Variance

Do you guys know what variance is and why it’s important in marketing?

Think of it like this. If you flip a coin 100 times and 55 out of 100 were heads (55%) then flip the coin again 10 more times and 7 out of 10 were heads (70%) then you just had variance occur.

It plays a big role when you are dealing with a small sample size, 10 vs. 100 for example.

The problem I see with a lot of marketers is that they tend to turn off a campaign because of short term variance. They’ll get 20 clicks on their ad and no conversions and assume that something must be messed up. That’s not necessarily true. A small sample could have a lot of variance. And, if you were to leave that same campaign on longer for another 20 clicks, you could have drastically different results because of the small sample. On the other hand, if you get really good results on a small sample size then that could also be due to variance in the small sample and not necessarily be scalable or repeatable.

I’ve also seen people get really good results spending $20/day on Facebook. But, then when they open it up and start spending $100/day then they start to realize that those small results were due to variance and that there campaign was not strong enough to open up to more people.

What’s the magic number? What kind of results do you need to ensure that it’s not just variance? Well, nobody really knows. But I will tell you that scaling is usually done most successfully in small increments. If you have a lot of extra money to spend you could take a chance and jump to a higher increment but you are taking a big risk usually.