The End of Last-Click: Meta's Robyn Creator Joins Fospha to Revolutionize MMM

NEW YORK, NY – October 07, 2026 — For over a decade, digital advertising has been addicted to a flawed metric: the last click. It is a system that rewards the final touchpoint before a sale, often ignoring the complex, multi-channel journey that actually persuaded the consumer to buy. But as privacy regulations tighten and artificial intelligence fundamentally alters how we shop, the last-click attribution model is facing an existential crisis.

Enter Marketing Mix Modelling (MMM), a decades-old statistical technique currently experiencing a massive, tech-driven renaissance. Today, Fospha, a leading enterprise marketing measurement platform, announced that Gufeng Zhou—the creator of Meta's pioneering open-source MMM framework, Robyn—has joined the company as a strategic advisor.

The appointment, announced during Adweek, signals a critical inflection point in the advertising technology sector. Zhou, who recently departed Meta to focus on independent measurement innovation, will advise on Fospha's product roadmap, specifically focusing on always-on, calibrated measurement. With Fospha already managing over $40 billion in optimized marketing spend for heavyweight brands like Dyson, Huel, and Callaway Golf, the alliance underscores a broader industry pivot: the transition of advanced, open-source marketing science into commercially deployed, real-time enterprise software.

From Big Tech Open Source to Enterprise Adtech

To understand the significance of Zhou's move, one must look back to 2021, when Robyn began as a hackathon project within Meta Marketing Science. At the time, MMM was largely the domain of expensive consultants who would deliver static, backward-looking reports on a quarterly or bi-annual basis. Robyn changed the paradigm. As an experimental, semi-automated, and open-source package, it democratized advanced econometric modelling.

By leveraging machine learning techniques like Ridge regression to handle multicollinearity among marketing channels, time-series decomposition for trend and seasonality analysis, and multi-objective evolutionary algorithms for hyperparameter optimization, Robyn put sophisticated measurement within reach of everyday advertisers and agencies. It proved that complex data science could be automated and scaled without requiring an army of PhDs.

However, creating an open-source framework is only the first step. The challenge for enterprise brands has been implementing and maintaining these complex models in a fast-paced commercial environment. Zhou's transition from a Big Tech engineering leader to a strategic advisor for a SaaS platform highlights the urgent need to bridge the gap between theoretical data science and practical, daily application.

"Robyn showed that MMM could run at speed and scale," Zhou noted regarding his new advisory role. "The next step is pushing the modelling capacity further towards actionability at the granularity marketers can use every day, week, month and quarter. Fospha's always-on infrastructure is the best place I know to start with."

The commercialization of open-source tools is a familiar trajectory in the tech frontier, but in the adtech space, it is moving at an unprecedented velocity. While Robyn paved the way, other open-source frameworks like Google Meridian and PyMC Marketing have since followed suit. Yet, the raw open-source code still requires a robust infrastructure to ingest daily data, calibrate experiments, and output actionable dashboards—a void that platforms like Fospha are aggressively filling.

AI Agents and the Final Nail in the Last-Click Coffin

The resurgence of MMM is not merely a technological evolution; it is a forced adaptation to a rapidly deteriorating digital landscape. For years, marketers relied on third-party cookies and granular user tracking to stitch together the customer journey. Today, that journey is heavily fragmented by stringent privacy laws, mobile operating system restrictions, and a growing ecosystem of walled gardens.

More disruptive, however, is the rise of AI shopping agents. As consumers increasingly rely on artificial intelligence to research, compare, and even execute purchases on their behalf, the traditional click is vanishing. When an AI assistant buys your weekly groceries, researches the best enterprise software, or books your flights based on a conversational voice prompt, there is no digital breadcrumb trail for a performance marketer to track. The AI acts as an intermediary, severing the direct link between the brand's advertising and the consumer's action. Measurement that depends solely on the click is rapidly becoming obsolete in this new paradigm.

This structural shift explains why Fospha is reporting a staggering 20-fold year-over-year increase in enterprise client demand for its MMM solutions. Brands are desperate for privacy-durable measurement methods that can prove the value of upper-funnel brand building and demand generation without relying on invasive user tracking.

"This is the most exciting moment for MMM in a generation," said Sam Carter, CEO of Fospha. "Marketing teams have never had more measurement, yet Monday's budget calls are still made on last click, because that is the number that refreshes every day. The demand we're seeing from clients for MMM alongside our incrementality testing and Media Mix Attribution solution has been extraordinary, with client demand for MMM alone up 20x year on year. Gufeng has done more than anyone to put MMM in marketers' hands. His decision to join forces with Fospha says a great deal about where this category is heading."

Bridging the Gap: The Rise of Daily, Always-On Measurement

The fundamental flaw of traditional MMM was its latency. A model that tells a Chief Marketing Officer in July that their February campaign was inefficient is essentially useless for modern, agile marketing. Enterprise teams require data at the speed of decision-making.

Fospha's architecture addresses this by providing an always-on system that measures the full funnel across every channel and sales destination. Crucially, this data is refreshed daily, drilling down to the performance of individual ads and creatives. By combining MMM with attribution, incrementality testing, forecasting, and brand impact analytics into a single connected system, the platform effectively replaces the outdated last-click dashboard with a holistic, privacy-safe source of truth.

Industry analysts point out that this shift is not just about better math; it is about changing corporate behavior. As one independent marketing attribution consultant observed, the hardest part of moving away from last-click is the cultural addiction to immediate, deterministic numbers, even when those numbers are fundamentally flawed. Executive boards have been conditioned to expect real-time ROI metrics. By providing daily-refreshed MMM powered by Bayesian calibration, platforms are finally offering a viable, scientifically rigorous alternative that fits into the weekly rhythm of corporate budget allocations.

The ROI of Privacy-Durable Architecture

For industries ranging from consumer electronics to direct-to-consumer nutrition, the return on investment for accurate measurement is massive. Misallocating a multi-million dollar advertising budget because of faulty attribution can severely damage a company's bottom line and cede ground to competitors. With over 11 years of enterprise retail expertise, Fospha's ability to optimize over $40 billion in marketing spend demonstrates the high stakes involved.

Brands like Dyson and Callaway Golf operate in highly competitive markets where the customer journey spans weeks or months, crossing physical retail and multiple digital touchpoints. For these enterprises, understanding the true incremental value of a YouTube ad versus an in-store promotion is the difference between profitable growth and wasted capital. Similarly, for direct-to-consumer pioneers like Huel, navigating the saturated digital ad market requires pinpoint accuracy in budget allocation to maintain sustainable customer acquisition costs.

As the industry looks ahead, the collaboration between open-source pioneers and enterprise software developers will define the next era of advertising technology. Later this month, Carter and Zhou are scheduled to host a joint industry discussion on the future of always-on measurement, a conversation that will likely set the tone for how major brands plan their 2027 budgets.

The era of easy, click-based tracking is definitively over. In its place, a more sophisticated, mathematical, and privacy-respecting framework is taking hold, ensuring that the future of marketing measurement relies on genuine business impact rather than fleeting digital signals.