, demand, and sales.
We were solving the wrong problem.
Because when we actually talked to them, advertisers were often the ones reaching out first. One care-services publisher told us inquiries came in steadily. The problem was what came next: there was no way to turn them into campaigns.
To run a single ad properly, you need more than you'd think — an ad server to deliver it, inventory to sell, measurement to confirm it's running, and reporting to show the results. Without that bundle, an inquiry simply can't be answered. That's why the reply got stuck.
"So just build it, right?"
We asked the same thing. But building it was the first wall.
Ads are unfamiliar territory for most product teams. Someone from the business side has to sit in early and explain how ads are actually sold, booked, delivered, and reported. At one publisher, just aligning on that took several people a few weeks. And it's expensive: in several conversations, teams described the in-house path as a year-long project with a six-figure cost — before maintenance even began. Outsourcing didn't fully solve it either. The upfront cost was still high, and every market change raised the same question: who maintains this?
Even if you build it, that's not the end
Even at teams that managed to build it, the real work started after.
Quoting advertisers, prepping campaigns, contracts, collecting and uploading creatives, swapping creatives mid-flight, talking after the campaign ends. One team said a single campaign meant 60–100 emails back and forth. In between, creative versions drift, schedules slip, and one small mistake turns into a trust problem with the advertiser. At one high-traffic service, a single person was effectively running ad ops and managing every creative by hand. Another team told us, "the more tools we added, the more complicated ops got."
The longer we watched, the clearer it got: selling ads isn't selling a placement. It's running a system, operations, and sales — all at once.
And even then, the math was uncomfortable. Direct ads needed ops time, sometimes sales time, and constant coordination — but revenue came in unevenly. So teams kept postponing it. Not because demand was missing, but because they weren't set up to catch it.
Everyone lands on the same answer
Facing all of this, most publishers reach the same conclusion: just plug in network ads — AdMob, AdSense. One line of code and you're live.
But that convenience has a cost. For many of the publishers we met, network ads meant low CPMs, limited control over pricing and placement, thin reporting, and very little say over how an ad affected the user experience. They'd love to sell premium direct deals, but they don't have the system to receive them — so they stay on low rates and compromised UX. A lot of them were there not by choice, but because there was no other way.
Here's the problem we defined
Step back, and the real problem comes into focus.
Every service was rebuilding the same thing — an ad-product management screen, a booking flow, reports, a settlement process. Different industries, different sizes, the same problem underneath. Everyone was carving the same wheel, each in a slightly different shape, and paying dearly to do it.
So we flipped the question: instead of re-carving that wheel every time, what if we standardized that layer once, for everyone?
Looking back, the answer feels obvious. At the time, it wasn't. For months, we thought the challenge was demand. It turned out demand was already there. The real challenge was everything that happened after someone said, "We want to buy an ad."
That realization became the starting point for Ad Control — the system publishers needed after demand showed up.
Next time
How we actually started solving it — what we focused on first, what the first version looked like, and the first "oh, this could work" moment. More in the next note.