My team over at Single Grain manages a lot of ad spend, so out of pure necessity we’ve developed simple systems to make analysis as easy as possible without sacrificing results.
If you work for an agency, you understand that every client is different – the size of their organization and available resources, what they want to accomplish, and how you can help them grow. For this reason, it’s absolutely critical to fully understand your clients’ major marketing objectives and goals during onboarding.
In the early stages of your client relationships, you have to make sure you clearly define how to measure success by asking the client what their number one KPI is.
Do they care about leads? Direct response sales? Engagement? Installs of their app? Knowing this early on will help set you up for success so you know where to focus to create maximum value.
One of my favorite traffic hacks is segmenting Facebook ad campaigns in different ways depending on the client’s primary goal. This allows me to do a couple of things:
1. Quickly see which ad is performing the best in a campaign based on the KPIs we established (leads, direct sales, engagement, app installs, etc.).
2. Structuring campaigns this way allows you to pit your ads against each other. If you’re using oCPM bidding (which you should be in most cases), Facebook will automatically give the best performers impression and budget priority. This makes it easy to run lots of campaigns for all types of different clients at scale.
Here are the 5 most common ways we segment campaigns for lightning-fast analysis:
For this setup, we would create a campaign with each ad targeting a different precise interest that has roughly the same audience size. When split testing ad creative, I recommend only testing one element at a time (e.g. either interest targets or creative) so you know what’s really making a difference in performance. In this case, you should leave all of your creative and messaging the same, changing only the precise interest being targeted.
For this campaign we used different video games as interest targets to promote a client that had just released a similar video game. With a few seconds of analysis, I can see that we should cut Skylanders and Minecraft since the other targets are converting at a lower CPA with many more conversions overall.
If you’re looking for large audiences that you can run ads for at scale, try segmenting your campaigns by different broad and partner category targets. Most of the category targets on Facebook have massive audiences in the millions, so when you find a winner you can typically run the campaign for months without touching it and still reach new people every day.
Again, I can quickly see that the category targeted ad labeled V2 has more conversions with a 47.4% lower CPA. In this case I’d cut the V1 ad and create a new ad for testing, leaving V2 as the control.
If you’re trying to reach customers internationally, segmenting your ads by country allows for quick analysis without digging into the reporting manager. This is one of my favorite ways to run ads for clients focusing on app install with international businesses. With international clients, I like start at the macro level by finding which countries are most profitable overall before I start drilling down into more granular ad targets.
Make sure you keep everything else the same as far as creative and ad targeting, only changing the country in each ad. Setting up your campaigns this way will allow you to quickly see which country is performing the best so you can easily note differences in performance at the country level. Facebook will give the country with the lowest CPA and most conversions priority over your other ads automatically so you don’t need to check the campaign constantly.
If you can’t tell by now, I’m a huge advocate of testing everything. I always hear about how the right-hand rail is ineffective and that I should only stick to News Feed placements. The funny thing about that is the data tells a very different story – I’ve had great luck with the right-hand rail depending on the vertical the client is in.
Segmenting out your ads based on ad type/unit (RHS, News Feed, Sponsored Story, etc.) lets you determine which converts at the lowest CPA with minimal effort. This campaign setup also allows you to analyze more granular campaign data for each ad type (like CTR and CPC) without using the reporting manager.
Ad Creative & Messaging
When testing different creative in a campaign (like visuals and messaging), use the exact same ad targets and change one element of the creative in each ad. If you make the creative and the targeting different in your tests, you won’t know which element impacted your conversion rate, so stick to testing one element at a time.
Ad image and headlines are the two primary drivers of CPA reduction, so start there with your testing.
There you have it! Use these different campaign segmentation options to make optimization faster and easier while maintaining the same results – no more spending hours doing analysis in Excel or the reporting manager!
Have you used any of these methods? Which are you going to try next?