What Is a Marketing Funnel — and Why the Standard Version Is Wrong
1 April 2026
The marketing funnel is one of marketing's most durable models. It is also one of its most misapplied. As a planning framework, it is genuinely useful. As a measurement model, it systematically drives brands to under-invest in the conditions that make growth possible.
The Ehrenberg-Bass Institute and LinkedIn B2B Institute found that 95% of potential buyers are not actively in-market at any given time, and that brand associations built before the buying process begins largely determine which brands get chosen (Ehrenberg-Bass / LinkedIn B2B Institute, 2023: https://business.linkedin.com/). If those associations are not built, the funnel has nothing to work with at the bottom.
What Is a Marketing Funnel
A marketing funnel is a model of the stages a buyer moves through on the way to a purchase decision. The original version, attributed to Elias St. Elmo Lewis in 1898, identified four stages: awareness, interest, desire, action. Most modern versions expand or rename these. Top of funnel covers awareness and consideration. Middle covers evaluation and intent. Bottom covers conversion and retention.
The purpose is allocation. By mapping where buyers are in their decision journey, marketers can match the right channel, message, and offer to each stage. That purpose is sound. The problem is what happens when the funnel shifts from a planning tool into a measurement framework.
When it becomes a measurement model, the funnel creates a systematic bias. The stages that are easiest to attribute get the most budget. The stages that are hardest to attribute get cut. The result is not a well-managed funnel; it is a funnel that has been hollowed out from the top down.
Where the Standard Funnel Model Breaks Down
The standard funnel assumes buyers move sequentially: first aware, then interested, then considering, then converting. Real buying behaviour does not follow this path.
Buyers enter at different points. They pause, leave, and return. They make decisions shaped by memory structures built months or years before any formal buying process begins. Research by Bain and Google found that 90% of B2B buyers have already formed a vendor shortlist before beginning active research, and 90% ultimately choose from that initial shortlist (Bain and Google, via Search Engine Land, 2024: https://searchengineland.com/). The buying decision is not made at the bottom of the funnel. It is largely made before the funnel starts.
The linear model also misrepresents how brand investment works. In a sequential funnel, brand building is stage one: something you do to fill the top, after which performance takes over. In reality, brand investment operates continuously across all stages, building the memory associations that determine which brands get considered at all. Treating it as a discrete stage to be completed before performance begins misunderstands the mechanism entirely.
How Buyers Actually Move Through a Purchase Decision
Buyers move through purchase decisions episodically. They are mostly dormant: not researching, not comparing, not actively in-market. Occasionally a trigger occurs. A budget approval. A pain point surfacing. A contract renewal. At that moment, they retrieve brands from memory.
The brands they retrieve are not necessarily the ones most recently researched. They are the ones most strongly associated with the problem being solved. Those associations were built through repeated brand exposure, often long before the buying episode began.
Ehrenberg-Bass Institute research shows that brands with higher mental market share — the share of category entry point associations held in buyers' minds — consistently hold higher actual market share (Ehrenberg-Bass Institute: https://marketingscience.info/). Which funnel stage a buyer is officially in matters less than the brand strength built before they entered it.
This is why two brands can run identical performance campaigns and see different results. The one with stronger pre-existing mental availability converts more efficiently, at lower CPA, because the audience already has a reason to choose it. The performance channel harvests demand. It does not create it.
Why Funnel Thinking as Measurement Drives the Wrong Allocation
The linear funnel creates a measurement bias toward the bottom. Bottom-of-funnel channels produce clean, immediate, attributable signals. A buyer clicks a search ad and converts. The channel gets credit. Upper-funnel channels, brand advertising, content, awareness spend, produce delayed and diffuse effects that are difficult to attribute in standard models. They rarely get credit.
Over time, this measurement bias drives budget toward the bottom. The brand investment making those conversions possible gets reduced or cut. CPA improves in the short term because the pool of warm, brand-aware audiences is still being drawn down. Eventually, that pool depletes. Performance metrics look healthy until the pipeline runs dry.
Binet and Field's analysis of the IPA Databank, covering over 1,400 case studies across categories, found that the optimal long-term split between brand and activation investment sits around 60% brand to 40% activation for most categories. Brands that consistently hit this ratio outperform those that tilt heavily toward activation (Binet and Field, IPA: https://ipa.co.uk/). Most brands running purely attribution-led allocation end up at the inverse: 20% or less in brand, 80% or more in activation. The funnel model does not cause this directly. The misapplication of funnel thinking as a measurement framework causes it.
A More Useful Planning Model: The 95-5 Framework
The funnel is not wrong as a planning tool. It is wrong as a measurement tool.
A more accurate model separates two distinct problems. The first is building the memory structures that determine which brands get considered when a buying moment arrives. This is the job of brand investment, and it is measured over months and years, not days. The second is converting buyers who are already in-market. This is the job of performance channels, and it is measured over days and weeks.
