Does Brand Purpose Move Buyers
Start with Why tells a firm to lead with the belief, and the research finds that announcing the intention is what makes buyers suspect the product was compromised to pay for it. The same benefit, arriving quietly as a side effect, cost nobody a sale.
Answer to the question
Does brand purpose actually move buyers?
A little, and least where it costs money. A fair trade label raised sales of two popular coffees by almost a tenth at an unchanged price, while a price rise of about a tenth cut sales of the cheaper one by roughly 30 per cent. In the largest meta-analysis, 744 effects from 165 articles, willingness to pay was the weakest response measured.
Key takeaways
- The largest synthesis of the question, published in 2026 by Rita Coelho do Vale and two co-authors, pooled 744 effect sizes from 165 articles and 67,270 participants, and found willingness to pay the weakest of every consumer response measured.
- In a United States grocery chain, Jens Hainmueller and two co-authors found a fair trade label raised coffee sales by almost 10 per cent at an unchanged price, while a 9 per cent price rise cut sales of the cheaper coffee by about 30 per cent.
- George Newman and two co-authors found in 2014 that a deliberate environmental benefit lowered purchase interest against the same benefit arriving as an unintended side effect, because buyers inferred that product quality had paid for it.
- Start with Why carries about 105 reference entries and no peer-reviewed journal article among them, with 36 of them websites and 14 of them the author's own interviews.
You are told to lead with the belief. Say why your company exists, then how it works, then what it sells, and buyers will follow the belief rather than the specification. That is the advice a generation of marketers took from Start with Why, and it is worth taking seriously enough to test. So here is the question this piece is about: does brand purpose move buyers, and in which direction? The published record has an answer, and it is not the one on the conference stage.
The advice at its strongest
Simon Sinek's Start with Why, published in 2009 and reprinted many times since, asks a company to describe itself from the inside out. The Golden Circle, his three rings of why, how and what, is a genuinely useful communication heuristic, and we should say so before taking anything apart. A message built from a belief outwards is easier to write, easier to repeat and easier for staff to recognise as their own. Plenty of good communication works exactly that way, and none of that depends on a single claim about anatomy.
Inside the advice, though, sits a proposition that can be checked. Firms that articulate a purpose outperform firms that do not. That has an obvious design, an obvious outcome measure and an obvious failure condition, which is the mark of a researchable claim rather than a slogan. The book never runs it. It never cites anybody who has.
So we went to look for the people who did, and what they found inverts the instruction. The research does not say that a stated brand purpose fails to register. It says that announcing the intention is the part that costs you in how buyers rate the product, and that the same benefit delivered quietly does not.
A chapter title as the evidence
The book's claim to scientific standing is made by a heading. Chapter 4 is titled "This Is Not Opinion, This Is Biology", and page 51 sets the terms before any of it is argued: do not take my word for it, none of this is my opinion, it is all firmly grounded in the tenets of biology. The chapter then maps the three rings onto parts of the brain. We are not going to restate that mapping, because the account it rests on was already out of favour in comparative neuroanatomy before the book appeared, and because nothing in 225 pages cites a source for it.
Four endnotes sit behind that chapter, and we checked all four. One points to a children's picture book. Two point to pages on a computer company's own website, one of them a press release. The fourth points to a trade title about the brain, cited for a different point about overthinking rather than for the mapping. Across the whole book, no anatomical, neuroimaging or psycholinguistic source appears anywhere.
The nearest thing to a scientific warrant is in the acknowledgements, where the author thanks a psychiatrist for helping him understand the neuroscience. A thank-you is not a citation. Our verification pass could not confirm that the man thanked endorsed the mapping, and we do not record him as endorsing it.
Set the book's two statements about its own evidence side by side, 174 pages apart.
> "None of this is my opinion, it is all firmly grounded in the tenets of biology" at page 51, and "there is precedence for this standard" at page 225. The second one is the accurate description.
The irony is that the advice would be in better shape if the chapter defending it had never been written. Nothing useful about describing yourself from the inside out needs a brain diagram, and the diagram is the only part a reader can check and find wanting.
One hundred and five entries
The apparatus is the more interesting failure, because it looks like diligence. There are about 105 reference entries across seven pages, organised by chapter and keyed by page and catchphrase, with no superscripts and no in-text citations anywhere in the running text. Counts rather than shares are printed below, because the classes we could sort come to 104 and the total carries a margin of roughly three either way, since several entries bundle sources under one keyed line.
| Class of source | Entries |
|---|---|
| Website | 36 |
| Newspaper or magazine | 28 |
| Trade book | 20 |
| Other, including the author's own interviews | 19 |
| Academic book | 1 |
| Peer-reviewed journal article | 0 |
One computer company's website is cited five separate times. Fourteen entries are the author's own interviews or correspondence. The single academic book is a trade reissue of Everett Rogers's Diffusion of Innovations. For a book whose central proposition is that purpose drives commercial outcomes, there is not one peer-reviewed article of any kind in the list.
An apparatus can make a book look more careful than a book with no apparatus at all, while carrying less. Compare a title in our vault with 43 entries and no in-text citations, where a reader can see instantly that there is nothing to check. Here the notes are arranged honestly by chapter, which is what makes it the cleaner case, because the reader who does the right thing and looks is the reader who is misled about what they have found.
There is a real distinction in how the sourcing is calibrated, and it deserves credit before it draws criticism. Hard adverse facts about companies are generally sourced: a falsified earnings report to a newspaper, a salary figure to a finance page, a wage settlement to a campaigning document. Claims about what somebody wanted, believed or intended are not. That is a theory about what needs evidence, and it is the wrong one, because a claim about a person's motives is exactly the kind a reader cannot check alone.
A populariser before an originator
The diffusion curve shows the same pattern in miniature. At page 115 the Law of Diffusion of Innovations arrives capitalised and unexplained, as a received thing with no origin, and Malcolm Gladwell is credited in the next paragraph on the same page for The Tipping Point. Rogers is credited properly one page later, with his title and the year 1962, and again at page 120. The populariser gets the byline before the originator does.
Two smaller things matter more than the ordering. Rogers did not call it a law. He described an observed distribution of adoption timings, synthesised from more than five hundred earlier studies, and capitalising it into a law gives an empirical generalisation the standing of physics. The tipping threshold of fifteen to eighteen per cent at page 120 is a hardening of a take-off region rather than a finding of his.
The lineage, for anyone who wants to credit it correctly, runs from Bryce Ryan and Neal Gross in 1943, whose study of hybrid seed corn in two Iowa communities produced the S-shaped curve, to Rogers in 1962, who synthesised it and set the adopter categories and their shares, to Geoffrey Moore in 1991, who owns the chasm and nothing else in the figure, and to Gladwell in 2000. The adopter shares as the book states them are Rogers's and are correct. The numbers are faithful and the status has been reshaped.
Selection on the dependent variable
Here is the structural problem, and it is not that the argument is unfalsifiable. The argument is perfectly testable. It is the evidence that cannot lose, because of how the cases were picked.
The explanatory variable is unobservable and gets read off the outcome. A company that succeeded is found to have had a clear sense of why it exists; a company that failed is found to have lacked one or lost one; and the reading is always done after the result is known.
We looked for the case that would break the pattern, a named company in the book that started with a clear purpose and failed anyway, and could not find one in 225 pages. The nearest candidate, a luxury saloon from a mass-market carmaker at pages 170 and 171, is granted a clear purpose and the car fails regardless, and the failure is then reclassified as a mismatch between what was sold and a purpose that stayed clear. The one case that could have tested the framework is absorbed by it.
A company that succeeded is found to have had a clear sense of why it exists, and the reading is always done after the result is known.
One hedge deserves credit. At page 50 the book concedes that knowing your why is not the only route to success, only the route to lasting success. That concedes purpose is not necessary. It does not concede that purpose might be insufficient, which is the possibility the evidence is arranged to exclude.
The general prior here is also unkind. When Alexander Coppock's team ran 59 randomised experiments testing 49 political advertisements on 34,000 people, the effects on favourability and vote were small, and they did not vary much by who was receiving the message. Values messaging is more intensively researched and more heavily funded in politics than anywhere in commerce, and that is the benchmark a brand purpose claim has to beat.
The distance between attitude and till
The largest synthesis of this question anywhere was published in 2026 by Rita Coelho do Vale and two co-authors, pooling 744 effect sizes from 165 research articles covering 67,270 participants, with articles from 1997 to 2021 and an effort to collect unpublished manuscripts so the file drawer would not flatter the result.
Their headline association between corporate social responsibility and consumer response is moderate and positive. Two details inside it matter more than the headline. Where the effect was estimated from models that hold other variables constant, it was roughly half the size it was in raw correlations, so the smaller of those two figures is the one an honest summary prints. And behavioural responses came out significantly lower than attitudinal ones, which is the attitude-behaviour gap arriving in a pooled estimate rather than in an anecdote.
Rank the outcomes and the ranking is the finding. Company image moves most. Loyalty follows. Perceived product quality comes out significantly lower than other responses. Willingness to pay is the floor, the weakest of every consumer response they measured, at well under half the size of the image effect.
Now the caveat that decides how much of this counts, and it comes from the authors' own coding table rather than from our reading of their prose. Their behavioural category is built from advocacy, engagement, loyalty, purchase intentions and willingness to pay. Purchase intentions and willingness to pay are coded as behaviour. So even the lower behavioural figure is substantially made of what people said they would do.
The largest meta-analysis of brand purpose and its cousins does not measure buying. It measures saying, and it is transparent about that.
The largest meta-analysis of this literature codes purchase intention and willingness to pay as behaviour, so even its lower figure is largely built from what people said they would do.
That distinction is not a technicality, as Pat Auger and Timothy Devinney showed in 2007. Nearly every study of ethical purchasing, they pointed out, asks unconstrained survey questions, where caring costs the respondent nothing and no competing option is on the table. Set those answers against estimates of what people would actually pay for the same social attribute and the survey questions turn out to be too noisy to be operationally meaningful, overstating intentions to a considerable extent. Every purpose statistic in the trade press is of that shape. None of it is citable at the published record, so we cite none of it.
The price of announcing the intention
This is where the instruction inverts, and the mechanism is specific rather than a matter of mood. George Newman and two co-authors at Yale found in 2014 that people were less likely to buy a green product when they believed the company had deliberately made it better for the environment, compared with the same environmental benefit arriving as an unintended side effect. The reason was a lay theory about resources: if effort went into the world, it came out of the product.
Say plainly what that evidence is. All three routes below are laboratory work measuring evaluation, willingness to pay and purchase intention, which is the currency the first half of this piece disqualified. The inversion is therefore a finding about what people say once the intention is disclosed, not about sales, and we hold it at that strength.
1. Inferred resource diversion. Announce the deliberate benefit and buyers infer that quality was sacrificed to pay for it, so purchase interest falls. Newman's team found the effect generalises to other perceived trade-offs, including healthiness against taste, and they report a boundary condition based on the kind of enhancement involved.
2. The category decides the sign. Michael Luchs and three co-authors showed in 2010 that buyers associate ethicality with gentleness and not with strength. Where the valued benefit is gentleness, purpose helps. Where the valued benefit is strength, it can reduce preference below the less sustainable alternative, which they named the sustainability liability. Explicit cues about strength soften it.
3. Motives read as insincere. Yeosun Yoon, with two co-authors, ran three experiments in 2006 on companies with bad reputations, in which social activity improved such a company's image when motives were read as sincere, did nothing when sincerity was ambiguous, and damaged the image when motives were read as insincere. Their load-bearing clause for anybody told to lead with the belief: high prominence of the cause hurts the company, particularly when buyers learn about it from the company itself. Their remedy is to spend more on the thing than on the advertising of the thing.
