What the research says about how brands actually grow
Most content briefs open with a question the evidence does not support: what makes us different? Decades of marketing science point somewhere else. Brands grow by reaching more buyers and by being easy to bring to mind when a purchase occasion arrives, and the measured effect of perceived differentiation is weak.
That matters more now that content is cheap to produce. Volume without consistency builds nothing.
The full citations are in the Brand Growth, Mental Availability & Distinctiveness research index. The papers below come largely from the Ehrenberg-Bass Institute, whose work is built on repeat-purchase panel data across categories and national markets.
Growth comes from more buyers, not deeper loyalty
The Dirichlet model predicts how repeat purchase behaves across competitive categories, and Ehrenberg, Uncles and Goodhardt (2004) report a consistent pattern in the Journal of Business Research: large and small brands differ greatly in how many buyers they have, and far less in how loyal those buyers are.
Loyalty is largely a consequence of size rather than a lever that produces it. The authors describe the pattern holding from soup to gasoline, across the United States, United Kingdom, Japan, Germany and Australasia.
For a content team, this reorders the brief. A campaign designed to deepen the relationship with existing customers is working on the variable the data says moves least.
Sharp and Sharp (1997) tested the point directly. They examined whether a large loyalty program bent repeat-purchase patterns away from their benchmarks and found market structure largely intact, with only weak excess-loyalty effects. The Double Jeopardy line held: small brands have fewer buyers, and those buyers are also somewhat less loyal.
Salience is breadth of recall, not volume of exposure
Romaniuk and Sharp (2004) argued in Marketing Theory that treating brand salience as "top of mind" is too narrow. They reconceptualized it as a brand's propensity to be noticed or come to mind in buying situations, reflecting the quantity and quality of the memory structures buyers hold.
The distinction is practical. Awareness asks whether someone knows the brand. Salience asks how readily it surfaces at the moment of choice, and across how many different occasions.
A brand linked to one situation very strongly is more fragile than a brand linked to eight situations moderately well. Widening that set is a different job from repeating one message louder, and it is the job most content programs are actually doing when they work.
Differentiation is weaker than the brief assumes
Romaniuk, Sharp and Ehrenberg (2007) assembled evidence in the Australasian Marketing Journal that perceived differentiation across competing brands is low, and that brands are bought regardless. Competing brands within a category tend to be similarly differentiated, with small asymmetries between large and small.
This is the finding that most often surprises marketers, so it is worth stating what it does not say. It does not say brands are interchangeable, and it does not say positioning is wasted effort. It says perceived difference is a weak explanation for why one brand outsells another.
What the evidence does support is distinctiveness: being recognizable as yourself. Colors, characters, phrasing and structure that let a buyer identify the brand without reading the logo. Distinctiveness is a memory property, and memory is the mechanism the growth research keeps pointing to.
| Property | What it asks | What the evidence says |
|---|---|---|
| Differentiation | Do buyers see us as different? | Low across competing brands; weak link to performance |
| Distinctiveness | Do buyers recognize this as us? | Supported; builds the memory links that drive retrieval |
| Salience | Do we come to mind when they buy? | Breadth of buying situations matters more than intensity |
| Penetration | How many people buy us at all? | The primary growth variable in Dirichlet benchmarks |
What this changes when AI produces the content
Consistency stops being a style preference and becomes the growth mechanism. If distinctiveness is what builds retrievable memory, then every piece that drifts on voice or drops the brand's recognizable cues is spending budget without accumulating anything.
That is a specific risk in AI-assisted production. A tool that starts each session with no memory of the brand will produce competent work that varies, and variance is precisely what erodes a distinctive asset. Copper Sun carries brand context across projects for this reason, so the twentieth piece still carries the same signals as the first.
The volume itself is not the problem. Undirected volume is. The research gives a directional test that costs nothing to apply: does this piece reinforce a recognizable brand cue, and does it attach the brand to a buying situation it was not already attached to?
If the answer to both is no, more of it will not help.
Frequently Asked Questions
Does this research say positioning does not matter?
It says perceived differentiation is a weaker driver than most brand strategy assumes. Romaniuk, Sharp and Ehrenberg (2007) found consumers perceive competing brands as similarly differentiated, and that brands sell anyway. Positioning still decides what a brand talks about and which buying situations it goes after. The evidence just does not support treating perceived difference as the engine of growth.
Does the Ehrenberg-Bass work apply to B2B marketing?
The Dirichlet and Double Jeopardy findings were established primarily in repeat-purchase consumer categories. Later work from the same institute extends mental availability and category entry points into B2B settings. The reasonable reading is that the memory mechanism generalizes, while the specific loyalty benchmarks came from consumer panels and should be applied to B2B with that caveat attached.
How is brand salience different from brand awareness?
Awareness asks whether a buyer recognizes or recalls the brand at all. Salience asks how readily it comes to mind in an actual buying situation, and across how many such situations. Romaniuk and Sharp (2004) separated the two deliberately, because a brand can post healthy awareness scores while the memory links that operate at the point of purchase stay thin.
What should a content team do differently after reading this?
Audit for consistency before adding volume. Check that the recognizable brand cues survive every piece, then widen the range of buying situations the content speaks to rather than restating one positioning line. The longer-term case for that discipline is in what the IPA Databank says about long-term marketing, which covers the budget split between brand building and activation.