Problem with Small Sample Size Results

Small sample sizes can be worse than no sample at all…

Too many advertisers overreact to small sample size results. A great start or a bad start? A good day or a bad day? A campaign that ran for two weeks on a $20 per day budget.

“This doesn’t work!” they tell me. It’s always possible that it doesn’t work for you, but be careful about making universal assumptions based on limited data. When you do, you close yourself off from options.

I appreciate that generating large sample sizes can be a challenge. It can be expensive. But YOU need to appreciate that, too. Understand that it’s an obstacle to evaluation and predictability.

Instead of saying “This doesn’t work”, say, “This wasn’t effective for me this time, and some of the potential reasons and causes are…” One of those causes is likely a small sample size.

Do not shut the door on trying that thing again. Increase your sample sizes when possible. Spend more when you can. Most importantly, have an open mind.

Be a student of the data, and recognize that when you have limited data, you have limited teaching experiences to learn from.

When you understand this, you’ll be a much better advertiser as a result.

Category: