Evidence-Based Marketing: the Science vs Hype Manifesto (2026)

In summary: Evidence-based marketing filters practices via triangulation: peer-review + public data + meta-analyses. Sharp, IPA, Cialdini and Kahneman converge on a few measurable principles. This manifesto debunks 12 widespread myths (Trout/Ries positioning, Sinek’s golden circle, Maslow’s pyramid) and lists 10 operational moves to allocate budget on data, not rhetoric.

Abstract / Summary

Marketing is overrun by gurus, non-replicated frameworks, “laws” turned into dogma. This manifesto separates what peer-reviewed research confirms from what is folklore sold for $49 in PDF. Seven points of method, twelve myths debunked with bibliographic references, an operational compass for entrepreneurs and marketers who want to allocate budget on data, not rhetoric.

1. What “evidence-based marketing” means

Evidence-based marketing is the approach that subjects every claim — about consumers, brands, advertising, pricing — to the same grid that evidence-based medicine applies to treatments: peer review, replication, effect size, control for confounders.

It does not mean rejecting the experienced marketer’s intuition. It means that, when an intuition contradicts 996 replicated case studies (IPA Databank), you update the intuition, not the data.

Three operational criteria to qualify a claim as evidence-based:

1. Publication in a peer-reviewed journal (Journal of Marketing, Journal of Marketing Research, Marketing Science, JCR, Applied Statistics) or industry databases with transparent method (IPA Databank, Nielsen, Kantar, System1)
2. Independent replication in at least one different population/category/year
3. Effect size declared (R², Cohen’s d, lift %) — not just “p<0.05” on inflated samples

Without these three, a claim is a sellable opinion, not science.

2. The five scientific sources of marketing 2026

Before debunking myths, here are the five sources we use as a compass — all peer-reviewed or with replicable public methodology.

2.1 Ehrenberg-Bass Institute (Adelaide)

Founded by Andrew Ehrenberg, today led by Byron Sharp and Jenni Romaniuk. Database of purchase patterns across hundreds of categories, decades, countries. From here derive:

• Double Jeopardy Law (Ehrenberg, 1959, Applied Statistics)
• Penetration > Loyalty (Sharp, How Brands Grow, 2010)
• Distinctive Brand Assets (Romaniuk, Building Distinctive Brand Assets, 2018)
• Light Buyer Dominance (Sharp & Romaniuk, How Brands Grow Part 2, 2016)

Replicated across >100 categories, >50 countries. Baseline standard for anyone measuring brand growth.

2.2 IPA Databank (Institute of Practitioners in Advertising, UK)

Collection of advertising case studies with measured business outcomes. Binet & Field, The Long and the Short of It (IPA, 2013) analyzed ~1,000 cases and produced the 60/40 rule (60% budget brand, 40% performance) — replicated in Effectiveness in Context (2018) and The Crisis in Creative Effectiveness (2019).

2.3 System1 + Kantar Test Your Ad

System1 has tested >100,000 creatives with emotional method (FaceTrace + Star Rating 1-5★). Published in collaboration with IPA. Results: emotionally charged creatives generate 3-5× the long-term ROI of rational creatives. Data replicated year over year.

2.4 Cialdini meta-analyses

Robert Cialdini, Influence (1984, updated 2006) and Pre-Suasion (2016). Seven principles (reciprocity, commitment, social proof, authority, liking, scarcity, unity) supported by hundreds of replicated randomized experiments. Use with care: large sample sizes but effects often smaller than pop-marketing suggests.

2.5 Kahneman & behavioral economics

Daniel Kahneman, Thinking, Fast and Slow (2011), summarizes 40 years of research with Tversky. Framing effects, prospect theory, anchoring: replicated in official replication labs (Open Science Collaboration, 2015 — less robust than expected, but framing/anchoring hold).

These five sources cover ~80% of operational marketing decisions. Everything else should be treated with skepticism until it satisfies the three criteria of § 1.

3. Twelve myths debunked with sources

Myth 1 — Maslow’s pyramid predicts consumer behavior

Popular claim: needs are satisfied in hierarchical order, from physiological to self-actualization. Verdict: false as a predictive model.

