Lookalike: definition and best practices
5min • Jul 12, 2024
Alexandra Augusti
Strategy & Operations Manager
Looking for effective ways to boost your website traffic, enhance your conversion rate, and acquire new customers? You've probably heard of the "lookalike" audience strategy, which relies on similarities between target groups.
But what exactly is a lookalike audience?
👉🏼 In this article, we'll explain what a lookalike audience is and provide our best tips and practices for making it a winning acquisition strategy.
What is a Lookalike Audience?
Defining lookalike
The term "lookalike," or similar audience, refers to a cornerstone strategy in digital marketing, aimed at identifying and attracting prospects who have profiles similar to those of your regular customers.
For a business, it involves selecting a group of its best consumers, known as the initial segment or source audience. The advertising platform (Meta Ads, Google As, TikTok Ads, etc.) then takes over, meticulously analyzing the attributes and behaviors of these customers—from demographic data to passions and consumption habits. By leveraging its advanced algorithms, the platform detects users with similar profiles, thus creating what we call a lookalike audience. This new audience then becomes the new target for advertising, with the hope that they will respond positively and similarly to the original audience.
Lookalike: how it works
đź’ˇ Check our best audience strategies for Meta Ads and Google Ads if your want to learn more on that subject
An example of lookalike strategy usage
Consider a company that sells books online and wants to boost its online sales. This company relies on its customer database, who regularly make purchases on its website.
Its strategy? To build a lookalike audience from this database, using Facebook as an advertising channel.
The company might try to find socio-demographic attributes (age, gender, geographic area, etc.) common to its best customers. However, looking at its best customers, there are conflicting signals: teenager vs. elderly person, living alone or in a household of 5, residing in Amiens vs. in Marseille, etc. Apparently, they are opposites. It is thus complicated for the company to establish its own system of “similarity.”
Facebook takes on the task of decrypting the profiles of these top customers to find similarities. Facebook can detect identical behaviors among individuals (same sites visited, same number of articles viewed, etc.) to define a “typical” profile. Facebook can then identify within its network profiles that share behavioral or socio-demographic similarities. Facebook creates an "affinity ranking," determining for each user whether they are likely to purchase or not.
By using Facebook’s lookalike audience, the company significantly enhances its chances of targeting qualified prospects with a high purchasing potential.
How to Implement a Lookalike Strategy?
The 3 Steps of a Lookalike Strategy
Developing a lookalike strategy necessarily involves these three key steps:
Step 1 : Identify the reference source for your audience. This step involves selecting the type of data used to forge your lookalike audience. You can choose to use a list of your existing customers (segmented or not), visitors to your website (tracked by a platform's tag), or people who have interacted with your brand on social media. Our recommendation is to use customer / lead segments (who have interacted with your brand), based on first-party data.
Step 2 : Build your reference audience. It is essential to segment your source to create a customized audience. You must provide your chosen advertising platform with a compilation of your most qualified users. This audience can be imported manually via CSV, tracked via Pixel/tag, or imported automatically by tools using a Reverse ETL process.
Illustration of Reverse ETL process
đź’ˇ We recommend building multiple reference audiences if you plan to conduct several lookalike campaigns, aiming to have the most refined audience possible. The platform then proceeds to analyze the characteristics and behaviors of this audience, including demographic data, interests, consumption habits, etc.
Step 3 : Generate the lookalike audience directly on the advertising platforms. This step involves asking the advertising platform to identify users who share similarities with your base audience. You have the freedom to determine the desired level of resemblance, often expressed as a percentage. A higher percentage translates to a broader but less precise lookalike audience. You can also specify the country or region targeted for your ad distribution.
Tips and Best Practices to Optimize Your Lookalike Strategy
Here are some strategic tips and best practices to maximize the results of your lookalike strategy:
Minimize overlaps between audiences When a user is present in multiple audiences used in different campaigns, the advertising platform generally opts to show them only one type of advertisement (from the globally most effective campaign). However, the most effective campaign may not necessarily be the most suitable for that user! To avoid overlaps, consider excluding your reference audience or using tools like the Facebook Audience Network or Google Display Network to examine your audience overlap.
A/B test with different audience sizes. Test the influence of your lookalike audience's size on the performance of your marketing efforts. Adopt a nested approach, consisting of developing several lookalike audiences of varying similarities, such as 1%, 2%, 3%, etc., and then compare the results in terms of coverage, clicks, conversions, etc.
Create lookalikes based on different segmentations. This involves designing lookalike audiences based on different criteria, such as the type of product purchased, the amount spent, purchase frequency, and more. This method refines your targeting, aligning your advertising messages to the needs and expectations of your prospects. You can also choose to have multiple reference audiences, based on different sources (e.g., recent buyers vs. web visitors).
Regularly renew your reference audiences. Make sure to regularly update your initial audiences, especially if the turnover of the audience is high. It is essential to remove users who have become less relevant to your business and incorporate new users matching the segmentation criteria. This ensures the relevance and efficiency of your lookalike audiences. To achieve this, we recommend using tools that allow for always up-to-date segmentation in advertising platforms, such as Reverse ETL or Customer Data Platforms.
The Impact of the End of Third-Party Cookies on Lookalike
Google's announcement of the phase-out of third-party cookies by early 2025 significantly impacts lookalike targeting strategies. Currently, third-party cookies are a central method for collecting user information across various sites. Without third-party cookies, it becomes challenging for advertising platforms to track users and understand their preferences.
Third-party cookies are used to track online behavior and retarget customers
In response to this evolution, how can you adapt effectively?
Let's explore a few adaptable strategies:
(Essential) Prioritize first-party data. This data, collected directly from your users with their consent, is of much higher quality and much more secure in terms of privacy than third-party data. Its use allows the creation of lookalike audiences based on criteria specific to your company, such as purchasing behavior, sign-ups, or interactions.
👉🏼 Learn more about first-party data and its importance
Increase the size of your lookalike audience. This method involves adjusting the similarity level between your base audience and the targeted one to include a broader spectrum of potential users. This could compensate for the decreased precision caused by the absence of third-party cookies. However, it is crucial to remain vigilant not to excessively expand the audience, which could lead to a loss of relevance and effectiveness.
Continuously update your audiences. This involves frequently renewing your audiences to reflect changes in user behaviors and preferences. Such an approach ensures the relevance, timeliness of data, and optimizes the results of your lookalike campaigns.
In summary, the end of third-party cookies does not mean the end of lookalike strategies but rather calls for a profound revision of the strategies and tools used.
Conclusion
Lookalike is a crucial strategy in the field of digital marketing, enabling precise targeting of prospects who share common traits with your existing customers.
However, the effectiveness of these strategies is threatened by the end of third-party cookies. To continue performing well, it is now essential to automatically send first-party data to platforms, to continue feeding them with quality data.
If you want to automatically transmit your audiences to all your platforms for your lookalike campaigns (and more!), don't hesitate to contact us.
