If reducing your cost per acquisition and bringing higher-quality leads into your funnel are part of your advertising goals, lookalike audiences (LAL audiences) are worth exploring.
In this article, we’ll explain what lookalike audiences are, how to build audience segments in Ortto and share them with ad platforms to build lookalike audiences, and best practices for lookalikes.
What is a lookalike audience?
A lookalike audience is an advertising targeting option available on platforms like Meta.
Rather than using a set of demographic and interest-based filters, lookalike audiences use a sophisticated algorithm to build a target audience with similar characteristics to your existing customers or community. This means targeting is less assumption-based, and more grounded in fact.
On most platforms, you can create a lookalike audience based on:
An existing custom audience that is uploaded via CSV or automatically synced via a customer data platform like Ortto
Conversion pixels you’ve created in the advertising platform in the past
People who like and follow your business page
Within each of these categories, you can choose the audience size, which will give you control over how ‘alike’ your lookalike audience really is. For example, if you are looking for a very similar audience, your available audience size will be smaller and you will likely pay more to reach that audience. The offshoot of that is a higher lead quality and less ad wastage.
Since ad platforms build lookalike audiences based on the data provided, it’s crucial to create a source audience that accurately reflects the kind of people you want to reach based on your goals. For more on how to do that, visit our blog on how to create audience segments for lookalike targeting.
How do ad platforms build their lookalike audience?
Most platforms are hush-hush when it comes to exactly how their algorithms work to build a lookalike audience, but many experts agree that the source audience provided is used to identify targeting criteria that may include data points like:
Purchasing habits
Interactions with content (likes, comments, shares)
Interests
Profile data including age, gender, and location
Page likes
Ads clicked
Pixel data
Activity in groups
With this information, the algorithm creates a net new audience that matches this criteria. This is why feeding the algorithm accurate source data is so important.
Benefits of using lookalike audiences
There are a whole host of benefits to using lookalike audiences, including:
Simplify your targeting
Lookalike audiences may seem like an advanced tactic, but they’re actually quite simple to use. Platforms like Facebook give you endless targeting capabilities which can become overwhelming. The algorithms used to create lookalike audiences remove the need to build an audience from scratch.
Save time
When your customer data platform and ads manager account are integrated, you can sync existing audience segments (like ‘customers’ or ‘users’) and use them to build your lookalike audience. This will save you hours that would otherwise be spent analyzing your audience to identify relevant characteristics and building audiences in your ad platforms.
Get higher quality leads
It’s easy to make assumptions about our target audience, even when the data is right there in front of us. The sophisticated algorithms used by ad platforms like Facebook will pick up on common characteristics that a human could miss. This means your target audience is based in fact, rather than assumptions, and you will be more likely to generate high-quality leads.
Share relevant content
If you use an existing audience segment or your following as source data for your lookalike audience, selecting relevant content for your ad creatives is simple. Look at the top-performing messaging and content pieces for the existing segment or your following, and use it to inspire the messaging, creative, and content shared in your ads.
Best practices for lookalike audiences
Before you send your lookalike campaigns off to the races, make sure you’re following these best practice guidelines.
1. Use the right custom audience for your goals
The trick here is to use high-quality data and sophisticated filters to drill down on the metrics that determine your best customers or audience members. If you’re not sure where to start, this blog includes five examples of audience segments that can be used to reach specific goals in your lookalike campaigns.
2. Test your audience size on Meta
A 1% match will be closely aligned to your first-party audience, where a 10% match will be less accurate, but will increase your potential reach exponentially. Your choice here will largely depend on how specific your product or offering is, how long your campaign is running, your overall budget, and your goals. For example, if you are a B2C or eCommerce company with a large addressable audience, big budget, and a long campaign period, optimizing for reach makes the most sense. If you are a B2B SaaS with a highly specific target market, you will be better off optimizing for similarity.
In Facebook’s advanced options, you can create up to 500 lookalike audiences from one source audience. This means you can test different targeting methods, like the ‘nesting’ strategy whereby you create a few segments (e.g. 1% most similar, 3% less similar) and optimize your bids accordingly. If you do test this, ensure you exclude your 1% audience from your 3% group to avoid audience duplication.
This strategy gives you the best of both worlds, allowing you to target a highly-similar audience with a higher bid, and a less-similar audience with a lower bid.
3. Keep your audiences up to date
Many marketing automation platforms (including Ortto) offer dynamic audience syncing, meaning you will not have to keep your lists up to date manually. If this is not the case, you need to ensure that the source audience you’ve uploaded (for example, as a CSV) is regularly refreshed In both cases, your lookalike audience will be updated dynamically every three to seven days.
4. Watch for audience overlap
Perhaps you built one audience segment of customers with a high CLV and another of customers who are subscribed to your newsletter. There’s a strong likelihood these two audiences will have some overlap, which means the same people could be in two different audience sets. This can become a problem if you have created specific messaging guides for each audience type — the individuals who are duplicated could receive conflicting or confusing messaging.
Meta Business Suite has a tool that will show you the audience overlap for your different audiences to help avoid this. In Ortto, you could also add a filter ‘Is a member of audience, is not newsletter subscriber’ to ensure your synced audiences do not overlap.
A powerful way to find your best customers
Lookalike audiences can be a powerful and easy-to-use tool for advertising. While the heavy lifting sits with the ad platform’s algorithms, it is crucial for marketers to do the upfront work of identifying a relevant and specific audience to use as source data. After all, an algorithm is only ever as good as the data it is fed.
Ortto makes this simple by allowing you to build highly specific custom audience segments and automatically sync them with your ad platforms. This ensures your audiences are always up to date and your leads can be quickly entered into relevant customer journeys.
