Why European B2B Marketers Are Losing, and Why AI Is About to Make It Worse

Why European B2B Marketers Are Losing, and Why AI Is About to Make It Worse

I take over B2B ad accounts that are paying five to ten times more than they need to for the results they're getting.

The pattern is almost always the same: the marketer running the account has been following industry advice from US-based sources, and that advice doesn't work the same way in European conditions.

I've watched this play out for ten years across Nordic and European B2B accounts. AI is now making it worse.

‍

TL;DR

European B2B marketers are wasting significant ad budget on advice built for the US market that doesn't work the same way in Europe.

Five forces cause this:

platforms turning into black boxes that need US-scale data,

EU privacy law reducing the signal available,

US competitors having a structural edge on the same platforms,

US content creators dominating the advice pool,

and AI models now repeating the same US-centric advice with confidence.

The way through it is discipline. Put real guardrails on the automation, get conversion tracking right upfront, and treat any confident-sounding advice with suspicion until you've checked whether it was written for a market like yours.

‍

The core problem

B2B advertising in Europe has become extremely difficult, and five different forces are pulling at once. I'll go through each of them below.

At the same time, marketers are under heavy pressure from leadership to prove that ad spend ties back to revenue. The sources they turn to for help, the ad platforms, the US content creators, and now the AI models, are giving them advice that fits a different market than the one they actually work in.

They follow the advice, the results underperform, leadership pushes harder, and they go back to the same sources looking for a better answer.

‍

Five forces are pulling the rug out from under European B2B marketers

The reason isn’t ONE clear problem, but rather SEVERAL factors that, together,

1. Platforms are becoming black boxes: Google, Meta, and the rest are moving from tools you configure to systems you feed data to.

2. EU privacy law changes the data supply: GDPR, cookie consent, and Apple's tracking changes cut the volume of data the algorithms receive.

3. The US/EU data asymmetry: Same platforms, different rules on top, different data volume underneath. US companies have a structural edge over EU companies on the same platforms.

4. US best practices don't translate: Most of the loudest voices in performance marketing operate in the US market, and their advice reflects that market's conditions.

5. AI advice inherits US bias: ChatGPT, Claude, and Gemini were trained on the same US-dominant content and repeat it back with a lot of confidence.

‍

All this while the leadership's demand for revenue attribution keeps going up while the ability to actually attribute anything keeps going down. I'll come back to that at the end.

‍

Force 1: Platforms are becoming black boxes

Ten years ago, running a Google Ads campaign meant configuring dozens of settings by hand. Bid adjustments by time of day, by device, by geography, by audience. You could see what was working and change it directly.

Most of that is gone now. Smart Bidding, Performance Max, Advantage+ campaigns. Google wants you to hand over the target and the budget and let the system decide the rest. The dashboards show less data because the platform is doing more of the work internally, which it doesn’t want you to see.

When everybody is using Automated Bidding (Google sets your bid on keywords), who are you bidding against? Ask this question to yourself and think about it.

This whole algorithmic way of optimizing performance isn't wrong in principle. If you have enough data, the algorithms can genuinely do a better job than a human can at optimizing across millions of signals. But that "if" is a big one. These systems are built for an environment with unrestricted signals, high traffic volume, and clean conversion feedback. When those conditions are present, they perform. When they aren't, they don't.

‍

Force 2: EU privacy is structurally different

GDPR gets talked about mostly in terms of consent banners. But that heavily understates what it actually does. It reduces the data the ad platforms receive, the data you base decisions on and how you conduct strategic analysis.

Take website visitor identification for instance. In the US, tools like RB2B and Customer.io let you identify anonymous visitors at the individual level. Which company they came from, often which decision maker, sometimes with a name and title. You can send targeted follow-up based on who actually visited. In the EU, that's illegal. You get aggregate patterns and nothing more, and that is if they clicked yes on the annoying consent banner. (They are planning to make this even worse with a global persistent browser checkbox instead in the future, good luck my fellow European marketers).

