‘The traditional marketing funnel is flatlining.’ That was the diagnosis of Dentsu’s social commerce director, Lindsay Dayanik, in a July article stating that the linear journey from discovery to purchase no longer maps how people buy.
Social commerce, she argued, has collapsed the sequence into a single scroll.
It’s a sentiment that’s been expressed throughout the year. In January, Boston Consulting Group suggested that the funnel ‘force-fits’ a complex array of behaviours into oversimplified stages. Unilever CEO Fernando Fernandez noted how Amazon has become a crucial channel for both sales and discovery on a recent earnings call — a sign of how blurred the lines between purchase and intent have become.
Dayanik has some numbers to support her assertion about the death of the funnel, too. In Dentsu’s survey of 1,000 UK adults, 58% said they had bought a product directly through a social platform. Among Gen Z and millennials, more than a third do so a couple of times a month.
Half of young consumers say social shopping is more convenient than the high street, and nearly as many rate it higher on personalisation and price. TikTok isn’t just a media space — it’s a storefront.
But even in Dentsu’s own data, the cracks show. Only 27% of respondents overall said they trusted social commerce more than traditional retail, and 40% still prefer established channels. Convenience has surged, but credibility lags behind.
Which makes the obituary for the funnel feel premature. This is, after all, a model that has survived for 125 years. Elias St. Elmo Lewis sketched out his tidy sequence of attention, interest, desire and action in 1898, and marketers are still drawing the same diagram today.
Rachel Coffey, chief strategy officer at Initiative, sees both sides of the shift. ‘I don’t think people move in straight lines anymore, where you see something, you consider it and then you convert,’ she told MediaCat UK in August. ‘Social platforms have turned scrolls into storefronts, AI is becoming this concierge of discovery, and 50% of beauty searches in the UK now happen outside Google.’
The money is already following the behaviour. Warc predicted that retail media spend would edge past that of linear TV by the end of 2024, with more growth forecast through 2025. TikTok is on track to generate more than $30bn in ad revenue this year.
Perhaps we should stop listening out for the funnel’s death rattle, and start questioning how much value compressing discovery and purchase leaves behind. If everything is frictionless, what sticks?
What’s changing
Consumers seem comfortable with immediacy. ‘[They] like point-of-sale ads because they’re relevant and don’t intrude into their lives,’ Gonca Bubani, global thought leadership director at Kantar, tells MediaCat UK.
Kantar’s 2024 Media Reactions report showed ad positivity at a record high across generations. Amazon and TikTok came out as consumers’ joint favourite platforms, even though marketers still preferred YouTube and Instagram.
Bubani argues these touchpoints shouldn’t be treated as a throwaway tactic: ‘Point-of-sale has long been considered a promotional tool, but this is missing a trick. When it leverages brand and campaign assets and keeps brand building in mind, it’s more likely to work in both the short and long term.’
This integration is exactly what Coffey has seen work in practice. ‘We did an amazing campaign for Deliveroo with Mob Kitchen last year where we served recipes on Instagram and users could swipe to have all the ingredients instantly added to their Deliveroo basket. The idea of a brand ad and a performance ad is redundant when we’re talking about something like that because every single thing existed in that one moment.’
It’s this blurring that fuels the collapsing-funnel rhetoric. Categories that used to be distinct — brand versus performance, upper versus lower funnel — now overlap in a single frame. What once took months (build awareness, nudge consideration, convert) can compress into seconds.
Platforms are hardwiring the behaviour. TikTok’s Market Scope promises a live view of how many users you have in awareness, consideration and conversion — and which touchpoints are nudging them.
Amazon’s Marketing Cloud is pitched as the ‘glue’ binding retail media, CTV and programmatic, giving buyers the closest thing to a full-funnel picture.
On the surface, this looks like collapse in action: journeys shortened, categories blurred, budgets redirected. Consumers buy in the same swipe they discover, and platforms monetise that comfort. But not everyone accepts the collapse of the funnel as gospel.
What’s not changing
For all the breathless talk of collapse, some strategists who work close to planning mechanics argue the fundamentals haven’t shifted. What looks like a broken funnel may actually be a misdiagnosed one.
On LinkedIn, Initiative’s senior vice president of strategic integration, Matthew Birkby, argued that marketers have mistaken funnels for tactics, confusing a diagnostic tool with customer journeys and touchpoints.
MediaCat UK reached out to Birkby to get a sense of what gets left out when people frame the compression of mental states as a total collapse.
‘They’re making a fundamental category error,’ he told us. ‘What looks like a collapsed funnel is actually multiple mental states being processed simultaneously in compressed time.’
In his analysis, Birkby references Netflix’s evolution. When the platform launched, its funnel measured streaming adoption: 47% of people knew about streaming video, and 49% of those who tried it converted to subscribers.
Today, with 85% understanding streaming, Netflix’s bottleneck has shifted entirely to password sharing tolerance and household restriction acceptance. The same company, same core product, but completely different mental states to diagnose and influence.
This isn’t funnel collapse — it’s funnel evolution based on strategic measurement.
The category caveat matters. Netflix’s subscription model allows clean mental state measurement, but compression works differently across sectors.
Beauty brands on TikTok benefit from visual-first discovery, while B2B software still requires multiple touchpoints and longer consideration cycles. The danger is applying TikTok Shop logic to enterprise software sales, or assuming automotive purchases follow lipstick patterns.
