Our October 2021 Facebook Ads Tests

Our October 2021 Facebook Ads Tests

Antoine Dalmas Facebook Ads Guide

Every month, our expert team of Media Buyers performs a series of Facebook Advertising Tests to offer you expert tips on how to optimize your campaigns and strategies.

To be the first to receive our test results, subscribe to our iOS 14 Squad Newsletter! Click here to subscribe for free and begin taking advantage of our expert tips.

Sales page : Facebook Funnel (October 4th)

It's Monday, October 4th, and one of the best tips we could give you for overcoming the issues brought about by iOS14 is to increase your average customer cart value.

If you’re going to generate ‘fewer’ attributed conversions, they may as well have a higher value in order to generate a better (on the surface) ROAS on Facebook’s platform (although your ROI is probably more important).

That's why one of the techniques we tested recently was to redirect users to a dedicated sales page to see if it would increase the average customer cart value.

Far from being a website with too many choices, too many buttons, and too many exits, a sales page can be the best way to focus your prospect's attention and allow them to take much fewer steps to complete a purchase.

To help you understand this better, let's imagine that your sales funnel is a video game. This is what a typical e-commerce site looks like (left), compared to a sales page (right).

ecommerce vs. salespage

ecommerce vs. salespage

On an e-commerce website, the user can take a number of possible paths (sometimes too many), while on a sales page, the path the user has to take is straightforward and simple (no detours!). Simply put: this is a sales page’s greatest strength.

Let’s use one of our e-commerce clients who sells hairbrushes (a flagship product) as an example. They're rather special as they address a problem that many parents are familiar with: brushing their daughters’ hair. The brush in question helps overcome this problem as it is specially made to detangle long hair without causing any pain.

Now, let's look at the client's site (which we normally redirect users to).

landing page facebook

Landing page

And the sales page.

funnel facebook ads

social proof facebook

In terms of the ads we ran for this client, we featured the best ones in both campaigns (site vs. sales page). The campaigns included UGC (very effective for this client) and more standard ads.

facebook ads examples

Facebook Ads examples

Below are the results for this client the month before we launched our Sales Page Campaign (in green). Our target CAC was €8 and below.

facebook ads manager

We launched our campaign the following month.

Facebook cost per purchase

We exceeded the target CAC by €1.41, so we began optimizing the campaign by running active ads. We then noticed the CPM decrease for our Evergreen campaign and its CAC decrease by €1.

Here are the results for the month after.

facebook ads test

The CPM was still lower than before, and our CAC began stabilizing, both for the new campaign and the account as a whole. In terms of ROAS, no drastic difference despite our average cart value now being €1 higher.

But in the end, the most interesting thing is that....

facebook ads scale

We increased the account’s acquisition spend by 88%, which was reflected in our retargeting campaigns in the first month.

facebook ads result

Let's summarize this experience. Our objective was, above all, to increase the average value of our customers’ carts. We did not achieve this, but:

  • We were able to drastically scale the account...
  • ...without visibly impacting our evergreen campaign...
  • ...and increasing the revenues of our retargeting campaigns.

With these results, we plan to continue optimizing our sales pages, adding upsell pages, and testing the efficiency of our sales page variations.

Dynamic Ads Broad Audience / DABA (October, 18th)

Here are our test results as of October 18, 2021.

Dynamic Ads Broad Audience or DABA.

If you’ve been following our tests since August, you’ve certainly already heard of this technique.

Recommended by our Account Manager Partner at Facebook, the technique consists of simply using a DPA campaign (not a Dynamic Creative Campaign) to advertise to a Broad audience, without any geographical targeting.

After our first test, we concluded that the most effective way to get the best results from a DABA is:

  • To be an e-commerce that sells a wide selection of products (more than 10)
  • Have coherent collections or product lines
  • To sell widely purchased products (ex: clothing), rather than niche items
  • To have attractive prices that are likely to get customers to make a spontaneous purchase

One of our Media Buyers has been following this same test for a long time and recently generated record results.

Their account, an e-commerce that sells shoes, often receives new models and collections that they need to promote. This can result in a lot of changes in a campaign, which often affect results.

