RivertrafficOfficial
Newbie
- Jan 20, 2021
- 18
- 3
Sometimes campaigns just need more data. Other times waiting only burns money. What's your minimum amount of data before making optimization decisions?
Are You Optimizing Too Early?
- Thread starter RivertrafficOfficial
- Start date
- Nov 7, 2021
- 98
- 24
I usually give it enough data to see a real pattern, not just a few clicks or conversions. There’s no magic number it depends on the budget and offer, but I try not to make decisions based on one bad day. Creatives usually are fastest to optimized, usually after 50 - 100 clicks you can already undarstand if a creative is going to be a successfull.
Wisdomjane
BANNED
- Jun 29, 2026
- 17
- 1
I prefer enough data to spot a real trend, not just to react to short term fluctuations.
ActiveRevenue
Registered Member
- Sep 12, 2024
- 99
- 21
i'd be careful with universal thresholds like x clicks or x impressions.
30 clicks on a cheap lead offer and 30 clicks on a high-payout funnel tell you very different amounts.
for me the better question is whether the data is enough to make a decision at the level you're optimizing — creative, source, placement, geo, etc. sometimes the overall campaign looks mediocre while one source already has enough data to clearly keep or kill.
30 clicks on a cheap lead offer and 30 clicks on a high-payout funnel tell you very different amounts.
for me the better question is whether the data is enough to make a decision at the level you're optimizing — creative, source, placement, geo, etc. sometimes the overall campaign looks mediocre while one source already has enough data to clearly keep or kill.
ActiveRevenue
Registered Member
- Sep 12, 2024
- 99
- 21
i'd be careful with universal thresholds like x clicks or x impressions. 30 clicks on a cheap lead offer and 30 clicks on a high-payout funnel tell you very different amounts. i'd base it more on whether you have enough data to make the decision at the level you're optimizing — creative, source, placement, geo, etc. sometimes the campaign-level data is still noisy while one placement has already spent enough to make the decision pretty obvious.