The aftershocks of the European Court of Justice’s Google Shopping judgement were still noticeable over the last couple of weeks. Unsurprising given the decade-long run up and built-up suspense to the judgement. There could be no better time to interview Albrecht von Sonntag, internet pioneer and co-founder of the Berlin-based online price comparison platform idealo who was at the forefront of the resistance against the Google’s self-preferencing practices since the beginning and even before the Commission initiated the investigation in 2010. We have spoken about the conclusion of the Google Shopping saga, the promises and perils of the DMA and learned about Albrecht’s visions of how to fix the internet, including a world where we think of individuals as citizens and human-beings, and less as consumers.
By Jasper van den Boom & Sarah Hinck
Can you tell us about your history at idealo and the early days of fighting Google?
I was one of idealo’s founders in 2000, two years before Google search arrived in Europe. Over the past 25 years, I was part of idealo’s management team and have acted as managing director. This June, I stepped down to focus full time on shaping the future of the internet and our battle against Big Tech. Trying to repair the Internet is a full-time job. It has not been an easy one, sometimes it feels like it is not even like David versus Goliath, but more like fruit fly versus Goliath.
“Trying to repair the Internet is a full-time job.”
I remember the times before Google came to Europe. It felt like we were still competing on the merits. This started changing in the early 2000s. Already then, investors’ ellipse was “what if Google does it?” If the answer was Google might do it, they would simply not invest. With this, competition really became about who has the biggest stick to beat with. Now, with the DMA, I think we really have something new.
You said things already changed around the early 2000s, when exactly did you become involved in antitrust proceedings? Did you reach out first to the European Commission or the German Competition Authority?
We started at the Commission. The first time we went there we did not leave our business cards, we were worried that if Google knew we were talking to the Commission they would treat us unfairly. There were others who went to the Commission first and openly, Foundem was the first one. We joined as a complainant in 2012, we started filing official complaints and were locked in in 2013. We still went through our daughter company to file these complaints, not idealo directly out of fear about retaliation. Since 2013, things got rougher. In any event, we were not subject of retaliation measures by Google. Coming forward may have even protected us from retaliation because Google seemed to have treated us more carefully now the Commission was involved. But that is just a guess.
What do you mean that it got rougher since 2013?
Well, first Google tried to compete equally through Froogle, the predecessor of Google Shopping. This did not work well enough for them. They already preferenced themselves by giving themselves one of the top spots in the search results display since 2010. This could be spot one, two, or five, but Google was always up there.
“There is no such thing as organic in search.”
In 2013-2014, they introduced their Panda algorithm. This is when they really started promoting themselves to number 1 while at the same time making sure that other comparison-shopping services were demoted in what people refer to as organic search results. The right term should be unpaid search results as there is no such thing as organic in search. After this competing got more difficult, and more users were diverted away from comparison shopping services.
You say that the behaviour actually got worse after the Commission started investigating. Then, in 2017, the Commission’s decision was out. What was this like for you? Did you think things would finally change? Did they change?
It was one of the most euphoric moments of my life – business life I should say. We had a champagne celebration. We were frantic, it was all clear! In the decision, the Commission said clearly it was about search and that everyone should be treated fairly in search generally. Then, Google’s lawyers made the case that the focus was on the Bronner criteria and access to the Google Shopping Box. This did not connect to the decision at all. Google pushed this “access” narrative, because it helped them to use the remedies to turn us, vertical search providers, into customers. We would become agents for Google who could give them our best offers. We could participate in the search box with one product offer but were still demoted in the other results. If you think about it in analogies this is crazy: this is as if someone would foreclose all the supermarkets but offer them to display a single product, not even the best offer, in its display of products. In this case we would never say that competition is restored. Google’s remedy meant either you are forced out of competition or you change business model to become one of Google’s agents. Many of the initial complainants are now agents for Google. They already had their back to the wall, and feeding into Google’s business model would put food on the table, but competing would not.
“You are forced out of competition or you change your business model to become one of Google’s agents.”
Did you have the impression that the Commission did not capture the importance of the remedy issue at the time?
I think you have to look at it in the spirit of its time. The Commission had to balance several considerations. First, self-preferencing was a novel concept in competition law, and it was uncertain whether the Court would accept it. With any new legal theory, there’s always an element of risk. Given their limited resources and broad enforcement responsibilities, the Commission needed to develop a solid legal precedent that would benefit them in future cases.
