The myth of 50 conversion, exploration & exploration, scale method

Put On Your Mask
J

Jack811

Active Member
Lately I got this thinking for SRN:

Upper-funnel events (trials) hit 50/week easily. However, I never see that the campaign become more stable, or CPA got better after hitting the number.
Purchases cost more and take longer, I barely got 10 in a week. Some campaign become less volatile after 10 & perform well, some doesn't.

So my says is 50 is just a ideal number. If the campaign do it well after 10, you don't need to increase the budget for 50.

Then the case is if the duplicated best campaign not doing well after being created, what would you do to scale the budget in your experience?

Also, Is there any real benefit to a big spend burst?
Say you spend $100k to push as many conversions as possible. Which of these actually happens?
A) The algorithm favors you. High spend earns you better delivery, and Meta starts finding higher-LTV users for you ( with higher CPI, CPA)
B) You're just buying data. More conversions give you enough volume to see who your top-LTV users are, and then you create the edge: value-based bidding, pLTV signals, lookalikes from top payers.
 
You don't necessarily need 50 but according to my exp, it's good for stability to be close to that number. I have seen campaigns with less fluctuations on the CPA side when they produce at least 30 events per week.

Regarding your question, I don't think an initial $100k bump will help to improve CPA or LTV if you are just focusing on events. When you run a CPA campaign, algorithms fully focus on bringing as many events as possible without looking at the value so if you want to improve value, you either play with signals on CPA setups o directly move to ROAS campaigns
 
and for campaigns that aren’t performing well in Week 1 and aren’t driving enough conversions, would you recommend letting them run until they reach 30+ conversions before making any changes, or testing new creatives right away?
 
ROAS can perform well if you send different values consistently so if you have a free trial, it's good that you created a proxy value for this event too.
Regarding the second question, if after 1 week the CPA is really far from my target, I don't wait for 30 conversions because the problem is visible before reaching that point: it's either on the creatives (look at the hook rate, hold rate, CTR) or on the install rate (App Store listing not converting well)
 
You must log in or register to reply here.
Back
Top