Table of contents of the article:
Il statement press and the measure Published in April 2026, the provisions leave no room for particularly flexible interpretations. The principle is clear: tracking can no longer occur silently, implicitly, or by default. Even seemingly innocuous tools, such as tracking pixels inserted in emails, fall squarely within the scope of technologies that require prior, free, specific, and informed consent.
For those who work in digital, this isn't a surprise. But it's further confirmation that the room for maneuver for data-driven marketing is progressively, steadily, and, above all, irreversibly shrinking.
The invisible pixel that is no longer so harmless
For years, tracking pixels have been a standard component of email marketing. Invisible to the user, simple to implement, and extremely effective. They let you know if an email was opened, when, from which device, and often even from which geographic area. This information was crucial for optimizing campaigns, segmenting audiences, and improving conversions.
But their use didn't stop at simply measuring openings. In many cases, that data was cross-referenced with other signals: purchase history, on-site behavior, previous interactions with other communications. The pixel thus became a piece of a broader system, capable of building increasingly detailed profiles over time. Even a single, seemingly innocuous email opening contributed to a predictive model: who was active, who was inactive, who was close to purchasing, and who was about to abandon.
From a technical standpoint, the mechanism is as simple as it is effective: loading a remote image generates an HTTP request to a server, and that request carries with it a set of metadata—IP address, user agent, timestamp—that can be collected, stored, and analyzed. No user-side code is required, and no explicit interaction is required. Simply open the email.
And it is precisely this asymmetry, according to the Guarantor for the protection of personal data, to represent the central crux of the problem. The average user has no concrete tools to detect this type of tracking, nor to understand its scope. There is no clear visual indicator, no interface that makes it clear what's happening when the email is opened. Everything happens in the background, without friction, without awareness.
In this context, talking about implicit consent or legitimate interest becomes extremely fragile. Without even the perception of the processing, the ability to exercise effective control is also lacking. And this is where the Data Protection Authority steps in, overturning an approach that for years has been considered established practice.
In practice, this also implies a change in the way companies design their communications. Tracking can no longer be a hidden layer, automatically integrated into every message. It must become an optional, separate, and documented choice. The information notice cannot be limited to generic formulas: it must clearly describe the type of data collected, its purposes, and any potential consequences.
From a legal perspective, this position is consistent with the GDPR framework, which places transparency and informational self-determination at its core. From an operational perspective, it represents a further element of friction in an already strained system, introducing complexity where previously there was automation, and requiring explicitness where previously implicitness sufficed.
A direction already traced for some time
This intervention does not come in a regulatory vacuum. It is the final step in a process that began well before, built through a progressive layering of measures that, over time, have substantially redefined the scope of digital profiling.
Over the years, the Guarantor for the protection of personal data acted with a certain coherence, intervening on multiple levels. In the first phase, Attention has focused on obvious tools like cookies, imposing disclosure and consent requirements that have radically transformed the browsing experience. The 2021 guidelines, in particular, marked a watershed moment: out with ambiguous or preconfigured banners, and in with more granular mechanisms, allowing users to specifically accept, reject, or select the different categories of processing.
But the banner itself wasn't the issue. The banner has always been merely the visible manifestation of a deeper problem: the systematic and often indiscriminate collection of data for profiling purposes.
In previous provisions, the Authority had clarified that profiling cannot be considered an ancillary or implicitly accepted activity. Instead, it requires a solid legal basis, based on consent that must be freely given, specific, and, above all, revocable without hindrance. Furthermore, it must also be separate from other purposes, avoiding the typical confusion between service provision and data use for other purposes.
Over time, this approach has led to a progressively narrowing of the operating space for those who base their model on extensive collection of behavioral information. Not through sudden or isolated interventions, but through a sequence of adjustments that have gradually eliminated gray areas.
The underlying theme has always been the same: reducing the information asymmetry between those who collect data and those who generate it. In other words, rebalancing a relationship that, by its very nature, tends to be unbalanced. On one side, there are entities with advanced technological capabilities and full visibility over the collection and analysis processes; on the other, users who often lack the tools or expertise to truly understand what's happening to their data.
The intervention on tracking pixels fits perfectly into this logic. It doesn't introduce a new principle, but rather extends already established principles to an area that, until now, had remained relatively less regulated than others.
Apple's Role and the End of the Era of Easy Tracking
The introduction of App Tracking Transparency features by Apple , which was made widely effective with the release of iOS 14.5 in April 2021, had an impact that, in some ways, was even more immediate and tangible than any regulatory measure.
With a simple permission request displayed to the user—a clear, direct, and difficult-to-circumvent pop-up—Apple has transformed tracking from an implicit and silent mechanism to an explicit choice. A seemingly trivial step, but one with profound consequences: when users are given a real choice, the majority choose not to be tracked.
