Walled gardens: do Google and Amazon control your ad budget?

In Italy, 82% of every euro spent on online advertising ends up in the coffers of Google, Meta and Amazon, the highest share among the major European markets. If you run campaigns, you're building on ground that doesn't belong to you.

Is it really smart to keep your entire budget on Google and Amazon?

It depends on how much you're willing to pay in ever-rising rent without ever owning the data. Concentrating everything on two or three closed platforms exposes your company to unilateral cost increases and opaque visibility into real results.

A walled garden, literally a fenced-off garden, is a platform that controls access to the audience, measurement data and auction rules, all at once. Google and Amazon fit this definition: you decide how much to spend, but you can't truly see what happens inside.

The problem isn't that these platforms are bad. The problem is that when 75% of your ad budget is concentrated there, you have no negotiating power, no independent first-party data and no alternatives ready if costs go up.

What did the CEO of The Trade Desk really say, and why does it matter?

Jeff Green, founder of The Trade Desk, published a long essay arguing that Google and Amazon are building a system where everyone else pays rent. What's interesting isn't who says it, but the data that backs up the claim.

The Trade Desk is a software platform that buys ad space on websites, apps and streaming TV outside the big platforms. Its founder has a direct stake in the issue, and he says so openly. But that doesn't make his analysis wrong.

In the essay published on PPC Land, Green writes that Google and Amazon "decide how content is monetized, share data that is a black box and invent measurement rules that serve their own interests." His proposed answer is the open internet, meaning the set of sites, apps and streaming platforms that don't belong to these giants, as a viable alternative.

The sharpest point is about AI: large language models (AI systems that read and summarize text, like those behind ChatGPT) are changing how people find content and products. Whoever controls these systems also controls the next layer of access to the audience.

What is really happening to the Italian advertising market?

Italy is more concentrated than the rest of Europe: Google has overtaken Rai as the top operator in the Italian media system, and local publishers' share of ad revenue has collapsed by 26 points in six years.

The numbers tell a clear story. According to AGCOM, in 2024 Google captured 12.4% of the total resources of Italy's Integrated Communications System (SIC), overtaking Rai (12%) for the first time. Meta ranks third with 9.7%.

Meanwhile, according to eMedia analysis cited by Il Sole 24 Ore, Italian publishers' share of advertising revenue fell from 67% in 2019 to 41% in 2025. Who took those 26 points? The global platforms.

For anyone running campaigns, this means one thing: the cost of reaching audiences on these platforms can only go up, because demand is growing and supply is controlled by a few.

What does this mean in practice for those running campaigns at an Italian company?

Concentrating your budget on Google and Amazon isn't wrong in itself, but without a diversification plan you're accepting a growing risk you can't measure.

Picture a fashion e-commerce store spending 90% of its budget on Google Shopping and Meta. When cost per click jumps 20% in a quarter, as happened in several sectors in 2025, it has no alternatives ready. No data on other channels, no audiences tested elsewhere, no leverage.

Diversifying doesn't mean abandoning Google or Meta. It means having at least one active alternative channel, with your own data, so you can shift budget in an informed way when needed. This applies to a manufacturing company doing B2B lead generation as much as to an online store.

Three concrete scenarios: what changes for an e-commerce, a professional firm, and a manufacturing company?

Diversification means different things depending on your industry and available budget. Here are three typical scenarios, with the reasoning applied.

Illustrative examples: numbers are meant to show the reasoning, they are not real customer data.

What tools exist outside the walled gardens, and how do they compare?

Programmatic outside the big platforms isn't one single thing: there are independent DSPs, third-party retail media and premium publisher networks. Each has different characteristics.

A DSP (Demand-Side Platform, a platform that lets you buy ad space automatically across multiple sites at once) like The Trade Desk lets you buy placements across thousands of publishers, including news sites, apps and streaming TV, with independent targeting data. It's not for everyone, but it's the most flexible tool for anyone who wants to step outside the walls.

If you run campaigns on Amazon, it's also worth reading how to truly measure Amazon Ads' contribution before increasing your budget: often the problem isn't the channel, but how it's measured.

Are there legal or regulatory risks to keep an eye on?

Yes: the regulatory landscape is changing fast, and some platform moves are already under scrutiny by European regulators.

