TL;DR — Marketing funnels are a pseudo-scientific model that works for those who sell them, not for those who buy them. Academic evidence demolishes their fundamental assumptions: consumers do not follow linear paths, small sample sizes make every statistical optimization a fiction, and the Double Jeopardy law explains why SMEs must focus on penetration — not funnel-based retention. This is the definitive proof.
The Problem with Marketing Funnels
Every year thousands of SMEs spend tens of thousands of euros on "marketing funnels": automated email sequences, optimized landing pages, multi-step retargeting, lead magnets followed by upsells and downsells. Every year, the majority of these SMEs cannot measure a clear return on this investment. And every year, the agencies and consultants selling these systems find a new explanation for why "the funnel hasn't worked yet" — and why another month, another test, another optimization is needed.
This article is not an attack on digital marketing. It is an evidence-based critique of a specific model — the linear conversion funnel — that is sold as a universal solution when the data shows it is a partial solution, often counterproductive for SMEs, and almost always in the economic interest of those selling it rather than those buying it.
Why Funnels Are Pseudo-Scientific for SMEs
The concept of the marketing funnel originated in 1898 with Elias St. Elmo Lewis's AIDA model (Awareness, Interest, Desire, Action). In 126 years, the model has never received robust empirical validation as an accurate description of real purchasing behavior. It remained a useful metaphor that became dogma.
The fundamental problem is epistemic: the funnel is a normative descriptive model (it says how consumers should behave), not a positive one (it describes how they actually behave). Consumer behavior research over the past 30 years consistently shows that real purchase journeys are non-linear, bidirectional, multi-channel, and highly idiosyncratic.
Three structural problems:
- The funnel assumes sequentiality: a consumer must pass through Awareness before Interest before Desire before Action. Consumer neuroscience shows that desire often precedes rational awareness (impulse), or that Action happens simultaneously with Awareness (impulse buying seen for the first time)
- The funnel assumes a single journey: one consumer, one product, one funnel. Reality is parallel and multiple — the same consumer simultaneously considers multiple solutions across different channels
- The funnel assumes linear optimizability: improving every step improves the final result. In complex non-linear systems, optimizing a sub-component can degrade the overall system (cobra effect)
The Statistical Insignificance of Small Samples
This is the most technical point — and the most destructive for the funnel industry. Most SMEs implementing funnels have insufficient traffic and conversion volumes to make statistically significant optimizations. Yet they continue to do so.
Suppose a landing page with a 2% conversion rate and 500 visitors per month. You want to test two variants (A/B test). To detect a 20% difference between variants (going from 2% to 2.4%) with 80% statistical power and a standard p-value <0.05, you need about 7,500 conversions per variant — 750,000 total visitors. With 500 visitors per month, that would take 125 years. No consultant tells you this.
The practical consequence: "optimizations" based on A/B tests with small samples produce random results — not causal improvements. The variation that appears to win is often statistical noise. The consultant interprets the noise as a signal, adjusts the funnel, and the cycle begins again.
Note: these calculations assume α=0.05, power=0.80, two-tailed test. The point is not that A/B testing is useless in absolute terms — it is useless with the typical volumes of SMEs being sold funnels at €3,000–€15,000.
Non-Linear Consumer Behavior: The Evidence
The most cited study on the non-linearity of the purchase journey is Google's "Zero Moment of Truth" (2011), which found that the average consumer considers an average of 10.4 information sources before a purchase (up from 5.3 in 2010). These touchpoints do not occur in sequence but in an overlapping and bidirectional way: a consumer can read a review ("consideration" phase) and return to generic search ("awareness") multiple times.
Research by Edelman & Singer (McKinsey, 2015) introduced the "loyalty loop" model as an alternative to the funnel: satisfied existing customers skip the entire funnel and return directly to purchase. This has enormous implications for SMEs: existing customers are 5–25× cheaper to convince to buy compared to new customers (Frederick Reichheld, Bain & Company). Building a funnel to acquire new customers while ignoring existing customers is often a misallocation of resources.
Double Jeopardy: Penetration vs Retention
The Double Jeopardy law, originally discovered by Andrew Ehrenberg in media (1959) and then applied to marketing by Ehrenberg-Bass, is probably the most robust empirical finding in scientific marketing: smaller brands have both fewer buyers and less loyal buyers. It is a double penalty tied to brand size, not quality or strategy.