A properly implemented Cookiebot or OneTrust cookie consent banner drops around 40 to 60 percent of your tracking, because that's roughly the share of European visitors who decline. Every declined consent might be a conversion you paid for that never gets fed back to the algorithm, and it gets labeled as Direct Traffic in HubSpot. This means that the bidding engine (Google, or Meta) doesn't know it got a conversion, so it doesn't learn that this keyword, this audience, or this creative drove pipeline. 

Apple's changes add another layer. App Tracking Transparency broke a lot of Meta attribution overnight. Mail Privacy Protection made email open rates unreliable. SKAdNetwork replaced deterministic mobile app attribution with a delayed, aggregated signal. Any one of these on its own is a pain, stacked together, they significantly reduce what the algorithms have to work with and it haunts every performance marketer at nights.

‍

Force 3: The US/EU asymmetry

Same Google Ads. Same Meta. Same LinkedIn. Different rules on top depending on geography, different data volume underneath, and the asymmetry favors US-based advertisers.

A US company advertising to the US market has full access to identification tools, richer conversion tracking, larger addressable audiences, and platform algorithms tuned on the highest-signal market in the world. A European company competing for the same buyers, even against a US competitor of the same size, is working with a fraction of that data because if you are based in the EU and markets to the US.

There's a version of this that almost nobody talks about. 

When a European company advertises into the US market, GDPR still governs how they can collect and process data from those US users. That's the correct reading of the law and it's how any competent DPO will guide you. So the European company targeting American customers still can't use the same identification and analytics stack the US competitor uses to reach the same audience. Same market, same platforms, way less data and ways to understand, act and make money from that audience.

‍

Force 4: US best practices don't translate

This is where most of the money gets lost. A marketer in Sweden or the Netherlands wants to run their first serious campaign, looks at what the loudest voices in the industry are teaching, and applies it directly. The result is usually bad, and it takes months or years to understand why.

Three specific reasons the US best-practice model breaks in Europe.

Fragmented markets: The US is a single, largely homogeneous, English-speaking market of over 300 million people. Europe has dozens of languages, dozens of buying cultures, and much smaller addressable audiences per country for your ad campaigns. A campaign structure that works in a US market with 300 conversions a quarter behaves very differently in a Swedish equivalent that might get 50. The algorithm has almost nothing to learn from.

Data starvation from regulation: Inside a single European market, GDPR and consent decline further reduce the share of signal the bidding algorithm actually sees. Smaller market plus smaller share of that market visible to the algorithm equals highly volatile performance.

Algorithm failure under low signal: A Smart Bidding campaign given incomplete data will learn the wrong patterns and bid confidently into the wrong audience, keyword, and creative, because it has decided on the basis of a few dozen data points that those are the winning combinations. Nothing in the dashboard flags this as a problem. The account continues to spend and the algorithm continues to be wrong. The cost of that mistake scales with the budget. This is how a European B2B company ends up paying five to ten times more than they need to.

‍

Force 5: AI advice inherits US bias

There's a straightforward reason this is about to get worse. The AI models people now consult for marketing advice, ChatGPT, Claude, Gemini, were trained on the public internet. The public marketing internet is overwhelmingly written in English by US practitioners and reflects US market conditions.

When a European marketer asks an AI model how to structure a Google Ads account, how to set up conversion tracking, or what bidding strategy to use for a new campaign, the answer they get is a competent summary of US-oriented advice. It sounds authoritative and detailed. Underneath, the model is often reproducing something from a Reddit thread or a US agency blog post. That original comment may have been correct for the person who wrote it in their market. It isn't automatically correct for a Nordic B2B SaaS company with a thousand visitors a month and cookie consent at 45 percent.

What tends to happen is that the marketer trusts the confident-sounding output, sets up the account accordingly, and spends six to twelve months and a five or six figure budget learning it doesn't work. By the time they figure out why, the money is gone and the credibility with leadership usually goes with it.

