marketing analytics
PR Newswire
Published on : Oct 8, 2026
Fospha has appointed Gufeng Zhou, creator of Meta's open-source Robyn marketing mix modelling framework, as a strategic advisor as the marketing measurement company pushes its MMM capabilities toward always-on, calibrated and more actionable measurement. The appointment gives Fospha access to one of the better-known figures behind modern open-source MMM at a time when marketers are looking beyond last-click attribution for ways to measure brand building, demand generation and cross-channel performance.
Marketing mix modelling is moving into a different phase.
Once largely associated with periodic measurement exercises used for major budget planning, MMM is increasingly being positioned as an operational layer that can inform decisions throughout the year. Fospha is betting on that shift, bringing Gufeng Zhou, creator of Meta's open-source Robyn framework, into the company as a strategic advisor.
Announced during Adweek and effective October 7, Zhou will advise Fospha on its MMM programme and product innovation around always-on and calibrated measurement. His work will include helping the company expand its MMM capabilities across its client base.
The appointment is significant because Robyn helped make modern MMM more accessible to advertisers, agencies and analysts. Developed inside Meta and released as an open-source project in 2021, Robyn combines automated modelling with techniques for exploring media effectiveness, adstock and saturation, and budget optimization. Its GitHub documentation describes it as a semi-automated MMM package designed particularly for digital and direct-response advertisers working with granular datasets.
Zhou's move to Fospha comes as the commercial measurement market is trying to solve a familiar problem: traditional attribution can be fast but narrow, while conventional MMM can provide a broader view of marketing effectiveness but has historically operated on slower reporting cycles.
Fospha's approach is to close that gap with what it calls Daily MMM. Its Core platform refreshes modelled measurement every 24 hours and provides outputs down to the ad level, while incorporating cross-channel and full-funnel signals. The company also combines MMM with incrementality testing, using experimental results to calibrate its modelling.
That architecture matters as marketers face increasingly fragmented customer journeys. Purchases can span paid media, organic discovery, marketplaces and other commerce environments, while privacy restrictions have reduced the reliability of user-level tracking.
Fospha argues that its measurement system can provide a broader view by combining aggregate modelling with deterministic signals and incrementality testing. Its own documentation distinguishes MMM's role in understanding broader channel contribution from incrementality testing's role in establishing causal evidence for specific budget decisions.
The company says demand for its MMM offering has increased 20-fold year over year, according to CEO Sam Carter. That is a company-reported figure rather than independent market research, but it illustrates the commercial opportunity Fospha sees in bringing MMM closer to day-to-day media decisions.
For Zhou, the next step for MMM is similarly about actionability. Robyn demonstrated that MMM could operate at greater speed and scale; the challenge now is making modelling outputs useful at the cadence marketers actually manage budgets.
That distinction could become increasingly important as AI changes both marketing execution and measurement. Automated systems can make optimization decisions continuously, but those systems still require reliable signals about incremental business impact. A model that updates quarterly is difficult to use as the foundation for automated budget decisions.
Fospha has already positioned its Daily MMM as part of that infrastructure. Its platform says models are retrained daily and that outputs can feed measurement, forecasting and optimization workflows.
The competitive landscape is also expanding. Google's Meridian and PyMC Marketing have joined Robyn in bringing modern MMM approaches to practitioners, increasing the availability of open-source and accessible modelling frameworks.
The bigger question is therefore no longer whether MMM can be made more accessible. It is whether MMM can become sufficiently frequent, granular and causally grounded to sit alongside the systems making everyday marketing decisions.
Zhou's appointment puts Fospha directly into that transition.
Rather than treating MMM as a quarterly reporting exercise, the company is attempting to turn it into an always-on measurement layer that connects strategic budget allocation with daily marketing optimization. If that model gains traction, the boundary between measurement, forecasting and media execution could become considerably less distinct.
The marketing measurement stack is increasingly being built around multiple methods rather than a single attribution model.
Fospha's own framework combines Daily MMM, incrementality testing and platform-level signals, with each serving a different measurement purpose. MMM provides the broader cross-channel view, incrementality testing supplies experimental evidence, and platform signals provide fast tactical information.
Open-source frameworks such as Meta's Robyn have helped expand access to MMM, while Google's Meridian and PyMC Marketing have added further alternatives for organizations building modelling capabilities.
The commercial opportunity is increasingly around operationalizing these techniques. Fospha's platform, for example, emphasizes daily model refreshes, ad-level granularity and the ability to connect measurement outputs to optimization workflows.
The strategic significance of Zhou joining Fospha is less about an individual hire than about where marketing measurement is heading.
As privacy constraints weaken user-level attribution and customer journeys spread across websites, marketplaces, social platforms and emerging AI-mediated experiences, marketers need measurement methods that do not depend entirely on deterministic click paths.
MMM provides one route, but its usefulness increasingly depends on speed and calibration. The next generation of measurement platforms will likely need to combine modelling, experimentation and high-frequency signals rather than force marketers to choose between attribution and MMM.
That creates a potential role for always-on MMM as the connective layer between strategic planning and operational decision-making.
Zhou helped popularize an accessible, automated approach to MMM through Robyn. His move to Fospha now places that expertise inside a commercial platform attempting to make MMM a daily operating capability.
The result could be a broader shift in how enterprise marketers think about measurement: from asking which channel received credit to asking which combination of marketing activity is actually generating incremental growth — and where the next dollar should go.
Get in touch with our MarTech Experts.
Looking to publish a press release, guest article, interview or podcast? Connect with us.
GET FEATURED