Ask How many views is a sufficient sample size to determine the conversion rate of PPV ads?

[Brajet]

Thought this was a straight forward statistics question but can't find a straight answer on the internet.

So assuming 2% conversion rate is the actual rate, according to this I would need 3012 clicks to have 95% confidence with margin of error at 0.5%.

Given the cost per view for my new campaign on is running at >£1.50, is the only way to find out burning through £4.5k first?

Context: Going into a new product category as a new small business, with no prior selling experience or data. I don't feel confident scaling up if I don't have a fairly good idea on CTA and AOVs. If this is the reality, I'm not sure how a small budget can work unless it's a "pay and pray".

 

[Anthony Joshua]

From my experience, the number of views you need really depends on how small your expected conversion rate is. If you're expecting something tiny, like 0.5% or 1%, you're going to need tens of thousands of views just to get a clear idea. For higher conversion rates, say 2–3%, 5,000; 10,000 views can start giving you some meaningful data. I usually run a small batch first to see how things go, then scale up once it looks predictable.

 

[Felix😎]

For me, it's about having enough views to see real patterns. Around 1,000 views is usually a safe number to start analyzing, but if your conversions are rare, you'll need more data—maybe a few thousand views. The bigger the sample, the better your results will reflect reality. Just make sure you're not drawing conclusions too early if the numbers aren't stable yet.