In the last note, we explained why publishers needed a selling layer of their own. This is the note where that idea met reality.
The customer who'd already tried
One of our first customers was Bukjeokbukjeok, a Korean app publisher. They weren't waiting for a tool. They'd already tried to build an ad system themselves. Like a lot of teams we'd talked to, they could feel the demand. Advertisers were asking, valuable inventory was sitting right there, and every campaign they couldn't run felt like revenue slipping away. So they did the reasonable thing and started building the machine to capture it.
It didn't go the way they hoped. The team spent real engineering time on it and ran into the parts nobody warns you about, like booking, scheduling, creative review, reporting, and billing. Eventually, they paused the effort. Not because the demand had dried up, but because building and maintaining the system to serve it looked more expensive than the revenue it could unlock. The demand was real. The economics of building for it weren't.
With Ad Control, they didn't restart that build. They switched it on. In minutes, the core setup was running, and direct deals became something they could actually operate, not just talk about. The layer they'd shelved as too costly to build was suddenly just… on.
Then the strange part
We'd braced for the dramatic version of success. A launch, a spike, a sales push, or at least someone calling at midnight because something broke. Founders are conditioned to expect a moment. The graph that turns upward, the customer who won't stop thanking you. That's not what happened.
We'd spent months talking about the problem, in interviews, in pitch decks, in our own heads. And then one day, there was simply… nothing to do.
The advertiser booked. The publisher approved. The campaign ran. Nobody called us.
No big launch, no heroic sales push. Just a publisher logging in, approving campaigns, and not needing us. It was almost anticlimactic. But that quiet was the whole point. That was the moment the sentence from note #1, "That's why the reply got stuck," stopped being a problem statement and became a before-and-after. The thing we'd only ever heard about in interviews was now running, on its own, on a real publisher's service.
What actually changed
The clearest way to see it is the before and after.
Before, a single advertiser inquiry set off a chain of manual steps, almost all of them landing on a person:
- find the rate card
- check the schedule for open inventory
- trade emails back and forth on price and dates
- collect the creative
- hand it to the dev team to launch
- send a report when it was over
Every one of those steps was a place the deal could stall, and every one needed someone paying attention. For a small team, that overhead was the reason "we should sell our own ads" stayed a someday project.
After, most of that chain moved into the product itself. The advertiser could complete the flow on the publisher's own branded platform:
- see available products
- choose a product and set a budget
- upload a creative
- submit the campaign
And the publisher? They approved. That was the job. The campaign moved forward without anyone holding every step together by hand. Pricing, availability, handoffs, and follow-up. All the coordination work that used to live in email was now handled by the platform.
We also started hearing about a second-order effect we hadn't fully expected. When booking was simple and reporting was clear, advertisers seemed more willing to come back. The friction that used to make direct deals feel "not worth it," for both sides, was mostly gone.
That was the shift we were really after. Not finding publishers more advertisers, but helping them capture the advertisers they already had, without us sitting in the middle of every deal.
Where it stands
Bukjeokbukjeok wasn't the last. Over time, more than 100 services switched it on, from large platforms and one-person studios to news apps and niche communities, across very different industries. Different sizes. Different categories. The same shape every time.
The advertiser booked. The publisher approved. The system handled the rest.
That's the part we're proudest of. The publisher runs their own ad business, and we're not standing in the middle of it. They take the advertisers, review, and approve, and the demand that used to leak away now lands somewhere. Ad Control isn't a middleman sitting in the center of every deal. It's the infrastructure quietly powering that business in the background.
Next time
We proved this in one market, with publishers who mostly shared a language and a context. The obvious question, the one we kept circling back to, was whether the same pattern would hold anywhere else.