None of this is new, which is the part that should sting. Sankar Sen and C. B. Bhattacharya asked the question in 2001 under a title that is already the answer, found that initiatives of this kind can decrease purchase intentions under identifiable conditions, and named product quality among the moderators thirteen years before the quality mechanism was isolated. Their own stated motivation was to challenge the marketplace polls, which is the same target Auger and Devinney took on six years later.
Real money and the honest gap
Two studies in this file watched money change hands, and both of them found something smaller and stranger than the slogan. In a United States grocery chain, Jens Hainmueller and two co-authors found that sales of the two most popular coffees rose by almost a tenth when they carried a fair trade label rather than a generic placebo label at the same price. At the premium end, a price rise of 8 per cent was absorbed with demand holding steady. At the value end, a rise of 9 per cent cut sales by about 30 per cent.
Read those three numbers in sequence. Purpose moved buyers when it was free to them, and wherever price mattered the people who would not pay for it outnumbered the people who would.
The second case narrows the mechanism further. Daniel Elfenbein and two co-authors studied eBay sellers who varied whether a listing carried a charitable donation, holding the product otherwise matched, and found that most of the benefit went to sellers without long trading histories, with fewer complaints among the charity-intensive ones. Charity was working as a substitute for a reputation, which is why it was worth most to sellers who had none. Our piece on public charity and who is watching covers the display side of the same behaviour, and our piece on going green where someone can see you covers the version with an audience, and on the brand side the pattern recalls why brand loyalty is mostly brand size.
Then there is the gap we have to report rather than fill. We searched for a peer-reviewed field study measuring a company's stated brand purpose against revenue, as opposed to a single label on one product, and found none. What came back instead was institute material and trade-press argument: a survey reporting that genuine awareness of a brand's purpose rarely exceeds a tenth of buyers, a rebuttal arguing that recall is a different question from effect, a reference to agency work finding purpose campaigns weaker than others, and a statement that a paper covering a thousand campaigns is in preparation.
### What we cannot show
We cannot say that brand purpose fails to move buyers. The pooled estimate is positive, the coffee label sold almost a tenth more at the same price, and charity-tied listings did better than matched ones. The claim available is narrower: the effect is real, smaller at the till than on the questionnaire, smallest of all on willingness to pay, and reversible under three identified conditions.
We also cannot count most of this. Seven of the ten published sources behind this article are closed papers whose abstracts give no sample size, so they can be quoted and not counted. For the fair trade experiment we do not know the number of stores, the duration or the chain. For the eBay study we do not know the number of listings or the size of the premium. Where a number is absent here, it is absent because we could not verify it rather than because it was inconvenient.
And the question that would settle the matter has not been asked in the way that would settle it. No meta-analysis we could find pools observed purchase rather than intention. That is a reportable fact about the field, and it means the stated-against-actual purchase gap is the real evidence a marketer has to work with.
So our position. Describing yourself from the inside out is good communication advice and we would keep it. Leading with the intention, as the proposition rather than the by-product, is the version the evidence does not support, because the announcement is what invites a buyer to wonder what it cost them, on the measures these studies took. Purpose sells best as a by-product and worst as the headline a buyer is asked to pay for.
Common questions
Does brand purpose actually move buyers?
A little, and least where it costs money. A fair trade label raised coffee sales by almost a tenth at an unchanged price in a United States grocery chain, but a price rise of 9 per cent cut sales of the cheaper coffee by roughly 30 per cent. In the largest meta-analysis, willingness to pay was the weakest of every consumer response measured.
Why can announcing a social benefit put buyers off?
Because buyers infer a trade-off. George Newman and colleagues found in 2014 that people said they were less likely to buy a green product when the environmental benefit looked deliberate than when the same benefit arrived as a side effect, since a deliberate benefit implies resources were diverted from quality. Those are stated responses rather than sales. Yeosun Yoon's experiments, on companies with bad reputations, found prominent causes hurt most when buyers hear about them from the company.
