Marketing Mix Modelling for FMCG has been the industry’s measurement workhorse for more than sixty years. But the questions it needs to answer in 2026 are not the questions it answered in 1965 – and most FMCG teams are still running it as if nothing has changed.
If you work in FMCG, you already know what MMM is. You’ve probably commissioned a project, seen the decomposition charts, and used the outputs to support next year’s budget with finance.
This guide is not “MMM 101”. (For that, see our complete guide to Marketing Mix Modelling.) It’s about what MMM needs to do differently for FMCG marketers right now – and how to make sure your model is actually built for the pressures your category is under.
FMCG is structurally different from most other categories using MMM. The pressures on the industry right now are also different from anything we’ve seen in the last twenty years.
Private label is now the default, not the “cheap option”.
In several Nordic and European grocery categories, private label exceeds 35–40% volume share. Branded FMCG is no longer competing on “premium vs budget” – it’s competing on relevance, distinctiveness, and brand equity. Yet a large share of brand budgets is still going into short‑term promotion.
Retailer media networks are pulling money out of brand‑building.
Trade investment used to fund slotting, joint promotions, and price support. Now a growing share goes into retailer‑owned media platforms – optimised on last‑click or last‑basket attribution. The bias towards short‑term, retailer‑attributable activity is strong, and most teams lack an independent view of whether that bias is actually profitable.
Inflation has compressed promotional ROI.
Higher base prices mean deeper discounts are needed to achieve the same uplift – and more of that uplift is borrowed from future periods, not incremental growth. Many FMCG teams are now running promotions that quietly destroy value. Without robust measurement, nobody can see which ones.
Brand equity erodes slowly – and then suddenly.
Years of promotional dependence don’t show up in the numbers quarter to quarter. They show up as a flat base sales line, then a declining one, then a market share problem. By the time the P&L flags it, you are years behind.
MMM is one of the few methods that can quantify all of this in one consistent framework. As digital attribution weakens under privacy constraints, it is also becoming the only reliable way to do so across channels.
These are the questions our FMCG clients come back to again and again – and that no other measurement framework answers with the same rigour.
Most FMCG businesses spend more on trade than on media. In many categories the ratio is 2:1 or higher. Trade is often measured in isolation by the trade team; media by the brand team. They rarely meet on the same axis.
MMM puts them on the same axis. For the same krona, it shows how much incremental sales came from a price promotion vs a TV flight vs paid social vs a retailer display. The first time most teams see this, the conclusion is uncomfortable: a meaningful share of trade spend is subsidising sales that would have happened anyway.
The simplest diagnostic for promotional dependence is the ratio of base vs incremental sales.
MMM is the only tool that gives you this number with credibility. Once you have it, you can have a different conversation with sales, trade and finance about how you grow.
Les Binet and Peter Field’s IPA work suggests that, for most categories, the long‑term optimum is around 60:40 in favour of brand building. Most FMCG brands sit closer to 30:70. Many don’t realise it.
Marketing Mix Modelling lets you:
Every channel has a point of diminishing returns. In FMCG the question is rarely “does TV work?” – it’s “where does TV stop paying back, and what should we do with the next incremental million?”
MMM saturation curves answer this directly. A frequent pattern we see:
New product launches are hard to measure cleanly.
Standard reporting can’t separate these effects. With the right portfolio‑level specification, MMM can.
Not every business is a good candidate for MMM. FMCG usually is.
You already have the data.
Most FMCG businesses have several years of weekly sales data by market and often by SKU, detailed trade spend, and media plans with reach and frequency. That is the foundation MMM needs. Many other categories don’t have it at this depth.
Your mix is broad and mostly non‑clickable.
TV, OOH, sponsorships, in‑store, trade promotion, retailer media, paid digital, organic – the mix is diverse, and a large part of it never produces click data. MMM puts all of it on the same footing.
Aggregate measurement fits how you operate.
