Most marketing science gets dragged into planning meetings as if it arrived with a warranty and a user manual.
Someone quotes a law, someone else turns it into a rule, and five minutes later the average behaviour of the market is being treated like a strategy. That’s a category error. It can make teams arrogant or fatalistic, sometimes both.
The prescription leak
The trap is the prescription leak: you take a descriptive pattern, then smuggle it into planning as a rule you must obey. You end up copying the average behaviour of the market, or despairing that the average behaviour of the market is fate. Both outcomes are very efficient ways to protect yourself from making a real call.
What the models actually do
A lot of the famous “laws” are built on models designed to describe what people already do at scale, across time, under normal conditions.
Descriptive models are useful because they show recurring patterns. They are weaker when you ask them to tell you how to become an outlier.
Markets with entrenched habits produce repeatable patterns because availability, recall, routines, distribution, and price architecture are doing much of the work.
So yes, in mature categories, share tends to drift slowly and fight you back towards equilibrium. Big brands usually have structural advantages in penetration, salience, distribution, and buying occasions.
A strong central tendency means growth usually requires time, unusual force, or both.
If you want a depressing stat, evidence summarised by WARC from Ehrenberg-Bass suggests 20% of very small brands grow in a single year, but only 3% sustain that growth over three years.
That’s an odds table, not destiny.
Why “descriptive” gets misread as “definitive”
Teams read the research as prescriptive because everyone in the room is paid to reduce uncertainty.
The CFO wants a number that implies control, agencies want repeatable playbooks, and marketing ops wants a dashboard that turns ambiguity into a dropdown.
So a descriptive generalisation becomes a compliance framework. Then you start confusing “normal” with “necessary”.
There is also a nastier second order effect. When enough teams apply the same “laws” in the same way, the category can become more homogenous, which strengthens the equilibrium they are trying to escape.
You get convergence disguised as discipline.
Where lift actually comes from
Lift shows up when you stop treating the average as the only acceptable version of reality and start looking for conditions where the forces change. Sometimes that means excess investment over long horizons, because share of voice and share of market are linked closely enough that sustained imbalance can move share, even if it moves slowly. Sometimes it means distinctiveness that travels, because people notice, remember, and repeat what is easy to carry in the head and share with other people. Sometimes it is distribution rather than comms, like a new route into buying situations that incumbents do not cover well.
Kantar’s work argues that brands that are meaningfully different to more people command 5x market penetration today, and their claim is directionally useful even if you want to argue about definitions.
The research tells you what the default outcome will probably be if you do nothing special. The opportunity still has to come from the plan.
A concrete operational example
Imagine a mid market B2B SaaS in a mature category, selling into a total addressable market of roughly 2,000 viable firms, with a six month sales cycle and a paid media budget of £40k a month.
The team reads penetration research and decides the “scientific” move is to go broad. They widen targeting, increase reach, and tell themselves the market will do the rest.
Three months later, the pipeline is noisy, sales says lead quality has dipped, and leadership decides the research was overhyped, because humans love one clean explanation for a messy system.
The fix is to treat reach as one condition in the plan rather than the strategy itself.
So they pick one lift hypothesis that changes the conditions rather than just the spend, and they operationalise it as a constraint-based plan.
They build a partner channel with three platforms already embedded in the buying workflow. They ship a product interface change that makes sharing an output the default action. They standardise two recognisable cues that appear in every shared artefact so recall compounds.
They track category coverage instead of weekly ROI, aiming for 60% of the 2,000 accounts to have at least one encounter per quarter. They accept the trade that short term efficiency will look worse while memory and distribution build.
After two quarters, the graph is still not miraculous. But the sales team starts hearing “I keep seeing you” and “you’re turning up in places you didn’t before”, which is what lift sounds like when it is still small.
Why this keeps happening
It keeps happening because “science” is used as a social weapon inside organisations.
Inside organisations, “evidence based” can become a way to shut down debate. “Growth is hard in mature markets” can become a way to avoid committing to a risky bet, which is why research can become an alibi instead of a decision aid.
There is also a measurement incentive. Teams prefer metrics that move weekly, so they optimise what twitches, not what accumulates.
Binet and Field’s work is largely an argument about how short-term and long-term effects behave differently, and why short-term metrics can pull teams away from the effects that accumulate.
The share of voice story, which many teams repeat as if it were a law of motion, is a model with assumptions and limits.
Decisions that respect the model
Choose whether you are playing a time game or a moment game. The measurement window and the creative approach will differ, and you cannot optimise both at once.
Decide what “coverage” means in your category, because broad reach without buying situation relevance is just expensive cardio.
Decide which constraint you will accept, because every lift mechanism has a cost. That might be slower payback, less precise targeting, or higher creative discipline.
Decide what you will standardise for recognition, because constant “refreshing” can trade short term novelty for long term forgettability.
Decide what would count as evidence you are finding lift, because otherwise you will keep flipping between cynicism and reinvention every time the numbers wobble.
What to watch if you want to call it science
Watch whether your plan assumes buyers are making fresh decisions or mostly repeating patterns with occasional disruption.
Watch whether your reporting forces a stop, start, or continue decision, or whether it mostly exists to keep everyone calm.
Watch whether your “strategy” is just the market average with nicer formatting, because the equilibrium loves a well designed slide.