This had an immediate impact on data availability. Access to the IDFA, which for years had been one of the primary tools for cross-app tracking in the mobile world, was drastically limited. Suddenly, entire targeting strategies based on reconstructing user behavior across different apps became unreliable.
But the impact didn't stop at the world of apps. With the introduction of Mail Privacy Protection — coming shortly thereafter, with iOS 15 in September 2021 Apple has extended this approach to email, targeting one of the most established digital marketing channels. The mechanism is as simple as it is effective: the preemptive and anonymous loading of remote images makes it impossible to determine with certainty whether, when, and by whom an email was actually opened.
The result is that historically crucial metrics, such as open rate, have lost much of their operational significance. Not because they've disappeared, but because they're no longer reliable. An email may appear "opened" even without the user ever viewing it, or it may fail to generate useful signals for precise segmentation.
For those who built campaigns based on this information, this was a significant change. Not so much because of the loss of a single piece of data, but because of the overall reduction in visibility into user behavior.
In this sense, Apple acted as an accelerator. It made concrete, on a large scale, a transformation that was already underway at the regulatory level but proceeded more gradually. Above all, it demonstrated that it is possible to drastically change the rules of the game simply by intervening at the primary access point between user and service.
Less data, less precision
At this point, the logical shift is inevitable. If the quantity and quality of available data is reduced, the ability to target precisely also decreases.
Profiling, regardless of the regulatory implications, has always had a very concrete function: identify the buyer persona more accurately, that is, that subset of users with a real likelihood of interest, interaction, and purchase. It's not just about "knowing more," but about building a decision-making framework based on behavioral signals: email openings, clicks, dwell time, purchase history, and interaction frequency.
This data allows you to progressively refine your audience, excluding those who aren't relevant and focusing your budget on those who are. In other words, it allows you to avoid shooting in the crowd.
When this mechanism works, the benefit is twofold. On the one hand, the user receives more relevant, less intrusive, and theoretically more useful communications. On the other, the company optimizes costs: less waste, fewer useless impressions, and a higher probability of conversion per euro spent.
This is where the concept of efficiency comes into play. An effective profiling system not only increases sales, but reduces cost per lead and cost per acquisition, because every action is driven by a higher probability of success.
But when data is missing—or becomes incomplete, delayed, or unreliable—this level of accuracy rapidly deteriorates.
And when targeting becomes less precise, marketing becomes less efficient.
It's not an ideological evaluation. It's an almost mechanical relationship.
If I can no longer reliably identify who is interested in a product, I'm forced to expand my target audience. Expanding my audience means including a growing share of uninterested users. And including uninterested users means increasing the number of impressions needed to achieve the same result.
In practical terms, this translates to this: more traffic to generate the same number of leads, more leads to get the same number of customers.
The result is a direct increase in the cost per lead and, in turn, in the cost of customer acquisition. Not because the market has suddenly changed, but because the ability to select it has changed.
The critical point: the increase in cost per lead is not proportional
Here comes an aspect often overlooked in public debate: the increase in advertising costs is not distributed proportionally to the value of the product.
It is an increase in large part absolute, not percentage.
This means that the increase in the cost of acquiring a customer tends to be similar in nominal terms, regardless of the price of the product sold. But the impact of that increase is completely different depending on the context.
For a €2.000 product, a €10 increase in acquisition costs is almost irrelevant. It falls within normal market fluctuations and can be offset with small optimizations or simply absorbed into the margin. In many cases, it doesn't significantly alter the overall sustainability of the model.
On a 20 euro product, the same increase completely changes the equation.
Because there, margins are structurally lower, often already compressed by logistics costs, commissions, returns, and operational management. In that context, advertising costs aren't an ancillary variable: they're a critical component.
Ten years ago, selling a low-priced product online could be as simple as two euros of advertising. The model worked precisely because the acquisition cost was aligned with the average cart size. There was room for margin, reinvestment, and scalability.
Today, in many cases, 8 or 10 are needed. Not because the product has worsened or is less competitive, but because Reaching the right person has become more difficult, more scattered, less predictableLoss of targeting precision translates directly into multiple attempts needed to achieve the same result.
At that point, the margin simply disappears.
It's not about earning less. It's not about reducing profitability. It's a more radical shift: the model is no longer viable.
When the cost of acquiring a customer approaches or exceeds the margin generated by the sale, no amount of optimization works. It's not a problem of operational efficiency or creative strategy. It's a structural problem.
And when the problem is structural, the consequence is only one: that product, under those conditions, is no longer sellable online.
When a product stops being sellable
When acquisition costs exceed a certain threshold, the issue is no longer optimization. It's sustainability.
As long as the cost remains within manageable margins, you can intervene: improve creative, optimize landing pages, work on the funnel, test new segments. But when the cost of acquiring a customer aligns with or exceeds the margin generated by that sale, any optimization efforts become marginal. It's no longer a performance problem, it's a model problem.
