Audit Insights From 27 Accounts Spending Millions on Facebook Ads

One of the benefits of being a part of the Founders Membership is a free monthly audit of members’ Facebook Ad Accounts. We see lots of accounts with lots of different strategies in the way accounts are run. We’re big believers that every account is different and there’s no ONLY ONE WAY to be successful on ads.

We’ve recently found some of the same overarching themes of findings in audits and wanted to share them. Our success is your success.

One caviat to keep in mind: if what you’re doing currently is working, there’s no use in rebuilding or restructuring an account for the sake of it. We’re big fans of: “If it ain’t broke, dont’ fix it.” That being said, these might be good ideas to slowly implement, or implement some, but not all. Use your own discretion. You’re the expert of your account.

Have questions? Consider joining the Founders Membership where you get one FREE audit per month and can learn from some of the absolute best ad buyers worldwide.

Account Structure

• Don’t do too much audience testing

  • With audience expansion as the default, there isn’t that much differentiation between a 3% and your 8% LLA, for example. If you have found a few audiences that work, lean on them and focus the rest of your efforts on creative testing

  • Remember that larger audiences are still working best for this exact reason, so consider leaning into 10% LLA stacks, totally broad audiences (with simple exclusions), or very large interest persona targeting

• Don’t have too many hero products or one product-only specific creative

  • Generally, you don’t want to have too many separate campaigns targeting the same audiences, with the only differentiation being product(s) advertised. HOWEVER, sometimes you have to because FB will often choose one ad/product and focus all the spend on that one, therefore not pushing different products to prospecting audiences. Try to focus different products in ads to no more than 2-3 to the same audience (determine your very best hero products, then upsell/cross-sell in retargeting campaigns)

    • An alternative could also be displaying multiple products in a carousels, collection ads, slideshows, etc., or even a video testimonial/UGC mashup with multiple products. Try to not get too segmented or pigeon-holed here

• While you can (and some ad buyers suggest to) use higher funnel campaign objectives in prospecting, one problem with this is that you tend to get a lot of people with low conversion intent. FB is very good at doing what you ask, so in this case they will look for people they think will click, but not necessarily people likely to convert

• For scaling campaigns, keep 2-3 ads in each ad set. When one starts to fatigue, turn it off and replace with what is hopefully a new ad that tested well in the testing campaign recently. Repeat as necessary

Creative

• Have a creative testing program in place so you establish a pipeline of ads that have tested well before they go in your main funnel

  • Try to use one of your very best performing TOF audiences for your creative testing campaigns so that the creative has the highest likelihood to be successful

  • Note that a separate creative testing campaign is most helpful and works best when you have a LOT of creative and an ongoing inflow of new creative that is too much to insert into existing campaigns daily or weekly. (This is likely when you have more than 10+ creatives to add into the account per week)

  • Aim to keep testing budget about 10% of your total budget

  • One benefit of a separate testing campaign is that ads will tend to perform much differently when their budget is $30/day vs $300/day vs. $3,000/day. By having one testing ad set, it keeps the daily budget it's being run at closer to those in my main funnel to mirror similar results

• Keep in mind that testing a lot of creative can definitely get pricy

  • To start, try very different creatives if you don't know what's going to move the needle, not small changes

  • Once you get a feel for what's working, you can start to iterate on it with medium and then smaller changes

  • Also be sure to test creative formats (i.e. carousels, collections, photos, videos, etc., and then smaller changes would be graphics vs. UGC images or produced video vs. UGC testimonial video, or 30-second video vs. 6-second video, etc.)

• Always test proven creative to new audiences and new creative to proven audiences

• For new creative/ad angles, run them in the same ad set as the current ads - both for testing and funnel - and let FB sort out who to show the ads to

  • Every so often, check to make sure both/all creatives are getting spend — doesn’t have to be exactly equal

• Don’t forget to test new ads and angles to your warm traffic as well as cold traffic

• Dynamic Creative Testing (DCT) can be helpful when just starting an account and I don't know what kind of creative, copy or headline will work (or any hypothesis of what is best). With DCT, you can test all of them at the same time to get learnings faster

Technical, Analysis & Reporting

• If you aren't seeing results from retargeting, try removing exclusions (even just starting with website visitor exclusions and leaving purchase/conversion audiences) and allow your prospecting ad sets to also do some retargeting for you

  • Do this especially if retargeting and prospecting ad creatives aren’t that much different

• Make sure conversions are reporting in-platform on FB

  • Even though we know that data is modeled and you need a third-party tracking software (like TripleWhale or NorthBeam) to ensure validity of the data, IN-PLATFORM DATA is what FB is using to optimize and deliver ads, so data does need to be tracking

  • If there's no way to fire a custom event or conversion, look into Offline Conversions for tracking

• In-platform reporting post-iOS14.5 has not been accurate, as we know it is all modeled data. Try TripleWhale (Shopify only for now) or NorthBeam to gather third-party tracking information as a true source of real data. These are additional costs (TripleWhale more inexpensive than NorthBeam), but can help with showing real data to make decisions and optimizations off of

  • Google Analytics UTM reporting is helpful, but still incomplete as it's last click only

• One benefit of using 7-day click/1-day view, especially when using a third-party software for conversion tracking and more accurate results is that you want FB to have the widest amount of data possible to optimize (even if it’s slightly incorrect)

• First time impression ratio (FTIR) can be a nice tool to diagnose when ads aren't performing well, but we very rarely look at it unless diagnosing a problem

  • Note that exclusions will lead more rapidly to a lower FTIR. BUT, this doesn’t have to be concerning if you don’t mind customers seeing your ads several times

• If you’re targeting multiple countries/territories, pull a breakdown report of results by country and see where you are having the most/least success, as well as if one location is getting significantly more or less delivery than others

Love these audit findings? Did you know you get one FREE audit per month when you join the Founders Membership?

A huge thank you and shout out to Foxwell Digital’s audit captain, Alex Afterman, on his input for this blog.