Is there a study showing purpose-led brands grow faster?
Not in peer review, as far as we could establish. We searched for a peer-reviewed field study measuring a stated brand purpose against revenue and found none. What exists is institute survey work and trade-press argument reporting unpublished analyses, which we have not cited. The published evidence measures attitudes, intentions and single labels rather than company purpose against revenue.
How should a marketer use brand purpose?
Do the thing before advertising the thing, which is also the remedy Yoon's team identified. Expect the gain where buyers value gentleness and expect a penalty where they value strength. Keep the claim specific to what you actually changed, and measure purchases rather than agreement, because the gap between the two is where most purpose research lives.
Sources
Coelho do Vale, R., Matos, P. V., & de Almeida, F. (2026). 'Meta-Analysis on the impact of Corporate Social Responsibility Initiatives on Consumer Attitudes and Behavior.' BRQ Business Research Quarterly, 29, article 23409444251368391. doi:10.1177/23409444251368391. [FULL TEXT READ, diamond open access at journals.sagepub.com, including methods, Tables 2 and 3 and all moderator results; author order confirmed at Crossref and OpenAlex, Coelho do Vale first, Universidade Catolica Portuguesa; Matos at ISEG, University of Lisbon. Online 21 August 2025, print issue January 2026; Crossref deposits an article number rather than a page range]
Auger, P., & Devinney, T. M. (2007). 'Do What Consumers Say Matter? The Misalignment of Preferences with Unconstrained Ethical Intentions.' Journal of Business Ethics, 76(4), 361-383. doi:10.1007/s10551-006-9287-y. [Abstract read in full at link.springer.com; author order confirmed at Crossref and OpenAlex, Auger first, University of Melbourne. Crossref records print publication November 2007 while OpenAlex dates it 2006; cite 2007. NEITHER Crossref NOR OpenAlex carries a deposited abstract, which was read on the publisher's landing page]
Hainmueller, J., Hiscox, M. J., & Sequeira, S. (2015). 'Consumer Demand for Fair Trade: Evidence from a Multistore Field Experiment.' The Review of Economics and Statistics, 97(2), 242-256. doi:10.1162/rest_a_00467. [Abstract read in full; author order confirmed at Crossref and OpenAlex, Hainmueller first, Stanford; Hiscox at Harvard; Sequeira at the London School of Economics. OpenAlex dates it 2014 online and Crossref's print issue is May 2015; cite 2015]
Elfenbein, D. W., Fisman, R., & McManus, B. (2012). 'Charity as a Substitute for Reputation: Evidence from an Online Marketplace.' The Review of Economic Studies, 79(4), 1441-1468. doi:10.1093/restud/rds012. [Abstract read in full; author order confirmed at Crossref and OpenAlex, Elfenbein first; Fisman second. METADATA WARNING: OpenAlex's record for the journal article names the third author 'Bruce M. McManus', which is wrong; Crossref deposits the initial only as 'B. Mcmanus', and OpenAlex's record for the NBER working-paper version gives 'Brian McManus'. Use the initial or Brian McManus and never print Bruce]
Newman, G. E., Gorlin, M., & Dhar, R. (2014). 'When Going Green Backfires: How Firm Intentions Shape the Evaluation of Socially Beneficial Product Enhancements.' Journal of Consumer Research, 41(3), 823-839. doi:10.1086/677841. [Abstract read in full; author order confirmed at Crossref and OpenAlex, Newman first, all three authors at Yale]
Luchs, M. G., Naylor, R. W., Irwin, J. R., & Raghunathan, R. (2010). 'The Sustainability Liability: Potential Negative Effects of Ethicality on Product Preference.' Journal of Marketing, 74(5), 18-31. doi:10.1509/jmkg.74.5.018. [Abstract read in full; author order confirmed at Crossref and OpenAlex, Luchs first, William & Mary; Naylor at Ohio State; Irwin and Raghunathan at UT Austin; green open access at scholarworks.wm.edu. NOTE the DOI's padded form, .018 and not .18]
