What Is a Custom Audience?
A custom audience is a targeting feature offered by advertising platforms like Meta Ads Manager and Google Ads that lets you use data you already own to reach specific, identified people with your paid campaigns. Rather than relying on the platform to guess who might be interested in your product based on broad interest categories, you provide a list of real people who have already had some connection to your business.
There are several ways to build a custom audience. The most direct method is uploading a customer list, a CSV file containing email addresses, phone numbers, or other identifiers from your CRM or loyalty programme. The platform hashes these identifiers and attempts to match them to user accounts. A second method uses your website pixel or Google tag to automatically build audiences based on site behaviour, such as everyone who visited a specific page in the last 30 days, or people who added a product to cart but did not purchase. A third method targets people who have engaged with your content on the platform itself, such as video viewers or Instagram page visitors.
Custom audiences are the foundation of effective retargeting and are also used as the seed for building lookalike audiences. Their power lies in precision: instead of wasting spend on people who have never heard of you, every rand goes towards reaching someone with a documented connection to your brand. For South African businesses where advertising budgets are often under R20,000 per month, this efficiency is not optional; it is what separates profitable campaigns from wasteful ones.
POPIA (the Protection of Personal Information Act) governs how South African businesses collect and use personal data. When building custom audiences from customer lists, your privacy policy should clearly disclose that you may use customer data for advertising purposes, including sharing with platforms like Meta and Google for matching. Getting this right protects your business and builds customer trust.
Custom Audience In Practice
The scenario below is an illustrative example, not a Juicy Designs client result. The figures indicate the scale of effect that custom audience targeting typically produces, so treat them as indicative rather than measured.
Picture a Pretoria gym with around 2,000 active members that exports its member list from its management software and uploads it to Meta Business Manager. Meta would typically match something in the region of 1,200 of those members to active Facebook and Instagram accounts. The gym could then create a campaign specifically for this custom audience promoting an upgraded membership package, knowing every person who sees the ad is already a paying member who understands the brand. The campaign would not need to explain what the gym is or why anyone should join. The messaging could go straight to the upgrade benefit and the price, which would typically produce a much lower cost per conversion than any cold-traffic campaign could achieve.
From the same custom audience, the gym could create a 1% lookalike audience to find South Africans who share characteristics with its members, giving it a qualified prospecting pool for new member acquisition campaigns. This combination, custom audience for retention and upsell, lookalike for acquisition, is the core of a modern paid social strategy for any South African business with an existing customer base.
Types of custom audience
A custom audience is a group you define from your own data to target with ads, and it comes in a few forms. Customer-list audiences are built by uploading contact details, hashed for privacy, to match against platform users, useful for reaching or excluding existing customers. Website audiences are built from visitors captured by a pixel or tag, such as everyone who viewed a product or abandoned a cart. Engagement audiences are built from people who interacted with your content on a platform, a video, a profile, a form. App audiences are built from app activity. Each turns first-party data or on-platform behaviour into a targetable group, which is the foundation of both remarketing and, through lookalikes, prospecting.
Custom audiences and lookalikes
Custom audiences do double duty: they let you reach people who already know you, and they seed lookalike audiences that find new people who resemble them. A platform analyses the traits of a strong custom audience, your best customers, say, and builds a larger audience of similar users you have never reached, extending targeting from retention into prospecting. The quality of the source audience shapes the result: a lookalike built from high-value customers finds better prospects than one built from all visitors. As third-party tracking fades, this first-party-data-driven approach, building custom audiences from your own data and expanding them with lookalikes, has become central to efficient advertising, which is why capturing your own audience data matters more than before.
FAQ
Is it safe to upload your customer list to Meta or Google?
Both Meta and Google hash customer data before using it for matching, meaning your raw email addresses and phone numbers are encrypted before upload and never stored in plain text. Ensure your privacy policy discloses that you share data with advertising platforms, as required by POPIA in South Africa and standard platform terms of service.
What match rate should you expect for a custom audience upload?
A typical customer list upload achieves a 40% to 60% match rate, meaning the platform finds accounts for roughly half your uploaded contacts. Match rates improve when you include multiple identifiers like email, phone number, and full name. South African mobile numbers formatted with the country code (+27) tend to match more reliably than local formats.
Is it safe to upload a customer list to Meta or Google?
The platforms hash contact details before matching, so the raw list is not exposed to other advertisers, and reputable platforms use it only to match against their users. You remain responsible under privacy law for having a lawful basis to use customer data this way, so consent and notice still apply.
What is a lookalike audience built from a custom audience?
A new, larger audience the platform builds by finding users who resemble a source custom audience, such as your best customers. It extends targeting from people who know you to similar prospects who do not, and its quality depends on how strong and well-defined the source audience is.