Two concrete signals from 2026. First: Meta changed its commercial terms, adding a clause that lets it scan and use for AI the websites linked to advertisers' campaigns. After publishers protested to EU and UK regulators, it pushed the effective date to November 20, without changing the text. If you run Meta campaigns, you link your site: it's worth reading the updated terms.

Second: organic search traffic to North American publishers' sites fell to 12% of pageviews in Q3 2026, from 21% a year earlier, according to Chartbeat data. Less organic traffic means more dependence on paid advertising, and therefore more power for the platforms that sell it.

If you run campaigns on Google, it's also useful to understand how Google Ads handles priorities and budget allocation by channel: the platform's internal logic influences where your money really ends up.

Why is this being discussed now: is the European market really at a turning point?

European digital advertising is worth €131 billion in 2025 and growing 10.5%, but the growth is concentrated in platform-controlled channels. Retail media and social advertising are growing at double-digit rates, while independent publishers lose share.

According to the IAB Europe AdEx Benchmark 2025 report, retail media, meaning advertising sold directly by retailers like Amazon, Zalando or Carrefour, grew 16.7% in Europe, reaching €13.3 billion. Social advertising grew 19.2% to €35.5 billion. Both segments are dominated by players that control both the channel and the measurement.

Meanwhile, the research by the Osservatorio Internet Media at Politecnico di Milano estimates that internet advertising in Italy is worth €6.2 billion in 2025 (+11%), or 53% of all advertising investment. A growing market, but increasingly concentrated.

Diversification isn't a new topic, but the speed at which AI is reshaping how people find content and products makes it urgent. Those who wait and see risk having no alternatives when costs rise or the rules change.

How do you actually get started diversifying your ad budget?

It's not about abandoning Google or Meta, but about building a gradual plan that reduces dependence without losing short-term performance.

What mistakes are most common when trying to diversify?

The three most common mistakes are testing too little, measuring poorly and expecting the same immediate results that platforms optimized for years deliver.

• A test budget that's too low. A €500 test on a programmatic DSP tells you nothing. You need enough volume to get statistically useful data.
• Measuring with the same metrics as the walled gardens. The ROAS (return on ad spend) a platform attributes is always optimistic. On new channels, look at conversions you can verify in your CRM or order system.
• Quitting after 2-3 weeks. New channels need time to optimize. Eight weeks is the minimum for a reliable signal.
• Starting without a clear goal. "I want to diversify" isn't a goal. "I want to acquire B2B leads on LinkedIn for under X euros by Q1" is.
• Ignoring creative quality. An ad built for Google won't work on a programmatic DSP or on LinkedIn. Every channel has its own format and context.

On this point, it's also worth considering how adapting the same content to multiple channels with AI can cut creative production costs as you expand your mix.

How do I know if diversification is really working?

There are specific indicators that show whether the new channel is contributing to the business, beyond what the platform itself reports.

So, is diversifying your ad budget worth it for your company?

Worth it if you have more than 70% of your budget concentrated on Google or Meta, your cost per click has risen over the past 12 months and you have no first-party data independent of the platforms. Every euro you spend today strengthens their position, not yours.

Somewhere in between you're already testing a few alternative channels but without a structured measurement plan. The risk is that the tests produce unreadable results and get abandoned too soon. It's worth systematizing before you expand.

Not yet if you already have an active channel mix, first-party data in your CRM and an independent attribution system. In that case, optimizing within your existing channels brings more value than further expansion.

How we see it at SHM

For us, technology is a means: the goal is for the company to keep control of its relationships with customers. When your entire ad budget flows through closed platforms, that control erodes, slowly and then suddenly.

We've worked with companies that had to rebuild their acquisition capability from scratch after an algorithm change or a sudden cost increase. The lesson is always the same: you build diversification when it isn't urgent, not when it already is.

This isn't ideological. A well-measured channel mix, with your own data, clear goals and an independent attribution system, is simply more resilient than any optimization within a single walled garden. That holds for Amazon campaigns too, where the new full-funnel features change the targeting logic but don't solve the problem of structural dependence.

If you want to understand where you really stand and how to build a concrete diversification plan, explore our strategic AI integration services, our Google Ads campaigns, our LinkedIn campaigns, or write to us directly via our contacts. You'll find more insights on the topic in the Marketing and ad performance section.

Frequently Asked Questions

The real question isn't whether Google and Amazon are too powerful, but whether your company has a plan B ready for when their prices go up again.