At that point, the product exits the market. Not gradually, not with a slow downsizing, but often with a sudden halt. Campaigns are shut down, investments are halted, and online distribution is simply no longer justifiable.
And when a product goes off the market, it's not just the seller who is stopped.
The supply chain stops.
The initial impact is visible on e-commerce, which stops selling, but cascading effects spread to all stakeholders. Suppliers see reduced orders. Distributors lose volume. E-commerce platforms see a decline in transactions. Logistics operators move less merchandise. Marketing agencies lose budget and clients. Developers see reduced demand for new features and optimizations.
It's an interconnected system, built over the years on one premise: the ability to reach latent demand relatively efficiently through digital. When that premise fails, not a single cog breaks, but the speed of the entire mechanism slows.
Reducing the efficiency of that system doesn't simply mean spending more to achieve the same result. In many cases, it means reducing the very number of economically viable transactions. And when this happens on a large scale, the consequences go far beyond the individual ad or campaign: they impact the entire market structure.
The risk of further concentration
Another side effect, less visible but potentially more relevant, concerns the structure of the market.
Large platforms possess enormous amounts of first-party data. It's not just about volume, but quality and continuity: data collected throughout the user's lifecycle, within controlled environments, where every interaction—search, purchase, viewing, stay—contributes to enriching an existing profile. In these contexts, tracking isn't perceived as something external or additional, but as an integral part of the service.
This means that the dependence on third-party tracking technologies is much lower. Even with regulatory or technical restrictions, these platforms maintain a high level of profiling capacity because the data originates and remains within their ecosystem.
For a small or medium-sized operator, the situation is radically different.
Those who don't have a large, consolidated user base, or who don't directly manage platforms with high traffic or frequent interactions, have historically relied on external tools to fill this gap: pixels, cookies, and tracking systems that allowed them to build a progressive understanding of their audience over time.
When these tools are limited or made less effective, that gap widens.
Not because small operators do anything differently, but because they start from a structurally weaker position. They have less data, fewer touchpoints, and less ability to observe user behavior over time. And without data, their ability to compete on the marketing front is drastically reduced.
The result is a progressive concentration.
Not immediate, not declared, but gradual. Fewer players are able to sustain effective campaigns, fewer operators are able to optimize acquisition costs, and fewer companies are able to scale.
Meanwhile, those who control platforms and data are strengthening their position. Not necessarily because they improve the quality of their service, but because they have the tools to continue operating in an environment where others are losing them.
It's a well-known effect, partly predicted, but rarely discussed with the same emphasis as privacy protection. This is because it manifests itself less clearly and more slowly, yet profoundly impacts the competitive balance of the digital market.
The paradox of tracking that doesn't disappear
There is one last consideration, more technical but no less relevant.
Restricting explicit tracking doesn't necessarily mean eliminating it. In many cases, it means pushing it toward less visible, less transparent forms, which are harder to control, and, above all, harder to explain to the average user.
When declared and relatively simple tools like cookies and pixels are restricted, the system doesn't stop. It adapts. And this adaptation often involves techniques that operate at a lower level of the technology stack, making them less accessible and intuitive.
Fingerprinting, for example, isn't based on a user-stored identifier, but on a combination of device and browser characteristics: screen resolution, installed fonts, system configurations, and rendering behavior. Taken individually, these elements don't identify anyone. Combined, they can generate a fingerprint unique enough to recognize a user over time.
Probabilistic tracking follows a different logic, but with the same goal. It doesn't seek a certain match, but a statistical correlation: multiple weak signals combined to estimate with a certain probability that two sessions belong to the same subject. It's not deterministic, but it's often "precise enough" to be useful.
Aggregate behavioral analysis also fits into this context. In the absence of direct identifiers, behavioral patterns are observed: sequences of actions, interaction times, navigation paths. The user isn't formally identified, but models are built that still allow for segmentation and prediction.
The point is that these techniques didn't emerge in a vacuum. They emerged precisely in response to regulatory and technological restrictions on more explicit tracking methods. Where one approach is restricted, another tends to take its place.
The risk, therefore, is that of replacing an imperfect but relatively transparent system—because it is at least declared, documented, and to some extent controllable—with a more complex and opaque one, where the boundary between what is permissible and what is not becomes less immediately clear.
And at that point, the inevitable question becomes: does the level of user control actually increase, or does it simply shift to a less visible level, where understanding requires technical skills that most users don't possess?
More privacy, less efficiency: a difficult balance
The role of the Guarantor for the protection of personal data It's clear and, in principle, difficult to dispute. Protecting citizens' fundamental rights, including the privacy of personal data, is an integral part of the European legal architecture.
