A walled garden is an advertising platform that keeps its users, its data and its measurement inside its own system. Google, Meta, Amazon, TikTok and Apple are the usual examples: you can buy ads and see results, but you cannot take the underlying user-level data away or check it independently.
How walled gardens work
Each of these companies has large numbers of logged-in users, which lets it recognise the same person across devices and services without relying on third-party cookies. That identity data is the platform’s main asset, so it stays inside the walls. Advertisers get three things:
- Targeting built on the platform’s own knowledge, such as interests, purchase history or search behaviour.
- Reporting in aggregate, showing clicks, conversions and spend, measured by the platform itself.
- Limited data exchange Where you can send your own customer data in, through tools such as a Conversions API or customer lists, but get only summaries back.
Some platforms offer data clean rooms, such as Google’s Ads Data Hub or Amazon Marketing Cloud, where larger advertisers can run analysis on combined data under strict privacy rules. Even there, individual records do not leave the platform.
Privacy changes have made the walls higher. Since Apple introduced App Tracking Transparency in 2021, iPhone apps must ask before tracking people across other companies’ apps and websites, and many people decline. Platforms with large logged-in audiences cope with this better than independent ad networks, because they can still observe what people do on their own services. That is part of why the biggest platforms have kept such a strong hold on advertising budgets.
Why it matters
The biggest consequence is that each platform marks its own homework. Google measures Google Ads, Meta measures Meta ads, and each uses its own attribution rules. Because neither can see the other’s activity, both can claim the same sale. Add the reported conversions from three walled gardens together and the total often exceeds the orders a business actually received.
It also affects how dependent you become. A Manchester retailer that sells mainly through Amazon and finds customers mainly through Meta ads holds very little information about who its customers are. If costs rise or an account is suspended, there is no audience to fall back on.
For UK businesses, walled gardens are not exempt from the rules. When you upload a customer list or send conversion data, you are still the controller of that personal data under UK GDPR, and you need a lawful basis, clear privacy information and, for the tracking tags, consent under PECR.
Common mistakes
- Trusting each platform’s reported return without comparing it with actual sales.
- Summing conversions across platforms to judge total marketing performance.
- Building the whole business on rented audiences, with no email list or customer data of your own.
- Uploading customer lists to ad platforms without checking that your privacy notice covers it.
How to act on it
Keep one independent source of truth for results, usually your shop platform, CRM or accounts, and judge every channel against it. Use GA4 as a neutral referee where it helps, knowing it has gaps of its own.
Build assets you own alongside the platforms: first-party data collected with consent, an email list, and organic search visibility. Where spend is large enough, test incrementality by pausing a channel in some areas and watching what happens to total sales.
Deciding how much to rely on each platform, and how to measure them fairly against one another, is central to the digital marketing strategy work I do with UK businesses.