Yoon, Y., Gurhan-Canli, Z., & Schwarz, N. (2006). 'The Effect of Corporate Social Responsibility (CSR) Activities on Companies With Bad Reputations.' Journal of Consumer Psychology, 16(4), 377-390. doi:10.1207/s15327663jcp1604_9. [Abstract read in full; author order confirmed at Crossref and OpenAlex, Yeosun Yoon first, Ewha Womans University; Gurhan-Canli at Koc; Norbert Schwarz third. Crossref records print year 2006 with an online date of 25 January 2008]
Sen, S., & Bhattacharya, C. B. (2001). 'Does Doing Good Always Lead to Doing Better? Consumer Reactions to Corporate Social Responsibility.' Journal of Marketing Research, 38(2), 225-243. doi:10.1509/jmkr.38.2.225.18838. [Abstract read in full; author order confirmed at Crossref and OpenAlex, Sen first, Baruch College, City University of New York; Bhattacharya second]
Coppock, A., Hill, S. J., & Vavreck, L. (2020). 'The small effects of political advertising are small regardless of context, message, sender, or receiver: Evidence from 59 real-time randomized experiments.' Science Advances, 6(36). doi:10.1126/sciadv.abc4046. [Abstract read in full; gold open access; author order confirmed, Coppock first, Yale; Hill at UC San Diego; Vavreck at UCLA]
Rogers, E. M. (1962). Diffusion of Innovations. New York: Free Press of Glencoe. NO DOI. [BIBLIOGRAPHIC RECORD ONLY, verified at openlibrary.org, which returns a first-publish year of 1962 with publishers recorded as Free Press of Glencoe, Free Press and Collier-Macmillan, and later editions in 1968, 1971, 1983, 1995 and 2003. THE BOOK HAS NOT BEEN READ, by this pass or by the vault, and that remains a standing gap recorded at the vault's diffusion note]
Ryan, B., & Gross, N. C. (1943). 'The diffusion of hybrid seed corn in two Iowa communities.' Rural Sociology, 8, 15-24. NO DOI EXISTS. [Published record verified at OpenAlex, which carries the year 1943, the journal Rural Sociology, volume 8 and pages 15-24. METADATA CAUTION: the OpenAlex record for the 1943 article lists only 'B. Ryan' as author. The two-author form is the universal citation, and OpenAlex separately carries a 1950 Iowa State University research bulletin, 'Acceptance and diffusion of hybrid corn seed in two Iowa communities', listing Bryce Ryan and Neal Crasilneck Gross together, which establishes the pair. NOT READ]
Sharp, B., Danenberg, N., & Bellman, S. (2018). 'Psychological targeting.' Proceedings of the National Academy of Sciences, 115(34), E7890. doi:10.1073/pnas.1810436115, with Correction, PNAS, 115(39), E9257, doi:10.1073/pnas.1814739115. [FULL TEXT of BOTH READ at pnas.org and PubMed Central, PMCID PMC6166798; author order confirmed at OpenAlex, Sharp first, all three at the Ehrenberg-Bass Institute, University of South Australia]
NOT CITABLE, recorded for honesty because the gap is itself the article's finding: NO peer-reviewed field study measuring the effect of a stated brand purpose on sales was located. One extended WebSearch returned only institute material and trade-press argument, namely an Ehrenberg-Bass survey of about 3,000 respondents reporting that genuine awareness of a brand's purpose rarely exceeds ten per cent across fourteen purpose-led brands; a rebuttal in The Drum arguing that recall is a different question from effect and that the institute's growth piece rested on a handpicked set of five brands; a secondhand reference to Peter Field's IPA work reporting purpose campaigns weaker than non-purpose ones; and a statement that Felipe Thomaz of Oxford's Said Business School has a peer-reviewed paper on a thousand campaigns in preparation. NONE of this is citable at the published record and all of it reached us through trade press reporting other people's unpublished analyses. Record that we looked, name the gap, cite none of it.