What Are Lookalike Audiences, and When Should You Use Them?

What Are Lookalike Audiences, and When Should You Use Them?

Back in early 2018, when Facebook announced its plans to end Partner Categories targeting, no one in the digital marketing world knew what would happen next. But our 9 Clouds team was sure of two things:

  1. There certainly would be a good — if not better — solution to make up for the Partner Categories’ absence.
  2. We would find it.

And we did — for now (since we know the digital world is always on its feet). Thanks to the many recent changes in Facebook Ads Manager, we’ve found an effective and arguably better tool to start taking advantage of in place of Partner Categories.

Hey, Lookalike Audiences, we see you. And while the feature has been around awhile, it’s finally getting its shining moment — at least with our automotive clients. 

Looka-what? ?

If you’ve never experimented with Lookalike Audiences in Ads Manager, here’s the rundown.

A Lookalike Audience is an audience targeted by Facebook that is based off of a custom list you’ve imported. Each time you import a Custom Audience, Ads Manager presents you with an option to create a Lookalike Audience from that list:

"Create a Lookalike Audience" screen in Facebook Ads Manager

So what exactly does that mean?

Here’s an example. Say you create a Custom Audience of each person who has purchased a RAM 1500 truck from your dealership in the past five years by pulling that data from your customer relationship management (CRM) system and importing it into Ads Manager.

When you create a Lookalike Audience based on that list, Facebook finds people with similar demographics and interests to those of your existing customers and uses them as the target audience.

Our Success with Lookalike Audiences

We’ve spent the past couple of months A/B testing Lookalike Audiences against the Partner Categories data before it goes away in its current form, and our findings have been remarkable.

In nearly every campaign in which we tested a Lookalike Audience against third-party data, our automotive clients saw a higher relevance score, lower cost per click, and more impressions for the ads shown to the Lookalike Audience.

More importantly, we also saw an increase in offline leads and offline purchases when using the Lookalike Audience.

I’ll break it down even further for the data nerds. Let’s take a look at a few specific tests our team ran.

For each of these campaigns, we ran two ad sets at the same budget, in the same location, and using the same creative. For Ad Set A, we used third-party data to target relevant in-market shoppers. For Ad Set B, we created a Lookalike Audience using a Custom Audience list of past purchasers from the client’s CRM.

Test 1 | New Hyundai Models

Facebook ads for Hyundai dealers

For this Midwestern auto client, we created a Facebook campaign promoting its new Hyundai inventory.

Partner Categories Data Results (Ad Set A)

  • 163 link clicks
  • 25,889 impressions
  • $0.98 CPC
  • Relevance score – 2
  • Offline leads – 5
  • Offline purchases – 1

Lookalike Audience Results (Ad Set B)

  • 476 link clicks
  • 40,102 impressions
  • $0.33 CPC
  • Relevance score – 4
  • Offline leads – 8
  • Offline purchases – 2

Test 2 | New RAM trucks

Facebook ads for RAM dealers

We again put Lookalike Audiences head-to-head with the third-party data for this rural dealership in a campaign promoting new RAM trucks.

Partner Categories Data Results (Ad Set A)

  • 54 landing page views
  • 23,423 impressions
  • $1.61 CPC
  • Relevance score – 2
  • Offline leads – 0
  • Offline purchases – 1

Lookalike Audience Results (Ad Set B)

  • 141 landing page views
  • 31,546 impressions
  • $0.68 CPC
  • Relevance score – 3
  • Offline leads – 6
  • Offline purchases – 3

Test 3 | New Ford Escape

Facebook ads for Ford dealers

In this third example, we paired the third-party data to a Lookalike Audience of past Escape customers.

Partner Categories Data Results (Ad Set A)

  • 68 landing page views
  • 29,924 impressions
  • $1.15 CPC
  • Relevance score – 2
  • Offline leads – 5
  • Offline purchases – 1

Lookalike Audience Results (Ad Set B)

  • 113 landing page views
  • 39,092 impressions
  • $0.69 CPC
  • Relevance score – 4
  • Offline leads – 8
  • Offline purchases – 1

And the Winner Is…

Lookalike Audiences — by a landslide.

I could go on with more data from the tests we’ve done, but you’d see a similar trend.

We’ve been pleased with the results of using Lookalike Audiences in our Facebook campaigns and believe it’s a valuable — while often overlooked — tool. Especially in light of the changes to Partner Categories data, this is definitely a tool you’ll want to become familiar with.

When You Should Use Lookalike Audiences

matching shoes

You should utilize Lookalike Audiences if you know what you’re selling (such as a specific vehicle) and have a detailed list of past customers who have purchased that thing (AKA a list from your CRM).

Before finding the Lookalike Audience, Facebook identifies specific user information to match your customer list to their actual Facebook profiles, so the more information you can import, the better. Here are the customer properties Facebook can use:

  • First name
  • Last name
  • Email addresses (up to three)
  • Phone numbers (up to three; include country code)
  • City
  • State/province
  • Zip/postal code
  • Country
  • Date of birth
  • Year of birth/age
  • Gender

This is why it’s advantageous for your sales team to actually gather this information from their leads and customers — because you can more accurately use it to find others who may be at the same point in the sales funnel.

What Next?

Don’t just take our word for it — test Lookalike Audiences for yourself! (We’re all about A/B testing over here.)

If you’ve never before created a Lookalike Audience in Ads Manager, give it a try, and see how it performs compared to other methods. While this is working beautifully for our automotive clients, you must find what works best for your business, your product, and your audience.

Just don’t be surprised if utilizing a Lookalike Audience helps you hit your ideal business metrics!

Improve Your Digital Marketing

As we help our clients’ ad accounts get whitelisted to use Oracle’s third-party data, we’ll continue taking advantage of Lookalike Audiences using our clients’ CRMs. In the meantime, we’ll be tuned into and listening for the ever-constant Facebook updates, so we can continue to adapt our marketing strategy.

Need help? Jump on board with us!

Get some more help with your Facebook ads and the rest of your online marketing strategy from the 9 Clouds team.



© 2019 9 Clouds | Privacy Policy