Walled gardens: should you really invest your entire budget in Google and Meta?

📝 In summary

• Google, Meta, and Amazon are walled gardens that are highly effective thanks to their data and algorithms.
• Their efficiency comes with dependency, a lack of transparency, and data silos.
• Avoid concentrating all your investments : diversifying your media channels is crucial.
• Regaining control of your data (first-party, CRM, tracking) allows for better campaign management.
• The challenge is to find the right balance between performance and control.

‍

In 2025, the digital advertising market in France grew by 11%.

However, this growth remains highly concentrated: 76% of investments and 83% of growth are captured by just 8 companies, including Google, Meta, and Amazon (35th E-pub Observation, SRI and UDECAM).

These platforms now dominate marketing strategies. This dominance is based on a very specific model: that of walled gardens.

‍

Meta, Google… are walled gardens, what does that mean?

A walled garden refers to a closed ecosystem in which a platform controls:

• the user data
• the advertising formats
• the distribution rules
• and often the measurement tools

👉 In practical terms, everything happens within the platform, without transparent access to the outside.

This is the case for Google (Search, YouTube), Meta (Facebook, Instagram), and even Amazon.

These players possess a complete and proprietary view of the user journey within their environment.

‍

Walled gardens: Why do these platforms overperform?

It is no coincidence that walled gardens capture so much investment.

1. A massive accumulation of first-party data

Platforms like Google or Meta continuously collect data from billions of interactions: browsing behaviors, intentions, and purchase or content consumption histories.

👉 This exclusive and constantly enriched first-party data allows them to build an extremely nuanced understanding of users, far beyond what is possible on the open web.

2. Ultra-high-performance optimization algorithms

This wealth of data powers advanced optimization models capable of adjusting targeting, bidding, and campaign delivery in real time.

Advertisers thus benefit from:

• highly precise targeting
• effective retargeting
• automated performance optimization

👉 Result: the ability to quickly generate conversions, making these platforms virtually essential for any acquisition strategy.

3. A fully controlled closed loop

Ultimately, their true advantage lies in their closed-loopoperation. Ad delivery, performance measurement, and optimization are all integrated within a single environment.

👉 This significantly limits data loss between tools and enables faster, more effective continuous optimization.

Furthermore, this controlled ecosystem reduces the risk of fraud and ensures a managed inventory, providing a more secure environment for brands in terms of brand safety.

The limitations and challenges of walled gardens…

The impressive performance of walled gardens does, however, have a downside.

1. An imbalance between investment and usage

Advertising investments do not always reflect the actual time spent by users. Studies show a disconnect between the share of time spent on these platforms and the investments made in them.

In France, GAFAM platforms captured approximately 76% of digital advertising spend in 2025. Yet, usage data shows a more limited concentration of attention: according to Médiamétrie, they accounted for 36% of the total time French people spent on the Internet.

For comparison, in the UK, Ofcom estimates that by 2025, Google and Meta will account for approximately 51% of online time, which is still lower than their share of advertising spend.

This discrepancy, also documented in the work of the CNIL, highlights an over-representation of investment in dominant platforms, relative to the actual attention captured.

👉 In other words: budgets are concentrated where it is easiest to buy… not necessarily where attention is highest.

2. Data silos

Each platform operates by its own rules, making it difficult to reconstruct a comprehensive and consistent customer journey.

👉 The result: a fragmented view of performance, which limits the ability of advertisers to effectively manage their overall strategy.

3. High dependency

By concentrating their investments on a few platforms, advertisers gradually become dependent on these ecosystems.

Data is rarely or never exportable, and the relationship with the user is largely mediated by the platform.

👉 Ultimately, this reduces the ability to activate one's own marketing levers independently.

4. A lack of transparency

Measurement remains largely opaque. KPIs are calculated using proprietary methodologies that are difficult to compare or audit.

👉 This makes performance difficult to compare, audit, and sometimes interpret with complete objectivity.

‍

…amplified by AI

In this new phase of walled gardens, platforms are going even further by offering campaigns generated from A to Z by their proprietary AI. On the Googleside, the conversational experience in Google Ads, now powered by Gemini, already allows for the creation of a complete Search campaign (keywords, ads, assets) from a simple URL and business objectives. On the Metaside, the stated goal by the end of 2026 is to allow brands to simply provide a product image and a budget, with the AI automatically generating visuals, videos, text, targeting, and budget recommendations across Facebook and Instagram.

The promise is attractive: massive time savings, instant reporting and recommendations, and optimized performance. But it comes with a twofold strengthening of walled gardens:

• on one hand, optimization remains focused on platform KPIs (clicks, conversions, CPA) rather than customer quality or brand image;
• on the other hand, control and transparency decrease as targeting, budget allocation, and creative are absorbed into algorithmic black boxes, making dependence on walled gardens even stronger… and now amplified by AI.

‍

Solutions

Faced with these limitations, the challenge is not to oppose walled gardens, but to better structure their use within a global strategy.

1. Diversify your channels

Today, these environments naturally attract investment due to their simplicity and performance. However, other levers exist—particularly on the open web (programmatic, DOOH, CTV)—that provide access to different formats, alternative attention contexts, and complementary audiences.

👉 This diversification not only helps spread risk, but also allows you to regain some control over your media activations, by partially breaking free from platform-specific logic.

2. Deploy a data-driven strategy

In an ecosystem where platforms control access to data, developing your own data is essential.

Implementing a strategy based on first-party data (CRM, tracking, unique identifiers) allows you to better understand your audiences and manage your campaigns with greater precision.

👉 By structuring data identification and activation frameworks, our clients have achieved significant performance gains, reaching up to +52% ROAS.

Beyond performance, the goal is clear: take back control of your marketing strategy, rather than delegating it entirely to the platforms.

‍

Conclusion

Walled gardens are currently the most effective environments for generating quick performance.

But this efficiency comes at a cost.

👉 The more platforms simplify media buying, the more they consolidate control.

👉 The better they perform, the more they reinforce dependency.

The arrival of AI only accelerates this dynamic, making systems even more powerful… but also more opaque.

Investing heavily in Meta or Google is not a mistake. In most cases, it is actually a logical decision.

However, a strategy focused solely on these platforms remains incomplete.

👉 The real challenge today is not choosing between walled gardens and the open web, but finding the right balance between performance and control.

‍

‍

FAQ

What is a walled garden?

A walled garden is a closed advertising ecosystem where a platform controls data collection, ad delivery, and performance measurement.

👉 Examples: Google, Meta, or Amazon.

What are the advantages of walled gardens?

The main advantages of walled gardens are:

• high performance thanks to rich data and powerful algorithms
• ease of use, enhanced by AI
• integrated measurement and real-time optimizations
• a more brand-safe environment (brand safety)

What are the downsides of walled gardens?

Walled gardens also have several limitations:

• a heavy reliance on platforms
• a lack of transparency regarding performance
• non-exportable data between platforms
• a fragmented view of the customer journey
• and, with AI, a gradual loss of control

Should you focus your entire budget on Google and Meta?

Not necessarily. These platforms offer fast and simple performance, but relying solely on them can limit control and audience diversification.

👉 The recommended approach: combine walled gardens and open web to find the right balance between performance and autonomy.

How can you get the most out of walled gardens?

• Diversify channels while leveraging walled gardens for their efficiency
• Implement a first-party data strategy to maintain control over your audiences
• Regularly measure and compare performance to optimize your media mix

What impact does AI have on walled gardens?

AI increases the speed and efficiency of campaigns by automating targeting, creative, and budget.

But it also increases platform dependency and reduces transparency regarding how campaigns are optimized.

‍