How to Micro-target Facebook Audiences Using Segment Personas

Facebook and Instagram ad costs are rising rapidly. There’s a trick to fight against this: micro-targeted audiences. This is something you can do within minutes using Segment. That’s what this post walks you through.

First, what’s a micro-targeted audience? It’s segmenting out your existing customer list by some demographic criteria or behavior they’ve taken in your apps, sites, or even offline interactions.

How can they drastically improve paid ad performance? Consider this: If you have 10 ad sets running — each to its own micro-audience — as opposed to one broad ad set targeting all your users, you can then tailor your ads and corresponding landing pages to each segment. Meaning, you can serve B2B customers an ad that appeals best to them, young women another ad, high-value past customers another, and so on.

Micro-segmented targeting, ads, and landing pages are a quick way to better optimize your paid acquisition funnels. And it’s easy to do.

That’s where Segment comes in. We’ll also show you how to use Segments audience segmenting feature to reduce churn and improve email marketing conversion rates. It’s kind of magical.

The bad way to do this

Traditionally, growth marketers upload a customer list to Facebook to create a Lookalike audience. They run Facebook and Instagram ads against it and call it a day.

That’s okay when you’re starting out, but it doesn’t scale efficiently with your growing customer base and product offerings. And it doesn’t take into consideration the differentiation among your leads and users. Not to mention the impracticality of constantly uploading new CSV’s to keep Lookalike audience up-to-date.

The right way to do this

In this tutorial, we’ll setup much better performing micro-segmented audiences (and Lookalikes) using Segment Personas. It works automatically, and Segment Personas is the pretty UI on top of it all. Segment Personas allows you to:

• Ingest your lead and user behavioral data from everywhere — across all devices, business tools, and channels. Segment soaks up everything people do with your business and cleanly centralizes it. It’s super easy to set up.
• With a drag-and-drop interface, you can segment your audience and leads any you want. Age, purchasing behavior, visit frequency, industry — anything.
• Then you can automatically sync these audiences to platforms like Facebook. They’ll be kept up-to-date in real-time.
• With this system in place, you can run better performing ads and email campaigns.
  

Almost every company we train at Demand Curve eventually adopts Segment because doing growth without these features is doing growth the silly old school way: treating every user the same.

Want help running facebook and instagram ads the right way? Our agency has grown 100+ startups with paid ads.

Quickly setting this up

Even if you’re not already using Segment, we recommend reading this post in full as it provides useful context plus additional growth marketing tactics for significantly reducing Facebook and Instagram cost of acquisition. Plus, pretty images..

Step 1: Authorize Facebook Custom Audiences

Pop into Segment to authorize Facebook Custom Audiences on your account.

The audiences you create using Segment Personas will sync into Facebook as a User-List. Segment keeps this updated in real-time. Now you can run ad campaigns in Facebook and Instagram without having to bug your data or engineering team for a manually segmented CSV export of users.

(Full detailed instructions are here. Read it later.)

Step 2: Build segmented audiences in Segment

Next, use Segment’s UI to segment users any way you want.

You can segment audiences based on the events they take on your app or site, their traits, and your pre-defined audiences. It’s an enormous pain to do this without Segment.

Let’s super quickly break down how each of those segmentation types can decrease your cost of acquisition on Facebook and Instagram.

2.1: Build a custom audience

Events

You can segment an audience from any of the events Segment detects in your app, site, or email activity. Anything users do you, you can collect that data automatically with Segment.

For example, you might want to segment users who’ve viewed one of your products above a certain price threshold — but not yet completed an order. That lets you run ads to them reminding them to check out.

In Segment’s Personas dashboard, you can do that in seconds:

You can use this same segmentation capability to hyper-personalize your marketing emails — not just your ads. (We cover email in this post.)

One more example. You can filter on a strict order of events within specific time windows. Get as granular as you want to serve users ads and emails at exactly the right time with exactly the right messaging needing to move them to the next step in your funnel:

‍

Here are two more data points you can segment an audience on.

Custom Traits

Custom traits are the extra pieces of information you collect (i.e. their industry, the company’s number of employees, etc) when users complete actions — like completing forms, signing up, talking to sales or support, and purchasing products.

Computed Traits

You can use computed traits to calculate special variables on the fly. For example, you can create an audience of big_spender customers that exceed a certain payment threshold.

Custom audiences

If you’re a B2B company, you might want to upsell your existing accounts. Segment makes automating that much easier. For example, you might want to re-engage a list of at-risk accounts defined as companies on a business tier plan where none of the users in that account have logged in recently. That’s the type of segment that can help you significantly reduce churn.

2.2: Review and create your audience

Once you’ve created your custom audience, click Preview Results to double-check if it makes sense.

Some key growth marketing insights

With your segmented audience, you can run ads (or emails) directly to those users, or you can build lookalikes off of them. The latter is important: Building a giant lookalike off of all your users isn’t helpful if your users have very different behaviors. Consider the case of an ecommerce store that sells to both men and women who shop at different price points. I can see at least four audiences that should be segmented out that should each lead to their own lookalike.

Other criteria for segmenting out Lookalikes: people who purchase from distinct product categories, the customer’s business industry and size, number of repeat purchases, and so on.

Speaking of business industry and size, you can use Segment to automatically source that information for you. All you need is a user or lead’s business email: Segment supports integrations like Clearbit Enrich that automatically source firmographic information so you can segment on their job title, location, and industry — even if you never got that data directly from your customer.

And, finally, Segment Personas makes your retargeting ads perform better too. For example:

• For eCommerce companies, target people who abandon carts and know the exact items they left in there. Run ads with the corresponding imagery.
• Retarget users with different messaging based on how far into your funnel they got: Did they full set up their account? Did they only make one purchase? Did they churn? Push them to the next step.
  

Micro-segmented audiences are the future. Segment is how we tell everyone to do it.