Most teams measure the channels that are easy to track and assume the rest. That over-credits digital and starves brand and offline. The fix is marketing mix modelling for the top-down view across every channel, with geo-lift and holdout experiments as the tie-breaker when the numbers disagree.
Your television campaign runs. Search volume climbs the same week. The search click takes the credit, the television spend looks weak, and next quarter someone proposes cutting the television budget. The channel that did the work loses the argument.
Why the easy-to-track channels win the budget
Measurement follows the click. Digital carries a cookie and a conversion event, so its contribution is easy to log and easy to defend in a meeting. Television, radio, print and out-of-home carry neither. They move people, then hand the visible action to a channel that can be tracked.
So the trackable slice gets over-credited. Search, retargeting and paid social absorb the credit for demand that brand and offline created. Then the budget follows the credit. You fund the channel that recorded the last click, not the channel that started the buying.
Digital-native teams have the sharpest blind spot. Many do not register offline influence at all, because their whole stack is built to count clicks. If a channel does not fire a pixel, it does not exist in the dashboard. What you cannot see, you cannot fund, so you defund it by default.
This is the same failure as multi-touch attribution, scaled up. Attribution mis-credits touches inside the digital funnel. This gap mis-credits whole channels that never touch the funnel at all.
The evidence
Only 32% of marketers globally, and 23% in Europe, measure spend across both digital and traditional channels (Nielsen, 2025, from a survey of 1,400 respondents). So roughly two in three are steering the whole media budget on a partial view. They optimise the trackable third and guess the rest.
That is not a small reporting gap. It is a structural bias in where the money goes. Every quarter the measurable channels compound their advantage, because they can prove a number and the others cannot. The budget drifts toward what is easy to count, not what works.
Is this you?
Five quick checks. Answer each yes or no.
- Does your reporting cover paid search and social in detail, but treat television, radio or out-of-home as a lump you cannot evaluate?
- When brand or television spend goes up, do you see search and direct traffic rise but credit those to their own channels?
- Has anyone proposed cutting an offline channel mainly because it is hard to measure?
- Do you run any controlled experiment, such as holding a channel back in some regions, or only before-and-after reads?
- Can you state, with evidence, what each channel contributed to last quarter’s sales?
Three or more uncomfortable answers means your budget is following your measurement, not your results.
What the gap costs
The cost is misallocation, and it hides because every number in the dashboard looks right on its own.
You over-fund the trackable channels because they can prove a return. You under-fund brand and offline because they cannot, so their contribution decays until the number finally looks as weak as the measurement always claimed. That becomes a slow, self-inflicted decline dressed up as data-driven discipline.
You also make confident decisions on a third of the picture. Cutting a channel that quietly drives demand elsewhere looks like a saving in the model and lands as a revenue drop you struggle to explain. And because the read is biased the same way every quarter, the error does not average out. It compounds.
Two ways to see the whole picture
The fix is not better click tracking. You cannot pixel your way to measuring television. You need methods built to work without user-level data.
Model the whole mix, top down. Marketing mix modelling relates spend across every channel, including offline and brand, to sales over time. It reads historical data, so it does not need a cookie or a login. It gives you a strategic view of what each channel contributed, together, which is the view the click-based stack can never produce. Open-source tooling such as Meta Robyn, Google Meridian and PyMC-Marketing has taken the entry cost close to zero.
Test with experiments when the numbers fight. Models disagree with attribution dashboards. When they do, do not argue, run a test. Geo-lift and holdout experiments hold a channel back in some regions and run it in others, then compare outcomes. That gives a causal read, not a correlation, and it is the tie-breaker that settles which method to trust.
One honest caveat. This is a strategic bet, not a quick tidy-up. Marketing mix modelling needs analytical skill and clean historical data, and if the history is patchy the model inherits the mess. That loops straight back to data quality, which is why measurement projects so often stall before they start.
Get the full cross-channel measurement playbook.
Fund what works, not what is easy to count
The channels that are hardest to measure are not the weakest. They are just the quietest in the dashboard. Measure only what is easy, and you will keep defunding the work that starts the buying, one confident quarter at a time.
Build a view that sees every channel at once. Then let experiments settle the arguments the models cannot. The goal is not a tidier report. It is putting the next pound where it actually earns.
Go deeper on customer data maximization
Three ways forward. Pick the one that fits where you are.
- Get the playbook. Practical notes on turning the customer data you already own into revenue, straight to your inbox. Join the newsletter at the foot of this page.
- Take the assessment. Score your customer data maximization in four minutes and see your top revenue blockers. Start the assessment →
- Book a meeting. Bring your data problem. Leave with a prioritised fix, not a platform pitch. Book a call →
Post 9 of 25 in the Customer Data Maximization series. Previous: What is the dark funnel, and how do you measure pipeline you cannot see?. Next: Why the MQL is costing you revenue, and what to measure instead.
Frequently asked questions
Why does digital marketing get over-credited?
Digital carries a click and a cookie, so its contribution is easy to log. Television, radio, print and out-of-home carry neither. When television drives a search and the search click takes the credit, the trackable channel looks stronger than it is and gets more budget than it earned.
What is marketing mix modelling?
Marketing mix modelling, or MMM, is a top-down statistical method that relates spend across every channel, including offline and brand, to sales over time. It measures each channel's contribution from historical data without needing user-level tracking, so it works where cookies and clicks do not reach.
How many marketers measure both digital and traditional channels?
Only 32% of marketers globally, and 23% in Europe, measure spend across both digital and traditional channels, according to Nielsen's 2025 survey of 1,400 respondents. Most measure the channels that are easy to track and assume the rest, which quietly starves brand and offline.
What is a geo-lift experiment?
A geo-lift experiment holds a channel back in some regions and runs it in others, then compares outcomes. It is a controlled test rather than a model, so it gives a causal read on what a channel actually caused. Use it as the tie-breaker when models and attribution disagree.
Do I need expensive software for marketing mix modelling?
No. Open-source tools such as Meta Robyn, Google Meridian and PyMC-Marketing have lowered the entry cost to near zero. The remaining cost is skill and clean historical data. The maths is free now. The analytical judgement and the data discipline are what you still have to invest in.