The implications for SME strategy are radical:
- Loyalty is a consequence of penetration, not vice versa. Brands with high penetration naturally have more loyal customers — not because they have better loyalty programs, but because loyalty correlates with brand awareness and mental and physical accessibility
- Investing in retention before achieving sufficient penetration is mathematically wrong. If your brand has a 3% market share, your loyalty ceiling is structurally low regardless of how good your email sequence is
- Growth for SMEs comes almost exclusively from acquiring new customers — not from increasing purchase frequency among existing customers. This is the opposite of what most funnel consultants preach
Binet & Field's 60/40 Rule: The Hierarchy of Objectives
Les Binet and Peter Field, in their analysis of over 1,400 advertising cases for the IPA (Institute of Practitioners in Advertising), demonstrated with data that sustainable growth requires a balance between brand building (60%) and direct activation (40%). The marketing funnel is a direct activation tool — it optimizes conversion for those already in the market. But if 60% of future potential customers do not yet know you, optimizing the conversion of the 40% who do know you is a niche operation.
The problem with SMEs that invest heavily in funnels is that they are often spending 90%+ of their budget on direct activation (paid performance, email sequences, retargeting) and almost nothing on brand building (content, PR, non-conversion video, sponsorships). The result: good efficiency in immediate conversion, zero structural growth.
Binet & Field also quantified the time-based returns:
- Direct activation: peak effect in 0–6 months, then decays rapidly if you stop spending
- Brand building: effects visible from 6–18 months, but compounding over time with effects that persist even when you stop spending
Reach vs Frequency: The Retargeting Trap
Retargeting — the backbone of most digital funnels — is based on a frequency principle: showing the message more times to those who have already interacted to "push" them toward conversion. The theory seems solid. The data less so.
Andrew Ehrenberg's research on purchase distributions shows that the optimal frequency of exposure is much lower than marketers assume. The "effective frequency" of 3× popularized by Michael Naples (1979) was based on TV studies with methods now considered insufficient. More recent research (Ephron, 1995; Jones, 1995; Reichel, 2017) suggests that in the majority of cases the marginal effect of each additional exposure beyond the 2nd–3rd is negative: it increases irritation (ad fatigue) without increasing the probability of conversion.
The consequence of aggressive retargeting: consumers who were close to converting are driven away by overexposure. This is measurable — there are documented cases of companies that increased sales by disabling retargeting on certain segments.
When Disabling Funnels Increases Sales
This is not a theoretical paradox: there are real, documented cases of companies that saw conversions increase by reducing or eliminating funnel elements. The mechanism is not magic — it is the removal of artificial friction.
The most common cases:
- Removing exit intent popups: in several A/B tests with sufficient volumes, exit intent popups (which intercept those leaving with a last-minute offer) showed neutral or negative effects on the LTV of acquired customers, while marginally increasing immediate conversion. Customers acquired via exit intent discounts have higher churn rates
- Simplifying checkout: Amazon built an enormous competitive advantage with 1-click purchasing. Research by the Baymard Institute (2024) shows that the average checkout funnel has 5.1 unnecessary steps. Removing them increases conversions by 20–35%
- Reducing follow-up emails: email sequences of 7–12 emails for unconverted leads often show diminishing returns after the 3rd follow-up, with increasing unsubscribe rates that damage future deliverability
The Funnel Industry as a Profit Center for Consultants
This is not a moralistic critique: it is an analysis of economic incentives. The funnel industry — courses, agencies, software (ClickFunnels, Kartra, GoHighLevel, ActiveCampaign with advanced automations) — has generated billions of dollars in global revenue. The business model is solid: the perceived complexity of the funnel creates dependency on experts, intermediate metrics (open rate, CTR, lead magnet downloads) show "progress" even when sales do not grow, and infinite customization ("your funnel must be unique for your sector") justifies unlimited consulting hours.
The consultant's incentive is not aligned with the SME client's incentive:
- The consultant benefits from complexity (more billable hours)
- The client benefits from simplicity (lower costs, less management)
- The consultant benefits from vague metrics (difficult to attribute failure)
- The client benefits from clarity (measuring real ROI)
- The consultant benefits from dependency (proprietary software, custom setup)
- The client benefits from autonomy (internally understandable systems)
This does not mean all marketing consultants are dishonest. It means the incentive model structurally gravitates toward complexity, regardless of individual intentions.