At the time, I was frustrated by the Commission’s approach, but in hindsight, I see the strategic move. While I would have preferred a different outcome as a competitor, their focus was on more than just the shopping market—they had to consider other sectors like travel and jobs. Our case served as a precedent for tackling issues in 10, 12, or even 20 other industries crucial to Europe’s digital economy. This was the birthplace of the Digital Markets Act (DMA), which provides a faster and more comprehensive solution.
After the DMA compliance workshops in March, maybe even after my speech of how Google’s presentation of its so-called compliance measures is a waste of time, the Commission promptly opened a non-compliance investigation. We will likely see preliminary findings soon. The Commission now has the Court’s ruling to support their efforts, strengthening their position in future actions.
Also, we now have a binding definition of what constitutes a price comparison service, which has far-reaching implications. Now, the criteria—images, product price, merchant information—are set, impacting other industries and aligning with the DMA’s requirements under Article 6(5). This clarity is beneficial for enforcing regulations across various markets, ensuring that there’s a solid framework everyone must follow.
Do you think that these definitions will help with things like damages claims and private action?
The CJEU’s ruling has created a new landscape for damages claims. It’s similar to around 10 years ago, when the Commission’s findings were automatically binding on local courts as a result of the EU Damages Directive. Now, many things that we previously had to prove are considered binding, which simplifies the process of pursuing damages claims. The CJEU confirmed in Google Shopping that there was significant traffic diversion, making many of these elements indisputable in court. The main focus now is on the counterfactual—what would the market have looked like if Google hadn’t engaged in anti-competitive behaviour?
We have been actively pursuing damages claims, and with Alphabet now generating about 60% of its revenue from Product Listing Ads (PLAs) and comparison shopping services capturing roughly 50% of that market share, this sets the stage for another significant damages claim, potentially reaching billions of dollars.
Do you also think this ruling will now shape the enforcement of the DMA? We already have an ongoing non-compliance investigation, but some parties have been asking why the preliminary findings for Google seem to take so long. Do you think this is related to CJEU judgement?
The judgment is indeed very significant. If the CJEU had ruled against the Commission, it would have been a major setback. Until the ruling, they couldn’t clearly define what constituted price comparison or self-preferencing, as the interpretation was so closely tied to this specific case. The uncertainty forced the Commission to prepare for both a win and a loss, which created complications for stakeholders. But now, the ruling provides much-needed clarity, which changes everything.
I have interacted with the Commission multiple times, though not as frequently as Google does due to the time and resources it invests. Despite this, every meeting felt a bit frustrating, as the officials seemed overly cautious. But after the ruling, I noticed a shift almost immediately. I attended three workshops during that time: one on Monday focused on Google Shopping, where the Commission seemed hesitant to delve into past issues. On Tuesday, the ruling came out: By the next morning, there was a noticeable change in the Commission’s stance during discussions on trains and flights comparison – areas idealo also is active in.
It’s evident that they now feel empowered to take a firmer approach. So, the ruling has provided a stronger legal foundation, and I believe it will indeed speed up the enforcement of Article 6(5) DMA. The Commission now has a solid precedent to back their actions, and we’re likely to see a more assertive and faster enforcement process moving forward.
Coming back to the disappointment with Google’s compliance measures which you expressed during the compliance workshops, what was the most problematic element in your view?
Everything Google presented was blatantly non-compliant, often with very problematic reasoning. Nothing has changed with their compliance measures. They said things like ‘well, we are free to do what we want where it concerns paid results, that was never the intention of the DMA to touch this’. Of course the DMA is about the types of results Google is showing and how these relate to results offered by other companies generally, and not about whether Google gets paid or how much it gets paid.
“Everything Google presented was blatantly non-compliant.”
Throughout the process, it initially felt like the Commission was just going through the motions—listening to concerns but not taking meaningful action. This was essentially Google’s approach as well, just ticking boxes without showing genuine commitment to compliance. The Commission’s role shifted over time. When the DMA first came into force, the Commission actively designated gatekeepers but then there was a lull. In the run-up to the compliance date and during the workshops in particular, the Commission took on more of an advisory role, providing guidance without enforcing any real change, more like moderators than regulators. After March 7, the Commission’s role shifted and it became the enforcer of new rules with powerful tools at its disposal. Google can no longer simply interpret the DMA as it sees fit. From this point forward, the Commission has to set clear expectations and ensure these companies follow through.
Has cooperation between you and the Commission changed after the compliance date and the announcement of the non-compliance investigations?
