Эксперт Trackier заявила, что установки приложений перестали быть ключевым показателем эффективности в эпоху приватности

As privacy regulations and platform-level restrictions continue to dismantle the identity-based foundations of digital advertising, mobile marketers are facing a fundamental rethink of how they attribute campaigns and measure performance. The shift is particularly pronounced on iOS, where Apple’s privacy framework has made individual-level attribution increasingly difficult and pushed marketers towards aggregated, privacy-preserving signals. The challenge, however, is not simply technical. It goes to the heart of a larger marketing question: if marketers cannot identify and track individual users in the way they once did, how should they determine whether their campaigns are actually creating business value? That was the central theme of Akansha Tiwari, Director – Sales, Trackier’s keynote at CMOs’ Charcha – Bengaluru Chapter 2026. Drawing on her own early encounter with Apple’s SKAdNetwork (SKAN), Tiwari made the case for moving beyond install-based attribution towards a more sophisticated understanding of user quality, conversion value and revenue. Her keynote address, in many ways, was less about decoding another industry acronym and more about challenging marketers to confront an uncomfortable measurement blind spot: the audience that is harder to measure may also be the audience that matters disproportionately to the business. The iOS paradox marketers cannot ignore Tiwari opened with a personal anecdote. When her enablement team first introduced her to SKAN several years ago, she decided to understand it later. That decision changed during a live client call, when she realised that Apple had fundamentally altered the rules of mobile attribution. The experience became a useful entry point into a wider industry transition. For years, mobile attribution largely depended on being able to connect an impression to a click, an install and eventually a purchase through persistent device-level identifiers. Apple’s privacy interventions disrupted that model and forced marketers to rethink how much of the consumer journey they could legitimately observe. Yet Tiwari argued that the common response among Indian marketers — that iOS is too small a market to become an immediate priority — overlooks a critical part of the equation. The familiar argument, she noted, is that “80% of our acquisition happens through Android”. Since iOS represents a smaller share of the Indian smartphone market and SKAN is more complex, marketers often conclude that it can be addressed later. But the more important metric is not necessarily the number of users acquired. It is the value those users generate. Tiwari cited figures indicating that iOS devices account for around 30% of the Indian market but approximately 70% of mobile advertising spend, while Android represents roughly 70% of devices and the corresponding inverse in spend. She also pointed to a significant difference in annual app spending, at approximately $63 on iOS compared with around $13 on Android. The implication is clear: a smaller iOS user base does not necessarily translate into a smaller business opportunity. For marketers, therefore, treating iOS as a secondary measurement problem simply because it represents fewer devices could mean overlooking a disproportionately valuable segment of their customer base. Why Apple changed the attribution game Tiwari also sought to reframe the conversation around SKAN. Rather than viewing Apple's privacy changes merely as an obstacle to advertising measurement, she argued that they represent a fundamental shift in the question marketers must ask. The old attribution model was built around identity. Marketers could connect a sequence of actions — an impression, click, install and subsequent purchase — to an individual device and use that information to assess campaign performance. Apple challenged that underlying architecture. The move formed part of a broader global shift towards privacy-first digital ecosystems, following developments such as GDPR and CCPA, Apple's App Tracking Transparency framework introduced in 2021, and subsequent iterations of SKAN, including SKAN 4.0. India, too, is moving towards a stronger privacy-oriented data environment through the Digital Personal Data Protection framework. For marketers, Tiwari's message was that privacy-first measurement is not a temporary disruption to work around. It represents the direction in which the industry is moving. The question, therefore, is no longer simply whether marketers can reproduce the old attribution model under new rules, but whether they can build measurement systems that work without relying on individual identity. ‘Installs have stopped being the point’ The strongest takeaway from Tiwari's keynote was encapsulated in one line: “Installs have stopped being the point.” Under the earlier model, an install could serve as a relatively straightforward indicator of acquisition performance. But an install says little about whether the user subsequently became a paying customer, returned to the app or generated meaningful revenue. SKAN changes that equation by providing aggregated signals rather than the same level of individual-level attribution marketers were accustomed to. The focus consequently shifts from “How many users did we acquire?” to “What did those users actually do?” Did they make a purchase? Did they return? Did they complete a high-value action? That distinction is increasingly important as acquisition costs rise and marketers move from volume-driven growth towards quality and lifetime value. Decoding SKAN without losing sight of the business Tiwari broke down the mechanics of SKAN while keeping the discussion anchored to the implications for marketers. SKAN operates across multiple attribution windows — zero to two days, three to seven days, and eight to 35 days — and provides marketers with different levels of conversion signals. These include coarse signals, broadly categorised as low, medium and high, as well as fine-grained conversion values that can represent up to 64 values defined by the advertiser. A brand, for example, could configure those values to represent different user actions or purchase-value ranges. But there is a fundamental trade-off. The more granular fine conversion information is primarily available in the earliest attribution window. As the measurement window progresses, marketers receive less detailed signals. SKAN also imposes thresholds around the volume of activity required for reporting. This means that simply dividing campaigns into a large number of smaller segments in the hope of obtaining more granular data can prove counterproductive. Campaigns without sufficient volume may return null values rather than the detailed information marketers are looking for. The broader lesson is that SKAN cannot simply be bolted onto an existing measurement framework. The way conversion values and campaign structures are designed becomes critical to what marketers can ultimately learn from the data. From compliance to predictive measurement This is where Tiwari's argument moved beyond SKAN implementation. She outlined a five-stage progression for marketers, beginning with organisations that have not implemented SKAN at all, moving through technically compliant but poorly designed schemas, and ultimately towards systems in which conversion values are aligned with genuine predictive indicators of business outcomes. At the most advanced stage, measurement itself begins influencing product and marketing decisions rather than merely reporting what has already happened. That represents a significant shift in mindset. Instead of treating SKAN as a compliance exercise or another attribution layer within the marketing stack, brands need to determine which user actions genuinely indicate future value and design their measurement architecture around those signals. Three questions for marketers Tiwari ended with three questions designed to take the discussion from the conference room back to marketing teams. The questions expose the difference between simply having SKAN implemented and actually using privacy-first measurement as a strategic tool. As mobile advertising continues to move away from device-level identity, marketers may have to accept that perfect visibility is no longer the objective. The more important task is to determine which signals matter, build measurement systems around them and connect those signals back to actual business outcomes. For Tiwari, the future of mobile attribution is therefore not about recovering the measurement capabilities of the past. It is about learning to measure value without identity — and recognising that in a privacy-first ecosystem, better measurement may ultimately mean measuring less, but measuring what matters more.