The compression test is simple: can your customer make an informed purchase decision within the attention span of a social scroll? If not, sequential touchpoints still matter more than algorithmic bundling.
Birkby points to a meta-study from Analytic Partners: of 51,232 campaigns reviewed, only 14 actually wore out. The problem wasn’t creative decay, so much as brands abandoning strategy before building sufficient mental availability.
Birkby warns that compressed journeys create attribution illusions. Bottom-funnel tactics show cleaner measurement because the results are immediate and visible, while upper-funnel investment shows diffused, delayed returns that are harder to track.
When TikTok shows you ‘awareness’ and ‘conversion’ metrics within the same dashboard, how do you know the awareness wasn’t just people who were already going to buy anyway?
Platform measurement conflates correlation with causation, making incrementality nearly impossible to prove. Most brands end up flying blind, trusting algorithms they can’t audit with budgets they can’t properly measure.
The result is sloppy resource allocation: brands over-weight the bottom end, chasing the most visible conversions. As Birkby warns, brands end up with ‘80% of budget chasing the 8% of people ready to buy today, while completely ignoring the 65% who don’t even know they have the problem you solve.’
Boston Consulting Group, for its part, has promoted ‘influence maps’ as an update to the funnel — not a rejection, but a way to map behaviours (streaming, scrolling, searching, shopping) that overlap and influence multiple states simultaneously. In practice, they still serve the same function: a diagnostic, not a literal journey.
Consumers may collapse time between discovery and purchase, but the mental processes of how brands grow haven’t evaporated.
Ehrenberg-Bass research remains clear that penetration and availability drive growth. Binet and Field’s 60/40 rule still reflects the relationship between brand activity and performance.
Shortening the timeline doesn’t rewrite those laws. It just makes it easier to forget them.
The reality of the collapsing funnel isn’t binary. Discovery and intent don’t simply merge into one neat stage, nor do they stay neatly apart. They rub against each other, creating tensions that marketers now have to work with.
Speed vs. stickiness
The collapsing funnel story is seductive: faster journeys, cleaner attribution, instant sales. Behavioural science has long argued for making things easy — Richard Shotton among others has shown how shaving off small frictions can drive big shifts in behaviour. Ease works in the moment.
But what gets you a click doesn’t always get you remembered. Elizabeth and Robert Bjork’s research into ‘desirable difficulties’ showed that memories form more deeply when the process demands a little effort. A pause, a ritual, a touch of friction makes a brand harder to forget.
Advertising research pushes the same point further. Orlando Wood argues that effectiveness comes from breaking autopilot: work that feels distinctive enough to slow us down. In a collapsing funnel, that kind of creative friction isn’t optional — it’s the only way a brand earns attention in compressed time.
Bubani makes a crucial distinction between types of friction. ‘Friction isn’t a good thing if it’s created by a mismatch or lack of connection between brand-building and point-of-sale activity,’ she says.
This kind of friction is structural friction: campaign messaging missing the purchase moment, or when brand identity gets lost in platform defaults.
‘But that’s different from friction created by distinctiveness in the point-of-sale environment. Just as we say ads need to stand out to be memorable, we also believe unique and interesting point-of-sale material has a greater impact on both short-term sales and long-term memorability.’
The danger in chasing speed isn’t just forgettable campaigns, it’s structural. Collapsed journeys tilt budgets towards conversion and away from the slower cues that build memory. They flatten distinctiveness into platform defaults, leaving brands to fight over price and placement.
Friction isn’t a nuisance in that system. It’s the last lever marketers have to create difference.
Platform vs. inevitability
Much of the collapsing-funnel rhetoric assumes brands can choose whether to embrace compression, but technology is making that choice for them.
Privacy changes have eliminated much of the data that enabled traditional funnel-stage targeting, while automated campaign structures like Google’s Performance Max and Meta’s Advantage+ are designed to optimise across multiple objectives simultaneously.
Automated bidding systems like Performance Max don’t let you target awareness separately from conversion. You’re buying access to audiences and trusting the algorithm to figure out whether they’re browsing or buying.
The smart response isn’t to resist this technological shift, but to recognise where friction still matters. Automated campaigns may bundle awareness and conversion, but creative execution still determines whether a brand gets remembered or ignored. The friction has moved from targeting to storytelling.
Looking ahead
If the last 18 months were about collapsing discovery and purchase into the same swipe, the next phase may take decision-making even further out of human hands.
Coffey predicts that large language models will be shoppable within six months, radically altering how brands think about discovery and performance. She points to ‘agentic AI’ — algorithms acting as personal shoppers — as the real break from the funnel: once consumers delegate choice, salience and distinctiveness may be filtered out by intermediaries.
For brands, that raises uncomfortable questions: how do you build memory when discovery is outsourced, or maintain identity when the interface decides what surfaces?
The paradox is that even as technology accelerates, strategy may need to slow down. Birkby warns that brands are often their own worst enemy here, abandoning campaigns long before audiences tire of them. What gets mistaken for creative wear-out is usually marketer impatience, deleting the very memory structures future sales depend on.
The brands that survive won’t be those that master every platform’s latest ‘full-funnel’ tool. They’ll be the ones that understand when compression helps and when it hurts — building memory structures that work whether the purchase happens in six seconds or six months.
Main image by Logan Voss on Unsplash