It is hard to say, the Commission officials really keep a sphinx-like face. This makes sense, the people in these institutions are not activists, they are trying to enforce the rules with the resources they have. Now I am optimistic, but at the time of the workshops, I was really frustrated. We are dealing with so many layoffs and challenges, and then you see Google and the Commission in the workshops sitting next to each other, leaning over and chatting with each other – almost sitting in each other’s chair! This is really hurtful when you have been dealing with these issues so long. I found the setup really strange. It almost felt like the Commission was representing Google as their lawyers. I think announcing the non-compliance investigations helped a lot. Now I have more trust that the Commission will use their tools.
As the DMA sets the standard of effective compliance, the Commission might now find that Google is not compliant. What should effective compliance look like according to you?
I really think the DMA is quite clear on this. Gatekeepers must be compliant by design which must be clear-cut and easy to monitor for the Commission but also for business users and competitors in order to be effective. For compliance to be effective, there can be no doubt about whether gatekeepers are compliant or not. If it is unclear, we already have a problem. The Commission should be really pro-active here. We don’t want a solution that barely complies. We want real compliance with pro-active efforts by the gatekeeper. If they are even toeing towards that line of non-compliance, entering a grey area, they should be risking significant fines.
The DMA is not about negotiating with gatekeepers again to see what they think is compliance. It is about enforcement. You want to go into that grey area, best beware, because one millimetre over the line and you receive a harsh punishment. We need a few landmark cases like this so that gatekeepers will not go anywhere near that line. If we do not get good landmark cases then there will be no interest in other jurisdictions to introduce similar laws.
You said in the beginning trying to repair the internet is a full-time job. Are there things outside the DMA you think should be done to achieve repair?
There are three important points beyond the DMA we need to look at in my view to make markets work for us as human-beings and citizens, not as consumers. We should go beyond calling ourselves consumers, we are human-beings and citizens. We should broaden the discourse to what Big Tech’s behaviour does with our ability to inform ourselves, to our democracy, and how we perceive reality.
The first point relates to the matter of vertical integration and how Big Tech takes over and exploits downstream markets. It is like a food chain. It starts with operating systems, then they take over web browsers, then search, then vertical search, then become a direct supplier. This is done by buying up other companies and integrating them but also by making exclusive deals with downstream companies or auctioning off this market to a limited circle of companies. Big Tech takes a big piece of the pie everywhere, and push everyone else out. Every downstream competitor has to pay Big Tech operating upstream, and the lower you are the more you have to pay. This price ultimately ends up with the citizen.
My second point relates to the quantity and quality of advertising we see. In broadcast television we had rules about advertisement quantity, limited to maybe 12 minutes per hour. There are no similar rules for online platforms. How much of Google’s search results can be advertisements? We could be working to a situation where just 100% of the results are advertisements, and it has nothing to do with relevant results anymore. This would mean that one, or more, of the channels you use to inform yourself about society are just completely bought and paid for.
On the quality of advertising, traditionally, we have had clear distinctions between content and advertising in television, but now we’re navigating a much more ambiguous space, especially with digital advertorials that blur these lines. People have argued that if it’s an advertorial, it must be clearly marked as such. However, this isn’t always the case.
Take the app store as an example—when you search for a game like Clash of Clans, you might see a very similar name appear at the top. It’s easy to mistake it for the original. The design is intentional: the real game might be further down, hidden among other Apple-related suggestions and ads. This tactic relies on misleading placement, where users, especially if they’re tired or distracted, may end up clicking the wrong option.
I believe it’s crucial to differentiate between two types of advertising: one that simply presents an alternative alongside the main product and another that actively obstructs the user’s path to the intended result. The latter is problematic—it’s designed to trick people who might not have the sharpness or media literacy to spot the deception.
The third point I want to make is the biggest problem in my view. We have these huge digital firms that are governing our access to information, and they can advertise however they want. They can advertise to you one-on-one. Before, advertising was for all: you put your ad on TV or in a magazine and everyone could see the same ad. If there was something problematic about it, it could be easily spotted it and an authority could step in.
Now, internet platform turned this to a one-to-one scenario. Think of TikTok, two children, brothers even, would have completely distinct reels on TikTok. I as a parent who has a legitimate interest to see what is going on has no idea what my sons were seeing. These alternative realities are starting to pop up, which we have already seen before with the Cambridge Analytica scandal, some people get information in a way and on topics that are completely unmonitored. With the development of AI tools, this is all going to get much more problematic.
The implemented algorithms want you to stay on the platform as long as possible; second, they want you to engage with your friends; third, they want you to keep engaging yourself; and fourth, they aim to maximize revenue from every second of your time.
