In short: AIDA (Attention, Interest, Desire, Action) is an 1898 model that survives in marketing textbooks more out of pedagogical inertia than empirical evidence. Ehrenberg-Bass, Google ("messy middle") and Nielsen document that real buying behaviour is non-linear: light buyers enter and exit the "funnel" in any order, often skipping entire stages. Penetration > funnel. AIDA remains useful as a teaching scheme, not as an operational tool to allocate budget or design campaigns.
Where AIDA comes from: Lewis 1898, door-to-door copywriting
The AIDA model is attributed to E. St. Elmo Lewis, an American copywriter, who formulated it around 1898 to train door-to-door life insurance salesmen. The aim was to sequence the sales conversation: capture attention (a hook), generate interest (product presentation), create desire (personalised benefits), drive action (the close).
It was an operational model for a specific scenario: 1-to-1 selling, high involvement, one-shot decision. It worked because the salesman had full visibility on the prospect's response and could adapt in real time. It was not designed as a model of generic buying behaviour, nor as a framework to design advertising campaigns at scale.
Why it is still taught: pedagogy vs evidence
AIDA is taught in practically every entry-level marketing course because it has three pedagogically winning properties: it is memorable (a 4-word acronym), it is intuitive (it follows the temporal sequence anyone imagines), it produces a diagram easy to draw on a whiteboard. The problem is that these properties have no relation to its empirical validity.
The "funnel" as a visual metaphor entered textbooks in the 1960s (St. Elmo Lewis himself never drew a funnel) and gave the illusion of a measurable physical flow: so many people at the top, so many at the bottom, conversion rate between stages. This visualisation is scientifically misleading because it assumes that buyers proceed in a fixed sequence, which decades of behavioural data refute.
What Ehrenberg-Bass says: penetration > funnel
The Ehrenberg-Bass Institute has documented for over 50 years (Andrew Ehrenberg, "Repeat Buying" 1972; Byron Sharp, "How Brands Grow" 2010) that the growth of a brand depends almost entirely on penetration (the number of distinct buyers) and only marginally on loyalty. Light buyers (those who buy once or twice a year) generate most of the revenue. These consumers do not pass through any "funnel": they enter the category when there is a Category Entry Point (CEP), buy the most mentally available brand on that occasion, and forget the decision until the next trigger.
Sharp explains that the funnel produces a classic strategic mistake: concentrating budget on the "consideration phase" (re-targeting, lead nurturing, email automation) of a small group that has already almost decided, while neglecting the 95-98% of non-buyers who need mental availability over the long term. The result is the pattern documented by Binet & Field: declining marketing ROI, brand erosion, dependence on promotions.
Google's "Messy Middle": real behaviour
In 2020 Google published "Decoding Decisions: Making Sense of the Messy Middle", based on 31,000 simulated purchase scenarios. The conclusion: between the initial trigger (an ad, an event, a recommendation) and the actual purchase, buyers oscillate non-linearly between two modes — exploration (looking for alternatives, reading reviews, comparing) and evaluation (narrowing the options). These cycles repeat in a disordered fashion, with interruptions, restarts, abandonments.
The study identifies 6 cognitive biases operating in the messy middle (category heuristics, social proof, scarcity, authority, power of free, power of now). None of them maps to an AIDA stage. Google's researchers explicitly conclude that "linear frameworks like AIDA do not describe the real behaviour of digital buyers".
Nielsen and tracking data: overlapping stages
Nielsen Catalina panel data (US, 2018) and Nielsen Media Lab behavioural tracking show a consistent pattern: in FMCG categories, more than 70% of buyers purchase the first brand they think of at the moment of need, with no "consideration" process at all. In high-involvement categories (cars, mortgages) the process is longer but still non-sequential: the consumer can jump from "discovery" to "purchase" with no intermediate steps, or return to "discovery" after signing a quote.
The funnel stages, when they exist, overlap. A consumer can be "Aware" and "In action" simultaneously for different brands. AIDA implies a one-way progression that the data do not confirm.
Why AIDA fails in real retail and e-commerce
In physical retail, the Procter & Gamble "First Moment of Truth" study (2005, A.G. Lafley) had already documented that the final decision happens at the shelf in 3-7 seconds, regardless of how much advertising the consumer has seen before. There is no "Desire" before "Action": they are the same event.
In e-commerce, Adobe and Shopify analytics (Adobe Digital Index 2023 report) show that 35-40% of purchases are "first session purchases": the consumer arrives on the page, evaluates, buys, all within the same session of a few minutes. AIDA, with its 4 separate stages, does not apply.
