Clicks vs Reach to determine winner of ad test?

[Sutra]

Que paso vato's!

When split testing ads on Facebook, is it better to use Reach to determine the winner, or should you use Clicks?

I've been going by clicks. However, I've heard a couple people say that after your ads each hit 1000 reach you can confidently choose a winner.

What say you?

 

[Calamari]

Cost per click or roi or cost per conversion depending on goals.

 

[Sutra]

I'm running a new campaign for leads (email optins) so it's cost per conversion for this.

How many optins would each test have to get before you would determine a winner?

Right now I have 3 splits tests running. One of them has almost twice as meany leads (15 vs 8 vs 8) for almost half the price. Wondering if that variance is enough to declare a winner once the Reach hits 1000 (it's about 900 right now).

 

[lion1978]

One of them has almost twice as meany leads (15 vs 8 vs 8) for almost half the price

I would call that a winner, and then move on to the next split test, asuming you only test one part of your ad copy at the time.

 

[Sutra]

Yeah, the only difference is the image. Everything else exactly the same.

Seemed like that was enough data to declare winner but wanted to be sure. Thank you Mr Lion.

Next stop: Split tests on audiences!

 

[Calamari]

You're better off using one of the calculators than your gut.

http://drpete.co/split-test-calculator

 

[Ryuzaki]

Definitely if one is producing twice the leads for half the cost... that's a 4x ROI over the others, assuming they are all in the black.

This is not something I've done and I'm eager to keep learning from this thread, so I want to share what I was thinking as I read this.

It depends on your goal. If it's conversions of some sort then it's basically case closed. That's what you should measure.

But if you're going for traffic and social signals, I'd think that clicks and reach are pointing at the same thing, as they are both interdependent. But reach represents a potentiality of engagement & traffic (likes, shares, visits) which should have a nice SEO and branding effect. Clicks on the other hand are only measuring traffic itself. An increased reach should indicate increased engagement, while clicks are only measuring one type of engagement. Again it depends on the goals.

Does that actually make sense outside of my head and including my lack of experience with social PPC?

 

[lion1978]

Yeah, the only difference is the image. Everything else exactly the same.

I guess an image can be woth a thousand words, as the old saying is saying LOL

 

[Sutra]

Definitely if one is producing twice the leads for half the cost... that's a 4x ROI over the others, assuming they are all in the black.

This is not something I've done and I'm eager to keep learning from this thread, so I want to share what I was thinking as I read this.

It depends on your goal. If it's conversions of some sort then it's basically case closed. That's what you should measure.

But if you're going for traffic and social signals, I'd think that clicks and reach are pointing at the same thing, as they are both interdependent. But reach represents a potentiality of engagement & traffic (likes, shares, visits) which should have a nice SEO and branding effect. Clicks on the other hand are only measuring traffic itself. An increased reach should indicate increased engagement, while clicks are only measuring one type of engagement. Again it depends on the goals.

Does that actually make sense outside of my head and including my lack of experience with social PPC?

Yeah, haha, I think I get what you're saying.

I probably should have been more clear in the beginning though. Because it's not clicks that I'm measuring in this instance; I'm measuring conversions (email optins). But as you said, reach includes engagement.

A nice by-product of more engagement is lower prices per conversion.

I guess an image can be woth a thousand words, as the old saying is saying LOL

Hah, truth!

 

[BogTrotter]

I've been going by clicks. However, I've heard a couple people say that after your ads each hit 1000 reach you can confidently choose a winner.

What say you?

When they said that, they were probably not suggesting that "reach" should be your goal. They were probably suggesting that you can use 1000 reach as an approximation for when your results and sample size are big enough to declare a reliable winner.

That is sometimes expressed as a "confidence percentage."

In this case, it would be the same as saying you'd let each ad run 1000 impressions before choosing a winner, for whatever your goal is.