Open Web vs. Walled Gardens: Where Is AdTech Headed

First-party data – the information platforms collect directly from their own users – has become the defining asset of digital advertising. When Google in 2025 confirmed it would not deprecate third-party cookies in Chrome, it simultaneously retired most Privacy Sandbox APIs, because the proposed alternatives never gained meaningful industry adoption.

Retail media tells the same story: Amazon, Walmart, and Target have built their ad products entirely around first-party purchase data, and that segment is now the fastest-growing in the market, forecast at $69 billion in the US in 2026. Money follows first-party data.

That is why the question of where budgets go – platforms or open web – cannot be separated from the question of who controls the data behind them. And increasingly, the advertisers asking that question are reconsidering just how much control they are willing to give away.

This article looks at where the balance stands today, why it got here, and what is shifting in 2026.

What the Split Actually Looks Like Today

Walled gardens are closed platforms that control their own inventory, audience data, and measurement: Meta, Google, Amazon, TikTok, Apple. Advertisers can buy inside these ecosystems, but the underlying data never leaves the platform.

The Open Web is everything else: publisher inventory accessed through programmatic pipes, ad exchanges, Demand-Side Platforms (DSPs), and Self-Service Platforms (SSPs).

The numbers have leaned toward the gardens. In 2024, walled gardens accounted for around 71.5% of US programmatic digital display ad spending, with advertisers spending roughly $2.50 inside platforms for every $1 on the open web. Retail media networks (Amazon, Walmart, Target) are classified within this group and are growing faster than almost any other ad channel, with the US retail media spend forecast at $69 billion in 2026.

But the picture is changing. Walled gardens lost programmatic market share in 2023 for the first time since tracking began, and the open web has recovered ground each year since. The split is not static, and the reasons why are worth understanding.

Why Walled Gardens Attract Budget – And What They Trade Away

Walled gardens have captured large advertising budgets for understandable reasons. Meta, Google, and Amazon accumulated massive user bases through their consumer products – search, social, e-commerce, and the behavioral data that came with them. For advertisers with no existing data infrastructure, buying into these ecosystems offered a ready-made targeting solution.

But “ready-made” has a price. Everything that makes walled gardens feel easy also makes them a dependency:

• Targeting parameters are defined by the platform, not the advertiser. You reach the audiences the platform allows you to reach, described in the terms the platform chooses.
• Attribution is reported by the same entity selling the inventory. The numbers close cleanly because the platform decides what counts.
• The audiences the platform builds on top of your data – lookalikes, behavioral segments, engagement signals – belong to the platform, not to you. You can take your own customer list elsewhere, but the targeting intelligence built around it stays behind. And because the platform both generates and reports those signals, you can’t independently verify them.

This is the fundamental tension. Walled gardens offer a powerful shortcut to reach. But advertisers building on that foundation are building on infrastructure they do not own.

The Hidden Cost of Overexposure: A New Look at Frequency Capping

What the Open Web Actually Offers

The open web covers inventory, audiences, and capabilities that walled gardens simply cannot match.

• Scale and diversity that platforms cannot replicate. The open web reaches audiences across thousands of independent publishers, niche content environments, news sites, and specialist communities. Audiences who do not spend meaningful time on Meta or Google are reachable here and nowhere else. No single platform owns the full attention of the internet.
• Independent measurement is a feature. On the open web, a buyer working with the right partners can access impression-level data, audit supply paths, and build a measurement picture from direct observation rather than platform-reported numbers. A campaign running across multiple publishers and tracked through an independent measurement layer gives advertisers something that no walled garden can offer: an unbiased view of what their spend actually did. This requires more infrastructure than clicking “boost post,” but it produces a picture that the advertiser owns and controls.
• Data portability and audience ownership. When an open-web campaign runs, the signals it generates can stay with the advertiser. Segment-level insights, audience overlap data, performance signals by context and format – these are assets the advertiser can carry into the next campaign. In a walled garden, the equivalent signals stay with the platform.

The Data Control Question

The walled garden model is, at its core, a data control model. The platforms accumulated first-party data by building products people use every day. The advertising business monetizes that data. Advertisers benefit from the targeting it enables during the campaign, but they do not retain the data after the campaign ends.

This matters most when a brand’s own customer data is involved. Running a campaign that matches a brand’s CRM list against a platform’s user base requires uploading that list to the platform. The advertiser gains targeting reach; the platform gains a signal about the advertiser’s customers. 