FMCG has never relied on user‑level tracking or logged‑in IDs. Privacy headwinds hit DTC and app‑based models much harder than grocery retail. MMM has always been the right tool here; the difference now is that alternatives are weakening.
Brand equity is your main defence.
In FMCG, the “base” component of MMM is essentially a financial expression of brand equity. In a world of strong private label, that is your moat. Knowing how choices in media and promotion affect that base is critical.
Across our FMCG clients, Nepa’s MMM projects typically deliver double‑digit improvements in marketing ROI from reallocation alone – before any extra budget is added.
The pattern is consistent:
Hemköp, one of Sweden’s leading grocery retailers, had high brand recognition but limited differentiation in a crowded market.
Working with Nepa, Hemköp combined Marketing Mix Modelling with a Brand Asset Study and Campaign Evaluation:
The outcome was a clearer picture of Hemköp’s position in shoppers’ minds, a rebalanced plan that protected equity‑building activity, and a budget argument anchored in both sales and brand metrics. Read the full case here.
Standard MMM, applied without category sensitivity, often produces misleading conclusions in FMCG. The main issues we see:
Trying to answer all three in a single model forces unhelpful compromises.
If you are commissioning your first Marketing Mix Modelling for FMCG, or upgrading an existing one, sequence matters.
The main gaps tend to be distribution data and competitor inputs – both solvable via panel data (e.g. Nielsen, IRI) or equivalent.
MMM is not new to FMCG. It is the foundation. But many organisations are still running a version designed for another era – fewer channels, simpler competitive dynamics, slower cycles.
The FMCG marketers getting ahead now are those who have upgraded:
The reward is what we typically see across our client base: measurable double‑digit ROI uplift, a defensible case for brand investment, and budget decisions made with the same confidence finance has when they read the P&L.
Nepa works with leading FMCG and CPG brands across Europe, combining Marketing Mix Modelling with brand tracking and campaign evaluation. Our FMCG clients typically see significant ROI improvements from better allocation alone.
How is MMM different in FMCG vs other industries?
FMCG businesses usually have richer historical data, more diverse channel mixes (incl. above‑the‑line and trade), and a stronger reliance on brand equity. MMM for FMCG needs to decompose trade promotions, account for distribution shifts, and isolate long‑term brand effects – standard “template” MMM rarely does this well.
Can MMM measure trade promotion ROI?
Yes. In FMCG this is often one of the most valuable outputs. A well‑specified model separates price, display and feature effects to show which mechanics drive incremental volume and which primarily subsidise existing sales.
Does MMM work at SKU level, or only at brand level?
Both, depending on the decision. Brand‑level modelling is usually right for media optimisation. SKU‑level is key for pricing, promotion design and cannibalisation within a portfolio. Many FMCG organisations run both, at different cadences.
How does MMM handle private label competition?
By including competitor variables – e.g. private label share, price gaps, category promotional intensity – MMM can isolate the impact of private label on your performance. Without these, private label effects are typically misattributed to your own marketing.
How often should an FMCG brand refresh its MMM?
At minimum annually; quarterly is better. Many FMCG businesses now move toward monthly refresh for key categories, treating MMM as an operational decision tool rather than a yearly exercise.
Can MMM measure the impact of an NPD launch?
Yes – with appropriate portfolio‑level modelling and launch flags, MMM can separate category growth from cannibalisation. It’s a challenging area and needs specific model design, but it is solvable.
How much data do we need?
Two to three years of weekly data is a sensible minimum; three to five is preferable in seasonal categories or when brand equity effects are in scope. Most FMCG companies already have this – the main challenge is integration, not availability.
Is MMM a replacement for shopper marketing measurement?
No, it’s complementary. MMM measures the contribution of media and trade to total brand sales. Shopper‑specific tools (e.g. in‑store tests, loyalty data analysis) measure behaviour at the point of purchase. Used together, they give a full‑funnel view of effectiveness.
WHY CHOOSE US
Leverage our advanced analytics to light up your strategic journey and turn complex data into clear, actionable insights that drive profitable outcomes.
Resource hub