This business model is not just about keeping people online—it is also about pushing them toward more extreme views. This is one reasons why we are seeing a surge in youth support for radical parties in some regions. The algorithms are tailored to exploit more extreme perspectives because those users are more likely to stay on and tend to spend more money.
Ultimately, the goal is to keep you consuming, not interacting meaningfully. If people spend less or adopt anti-consumerist values, platforms lose revenue. The conflict of interest is clear—business models are harming society by pushing people to passively consume rather than actively engage.
Do you have any suggestions how these fundamental issues should be fixed?
On breaking companies up, we need to focus on the conflict of interest. If I think of Google, I would say Search is the core of the harms, we should focus there. However, we can think more broadly. The problems with integrating from upstream to downstream shows a fundamental issue: companies shouldn’t be able to build these massive vertical stacks and monetize downstream markets to the extent that everyone else has to pay to access each layer. The higher up the stack you are, the more you can extract value from the entire chain.
“The higher up the stack you are, the more you can extract value from the entire chain.”
In Google’s case, their need to sustain such a vertically integrated structure forces them into anti-competitive behavior. For example, they reportedly pay Apple around $20 billion in 2022 just to be the default search engine on iOS. It’s like Mafia. But to cover that expense, they have to extract more revenue from their own downstream services, which can lead to further market abuses.
This shows the cycle of dominance: one layer feeds the next, and breaking up these integrated structures is essential. Regulation should focus on preventing companies from using their control over a core service—like an operating system or browser—to dominate downstream markets. A price comparison service, for instance, shouldn’t be allowed to control direct supply chains and use that to secure a monopolistic advantage.
So, breaking up is the solution of choice for you?
In essence, Google’s stack needs to be dismantled: Android should be separate from Chrome, Chrome should be separate from Search, and Search should be distinct from all the downstream services. This type of restructuring is necessary to ensure fair competition. Otherwise, the Big Tech will continue to dominate every segment, consolidating power at each level of the market, and leaving very little room for genuine competition.
Besides breaking up companies, there is another issue I really want to emphasise because it is not talked about enough. It refers to the underlying business model itself. People by now have such a personal relationship with these tech platforms, they spend hours on these platforms, and they think it is somewhat neutral. Tech savvy people may understand when they are being tricked or manipulated, but a huge part of the population simply does not see it. They do not know how much data is collected, how much companies like Google know about them already, that the algorithms know when they are emotional, outraged or dissatisfied, and that they know how to get them to buy things in these states.
We need to redefine our relationship with these platforms and ourselves and expect to be treated as citizens and human beings, we should not be manipulated at this level. So, I would suggest, if the relationship is this deep, maybe Big Tech should not be allowed to monetise their services by advertising. If you are this big and this important, and this intertwined in users’ lives, maybe you need to make it a subscription services. Their impact on how we view society and democracy is so fundamental, they push us towards politics and beliefs that sustain their business models, it is hugely problematic, we should expect of them to treat their position with a certain sense of responsibility.
Is this something that the DMA would be equipped to deal with?
My comments on the advertising issues don’t specifically fall under the DMA but there is a mechanism in Section 19a para. 2(b) of the German Competition Act that can help address exploitative advertising practices. It would be beneficial to clarify this further, and I’m in touch with the relevant authorities to offer some suggestions.
One challenge we face is that, because we are dealing with these global gatekeepers, we tend to rely solely on the Commission as the main authority capable of addressing them. However, we also need to recognize that local markets have their own unique dynamics. My concern is that the Commission might create solutions that work well in 10 or 12 countries but neglect the needs of smaller markets, deeming them less significant. That’s why it is essential to empower national authorities as well.
“My concern is that the Commission might create solutions that work well in 10 or 12 countries but neglect the needs of smaller markets, deeming them less significant.“
In this context, Sec. 19a GWB is a crucial tool, but it currently overlaps with European regulations, creating a grey area. As a result, Germany’s Bundeskartellamt often hesitates to act, fearing it could conflict with broader EU regulations. It’s in the lawmaker’s interest to make it clearer that national authorities still have a role to play and are not merely subordinate to EU directives. Clarifying these boundaries would enable local agencies to address issues specific to their markets more effectively, while still maintaining coherence with EU-wide efforts.
Albrecht von Sonntag was interviewed by Jasper van den Boom and Sarah Hinck of our SCiDA team on 1 October 2024. We thank Albrecht for his insightful views on the issues around Google search and digital platform regulation. If you do not want to miss any of our future blog posts you can subscribe to the SCiDA-Newsletter here.