Three clocks, not one
The confusion usually starts with the word "work." Work like what? Like you can see where your money went? Like the cost per conversion has stabilized? Like you're confident enough in the economics to double the budget?
Those are three different things. They happen on three different timelines.
Clock one: direction
The first clock gives you direction. Within the first week or two, you'll know a few things: whether an audience exists that responds to your ads at all, roughly what it costs to get someone through the door, and which creative direction is generating any engagement.
That's a pulse, not a result. Don't reallocate your life savings off of week one. What you have is a starting point and a direction. That's still valuable. Just don't confuse it for validation.
Clock two: stable economics
The second clock is when the numbers start to mean something. The platform has to climb out of its learning phase, which requires roughly 50 conversions per ad set per week. You then need at least two cohorts to compare before you can trust that a good number isn't a fluke.
That usually puts you in weeks four through eight. This is the first moment when the cost per event that actually matters — a trial start, a paid conversion, a qualified opportunity — carries real information.
Clock three: scalable stability
The third clock is the one that justifies scaling. That means you know your LTV/CAC well enough to push budget aggressively and expect the math to hold. For a subscription app, you literally have to wait for cohorts to retain. There's no dashboard shortcut. You need paying users to stay paying long enough to know what they're worth.
That stability shows up over months. Sometimes over a full quarter. The most avoidable damage in paid media comes from reading the wrong clock. Judging week two by clock-three standards, or calling a campaign dead in week six because it doesn't look like clock-one speed.
What to do with this
Match your expectations to the clock you're actually on. In week two, the question is: do we have direction? In week six, the question is: is the cost stabilizing on the event that matters? In month four, the question is: does the LTV math hold well enough to scale?
Different questions. Different timeframes. Same campaign.