What to use instead: Category Entry Points (CEP)
The Ehrenberg-Bass Institute (Romaniuk, "Building Distinctive Brand Assets", 2018) proposes Category Entry Points as an alternative framework. A CEP is a contextual trigger (occasion, time of day, functional need, social situation) that activates thought of a category. Example in coffee: "I want a break", "I need to wake up", "I'm offering something to a guest", "I want to treat myself".
The goal of marketing becomes: identify the 5-10 most frequent CEPs in the category, and build mental associations between brand and CEP through coherent and repeated communication. No longer "push the consumer down the funnel", but "be the mentally available brand when the need arises".
This framework explains why a brand can grow without re-targeting, without lead nurturing, without marketing automation: if its CEPs are well built, the consumer arrives ready to buy when the moment comes. It is what Coca-Cola, Mars and Galbusera have been doing for decades with their brand investments.
When AIDA is still useful (rarely)
The model retains usefulness in three specific contexts: (1) entry-level pedagogy to explain to students that communication has multiple functions (capture attention, generate interest) before the transaction; (2) one-on-one sales coaching, which was the original use case: a salesperson handling an individual conversation can effectively sequence attention → interest → desire → action; (3) staged email funnels for high-ticket B2B products with long sales cycles (3-12 months), where there is a formalised sales process that resembles a funnel.
In every other context — brand campaigns, broadcast advertising, social media, FMCG retail, mainstream e-commerce — AIDA is a misleading simplification. Treating it as an operational framework produces wrong budget decisions.
FAQ
Is AIDA "wrong" or just "limited"?
It is more accurate to say "limited to a narrow use case". For 1-to-1 high-involvement selling — Lewis's original 1898 case — it is a sensible model. For mass marketing, advertising, FMCG, e-commerce, it is a misleading simplification that leads to miscalibrated budget and creative decisions.
If AIDA is limited, why do all agencies use it in pitches?
For two reasons: (1) familiarity with clients — those without a marketing background find AIDA intuitive, and this makes services easier to sell; (2) educational inertia — the people managing budgets today studied AIDA in the 1990s-2000s. Replacing it requires an educational effort that many agencies avoid. It does not mean it is a valid tool to allocate resources.
Can I still use a "funnel" to measure conversion rate?
Yes, distinguishing the two uses. As an analytics tool — mapping how many people arrive on the site, how many convert, where they drop off — a funnel is useful (it is a conversion chart, not a behavioural model). As a strategic tool — "we must invest X% in awareness, Y% in consideration, Z% in conversion" — it is misleading because it assumes resources follow the same logic as the visible flow, which is often not true.
What do I tell a client who says "I want to activate the consideration funnel"?
Reframe: "what is the business decision behind this request?" Often the client wants more revenue, more qualified leads, or more recognition. The operational solution is usually 60-70% building mental availability over the long term (brand, content, distribution) and 30-40% conversion of active CEPs. Not "activating the middle stage".
Are there modern models that have replaced AIDA in companies?
Yes. McKinsey's "Consumer Decision Journey" (2009, updated 2015) supersedes AIDA with a circular framework. Forrester's "Customer Lifecycle" is similar. Google's "Messy Middle" (2020) is the closest to behavioural data. The Ehrenberg-Bass framework (CEP + DBA + mental/physical availability) is the most scientifically rigorous.
Should AIDA be removed from university courses?
Not removed, but contextualised as a historical model. Students should know its origin (Lewis 1898), its limited usefulness (1-to-1 sales), and the more valid alternative frameworks (Ehrenberg-Bass, Messy Middle, CDJ). Teaching AIDA as a current model of buying behaviour is pedagogically incorrect.
Sources and references
- St. Elmo Lewis, E. — original AIDA formulation (1898), historical reference
- Ehrenberg, A. — "Repeat Buying: Facts, Theory and Applications" (1972, North Holland)
- Sharp, B. — "How Brands Grow" (2010, Oxford University Press) and "Part 2" (2015)
- Romaniuk, J. — "Building Distinctive Brand Assets" (2018) and "Better Brand Health" (2023)
- Google & Behavioral Architects — "Decoding Decisions: Making Sense of the Messy Middle" (2020): thinkwithgoogle.com
- Binet, L. & Field, P. — "The Long and the Short of It" (IPA, 2013)
- Nielsen Catalina Solutions — Brand Effect studies (2018-2023)
- Procter & Gamble — A.G. Lafley "First Moment of Truth" framework (2005)
- McKinsey — "The Consumer Decision Journey" (2009, updated 2015)
- Adobe Digital Index — annual ecommerce conversion reports