‍

What this actually costs

I've inherited accounts running on this failure mode many times. The setup is usually a version of the same thing. Aggressive automated bidding, minimal negative keywords, broad match everywhere, no geographic or language controls, conversion tracking firing on the wrong events or on nothing at all. The account technically works. It spends the budget every month, shows conversions in the dashboard, and wastes somewhere between half and eighty percent of that budget.

The cleanup usually isn't complicated. It's putting the guardrails back on that the "best practice" content told the client to remove. Cap the bids. Restrict the geography. Add real negatives. Reroute the conversion signals to something that actually correlates with pipeline. Pull automation back to the parts of the account where it has enough data to work. And a huge bunch of intuition, and I say that because when you cannot rely on inconsistent and fragmented data, you need to rely on what your gut is telling you and real research - how are people actually navigating this product category, what are their thoughts, who actually sits down and does the short-list.

What to do instead

Going back to spreadsheets and manual bid adjustments doesn't work at this point either, so this isn't an argument against automation. The argument is for a different discipline around it. You use the algorithms, because you have to, but you don't hand them the wheel until they've earned it.

Here are the guardrails I use across every European B2B account I run.

• The algorithms won’t work with your limited conversion data, so we need to create more signals manually.
• Start every campaign with as manual settings as possible and watch closely the data you get - is it sufficient, stick with manual.
• Google is pushing automations, algorithms and AI's hard. Resist these as long as you have volume to live on manual settings.
• Don’t apply any recommendation offered by the platforms, ever, before consulting with someone that knows.
• Implement soft conversion signals (timers, scroll depths, engagement metrics) together with your qualified leads conversion data (MQLs, SQLs, Opps) and Deals - all these need to flow into your ad platforms.
• Aggressive geo/language/audience/competitor exclusions from day one.
• Manual negative keyword baseline before the first click, along with heavy pruning every month.
• Implement your own additional tracking fields to submit forms, HubSpot often misses the source during sessions and you end up with too many Direct traffic-converters.

We have it harder as EU marketers, so we need to work smarter; GDPR, privacy and algoritms doesn't affects all aspects of our jobs as B2B marketers.

The customer journey design still sits with a human. The algorithms decides how to bid inside a single campaign. It can't decide how to split the budget between Google, LinkedIn, and Meta based on where your buyers actually spend their time, because it can only see the account it's running in.

A B2B buying cycle of 180 days or more needs different touchpoints at different stages, only around 5 percent of your target market is actively buying in any given month, and the algorithm has no view on either of those facts. 

Landing pages have to be built and reworked as search intent shifts. Conversion tracking has to be set up correctly upfront, or the algorithm never gets a clean signal in the first place. Audience behaviour in a market can shift months before the campaign metrics catch up, and knowing when to scale an account versus when to hold is a judgment call the systems aren't equipped to make.

As platforms are harder to manage the outer work becomes more of the difference between companies that grow and the ones that doesn't.

Closing thoughts

The confident advice you're getting from the ad platforms, the US content creators, and the AI models isn't neutral information. It was written for a US market with different rules and much more data than you have, and if you apply it directly to a European B2B account it will cost you real money this month.

The paradox in all of this: the same AI tools you can't fully trust to set up your European campaigns are the tools your leadership is telling you to integrate deeper into the marketing function. You're being asked to hand more of your work to a system that will make your job harder, because it doesn't know your market and doesn't know that it doesn't know.

That's the pressure loop underneath everything above. Leadership wants better attribution. The marketer turns to the platforms and the AI for guidance. Both are optimized for US high-data conditions. The European numbers come in worse than they should. Leadership pushes harder. The marketer goes back to the same sources. The cycle repeats.

The way out is discipline. That means using fewer of the platform's default settings, putting more guardrails around the automation, and treating any confident-sounding advice with suspicion until you've checked whether it was written for a market like yours.

‍