How Integrating AI Into Our Advert Technique Helped Us Triple Our Sales

It’s always hard to give up control as an entrepreneur, especially when there’s money on the line. But, I’ve found that in some regards, the more control I give up when it comes to the advertising strategy for my beauty brand, Ogee, the more success the business has seen. 

This has never been more apparent than when my team started integrating Meta’s Advantage+ shopping campaigns (ASC+) into our social media advertising strategy.

If you’re unfamiliar, Advantage+ shopping campaigns, which are a newer option in the Facebook Ads Manager, use AI to determine which of your ads should go in front of which audience segments and how to best allocate your ad spend.

The tool does an amazing job of understanding which ad creative will appeal to whom, finding untapped pockets of potential customers, and meeting both new and existing customers at the right time to inspire them to purchase. 

Overall, since implementing these campaigns across Facebook and Instagram, we’ve seen a 50 percent increase in return on ad spend and a 45 percent improvement in cost per purchase.

We tripled our sales last year and are on track for a strong performance this year as well, and I attribute much of this success to improved ad performance.

In the end, the main thing I want as a business owner is better results. And if achieving that means letting go of the wheel a bit, then I’m all in. That said, there are a few things you can control with Advantage+ shopping campaigns to get the best results.

Read on to learn a few of the best practices we’ve picked up along the way.

Give it a wide variety of creative

There’s been a common catchphrase in advertising lately that “your creative does the targeting.” In other words, you want to ensure your ads are appealing to the right audiences rather than focusing your efforts on getting your targeting right in the back end. 

When it comes to Advantage+ shopping campaigns, your only focus should be on the creative, because that is truly your main performance leaver with this tool.

You load ad options into a campaign, choose which country you want to target, give it a budget and schedule, set what percent of the budget you want to be spent on existing versus new customers, and choose whether you want to send customers to your website and/or Meta Shops (we’ve found the most success when including Shops). 

That’s about it. No nuanced audience targeting, no micromanaging of how much budget goes to which ads.

Rather than just A/B testing two ads in a campaign, we’ve found we get the best results when we give the algorithm a lot of options — both plenty of volume and plenty of variety.

We’ve tried up to 150 different creative options for a single campaign, then let AI decide which will perform the best for which audiences. 

We upload ads in a variety of different formats (long-form video, short-form video, GIFs, still images) and creative styles (lo-fi, hi-fi, UGC, in-house creative). Within these, we make sure we’re using models that represent a diversity of people. The key is to try to cover all of your bases, aiming to have a specific creative that will resonate with every potential customer cohort you can think of. 

ASC+ does a much better job of targeting the right people with the right creatives for them than we could ever dream of doing with manual targeting. Meta’s machine learning knows better what each individual wants to see for an ad to generate the action we want to achieve: what will attract a top-of-funnel customer versus someone lower in the funnel, who wants to see a GIF versus a still image, what type of person each viewer wants to see in their ads. 

Even if it’s just ten older customers who are more likely to click on an ad with an older model, that’s ten new customers brought into our ecosystem that we could never have found with manual targeting. Meta is incredible at speaking to each user on their platform in the way they want to be spoken to.

To help generate all these ads, we have a robust influencer program for creating UGC (user-generated content), an in-house video editor who can remix content to no end, an in-house graphic design team that can create more polished ads quickly, and we also work with an outsourced performance creative agency that helps provide even more creative diversity to the mix. 

We’re also starting to experiment with AI content creation tools to help us come up with new ad varieties or give us options for copy so we can experiment even more efficiently. We are very excited about the creative AI editing tools that we hear Meta is rolling out soon.

Pay attention to existing versus new customers

One other thing you can control with Advantage+ shopping campaigns is what percentage of your ad spend you want to go to new versus existing customers. 

While we generally set the percentage of existing customers very low, we have found a few instances where it’s beneficial to target them.

For example, when trying to bump up sales on our less-popular products, we’ll sometimes set the existing customer spend to 100 percent to essentially create a retargeting campaign. During promo periods, we’ll do something similar to ensure existing customers see the promotions we’re running.

Whether you set limits on customer type or not, it’s important to break out new versus existing customers when you look at your performance metrics. I’ve noticed that sometimes the results of our campaigns are bolstered by existing customers, hiding the fact that the new customer acquisition isn’t as strong as we’d like it to be. When this happens, I reduce the existing customer percentage and refocus on getting new people in the door. 

Use the AI lessons in your other ads — and let yourself be surprised

Advantage+ shopping campaigns have been impressive, but we’re still keeping a diversified ad strategy. So far, we’ve found the best results when we have a mix of old-style Meta campaigns and Advantage+ shopping campaigns, so we are currently splitting our budget about 50/50.

That said, the lessons we’re learning from the results of Advantage+ shopping campaigns are informing and improving how we approach our other ads. For example, on our more traditional Facebook ad campaigns, we’re still leaving the targeting very open to let the algorithms find the right audiences. 

We’ve also done tests where we load all of our hero products into one Advantage+ shopping campaign to learn which gets the best return on ad spend. That can inform which we should be putting the most resources behind when advertising on other platforms and which products are worthy of their own old-school style CBO/ABO campaigns on Meta.

I recommend using these tools to support your market research and letting yourself be surprised by what you learn. A lot of times, we, as founders and marketers, think we know what will work with ads, but we are frequently wrong (or at least I am).

It’s important not to let our personal opinions or fear of trying something new make us miss out on some new game-changing strategy. Let yourself roll with what the results show, and the consumers tell you.

My final piece of advice when it comes to Meta’s Advantage+ shopping campaigns? Just test them. They’re one of the simplest campaigns to set up, all you have to do is:

• Go to your ads manager and click “create”
• Select “Sales” as your campaign objective and click “Continue”
• For your campaign setup, select “Advantage+ shopping campaign”

Then, fill in the details it asks for (audience location, budget, etc.), load in some of your best creative (you can select ads from existing campaigns or load in new ones), and learn from the results you get.

Don’t overthink it — I bet you’ll be surprised how much good can come from giving up a little control.