The open web dynamic is fundamentally different. DSPs and data partners in the open ecosystem operate in a more distributed architecture. The advertiser has real choices about where their data goes, who they partner with, and what they retain after a campaign runs. For brands that treat first-party customer data as a strategic asset – and most enterprise advertisers now do, given the direction of privacy regulation – this distinction is a core part of the budget conversation.

There is also a dependency risk that walled garden concentration creates. 

Platform policy changes, algorithm updates, and auction dynamics are outside the advertiser’s control. A targeting capability that drives strong results today can be deprecated without notice. The iOS 14 update in 2021, which restricted app tracking and significantly degraded Meta’s targeting signal, showed clearly what it means to build an advertising strategy on infrastructure you do not control. Advertisers with diversified open-web programs were insulated. Those who had concentrated their budget in platform ecosystems were not.

Ad Tech Security for Brands and Advertisers: A Practical Checklist

Where the Balance Is Shifting

Three developments are actively reshaping this debate in favor of the open web.

• Data clean rooms have emerged as the primary technical response to the measurement gap that historically drove budget into walled gardens. A clean room allows two parties to match their datasets in a privacy-safe environment without either party seeing the other’s underlying user records. Retailers, publishers, and open-web buyers are increasingly offering clean room access as a standard feature of their ad products. Clean rooms, curation, and AI are now among the defining adtech trends of 2026. The measurement quality gap is closing – on the open web’s terms.
• Retail media fragmentation is creating a structural problem for the walled garden model. More closed ecosystems mean more fragmentation for advertisers who rely on platform-native measurement. Each new retail media network is another siloed reporting environment, another attribution methodology, another set of numbers that cannot be reconciled with the rest. The open web, by contrast, operates on shared infrastructure where a single independent measurement layer can work across environments.
• Privacy regulations are raising the bar on data practices across the industry. GDPR, evolving US state-level privacy laws, and the ongoing deprecation of third-party signals are forcing every participant in the ecosystem to demonstrate a defensible data infrastructure. The open web is adapting well: publishers who have invested in direct audience relationships and consent infrastructure are strengthening their inventory quality story, and the programmatic ecosystem is building toward first-party-compatible solutions that do not require surrendering data to a closed platform to function.

The direction is toward an ecosystem where data infrastructure, measurement quality, and regulatory compliance determine competitive position, and the open web is building on all three.

FAQ

What are walled gardens in advertising?

Walled gardens are closed digital platforms that control their own inventory, audience data, and measurement systems. The major examples are Meta, Google, Amazon, TikTok, and Apple. Advertisers can buy inside these ecosystems, but they cannot export the underlying audience data or verify delivery and attribution independently. The term contrasts with the open web, where inventory is bought through open programmatic infrastructure, measurement is independent, and more data is visible to buyers.

Why do advertisers keep spending in walled gardens if the data is locked in?

Because the targeting and operational simplicity inside these platforms is real – particularly for advertisers without existing data infrastructure. Walled gardens offer a ready-made path to reach at scale. The tradeoff is that the data powering those campaigns stays with the platform, measurement is platform-controlled, and the advertiser’s strategy becomes dependent on infrastructure they do not own or control.

Is the open web losing ground to walled gardens?

It was. Walled gardens grew their share of digital ad spend consistently through the 2010s and into the early 2020s. That trend reversed in 2023, with walled gardens losing programmatic market share for the first time. The open web has recovered ground each year since, driven by programmatic infrastructure improvements, the growth of clean room adoption, and advertiser diversification away from platform concentration risk.

What is a data clean room, and why does it matter?

A data clean room is a privacy-safe environment where two parties can match their datasets without either side seeing the other’s underlying user records. They matter because they bring first-party data precision to the open web without requiring advertisers to surrender their audience data to a closed platform. Publishers, retailers, and open-web buyers are increasingly offering clean room access as standard infrastructure in 2026, narrowing the measurement quality gap that once favored walled gardens.

How should advertisers think about splitting the budget between walled gardens and the open web?

The right split depends on category, audience, and measurement maturity – but the framing should be about what each environment returns in terms of data ownership and long-term strategic value. Advertisers who concentrate entirely in walled gardens are working with a picture the platform constructs and building on infrastructure they do not control. A diversified strategy that uses the open web for reach, transparency, and audience ownership – with platforms playing a supplementary role – gives advertisers something platforms cannot offer: a durable, portable, measurable advertising program they actually own.