What Is a Marketing Funnel?

A marketing funnel maps the path a person takes on the way from not knowing a brand exists to becoming a customer. It's drawn as a funnel because the number of people shrinks at each stage: a wide crowd with little commitment enters at the top, and a small group ready to buy remains at the bottom. That shape, more than any measured figure, is where the model gets its name.

The version most used in content marketing splits the path into three blocks, known by their English abbreviations. TOFU (top of the funnel) corresponds to awareness: the person notices a problem or a question and searches for general information, without thinking about specific brands yet. This is where educational articles, guides, and organic search content with informational intent do their work. MOFU (middle of the funnel) corresponds to consideration: the person already knows what kind of solution they're after and compares options, through comparison posts, use cases, and downloadable content. BOFU (bottom of the funnel) corresponds to conversion: the person is ready to decide, and that's where product pages, demos, and free trials come in.

Some versions add a fourth stage after the purchase, focused on retention and turning the customer into someone who recommends the brand. There's no single correct version. What all of them share is the same logic: moving a wide group of strangers toward a small group of buyers, with a different type of content for each stretch.

A marketing funnel shouldn't be confused with a conversion funnel. The marketing funnel is the strategic framework a content or campaign team uses to decide what to publish and for whom. A conversion funnel applies that same shape to a specific, measurable process, with real analytics data, such as how many people move from viewing a product page to completing a purchase. The first is a planning model, the second a measurement.

A marketing funnel and a customer journey describe the same process from different angles. The funnel simplifies the path into clean stages so content can be planned in an orderly way; the customer journey documents what actually happens, including the repeats and jumps the funnel doesn't account for. Both models are useful, but they answer different questions: one organizes the team's work, the other describes user behavior.

The funnel has been the default image of marketing for decades, and it's been questioned for just as long, for the same reason: it assumes people move in a single direction, top to bottom, without going back. The available research shows that rarely happens that way.

In 2009, McKinsey published a study of nearly twenty thousand purchase decisions across five industries and three continents, written by David Court, Dave Elzinga, Susan Mulder, and Ole Jørgen Vetvik. Their conclusion was blunt: for years, touch points had been understood through the metaphor of a funnel, but "the funnel concept fails to capture all the touch points and key buying factors resulting from the explosion of product choices and digital channels." In its place, McKinsey proposed a circular model with four phases: initial consideration, active evaluation, closure (the purchase), and postpurchase, in which buyers can return to brands they'd already ruled out, even after buying. They called that repeated return the "loyalty loop."

In 2020, the Think with Google team reached a similar conclusion by a different route. Their research describes a middle zone between the trigger that sparks interest and the final purchase, which they call the "messy middle": a space where a person alternates between two mental modes, exploring new options and evaluating ones they already know, looping through that cycle as many times as they need before deciding. In their own words, "what happens between trigger and purchase decision-making is not linear."

Neither source argues for dropping the funnel. McKinsey acknowledges that the metaphor "does help a good deal," for instance to compare a brand's strength against competitors at each stage. What both point out is that the funnel poorly describes the real behavior of a person who researches, compares, gets distracted, and starts over, often within the same browsing session. As a planning framework, the funnel remains useful; as a picture of actual behavior, it's incomplete.

For a content team, the practical consequence isn't dropping stages, it's dropping the assumption that a user only moves through them in one direction. Someone can land directly on a product page (BOFU) from an ad, without ever having read an awareness article, while someone else goes back to reading informational TOFU content after already buying, to confirm they made the right call. That second case is common: in the facial skin care category, McKinsey's research found that more than 60 percent of buyers keep researching online after the purchase, something the classic funnel doesn't account for since it ends at conversion.

In practice, this means building content with entry points at every stage, not just at the top. A product page needs to answer questions from someone who's still comparing, and an educational article should link to decision-stage content for a reader who, after finishing it, is already ready to buy. Internal linking does the actual work here: it connects pieces from different stages instead of leaving each one isolated in its own box.

Measurement changes too. If real behavior is a loop rather than a straight line, a last-click attribution model undersells the role of awareness content that planted the decision weeks earlier. And remarketing campaigns stop being just a final push toward purchase, and start serving another purpose: bringing back someone who, after evaluating several options, drifted back into exploration without deciding yet.

That also changes how budget gets split: not every stage deserves the same share of ad spend, and that share shifts once a team can see where most users bounce back and forth between stages. A funnel with a particularly turbulent consideration stage needs more content and more retargeting there than one where that stage already runs smoothly.

• Assign a content type to each stage (TOFU, MOFU, BOFU) based on search intent, not publication date.
• Link content across stages: an awareness article should lead to decision-stage content for readers who are ready, and a product page should answer questions from people still comparing.
• Build entry points at every stage, not just the top: a good share of traffic lands directly on comparison or product pages from a search or an ad.
• Use remarketing to bring back people who drift back into exploration after comparing options, not just as a final nudge before purchase.
• Review post-purchase content: a meaningful share of buyers keep researching after deciding, and that content shapes retention.
• Combine the funnel with multi-touch attribution data, so the conversion isn't credited entirely to the last touch point.

• Treating the funnel as an accurate description of user behavior instead of a simplified planning model.
• Publishing only for the top or only for the bottom stage, leaving the consideration stage uncovered, where much of the purchase decision actually gets made.
• Measuring success purely with last-click attribution, which ignores the role of awareness content in decisions that take weeks to mature.
• Confusing the marketing funnel with the conversion funnel, and expecting a planning model to produce the same exact figures as an analytics tool.
• Forgetting the post-purchase stage, when a significant share of customers keep researching and can become a source of recommendations.