Scaling Marketing Mix model to fit consumer journey, not media planning timelines

Growing e-commerce wanted to maximize investment in frequent 360 campaigns and improve media planning method that no longer matched how consumers actually moved across channels. Decisions were locked in months before the audience signal arrived.

We rebuilt the planning cycle around consumer journey states rather than media flight windows, introduced weekly reallocation rituals between brand and performance, and aligned finance to release budget against tested elasticities.

Within two quarters, 15% of the working media budget was reinvested into higher-return touchpoints — without increasing total spend. More importantly, the marketing, data, and finance teams finally spoke the same language.

Maybe your brand managers and data platform are not on the same team yet, but your marketing channels are playing together, even if you forbid it.

What do you think about co-dependency between marketing channels? Why Your Best Marketing Channel Is Probably Not the One Getting Credit, and Why the Best is not always the same one.

Marketing teams love attribution, clickchains, scenarios, modern technical toys for tracking... Executives ask simple questions in the middle of the meeting when pressured to bring millions EUR worth decision on the table. "Which campaign generated the sale?" "What is our best-performing channel?" "Should we invest more in M or G or some AI?" These questions seem logical. Unfortunately, they are often the wrong questions. Consumers rarely purchase because of a single advertisement. They purchase because dozens of interactions gradually reduce uncertainty until buying feels like the obvious decision that will solve some of their real or imaginary problems.

The sale or mid conversion is not the result of one touchpoint. It is the result of an entire system.

Marketing works like a chain of events, not independent events. Imagine a customer buying a new running shoe. The journey might look like this: listen to a podcast about running and start thinking about it, watch a YouTube review, see an Instagram ad several days later, receive a CRM email from a shop they used ages ago, search Google for reviews, visit the brand website, leave without purchasing, receive a remarketing display ad, walk into a physical store, try the shoes, purchase online two days later.

Which channel deserves the credit? Google Search? Instagram? The retail store? The YouTube creator? The remarketing campaign? The email CRM? The answer is all of them. Could we skip one? Maybe. Maybe not. Only one obvious example is that CRM email won't work if we don't have emails gathered by other channels. Other relations are there, just not so obvious. If a competitor shoe kept the customer in their YouTube podcast, or if remarketing didn't happen, would they convert? We don't know except we scientifically research all the cases.

In an imaginary world we could take unlimited time and money and talk to each customer facing a psychologist to realize if they would buy anyway or never. But in a limited-resources world we need to calculate, extrapolate and predict the probability of scenarios. History is the teacher of life. In this case, teacher of probability models that are based on what happened with previous sales and what could or would happen. We need to go and write down all over-thinking crazy possibilities and analyse. Every interaction reduced uncertainty just enough for the next interaction to matter. Marketing is rarely additive. It is multiplicative.

The Dependency Between Channels. One of the biggest misconceptions in performance marketing is treating every channel as an isolated investment. Silos channels is usually the status when a company opens a new team for marketing analytics. With reason. Silos marketing is like independently preparing ingredients for one meal — not thinking how it fits together, what amount, what form. Try adding favorite food ingredients of ten people on the table, independently, in the amount they eat for snack or TV binge, and preparing a fancy healthy salad or any quality meal. Won't work. Same reason: the secret is in combination and relation, not just all good stuff randomly.

In reality, channels constantly influence each other. Television increases branded search. Social media improves email open rates. Outdoor advertising increases direct website traffic. Influencer campaigns improve paid search conversion rates. Retail presence increases online conversion because customers have already seen the product in person. The performance of one channel often depends on another channel existing. This creates hidden value that traditional attribution models rarely capture.

The Click Is Not the Decision. Digital analytics often gives us a false sense of certainty. A dashboard tells us: "Google Ads generated 1,500 conversions." What the dashboard cannot show is everything that happened before the click. Perhaps the customer had already watched three product videos, discussed the purchase with friends, compared prices, visited the store twice, seen five display ads, received two promotional emails. The final click simply happened to be measurable. It was not necessarily influential. This is known as the "last-click illusion."

The Purchase Moment Happens Earlier Than We Think. One fascinating observation from consumer behavior research is that the psychological purchase often happens before the transaction. Only afterwards do they determine: where, when, at what price, through which device. At this stage, marketing shifts from persuasion to convenience. Search, promotions and checkout optimisation become less about changing the decision and more about reducing friction. Understanding this distinction changes how we evaluate marketing investments. Different scenarios we can see when buying a Ferrari, when the decision was building up for years... and when buying chocolate in the store because it occupied our full visual space of attention, or an online gadget just because we don't have an idea what it is, so we need to have this thing first in class.

Online and Offline Are No Longer Separate Worlds. Consumers do not think in channels. Only marketers do. Neither do they think much about many brands... in FMCG we can see strong advertising fighting for mind space... being top of the mind in a supermarket full of similar things. Would you wait in line to buy a new iPhone, or a new fashion brand dress... but likely not detergent. Convenience wins. It's product, but also convenience of store, delivery, payment. A customer may discover a product on TikTok, research it on Google, inspect it in a physical store, compare prices on Amazon and finally purchase through the brand's mobile app because they already had their login details and don't need to type their card number again. Or because they want to have it delivered from the store and not from a courier service or reseller.

From the consumer's perspective, this is one continuous experience. From the company's perspective, it often appears as five disconnected datasets. The challenge is not creating more campaigns. The challenge is connecting customer behavior across environments — making a proper customer journey.

Measuring Dependencies Instead of Channels. Rather than asking: "Which channel performs best?" A more valuable question is: "Which combinations of channels create the strongest customer journeys?" This shifts analysis from attribution toward dependency. Examples include: Does paid search perform better after TV campaigns? Does email become more effective after website visits? Does social media increase branded search volume? Do customers exposed to both online and offline advertising convert faster? Which sequences consistently lead to purchase? These questions reveal interactions that traditional reporting often misses.

Marketing Mix Is a Network. Marketing mix modelling has traditionally focused on estimating the contribution of different media channels. Today, richer customer-level data allows us to go further. Instead of viewing channels as independent variables, we can model them as a connected network where every touchpoint influences the effectiveness of others. The objective is no longer identifying "the best channel." The objective is understanding the ecosystem that creates demand. First decide what purchase is targeted to occur — impulsive or thoughtful? If thoughtful, we will have data collected.

Consumers do not wake up one morning, see a single advertisement and immediately make a purchase. Trust accumulates. Confidence builds. Risk decreases. Every interaction contributes a small piece of evidence until the purchase feels inevitable. The companies that win are not necessarily those with the biggest advertising budgets. They are the ones that understand how different marketing activities reinforce one another and invest in the customer journey as a whole rather than chasing credit for the final click. In my experience, this strategy saves 10%-20% of marketing budget. In the end, marketing is less about individual campaigns and more about designing a connected system that helps people move naturally from awareness to confidence to action. And MarTech data are about mapping that process precisely and reinvesting budget where it matters the most. Stay tuned to read more about how it actually works through Marketing Mix Modeling (MMM), incrementality testing, multi-touch attribution (MTA), Markov chains, Shapley value attribution and more.