Skip to main content

A-B Testing

Lesson 35 from: Instagram Ads Fundamentals

Sharon Lee Thony

A-B Testing

Lesson 35 from: Instagram Ads Fundamentals

Sharon Lee Thony

buy this class

$00

$00
Sale Ends Soon!

starting under

$13/month*

Unlock this classplus 2000+ more >

Lesson Info

35. A-B Testing

Lesson Info

A-B Testing

reviewing your client's goals and using reports to ensure that you're delivering against them are highly important when it comes to campaign management and reporting different clients will have different needs. Some campaigns will aim to generate leads and others might be to generate traffic. Having a clear sense of your clients. KPI S or key performance indicators will help you to focus your efforts and report out on successes as well as areas of improvement reports. Allow us to optimize campaign activity to continue doing more of what's working and to stop or reduce the campaigns that are not performing in ads manager. You'll also be able to a B test various versions of your campaigns to determine what is working best. It's important to remember that clients may have different needs for reporting cadence or structure. Regardless of the project or client, ads manager will be able to provide you with the insights that will help you to run effective campaigns. The first type of optimiza...

tion we're going to talk about is A. B testing and ads manager A. B testing is simply testing one variable that's different between two different versions of an ad. Here's a case study of what Zara did with their A. B test. They decided to Test two different images showing the same skirt. One image on the left was user generated content. It was a piece of content that one of their followers had posted of her wearing the skirt that they wanted to feature in their promotion on the left, they chose what I'm going to call stock photography but probably it was custom photography that was developed for other aspects of the business. Maybe a brochure, maybe a piece of collateral, but they had chosen to use a more generic image to showcase that same exact skirt. What they found after they ran the instagram ad Was that the user generated content add performed better. And you can see here that it had a click through rate of .62% versus a click through rate of .24% for the stock image. Everything else in this ad was the same. Their caption was the same. Their call to action was the same. They're targeting was the same. Their placement and their budget were the same. The only difference was the image itself and this is a true A B test in that the variable here, which happened to be this winning image on the bottom one, the A. B test. What other kinds of things can you test when it comes to a be optimization. You can you can test any other piece of the content itself. So it could be testing, copy the words, the caption, the call to action. You might decide to test any other images or video that you show as well. You might test. But audience. So you could test who you're targeting against. You can test perhaps a different type of interest that they might have. Perhaps a different definition of the audience. Maybe you're testing men against women, maybe you're testing two different age groups. Perhaps you're testing people in two different locations. Another thing you could test his budget so spending or bidding on cost per click might be a different result than bidding on impressions. Optimizing for landing page views might be different for than optimizing for conversions and testing different budgeting strategies might be a way that you want to go about it. You can also choose the test placement. So an A B test could be running that same exact add the same content the same audience but running one ad on facebook one ad on instagram or maybe specific to instagram. You're running that same exact ad, same audience, same budget. But you're running one version in the instagram feed and the other version and only instagram stories to test what parts of instagram you're getting the best results from and then finally testing the call to action. So as mentioned, ads manager offers you many different choices, Testing something like learn more. Might be interesting to test against sign up or perhaps testing book now could be different from buy tickets, playing with those various ways of articulating the phrase and the calls to action can be a valid A B test as well when it comes to generating an actual split test or an A. B test in as manager. It's quite simple. The first step of creating any campaign is selecting your objective and after you select your objective, you'll be given the opportunity to indicate whether or not you want to create an A. B test and all you need to do is just slide this little marker over from left to right and by doing so it will automatically prompt the system to then create two different versions of that campaign for you where you determine whether or not you want to test creative, you wanted to test delivery, optimization, audience or placement and as you're building out the campaign, they'll ask you for two different versions of that. One thing that you've indicated that you want to test so it's pretty turnkey. What's also great about ads manager and optimization is that as you run your A. B test, the platform will automatically tell you when you you've been able to declare a winner. So no longer do you have to check in look at results and compare them once the results are different enough so that it's clear that one version is performing better than another. You'll get an email, you'll also get a notification in facebook telling you which version is winning and you'll have the option then to either accept that and and to continue with just that one version or to pause the campaign entirely and perhaps reevaluate and start all over. So it's very simple and completely turnkey and automatic Some tips on that is that you want to run ads for at least four days to give it enough time but no more than 14. So no more than two weeks. Because after that, if you don't have a winner after two weeks, then that just shows that there might not be anything different enough about your ads that's actually generating different responses. You also want to ensure that everything else remains the same for both of those scenarios. So do not use the image that you're testing in other any other campaigns or in any other digital marketing channels. So if you are testing something like what Zara was looking at, where you have a piece of user generated content and it's a great image. Don't then take that image and put it on the home page of your website because it's going to ruin the results of your test. People are going to be seeing it elsewhere and you're not going to know if it's truly that one image that's generating a lot of engagement or if it's because they've seen that image somewhere else and they've recognized it and then last, just set it and forget it. So, figure out your testing parameters, set that up and then sit back and let the platform figure it out for you. If you go in there and you constantly tweak budgeting or audience or placement during the first couple of days, you could also then screw up or ruin or affect the test results