Maslow (1943, Psychological Review) proposed the theory on anecdotal basis, without quantitative data. Wahba & Bridwell (1976, Organizational Behavior and Human Performance 15(2), 212-240) reviewed 10 empirical studies and found zero evidence of the hierarchy: needs are pursued simultaneously, not in sequence. Popular in business school courses by didactic inertia, not for validity.

Use instead: Jobs To Be Done (Christensen) or Category Entry Points (Romaniuk) — both with empirical foundation.

Myth 2 — The 22 Immutable Laws of Marketing (Ries & Trout)

Popular claim: 22 universal and immutable principles govern marketing. Verdict: aphorisms, not scientific laws.

Ries & Trout (The 22 Immutable Laws of Marketing, 1993) present rules like “the law of focus” or “law of the opposite” as universal. Sharp (How Brands Grow, 2010, ch. 2-3) empirically demonstrates that the “law of leadership” (being first matters more than being better) does not replicate on market data — brands grow through mental and physical availability, not entry order.

Myth 3 — The AIDA funnel describes the decision process

Popular claim: the consumer goes through Attention → Interest → Desire → Action. Verdict: didactically useful, operationally false.

E. St. Elmo Lewis published AIDA in 1898. It has never been empirically validated as a description of real behavior. Google (Decoding Decisions: Making Sense of the Messy Middle, 2020) tracked 31,000 purchase paths: the consumer executes non-linear cycles of exploration + evaluation, on average 95 touchpoints per considerable category.

Operational: use CEPs (Category Entry Points) to plan communication, not funnels.

Myth 4 — Buyer Personas predict behavior

Popular claim: an archetypal persona (“Mark, 35, manager”) drives targeting. Verdict: useful for UX, harmful for advertising/segmentation.

Cooper (The Inmates Are Running the Asylum, 1999) introduced personas for software design, not for marketing. Sharp & Romaniuk (How Brands Grow Part 2, 2016) empirically demonstrate that light buyers (~80% of revenue) do not fit any persona — they are heterogeneous and occasional. Persona-based targeting reduces effective reach.

Operational: CEPs + Distinctive Assets, not personas.

Myth 5 — Loyalty programs drive growth

Popular claim: retention costs less than acquisition, so loyalty programs grow revenue. Verdict: false. Growth comes from new customers, not loyal ones.

Sharp (How Brands Grow, ch. 9-10) replicated on FMCG, retail, banking: loyalty programs reward already-loyal customers, they do not increase market share. Effects measured on penetration: 0-3% (noise). Top 20 brands grow through penetration, not loyalty.

Myth 6 — Net Promoter Score (NPS) predicts growth

Popular claim: NPS > 50 = guaranteed growth (Reichheld, HBR, 2003). Verdict: the “ultimate question” promise does not replicate.

Keiningham, Cooil, Andreassen, Aksoy (2007, Journal of Marketing 71(3), 39-51) replicated Reichheld’s original claim across multiple categories: NPS-growth correlation varies from R² = 0.04 to R² = 0.24 — mediocre, not predictive. Morgan & Rego (2006, JM 70, 426-39) confirm: customer satisfaction predicts growth better than NPS.

Operational: use NPS only as a relative thermometer, never as “ultimate KPI”.

Myth 7 — Storytelling always sells

Popular claim: stories activate oxytocin and drive purchases (Paul Zak). Verdict: true for long-term brand building, false for direct response.

Zak (2015) proposed the oxytocin-narrative mechanism with small samples. Replication attempts (e.g. Lin et al., 2013, neuro-marketing variations) find smaller effects. IPA Databank: emotional storytelling works for brand campaigns at 6+ months horizon. For direct response (immediate sale), rational-functional creative beats narrative.

Myth 8 — fMRI neuromarketing is the “real” consumer test

Popular claim: fMRI reveals what consumers “really think” beyond declared. Verdict: often oversold compared to simpler methods.

Plassmann, Ramsøy, Milosavljevic (2012, J. of Consumer Psychology 22(1), 18-36) reviewed the literature: fMRI has scientific relevance but its cost/decision-utility ratio is low vs eye-tracking, FaceTrace System1, A/B creative tests. Global neuromarketing market 2026 ~$3.8 billion (Statista) — heavily vendor-driven.