The new restrictions on pixel-based tracking fit perfectly into this framework. They aren't an anomaly, but a natural extension of already established principles: transparency, consent, data minimization, and data subject control. From this perspective, the direction is clear.
The problem is that the system on which these rules are grafted is not neutral.
The commercial web of the last twenty years has developed on an implicit premise: the ability to collect, analyze, and use behavioral data to improve the effectiveness of marketing activities. Not as an afterthought, but as a structural component. Entire business models have been built on this foundation, progressively optimizing each step of the funnel thanks to the availability of increasingly granular information.
Intervening in this mechanism doesn't simply mean "correcting" certain practices. It means modifying one of the central mechanisms of the system.
And this is where the side effects emerge. Not theoretical, but economic, concrete, and measurable in the short term.
Costs increase because more resources are needed to achieve the same result. Efficiency decreases because audience selection becomes less precise. Market balances shift because not all operators are affected equally by these transformations.
The point isn't to deny the need for greater protection, nor to question the right to privacy. The point is to recognize that there is a real tension between these principles and the current functioning of the digital economy.
A tension that, at least for now, does not seem to have a simple synthesis.
Questioning the extent to which the digital economic system is compatible with an increasingly restrictive privacy model doesn't mean taking a stand against data protection. Rather, it means asking whether the current model is destined to adapt, transform, or, in some cases, no longer sustain the burden of these new conditions.
How to comply with the new directives?
In light of the recent intervention of the Guarantor for the protection of personal data For email tracking systems, compliance can no longer be treated as a formal or document-based requirement. Instead, it requires a technical and operational rethinking of how communications are designed, sent, and monitored.
The first point is to eliminate all forms of implicit tracking. Automatically inserting pixels into emails without a solid legal basis is no longer sustainable. If you intend to continue using tools to track openings or interactions, you must obtain prior, voluntary, and specific consent. This requires clearly separating the purpose of sending communications from the purpose of behavioral analysis.
Technically, this means implementing a consent management system that operates before sending the data. It's not enough to simply include a generic clause in the privacy policy: it's necessary to demonstrate that the user has explicitly consented to tracking. In the absence of such consent, emails must be sent in "no tracking" mode, avoiding the loading of remote resources used for tracking purposes.
A second aspect concerns transparency. The privacy policy must clearly and comprehensibly describe which data is collected through the pixel, for what purposes, and for how long it is retained. Vague or overly technical wording should be avoided, as it prevents users from truly understanding the impact of the processing.
The possibility of revocation must also be guaranteed. Users must be able to change their preferences at any time, as easily as they expressed them. This involves maintaining an updated preference management system and ensuring that changes are consistently propagated across all sending and tracking systems.
From an infrastructure perspective, it is appropriate to distinguish between components necessary for service provision and components used for analysis or profiling purposes. This separation principle, consistent with the concept of "privacy by design," helps reduce the risk of processing exceeding the stated purposes.
Finally, it is essential to document the choices made. Processing records, impact assessments (DPIAs, if necessary), technical configurations, and operating logic must be traceable and verifiable. This is not only for audit purposes, but also to concretely demonstrate the adoption of a compliant approach.
In short, compliance with the new directives isn't achieved by simply "deactivating a pixel," but by rethinking the entire data collection and usage flow. Tracking, from a standard and invisible element, becomes an optional feature, subject to a conscious user choice and managed with appropriate technical tools.
The question that remains open
Ultimately, the question is not whether it is right or wrong to limit tracking.
The question is whether we are ready to accept the consequences.
Because privacy comes at a cost. And that cost doesn't dissolve, it's not absorbed by an abstract entity. It's redistributed along the entire value chain, often silently but inevitably.
Part of this impacts companies, in the form of compressed margins, less predictability, and greater difficulty in planning investments. Part is passed on to consumers, through higher prices or a reduction in available supply. Another, less visible but perhaps more significant, part translates into businesses that simply cease to exist because they are no longer sustainable.
It's not an immediate or uniform process. It's gradual, fragmented, but cumulative.
The provision of the Guarantor for the protection of personal data It represents a further step in a specific direction: more control for the user, less freedom for tracking. A direction that, in the European regulatory context, now appears consolidated.
The question, however, remains open on the economic level.
Is this balance sustainable in the long term? Will the system be able to adapt by finding new forms of efficiency, or will we witness a progressive contraction of some of the dynamics that have characterized the growth of digital over the last twenty years?
And again: will the value generated by greater privacy protection be perceived as sufficient to offset the indirect costs that arise elsewhere?
In attempting to build an ecosystem more respectful of individual rights, we risk—at least in part—eroding the economic base that has made the widespread diffusion of accessible services, products, and models possible.
It's an uncomfortable question because it offers no simple answers and puts two legitimate needs at odds.
But it's a question that's becoming increasingly difficult to ignore, as the effects of these transformations become visible in the day-to-day functioning of the digital market.
