What are Lookalike Audiences?
A Lookalike Audience is a targeting tool you'll find in several digital advertising platforms. It is based on machine learning that uses an existing group of users – often called a source audience – to find new people who resemble them. The source audience could be your current customers, people who have visited your website, or users who have previously interacted with your ads on social media.
The platform analyses a wealth of data points from your source audience. These can include gender, age, interests, location, purchasing behaviour, scrolling behaviour or even device type. Based on these signals, the system builds a mathematical model that finds new users across the platform's entire database who share the same characteristics. The better the quality of your source audience, the more precise the result. The same applies if the data is fresh and reflects users' current behaviour.
You can use Lookalike Audiences on platforms such as Meta (Facebook and Instagram), Google Ads, LinkedIn, HubSpot and TikTok. Although the method is essentially the same, each platform has its own algorithms and options for fine-tuning. This means you can work strategically with precision, reach and geographic targeting depending on the campaign's purpose.
How do you use Lookalike Audiences?
When you create a Lookalike Audience, you start by choosing your source audience. This could be a remarketing list, an email list taken from your CRM system or the users who have filled in a contact form. You then choose the area you want to target – for example Denmark or Scandinavia – and define how close a match you want to your source. Many platforms let you choose percentage levels from 1 to 10, where lower levels give higher precision and less reach, while higher levels scale up your visibility.
In other words, you can choose whether to work with a narrow audience with high conversion potential or a broader audience for branding and awareness. Small adjustments to the percentage have a big impact on performance, so it pays to experiment systematically. However, this requires a well-structured data foundation – typically from a CRM such as HubSpot or through pixel data from your website – and that you work in a disciplined way to keep your data consistent across channels such as social media marketing and Google Ads.
An example of how the platforms use data:
Why you should use Lookalike Audiences
The main purpose of Lookalike Audiences is to increase reach without losing relevance. You avoid wasting ad spend on overly broad segments, because the algorithm identifies users who resemble your best customers. This means you can scale your campaigns effectively while maintaining high conversion potential. This is particularly valuable when you want to expand your customer base or increase the number of leads without significantly changing your existing strategy.
A Lookalike Audience therefore complements both your SEO work and your social ads. While SEO works on organic traffic in the long term, Lookalike Audiences help you scale quickly in paid channels. The combination gives you data you can use for better strategy and budget allocation across your digital channels.
What types and variants are there?
Although the concept is the same, there are different versions depending on the platform. Meta works with precision percentages from 1 to 10, where 1% corresponds to a very narrow match group and is used for performance campaigns, while 10% covers a broad reach segment. In Denmark, 1% typically corresponds to around 48,000 users, while 10% can cover close to half a million. Google Ads uses Similar Segments, which are generated automatically from your existing remarketing lists in campaign types such as Demand Gen. In HubSpot, you can build Lookalike Lists or Segments based on CRM data, website visitors or previous campaign engagement, while LinkedIn and TikTok offer similar features aimed at B2B and lifestyle segments respectively.
Example of combining platforms:
How do you use Lookalike Audiences in practice?
In practice, it's about letting the data work for you. Start with a high-quality source – this could be a list of people who have completed a purchase or sent a qualified enquiry. Check that your data is up to date and that you comply with current GDPR rules on consent and data storage. Then build lookalike segments at different percentage levels so you can test performance across campaign types. For smaller budgets in Denmark, 1–3% typically gives the best results.
You can use the new audiences in several contexts. In social media marketing, you can test ad formats such as video or images, while in SEM campaigns you can use the segment to optimise display targeting in Search or Demand Gen. Whatever the channel, you should continuously evaluate the segment's performance using metrics such as CTR, CPC and conversion rate, and then adjust your strategy so you're always working towards the best-performing data foundation.
What should you be aware of?
A Lookalike Audience is only as good as the data you feed it. If your source contains users who don't represent your core business, the algorithm learns from the wrong data and your ad effectiveness drops. You should therefore always keep your source audiences up to date and segment them by actual value, such as purchases, leads or interaction with specific pages. Also avoid overlap between your existing remarketing and lookalike segments, so your campaigns don't compete with each other in the auction.
Finally, it's important to evaluate how Lookalike Audiences work alongside the rest of your digital strategy. Used wisely, they can be an effective way to find new customers while you continue to build on your existing performance. This is exactly where you get the most value from combining technical data expertise with strategic marketing insight in everything from design and brand strategy to paid social and Google Ads.