Myth 9 — Pricing always ends in 9 (charm pricing)

Popular claim: $9.99 sells much more than $10. Verdict: real but small effect, not universal.

Anderson & Simester (2003, Quantitative Marketing and Economics 1(1), 93-110) — field experiment on a catalog: prices ending in 9 increased sales ~24% only in some segments, neutral or negative in others (luxury, B2B). The effect depends on context, not automatic.

Myth 10 — Social proof always works

Popular claim: “9 out of 10 doctors recommend” is enough to persuade. Verdict: depends on the fit between proof type and context.

Cialdini (Influence, 2006) warns: social proof works when the audience is uncertain AND perceives “others” as similar. In expertise contexts (technical B2B), generic social proof can reduce credibility (Aaker & Keller, JM, 1990). Trust signals (peer review, certifications) beat generic social proof.

Myth 11 — Creativity is “soft”, you can’t measure it

Popular claim: “ROI of creativity is impossible to quantify”. Verdict: it is measured regularly. See System1 + IPA.

Binet & Field (The Crisis in Creative Effectiveness, 2019) measure systematic decline in creative effectiveness post-2008: campaigns winning Cannes Lions today generate ~50% lower effect size compared to 2002-2008. Creativity is measured — the problem is that it is no longer evaluated with rigorous methods.

Myth 12 — Brand is “intangible”, performance is “data-driven”

Popular claim: brand investing is an act of faith, performance is science. Verdict: inverted. Brand investing has more empirical literature than pure performance.

Field (2017, IPA Marketing Effectiveness in the Digital Era) and Binet (Effectiveness in Context, 2018): allocation 60% brand / 40% performance maximizes 3-year effect in 96% of the 996 cases of the IPA database. Pure performance under-invests on the asset (mental availability) that fuels future campaigns.

4. What actually works (the six evidence-based rules)

Debunking myths is half the work. The other half is stating what research confirms with robust replication.

1. Penetration > Loyalty — growth comes almost entirely from newly acquired customers. (Sharp, How Brands Grow, 2010)
2. Mental Availability + Physical Availability — brand wins when remembered at purchase occasion AND available at point of sale/channel. (Romaniuk, 2018)
3. Distinctive Brand Assets (logos, jingles, mascots, palettes) — encode the brand in memory; consistency over 5-10 years amplifies ROI. (Romaniuk, Building Distinctive Brand Assets, 2018)
4. 60/40 Brand/Performance Split — maximizes 3-year effect size in 96% of 996 IPA cases. (Binet & Field, 2013, 2018)
5. Light Buyer Dominance — 50%+ of category buyers purchase 1× or 2× per year; segmentation that ignores them cuts half the revenue. (Sharp & Romaniuk, 2016)
6. Emotional creative > rational for long-term, rational-functional > emotional for short-term — System1 + IPA replicated 10+ years.

Allocate 80% of budget and attention on these six patterns; the remaining 20% is category-and-context specific.

5. How to recognize a reliable source

Four operational filters to separate research from hype.

Filter 1 — Explicit peer review A claim in a free PDF, guru e-book, podcast: zero weight. Same claim in Journal of Marketing or Marketing Science: scientific weight. Look for DOI, journal, year.

Filter 2 — Sample size + declared method “We analyzed 1,000 brands” — which? Random sampling or cherry-picking? Replicable method? Accessible database? IPA, Nielsen, Kantar publish methodology. Guru-PDFs usually don’t.

Filter 3 — Independent replication A solid claim has been replicated by other teams, in other countries/categories. “Immutable laws” that exist only in the original book, never replicated, are aphorisms.

Filter 4 — Effect size + confidence interval “Statistically significant” is not enough — on large samples every difference becomes significant. Look for R², lift %, Cohen’s d. If the source declares only “p<0.05” it is uninformative.