What Works Instead of Funnels (For SMEs)
Demolishing a model without proposing alternatives would be pointless. The evidence suggests that for SMEs with limited budgets, the strategies with the best documented cost/benefit ratio are:
- Market penetration: invest in brand awareness to reach "light buyers" (occasional buyers) who are the primary source of growth according to the Ehrenberg-Bass law. Channels: SEO, content, PR, organic social.
- Mental Availability: be present in the consumer's mind at Category Entry Points (Sharp, 2010). Prioritize brand memorability over funnel sophistication.
- Physical Availability: make the product/service as easy as possible to purchase. CRO on checkout, friction reduction, marketplace presence, verifiable reviews.
- Customer experience: satisfied customers are the most efficient funnel that exists. Net Promoter Score as the primary metric for sustainable growth.
In practice: for many SMEs, doing 4 simple things well (SEO, basic Google Ads, Google reviews, referral program) produces higher ROI than a 10-step funnel costing €15,000.
FAQ — Frequently Asked Questions
Do funnels never work?
Funnels work well in specific contexts: e-commerce with high volumes (where A/B testing is statistically valid), digital products with high margins (where even marginally improved CR is justified), and companies with an already established brand (where direct activation works on an already warm audience). The problem is selling them as a universal solution even to B2B SMEs with 50 leads per month.
Is Content Marketing better than funnels?
Content marketing is a brand building tool (See/Think phase) that builds Mental Availability over time. It does not replace direct activation — it complements it. The optimal strategy for an SME is not to choose between content and funnels, but to allocate budget in the 60/40 ratio between brand building and activation, using the simplest and most economical tools for each.
How do you evaluate whether a marketing consultant is worth the price?
Three questions: (1) Can they show you case studies with real business metrics (revenue, CAC, LTV) — not just vanity metrics (impressions, followers)? (2) Do their recommendations simplify or complicate your marketing stack? (3) Can you manage internally what they build, or do you create permanent dependency? If the answers are no, no, yes — renegotiate or change consultant.
Does the Double Jeopardy law apply to digital?
Yes. Subsequent Ehrenberg-Bass studies have replicated Double Jeopardy on social media, e-commerce, and mobile app data, finding the same structure: smaller brands/apps/accounts have both fewer users and less loyal users. The laws of consumer behavior do not appear to respect the online/offline boundary.
How do you measure whether your funnel is actually working?
The only metric that matters is CAC (Customer Acquisition Cost) compared to LTV (Lifetime Value). If CAC is less than LTV/3, the system is sustainable. If you are optimizing open rates and CTR but do not know your CAC, you are measuring the wrong things. Calculate CAC including: tool cost, consultant management hours, paid budget, and divide by real paying customers acquired in the period.
Sources and References
- Sharp, B. (2010). How Brands Grow: What Marketers Don't Know. Oxford University Press.
- Binet, L., & Field, P. (2013). The Long and Short of It. IPA.
- Binet, L., & Field, P. (2017). Media in Focus: Marketing Effectiveness in the Digital Era. IPA.
- Ehrenberg, A.S.C. (1959). The pattern of consumer purchases. Applied Statistics, 8(1), 26–41.
- Reichheld, F. (2001). Loyalty Rules!. Harvard Business School Press.
- Edelman, D.C., & Singer, M. (2015). Competing on customer journeys. Harvard Business Review, November.
- Google ZMOT. (2011). Winning the Zero Moment of Truth. Google Think.
- Naples, M.J. (1979). Effective Frequency: The Relationship Between Frequency and Advertising Effectiveness. Association of National Advertisers.
- Jones, J.P. (1995). Single-source research begins to fulfill its promise. Journal of Advertising Research, 35(3), 9–16.
- Baymard Institute. (2024). E-Commerce Checkout Usability Study.
- Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus and Giroux.
- Lewis, E.St.E. (1898). AIDA model. Historical reference in advertising literature.
- Ehrenberg-Bass Institute for Marketing Science. (2014–2024). Various papers on Double Jeopardy, Mental Availability, Physical Availability.
