If you have a business and you advertise online, you probably hope to reach new customers.
The only way for a business to grow is to win more customers, and Facebook is the most popular and best advertising platform for reaching new customers (together with Google).
The problem with targeting unfamiliar audiences is that you don’t know them, and therefore have no idea what are your chances to drive conversions.
In order to save time and money, you have to find methods to make your full-funnel targeting strategy more efficient. Luckily, Facebook has been working hard for you and has provided you with an intelligent capability for improving the targeting precision of prospecting customers: Facebook Lookalike Audiences.
A lookalike audience provides you with an opportunity to reach prospective customers, who share similar characteristics with your existing ones or with people who showed interest in your brand.
In order to create it, you first need to choose a source audience that will be the seed audience to lookalike from. Typically, this would be a custom audience created based on Facebook pixel data (website or app via SDK) or from your email list. Facebook then identifies the common qualities of the users included in your custom audience, like demographic information, interests, groups, friends lists, and other historical data, and creates a lookalike audience: a collection of users sharing those common characteristics.
You can define the size of your lookalike audience. While making this decision, you need to keep in mind that although creating a larger audience increases your potential reach, it decreases its level of similarity to your source audience.
Dictating the size of your lookalike is done by defining a percentage of the population in your target country. In Facebook Ads Manager you can create a lookalike audience of up to 10%.
Currently, Facebook lets you select countries to include when creating your lookalike audience, but this might change in the near future. Note that Facebook requires that your source audience includes at least 100 people from the same country.
In order to get the best results possible, it is recommended that you use a source audience that includes between 500-50,000 people. According to Facebook, people in your source audience will be excluded from your lookalike.
Lookalike audiences are updated every three to seven days, so they automatically keep up with the changes made in their source audience.
Do you want to reach the people who will most likely be your next customers? If the answer is yes, you should use Facebook Lookalike Audiences – it’s that simple!
Why are they more useful than age-, gender-, or interest-based targeting? First of all, Facebook interests might be outdated or incorrect because of a misinterpretation of a single data point.
Check the interests that Facebook defined for you and see for yourself:
Furthermore, your assumptions concerning your potential customers might be inaccurate.
Let’s say you’re selling lingerie, and you think your best audience is women between the ages of 20-50, but your top purchasers are actually men who buy your products as a gift for their girlfriends or wives. Lookalike audiences are made to find these unexpected commonalities among your current customers and create a more accurate profile of your prospecting ones.
Targeting unexpected commonalities will help you customize your creatives and tailoring content to audiences, which can lead to a major uplift in conversions.
Lookalike audiences are especially useful when expanding your business to new and unfamiliar territories. You can leverage customers’ data from your current market to create a lookalike for your target region/country either using the default lookalike creation or the international lookalike creation via the API. That can be a very good starting point if you don’t yet know which interests to target in a new country.
Although Facebook doesn’t explain how exactly they create lookalike audiences, Facebook advertising experts assume the data used includes:
With the Facebook pixel installed on millions of websites and all the other parameters mentioned above, it’s easy to see that Facebook has an enormous amount of data it can analyze when creating a lookalike audience, which makes it a very reliable solution.
In the last couple of years, lookalike audiences have proven to target people who are more likely to subscribe to your email list, and about 90% of the time, average a higher CTR than other audiences.
But the greatest advantage of lookalike audiences is simply building a larger audience of customers for you, letting Facebook better predict who’s more likely to buy, subscribe, and convert. As your audience grows, so does your ability to build a bigger audience.
There are many kinds of Facebook lookalike audiences available, and you should try as many as you can in order to find the ones who drive the best performance for your business.
This is a relatively new kind of lookalike audience that is meant to help you reach people who resemble your current most valuable customers.
You can use an arbitrary number to explain Facebook how much a customer is worth for your business in terms of lifetime value, and Facebook’s algorithms will locate similar users. In order to do so, add a customer lifetime value column to a list of customers, upload it (.txt or csv.) to Facebook, and define the value column.
A good example of a solid value-based audience to lookalike from is your top spenders. It tends to produce more website purchases and a higher return on ad spend. It is recommended that you use your current highest spenders, so make sure you filter your data to include people who have purchased within the last six months, for example.
Alternatively, you can choose your Facebook pixel, a mobile app, or an SDK as a value-based source to automatically create the lookalike audience. Your pixel, for example, captures and updates every purchase event on your website in real time. Having the data concerning all the purchases on your website allows your pixel to value your customers’ worth for your business and rank them accordingly.
A good way to identify people who are really interested in your brand is limiting your audience to people who watched at least 75% of your video. Unfortunately, it also includes the ones who skipped to the 75% point in your video’s length. Depending on your market and business, 50% and 95% viewing metrics may also be worth split testing.
This 75% viewing metric enables you to qualify a cold audience and use them to build a high-performing lookalike audience. This audience is a useful solution if you don’t have enough bottom-of-the-funnel events to use for lookalikes yet.
This is one of the best ways to double down on your brand’s success. It’s a lookalike audience based on a website custom audience of people who have completed a specific conversion event: purchase, add to cart, initiate checkout, lead, subscribe, and so on. Before you can create a custom website audience, you need to install the Facebook pixel and let it track this specific type of conversion.
Email list based lookalike audiences tend to deliver the best results because it’s almost like cloning your existing customer list. Furthermore, with this method you can unlock unlimited historical data, compared to the limited time range available when creating conversion lookalike audiences.
Although this kind of lookalike audience usually performs very well, it also tends to wear out its effectiveness and miss KPIs after a while. If you want to avoid it, you better upload new lists and create new lookalikes regularly – once a month or quarterly. This way, Facebook’s algorithm will be up-to-date with your relevant customer’s profile.
A lookalike audience based on your page likes is the easiest one to create. It’s probably the most straightforward way to leverage your page’s success and find people who are similar to your Facebook fans.
This audience is so simple to set up, you don’t even need to create a custom audience to lookalike from. All you need to do is choose your page name as the source for your lookalike audience.
This source audience is unfortunately often overlooked by advertisers, although it contains high-quality “warm” website visitors.
The common strategy concerning website visitors lookalikes is creating a source audience based on people who visited your website or specific website pages within a certain period of time, commonly 30 days. Although the Time Spent parameter is perhaps the least used, it can be very efficient as part of a retargeting strategy.
The Time Spent parameter is especially efficient when paired with specifying your product pages. For example, creating a custom audience of the top 25% of visitors by time spent on one of your product pages will provide you with an audience that has high potential because it’s sizable but specific.
Please note that for this to work, you must have the Facebook pixel installed and activated on your product pages.
Lookalike audiences are currently the best way to scale your campaigns and quickly and easily reach prospective customers. With the large amount of information Facebook gathers on its users, it’s hard to beat its algorithm when it comes to targeting new customers. By combining those lookalike audiences you can achieve multiple objectives, such as increasing reach, sales, conversions, and traffic.
Read more about other creative ways to target your customers and discover valuable insights.
Yahav Hartman is the CEO and Co-Founder of Madgicx, an all-in-one ad-buying platform for Facebook and Google Ads. Yahav has also been developing innovative solutions in Facebook advertising for over 8 years.
YouTube is the #1 place to share video content, hands down.
It seems like it’s harder than ever to capture the attention of buyers.
Customers are looking for your business online, are you making it easy to find what they need?
YouTube is the #1 place to share video content, hands down.
It seems like it’s harder than ever to capture the attention of buyers.
Never miss a post.
Subscribe to keep your fingers on the tech pulse.