6. Base bibliography (15 books + 10 papers)

Books (15)

1. Sharp, B. (2010). How Brands Grow. Oxford University Press.
2. Sharp, B. & Romaniuk, J. (2016). How Brands Grow Part 2. Oxford UP.
3. Romaniuk, J. (2018). Building Distinctive Brand Assets. Oxford UP.
4. Binet, L. & Field, P. (2013). The Long and the Short of It. IPA.
5. Binet, L. & Field, P. (2018). Effectiveness in Context. IPA.
6. Cialdini, R. (2006). Influence: The Psychology of Persuasion. Harper.
7. Cialdini, R. (2016). Pre-Suasion. Simon & Schuster.
8. Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus & Giroux.
9. Thaler, R. & Sunstein, C. (2008). Nudge. Yale UP.
10. Christensen, C. (2016). Competing Against Luck (Jobs To Be Done). Harper.
11. Heath, R. (2012). Seducing the Subconscious. Wiley-Blackwell.
12. Ariely, D. (2008). Predictably Irrational. Harper.
13. Sutherland, R. (2019). Alchemy. WH Allen.
14. Roberts, K. (2004). Lovemarks (critical reading as counterpoint).
15. Reeves, R. (1961). Reality in Advertising (historical classic).

Papers / reports (10)

1. Ehrenberg, A.S.C. (1959). The Pattern of Consumer Purchases. Applied Statistics 8(1), 26-41.
2. Wahba, M.A. & Bridwell, L.G. (1976). Maslow Reconsidered. Organizational Behavior and Human Performance 15(2), 212-240.
3. Tversky, A. & Kahneman, D. (1981). The Framing of Decisions. Science 211, 453-458.
4. Anderson, E.T. & Simester, D.I. (2003). Effects of $9 Price Endings. Quantitative Marketing and Economics 1(1), 93-110.
5. Keiningham, T.L. et al. (2007). A Longitudinal Examination of NPS-Customer Loyalty. Journal of Marketing 71(3), 39-51.
6. Morgan, N.A. & Rego, L.L. (2006). The Value of Different Customer Satisfaction and Loyalty Metrics. JM 70, 426-439.
7. Plassmann, H., Ramsøy, T.Z., Milosavljevic, M. (2012). Branding the Brain. J. of Consumer Psychology 22(1), 18-36.
8. Open Science Collaboration (2015). Estimating the reproducibility of psychological science. Science 349.
9. Google (2020). Decoding Decisions: Making Sense of the Messy Middle. think with Google.
10. Binet, L. & Field, P. (2019). The Crisis in Creative Effectiveness. IPA.

7. FAQ

Frequently Asked Questions

What does evidence-based marketing mean in practice?

It means justifying every marketing decision (budget allocation, channel choice, creative brief) with at least one peer-reviewed source or replicable data, and abandoning models (Maslow, AIDA, predictive NPS) that research has already falsified.

Is Byron Sharp always right?

No. Sharp is excellent on mass-market FMCG, less robust on B2B, luxury, niche. Even his statements should be treated as strong hypotheses to be contextualized. Evidence-based ≠ Sharp-first.

Should I eliminate buyer personas from my process?

From advertising and segmentation, yes. From UX research and product design, no. Personas remain useful where they were born (Cooper, software design).

Is fMRI neuromarketing useless?

Not useless. It has academic relevance and niche B2B applications. In SME marketing the cost/decision ratio is low vs eye-tracking, FaceTrace, A/B creative tests.

How do I measure creativity if Cannes Lions don’t suffice?

System1 (Test Your Ad), Kantar, Millward Brown publish Star Rating benchmarks with year-over-year replication. Winning Cannes ≠ effective campaign.

How much budget should I allocate to brand vs performance?

60% brand / 40% performance maximizes 3-year ROI in 96% of 996 IPA cases. Adjustable per brand stage: early-stage startup can start at 50/50, mature brand can rise to 65/35.

8. Synthetic manifesto (10 points)

For those who want a pocket compass.

1. A claim is not science until it is peer-reviewed AND replicated AND with declared effect size.
2. Growth comes from penetration, not loyalty.
3. Mental + Physical Availability are the two assets that fuel future sales.
4. Distinctive Brand Assets must be built over 5-10 years, not renewed at every rebrand.
5. Allocate 60% brand, 40% performance — until your category empirically refutes it.
6. The light buyer is 50%+ of revenue; you don’t segment them with buyer personas.
7. Emotional creative maximizes long-term ROI; rational maximizes short-term DR.
8. AIDA, Maslow, predictive-NPS, “immutable laws”: didactic, not operational.
9. A free guru PDF source does not replace Journal of Marketing.
10. When personal experience contradicts 996 case studies, the error is in the experience, not the data.