How do you decide whether to scale, fix, wait or stop?
Decide from data, not feelings. If results beat your target and the budget isn't used up, scale. If there's a clear sign that the creative or funnel is stuck, fix it. If the system is still learning or you don't have much data yet, wait. And if results stay poor with no reason to keep waiting, stop. The rule: change one thing at a time, then wait 48 to 72 hours.
- 4 options
- Scale, Fix, Wait, Stop
- Wait for results
- 48 to 72 hours
- Change budget by no more than
- 20 to 30 percent at a time
What M9 solves
- You decide too early, while the system is still learning, and end up stopping something that was about to work.
- You change several variables at once and can't tell what made the difference.
- You leave ads that aren't paying off running for too long and waste the budget.
What you'll do in this module
- 1
Observe
Wait for the system to learn and for enough data to build up before you decide.
- 2
Decide
Choose whether to scale, swap the creative, wait or stop, based on the evidence you have.
- 3
Fix one thing at a time
Change one variable at a time, so you know what caused the change.
- 4
Read the results, then loop back
Go back to M6 when the creative needs to change, and start a new cycle.
The M9 prompt
This module's prompt is part of the Skill File that students receive when they finish the course, because it only works properly together with the output of the earlier modules.
Right now, 3 modules that work on their own are free to use: M1 Business analysis, M3 Competitor analysis and M10 Tracking check
FAQ
How many days should I wait before judging whether an ad is good?
Wait 7 to 14 days. A newly launched ad hasn't found the right people yet, so results in the first 1 to 3 days swing a lot and costs are usually higher than normal. Watch the trend for at least 7 days, then make the real call at 14.
How big a budget change makes the system start learning again?
A sudden jump of more than roughly 20 to 30 percent usually sends the system back into a new learning phase, and so does changing the audience, the optimization goal or the creative.
At what frequency should I change the creative?
If it goes above roughly 3 to 4, the same people are seeing your ad very often. Together with a falling click-through rate and rising costs, that's the signal to add or swap in new pieces.
My ads are doing well. Should I raise the budget a lot at once?
No, because a sudden jump usually sends the system back into learning and results drop. A safer approach is to raise it by 20 to 30 percent at a time, then wait 48 to 72 hours before the next increase.
Why shouldn't I switch ads off and on every other day?
Because every time you switch an ad off and on again, the system starts collecting data from scratch, so no cycle ever builds enough data to work at full strength. The result is higher costs and results you can't read at all.