Some time ago, I held a webinar about the way we structure Facebook Ads accounts and the most up to date strategies. I explained our approach and showed how it evolved in 2020.

So, if you want to catch up on this, feel free to watch the full webinar. It includes my presentation and comments from the audience. You can watch it here.

If you prefer to read, rather than listen, don’t worry! Here is a summary of the key points of my talk as well as the most important advice. Read this, and you can be sure you’re up to date with our strategies and Facebook’s changes.

Okay, let’s dive in by starting from the biggest challenges waiting for you in the Facebook Ads world!

Problem 1: Constant changes

The biggest problem with Facebook Ads is the fact that it’s always changing. Everybody keeps watching courses and following Facebook Ads gurus, but changes happen so quickly, that things that you learned a few months back are quite often already outdated.

That’s why it’s important to regularly refresh knowledge, exchange insights with other advertisers, and test this knowledge right away. This is one of the key reasons why we created this group. We want it to be a platform for sharing information and ideas. My mission is to bring to you everything we’ve tested and tell you what works and what doesn’t.

Problem 2: The strategy evolves with your business

You need to remember that strategies that took you from point A to point B are not always the strategies that will take you from B to C. Things evolve as your business scales. And you need to adopt different structures and strategies on various stages of an e-commerce business.

Most common account structure mistakes

• Not knowing how to analyze your data – It’s very common, so I am sharing a resource I prepared to make it easier. It’s a list of the report columns that I use when I analyze the ads account. They give me the full picture of the campaign so that I can diagnose where in the funnel my ads are not working and what decisions I should take based on that.
• Messy naming convention for campaigns & ads – If you’re getting started, it probably doesn’t matter that much. But if your ad account is massive, and you’re testing a lot, then you need to make sure that you keep everything organized. I also include a link to our internal research on how our media buyers name their campaigns.
• Not knowing your numbers – It’s super common for people to just come up with arbitrary targets such as 3x or 4x ROAS without looking at the break-even targets with your margins. This way, you don’t optimize for the ultimate volume of profits, or you make it much harder for yourself to scale.
• No exclusions between steps of the funnel – That’s probably the biggest mistake. Even if you have campaigns that seem to be doing well, it may not be the case. If they don’t exclude, for example, your past customers or people who are lower in your funnel, the ROAS are going to sink with scaling the campaign.  You will simply exhaust the audience.
• No loyalty campaigns and missing retargeting steps – There’s no simple solution to these mistakes. Type of loyalty campaigns and the importance of retargeting measures depend on product qualities such as price and specificity.
• Improper use of CBOs (Campaign Budget Optimisation).
• Too narrow audiences.
• Forgetting about structure levels – As I mentioned, there’s no such thing as “one size fits all” structure optimization. Your structure is going to evolve.

How do we work with new accounts in SalesGenomics?

• We set up three types of campaigns:
  • Testing campaigns
    • Audience testing campaigns
    • Creative testing campaigns
    • Retargeting testing campaign
  • Prospecting campaigns
  • Hybrid scaling funnel (our biggest discovery of 2020).

   

  Initially, we test some audiences and creatives at the top of the funnel campaigns. Once we are getting some data, we start to run retargeting campaigns, separate for each funnel (top, middle, and bottom). Sometimes we already test retention campaigns at this point.

  Then we test some look-alikes from different funnel stages (in separate campaigns) and bundle them together into bigger CBOs. Next, we bundle all the retargeting campaigns into the funnel campaign. I call it an early stage scaling structure.

  In a very mature account structure, we literally have just two or three campaigns. One campaign that contains everything – retargeting and prospecting together, and a separate DPA campaign with the catalog sales objective

  Here is the structure of our testing campaigns.

First testing campaign (new store)

In the first testing campaign, we do:

• 1 CBO campaign (optimized for purchase);
• Number of Ad Sets depends on a daily budget (a minimum is half of your target CPA per ad set per day, but the more, the better);
• We usually start a single interest ad sets, sometimes we narrow them with engaged shoppers;
• Auto bidding;
• Only news feeds and stories placements at this point;
• We set up the minimum spends per ad sets in this initial campaigns to half of the assigned budget per ad set (so total campaign budget over a number of ad sets and divided by half is going to be your minimum spend per ad set);
• Two to four creatives per ad set.
• Funnel step exclusions – It’s super important. Make sure you exclude as much as you can. Most of the time, we exclude purchases 180 days. I would exclude even broader audiences, as much as possible.