In order to maintain a solid ROAS, they decided to launch a DABA campaign in addition to a standard acquisition campaign.

Let’s take a look at the campaigns.

dynamic ads broad audience

At this point in July, we had a standard acquisition campaign and a DPA retargeting campaign.

Acquisition Audiences:

  • Lookalike Website Visitors 6%

Retargeting Audiences:

  • View content & Add to cart 60 days

100% CBO with a total budget of $230/day.


We then launched a DABA campaign by taking 20% of the budget from our Standard TOF. We used the same creatives as the standard campaign but turned them into Collection ads. Here are the results for July and August.

ads manager

Dynamic Ads facebook

In August, we dropped our budget for both the standard campaign and DABA campaign to $125.

July and August were pretty weak months as we were just coming off of a big summer promotion in June. But in September, we launched the new Autumn/Winter Collection Campaign. Here’s what our results looked like:

facebook ads results

In addition to an improvement in our ROAS, which simply exploded, all 3 campaigns generated the same CA ($33,000) as August and July combined, bringing the CAC down from $46.74 to $33.

Over the last 7 days, as we write this email (October 14th), we’ve continued to keep an eye on the DABA campaign. Here are the results.

test facebooka ds

Since October, our Media Buyer has been using different creatives for the two TOF campaigns and has assured us that they spend significantly less time optimizing the DABA campaign than the Standard campaign (whose results are constantly decreasing).

All in all, this is a campaign we highly recommend, especially if you want to maintain a solid ROAS to manage more complex campaigns on the side.

E-commerce account structure (October, 25th)

Facebook Ads: A Campaign with a 3.57 ROAS

Here are the results of our latest test as of October 25th, 2021.

We recently received the following question on our Facebook group, Facebook Ads pour Francophones (Facebook Ads for Francophones).

facebook comment

Facebook Comment

For some Media Buyers, the Learning Phase is a source of anxiety, especially if a campaign is working well and you’re afraid of affecting your results.

But, this is rarely the case.

Because if your campaign’s results decrease, it’s very probable that the campaign was already on the decline before you introduced a new element.

If your results are stable, the worst that can happen is that your new ads do not get broadcast.

Simply because these new ads have no ‘experience’ versus other ads with more historical data.

This happened to one of our Media Buyers, bringing us to the topic of this week’s test: account structure.

The client in question is a jewelry e-commerce company with a huge product repertoire. It was important to run many tests to determine the next bestseller, but each time our Media Buyer tried to add new ads to our main campaign: 0 spend, learning limit, etc...

The main campaign had a budget of €250/day, so they launched a second campaign at only €50/day. The goal was to test new creatives without impacting the main campaign, which worked after a test budget of €290.

ads manager facebook

We ran this test in September and, and at the time, we were only using a BROAD audience. We maintained the following results for 7 days.

ROAS facebook Ads

But, we then noticed an apparent drop in the results, which prompted us to react by simply targeting the same BROAD audience, but with Dynamic Creatives this time.

facebook ads results

Thanks to this, our results began to stabilize, and the budget was fairly distributed.

Overall, here is how the account’s results evolved over the first 5, 7, and 14 days.

campaign dynamic ads facebook

test campaign facebook ads

facebook ads

For the last image, acquisition dropped by 11%, but the account still shows an increasing ROAS of 7.34%, which was partly made possible by this new campaign!

By using this strategy, our Media Buyer gained the following advantages:

  • The ability to test an alternative campaign without risk and then transfer the best creatives to the main campaign
  • A small drop in an already high ROAS thanks to the test campaign’s humble budget, which will alleviate a drop in ROAS elsewhere
  • More qualified visitors (the goal is to scale) and thus more retargeting conversions.

If you need to test new creatives, but are scared that doing so will negatively affect results, launch a side campaign. Using a small budget should not impact your Evergreen campaign and what you learn from this side campaign will benefit your entire account.

Almost every e-commerce account at J7 Media has a test campaign, so don't wait to try it too!

If you know someone else who’d be interested in our newsletter, feel free to share this email with them or direct them to our subscription page here.