How to Build Advertising Lookalike Audiences Using Customer Behavioral Data

It would be quite awesome to be able to clone your best customers. Pop them in a machine, press a few buttons, and Hey Presto! an infinite supply of revenue.

Unfortunately, while you can buy cloning machines on Amazon, they’re only for plants. And plants are notoriously fickle customers. So you have to get more inventive in finding customers like the best ones you currently have.

This is where the idea of lookalike audiences comes in. Lookalike audiences are people who look like your current customers. Not physically, but behaviorally or demographically. They might visit the same sites, or have the same type of job, or search for the same types of things online. Effectively, they are clones of your best customers.

Let's look at how creating lookalike audiences helps your business, and how you can leverage your Freshpaint data to build these audiences and create real time, personalized marketing ad campaigns.

Lookalikes are about finding your best customers

Audience is everything in effectively using ad platforms like Facebook. Without defining an audience, none of your ads are targeted and your ad spend can balloon with no return.

Defining an audience lets you target your ad spend. There are a few ways you can do this:

1. You can create an arbitrary audience. Show my ad to all 27-year-old flutists in Western Iowa.
2. You can create a custom audience. These are people who have already interacted with your product in some way, or are already customers. If you’ve ever been to a website and then suddenly seen an ad for that product everywhere, congratulations, you are now in a custom audience. This type of targeting is also known as retargeting.
3. You can create a lookalike audience. This is an audience that shares similarities to your custom audience, but isn’t them. The idea being that if your best customers are all ABC or all do XYZ before signing up for your product, any person out there being ABC or doing XYZ is likely to find value with your product too.

You can do any of these, but lookalike audiences are the most valuable, because they have the ability to reduce costs while increasing revenue.

Lookalikes reduce CAC

The Greek playwright Plautus wrote “you have to spend money to make money.” But, before he was a playwright he tanked his merchant shipping company, so he might not be the galaxy business brain the quote suggests.

But customers rarely do come for free. There is always some cost for acquiring customers. This customer acquisition cost, or CAC, is an important metric for businesses to help understand the efficiency of their marketing and sales efforts. By comparing CAC to the lifetime value of a customer (LTV), businesses can determine the profitability of acquiring new customers and make informed decisions about how much to invest in customer acquisition.

This is one way lookalike audiences can help businesses. Lookalike audiences help you target marketing efforts more effectively and efficiently, reducing CAC. By focusing on individuals who are similar to existing customers, you can reach potential new customers who are more likely to be interested in your products or services. So you can target your ad spend on fewer people to keep your budget lower, and those people have a higher rate of conversion.

Lookalikes increase LTV

But really, lookalikes are about a wider problem than just reducing CAC—finding great customers. Great customers are often less expensive, because they are already looking for your solution and the right person to use it, so don’t need a ton of sales and marketing. But they are also more valuable, because:

• They stay longer. If you are basing your lookalike audiences on customers that have got value out of your product, there is a higher likelihood that these new customers will as well. They will be activated and retained easier.
• They upgrade. The same goes for increasing revenue from these customers. If a customer finds value, they are going to upgrade as they become more successful with your product (and therefore need more of it). This also feeds into a better LTV:CAC ratio.

So working out your best customers from your product, and then heading out to ad platforms like Facebook and finding more of them is critical for growth.

How do you do that?

Build a lookalike audience from behavioral data

It’s actually pretty easy.

Let’s continue our video sharing app example from our previous posts. From working on our activation thesis and personalized messaging, we already know that the best customers in our product are those that record and share a video.

These behaviors are our starting point. To build a lookalike audience using behavioral data, you would need to gather data on the behaviors of your existing customers. We want to start collecting data on any user that exhibits these recording and sharing behaviors. This is going to be our custom audience.

As we said, a custom audience is the group of people you are basing your lookalike audience on. We could retarget to this audience (”Have you thought about sharing more videos with your colleagues?”), but what we really want is to find more of them. We want to identify common behaviors or characteristics among our existing customers and create an ideal customer profile (ICP) that the ad platform can use.

In Freshpaint we can do this by connecting an ad platform (here we’re using Facebook) and sharing the first-party data across. Here are the steps you need to take:

1. Get your pixel ID and conversions API access token from your events manager dashboard in Facebook.
2. Set up the required event parameters to make sure Facebook is getting the correct information. These will be things like the action_source for the event, but also user traits.
3. You then want to convert your Freshpaint events into Facebook events so Facebook can understand the most important event to track. Facebook’s audiences algorithm expects the most important event to be a purchase, but in our case it is sharing a video, so we’d have to map our shareVideo event to the Purchase event in Facebook. We then need to assign a value to it (again, because it is expecting a purchase). You can do all of this within Transformations.

To do all of the above you should follow the detailed instructions in our documentation.

That’s it from the Freshpaint side. In Facebook, you also only have a few things to set up:

1. Go to ‘Audiences’ in your ads manager and choose ‘Create Audience.’
2. First create your ‘Custom Audience’ using the data that is being shared from Freshpaint. You can choose ‘Website’ and it will track the Pixel ID you shared with Freshpaint.
3. Once that has been created, create a Lookalike Audience.

1. Choose your custom audience from the Pixel as your source.
2. Select the Purchase event (which will actually be the event you mapped in the Transformation)
3. Select the region
4. Select the audience size. A smaller audience is likely to be higher value, but fewer people.
5. Select Create

And you’re done (bar all the ad design and copy and budgeting, but that’s all simple 🙃)

There are two variations you can do to this workflow using Freshpaint:

1. You can pull in data from other sources, such as customer relationship management (CRM) systems or marketing automation platforms, to create a more comprehensive profile of your ideal customer and build a more targeted lookalike audience. The better that initial ICP, the better your targeting will be.
2. There is also a HIPAA-compliant way to do this. With ID Masking, we can strip out identifying information and hide the identity of the user so you can safely build cohorts in Freshpaint and then send them to the ad platform.

Using current customers to find new customers

Using first party data to create lookalike audiences is an effective way to narrow targeting and find great customers. When you are using a CDP like Freshpaint you can easily share this data (or a masked version of this data) with ad platforms to create the audiences these platforms need to perform optimally.

Doing so will not just reduce your costs and improve efficiency, but it will just plain get you good customers. These customers are going to find the same value in your product as your currently successful ones and make it easier to grow your business.

If you’re interested in building better ad campaigns using first-party data or trying to do it in a HIPAA-compliant way, reach out to us to learn more about how Freshpaint can help.

‍