Here is a guide on what kind of audiences to exclude from different stages of your funnel.

Creative testing campaign

We used to wonder how to test creatives. In the past, we inserted new creatives into the existing ad sets, but it didn’t work well. We had to wait for ages for the results. Now, we have a separate campaign for testing creatives.

Here are the rules we follow in the creative testing campaign:

• CBO;
• Conversion objective: purchase;
• Minimum spend per ad set;
• One creative per ad set;
• Ideally use no targeting or your best audience;
• Use control creative;
• Run multiple duplicates of each ad set
• Don’t use engagement or traffic objectives to test creatives – they used to work but now they don’t correlate well with results when transfered into conversion campaigns, plus this method is no longer cost effective, as CPMs on engagement campaigns increased, especially in the US.

Check out our webinars about creatives, here.

Audience testing campaign

The goal of this campaign is to test the audiences gathered after the initial campaign that we’ve launched.

The way we do that is similar to the first campaign. The only difference is that this time we don’t set minimum budgets per ad set, and we use multiple creatives. So we have:

• CBO;
• Conversion objective: purchase;
• No minimum spend on ad sets;
• Multiple creatives per ad set;
• We usually use single interest ad sets of min. 500k-1mln audience size;
• Avoid “laser-targeted” narrow interests, which are not in line with Facebook algorithms. Small audiences exhaust fast and don’t leave space for scaling.

Extra tip: When to kill ad sets?

• Again, no one size fits all.
• Rule of thumb – if an ad set spends 1x break-even CPA without a purchase, just try it.
• For audiences, you really can’t afford to waste money on something that doesn’t give you good results from the beginning.

Retargeting testing campaigns

I call it retargeting testing because we want to test what timeframes and what kind of stages of the funnel are going to be promising for a particular product. So we start with a very granular set up for each funnel.

• First of all two types of campaigns – Dynamic Product Ads (catalog sales campaigns) and conversion campaigns;
• Each is divided into the middle of the funnel (MOF) and bottom of the funnel (BOF)
• Different times timeframes for each of them;
• We test quite a few creatives on each funnel stage (2-10);
• Check what timeframes, funnels, and objectives are immediately profitable, and then you can test personalizing content for different stages. Don’t fixate over targeting if it’s not working very well. Focus on what generates results. Then bundle all the profitable or just marginally profitable retargeting ads together into single ad sets (if you don’t want to personalize them) or keep them separate (if you do want to customize the content).

Hybrid scaling funnel campaign

This is the ultimate purpose of the Campaign Budget Optimization. I realized some time ago that CBO is actually brilliant. When Facebook started implementing it, everybody was angry. They treated CBOs as a way to take away their control. Also, it wasn’t working as it was supposed to, because people didn’t know how to use it. But when you take time to really dig into CBOs, it makes so much sense.

The structure that we use is by far the best way to leverage Facebook Algorithm. It’s about giving Facebook’s Algorithm the most control you can imagine, and it pays us back.

This strategy works best for stores with very educated Pixel, but anyone can benefit from it once they see how it works.

Here’s an example of one of our accounts. By implementing this structure, their ROAS climbed from 2 to 3 in one month, with no other changes from our side.

In this structure, Facebook decides how to distribute the budget between prospecting and retargeting. If there is a good day for prospecting for a particular audience, Facebook is going to pump more budgets towards prospecting. If it’s a better day to retarget, it’s going to push more budget into retargeting. So all you do as an advertiser is deciding on the total budget.

So, we implement this hybrid scaling funnel campaign after testing and prospecting campaigns, once we bundle them together into single CBO. You can add retention campaigns to the bundle if they apply for the product.

Summary of the structure evolution

I would say as you increase the budget, theaudience size should increase as well (by bundling) on every level. But in general, it’s difficult to give specific advice on when to do what, because Facebook Ads 50% science and 50% art J. I guess there’s no fixed spend to audience size ratio. So, you have to find your own solution, see what works best for you.

For more actionable ecommerce scaling tips, join our Facebook community: ecommerce scaling strategies by SalesGenomics