The 95-5 framework offers a cleaner planning lens. At any given time, 95% of potential buyers are not in-market. Brand investment targets that 95%. Performance investment targets the 5% actively buying. The share of budget allocated between the two should reflect this split, not be inverted by attribution models that can only see the bottom 5%.
This does not mean abandoning ROI discipline in brand investment. It means applying the right metrics to each job. Brand is measured through branded search trend, direct traffic trend, and consideration scores tracked quarterly. Activation is measured through CPA, ROAS, and conversion rate tracked weekly. Both are accountable. They are accountable on different timeframes because they produce effects on different timeframes.
How This Changes Campaign Strategy in Practice
Three decisions shift when planning moves off the linear funnel and onto a more accurate model of buyer behaviour.
Channel weighting. Upper-funnel investment deserves more budget weight than last-click attribution suggests. A starting calibration: compare your branded search volume trend over the past 12 months against your brand spend trend. If branded search is flat or declining as brand spend falls, the contribution is visible. The gap tells you how much upper-funnel work the performance channels are missing, and that gap compounds over time.
Content orientation. Content built around category entry points outperforms content built around product features for building mental availability. Start by mapping the five to seven situations that cause buyers in your category to begin looking. Write content for each situation in the buyer's language, not the brand's. These generate the associations that brand investment builds over time and that performance investment later harvests.
Measurement cadence. Brand effects materialise over months, not days. Set a separate measurement cadence for brand metrics: branded search trend, direct traffic trend, and consideration scores, reviewed quarterly rather than weekly. These signals tell you whether the demand pipeline is being replenished. Weekly performance dashboards cannot answer that question. They can only tell you how efficiently you are drawing down what already exists.
The Funnel's Real Job
The marketing funnel started as a planning heuristic, a way of thinking about the different jobs marketing needs to do at different points in the buyer journey. That remains its best use. The mistake is treating it as a measurement model, where attribution mechanics systematically reward the bottom and starve the top.
The brands that compound their growth over time do not abandon the funnel. They use it to plan across the full range of buyer moments, and they measure results on the timeframe each job actually requires. Those that optimise purely for the bottom tend to produce impressive short-term performance metrics while quietly hollowing out the demand pool those metrics depend on.
The opening point stands: the funnel is only as good as the brand investment feeding it. When that investment is cut because it cannot be directly attributed, the model is not working. The measurement system is.
If you are reassessing how budget is allocated across the funnel, or questioning whether your current attribution setup is giving you an accurate picture of where growth is actually coming from, Kaliber works with brands to build measurement frameworks that account for both the bottom and the conditions that make the bottom possible. Start the conversation at kaliber.asia/contact.
Frequently Asked Questions
What are the stages of the marketing funnel?
The original marketing funnel, attributed to Elias St. Elmo Lewis in 1898, has four stages: awareness, interest, desire, action. Most modern frameworks expand this to include consideration, evaluation, intent, and retention. The specific labels matter less than understanding that different stages require different channels, messages, and measurement timeframes. Treating all stages as equally attributable produces a systematic misallocation of budget toward the bottom.
Why do people say the marketing funnel is dead?
The funnel is not dead. The linear, sequential version of it is increasingly inaccurate as a description of real buyer behaviour. Research shows buyers form shortlists before beginning active research, make decisions based on memory structures built long before any formal funnel begins, and move episodically rather than sequentially. The model is still useful for planning. It is unreliable as a measurement framework.
How is the marketing funnel different for B2B versus B2C?
B2B buying cycles are typically longer, involve multiple decision-makers, and carry larger individual contract values. This makes pre-funnel brand investment even more critical. Bain and Google's research shows that B2B vendor shortlists are often formed before active research begins. Brands absent from buyers' mental availability at that point rarely make the shortlist, regardless of how much is spent on performance channels during the active buying window.
What is the 95-5 rule in marketing?
The 95-5 rule, developed through research by the Ehrenberg-Bass Institute and the LinkedIn B2B Institute, states that 95% of potential buyers are not actively in-market at any given time. Brand investment targets this dormant majority by building the mental availability that shapes which brands get recalled when a buying moment arrives. Performance investment targets the 5% who are actively buying. Budget allocation between the two should reflect this split, not be determined by attribution systems that can only measure the bottom 5%.
How do I measure upper-funnel marketing activity?
Upper-funnel brand effects typically take months to register in revenue metrics. The most reliable leading indicators are branded search volume trend, direct traffic trend, and brand consideration scores from survey or panel data. These should be tracked quarterly, not weekly, and kept separate from the conversion metrics used to evaluate performance channels. A declining branded search trend while performance metrics remain stable is a leading indicator that the demand pool is depleting.
