MiQ study of 53 million households finds the marketing funnel is broken

MiQ, the global programmatic media company, released on April 22, 2026 a major research report arguing that the century-old marketing fgunnel no longer reflects how consumers actually make purchasing decisions. The report, titled "From Funnel to Flexibility," draws on proprietary data covering 53 million distinct households and two commissioned consumer surveys spanning four countries.

The scale of the data set is notable. According to MiQ, its Sigma platform processed over 700 trillion signals to analyze 53 million distinct households over a seven-day period in January 2026. Alongside that proprietary analysis, MiQ partnered with Censuswide to field a consumer survey of more than 4,000 adults aged 18 and over, split evenly across Australia, Canada, the United Kingdom, and the United States. A second survey of 600 marketing professionals across the same four markets was conducted in parallel. The data from those surveys was collected in December 2025 and January 2026.

The core finding: paths to purchase are now non-linear

According to the report, 87% of consumers switch between digital activities at least once per hour. More strikingly, 42% describe their own path to purchase as feeling entirely random. That figure undermines a foundational assumption of the traditional funnel model - namely, that consumers progress in an orderly sequence from awareness through to purchase. According to MiQ's analysis of the Sigma data, within any given 30-minute window, 43% of consumers are in a watching state, 23% are browsing, and 34% are in an active buying mode. But those states are fluid. Within that same 30-minute window, 66% of people who were browsing moved to a buying state, while 50% of those who were buying shifted back to watching.

The data also captures transitions within each hour. According to the report, 85% of consumers remain in a watching state within any given hour, but 10% move to buying and 5% move to browsing during that same period. It is a pattern that the report describes as fundamentally incompatible with linear funnel frameworks, which assume a consumer progresses through defined stages rather than cycling rapidly between them.

The funnel model has a history stretching back more than 100 years. According to the report, the first version was developed by American advertiser Elias St. Elmo Lewis, who divided consumers into discrete stages - most commonly described as awareness, consideration, and purchase. That framework has been adapted many times since, but its core logic - that consumers pass sequentially from one stage to the next - has remained largely intact. MiQ's position, supported by the new data, is that the internet has made that logic obsolete.

Second-screening is now the default behavior

One of the clearest signals in the data concerns simultaneous device use. According to MiQ, 91% of consumers use a second device while watching television. Of that group, half juggle multiple devices most of the time. The breakdown of second-screen activity is instructive: 43% use a second device to engage with social media, the news, or secondary video content, while 22% are researching products related to what they are already watching. The most common device pairing, according to the report, is connected television combined with a mobile phone, which covers 69% of respondents.

This matters to the CTV measurement discussion that has occupied much of the programmatic industry. If nearly all CTV viewers are simultaneously active on a second device, the question of which screen drives a conversion becomes substantially harder to answer. The challenge is compounded by the fact that 71% of consumers say they get distracted during product research and only return to a purchase when prompted by something external - turning retargeting into a structural challenge rather than an optional tactic.

"The funnel was built for a world where the customer journey is linear," said Moe Chughtai, Global VP of Strategy and Partnerships at MiQ. "Today a consumer can see an ad on CTV, price-check the product on their phone, and buy it on social - all before the next commercial break."

The data on social commerce is among the most striking in the report. According to MiQ, 72% of consumers under 34 have completed a purchase inside a social application - an entire purchase journey, from discovery to transaction, without switching screens. More than 50% of all consumers use social apps for different reasons in the same day, though that number rises sharply among younger groups: 82% of 18-to-24 year-olds and 81% of 25-to-34 year-olds use social platforms for multiple purposes simultaneously.

AI tools have entered the shopping process faster than most forecasts anticipated. According to the report, 45% of consumers already use AI chatbots when shopping - for tasks including comparing products, summarizing reviews, and getting recommendations. That figure rises to 73% among 18-to-24 year-olds and 67% among 25-to-34 year-olds. The report notes that these tools were barely used by mainstream consumers two years ago, which frames the speed of adoption in sharp relief. Looking ahead, 20% of all respondents say they plan to use AI shopping assistants over the next year.

The marketing funnel framework debate has run through the industry for several years, but AI-native tools as purchase intermediaries represent a new variable that existing measurement infrastructure was not built to track. When a consumer asks a chatbot to recommend a product and then buys directly through the chatbot interface, that transaction may not appear in any of the standard programmatic attribution models.

Measurement confidence remains low among marketers

The marketer survey findings reveal a significant disconnect between the complexity of current consumer behavior and the tools most marketers have available to understand it. According to MiQ, only 43% of marketing professionals feel confident in their measurement capabilities. Fragmented screens and fragmented supply paths are identified in the report as the primary blockers. Additionally, 42% of marketers say their retargeting campaigns are inconsistent or ineffective - a figure that connects directly to the finding that 71% of consumers routinely abandon research mid-journey and only return when prompted.

The cross-media measurement challenge is one that the industry has been working to address through multiple initiatives. But the MiQ data suggests the gap between what measurement systems can see and what consumers actually do is widening rather than narrowing. When 76% of consumers say they will abandon a purchase if a website is slow or checkout is complicated, the window for retargeting that lapsed intent is already very short - and existing systems frequently miss it entirely.

The report identifies five structural responses for marketers, framed around consumer states rather than funnel stages. The first is treating passive moments as commercially significant. According to MiQ's analysis, the watching state is actually the one where consumers are most likely to transition directly into buying. This cuts against the conventional logic of reserving performance budgets for moments of explicit intent, and points toward channels like YouTube and CTV as appropriate vehicles for what has traditionally been called brand advertising but is now functioning closer to the top of a compressed purchase arc.

The second structural response concerns social advertising. According to the report, because 72% of under-34 consumers have completed full purchase journeys inside social apps, pairing social ads with other advertising formats becomes a way to build credibility rather than just reach. Third, the report argues for audience tracking that follows consumers as their viewing and device choices shift rather than targeting fixed demographic or behavioral segments. For consumers under 24, paid streaming apps like Netflix, Hulu, and Disney+ lead viewing preferences at 33%, followed by YouTube at 29% and social platforms at 22%. Those preferences differ substantially from the over-55 group, which still leans toward traditional linear television.

The fourth structural response addresses second-screen behavior directly. Since 69% of consumers pair CTV with mobile as their most common device combination, the report argues for aligning creative and media strategies across that specific pairing rather than treating each screen independently. The fifth concerns measurement frameworks themselves. According to the report, 60% of marketers find standard measurement approaches limited, and over 80% want more customized planning and measurement tools. The report argues for state-based measurement models that can track performance from sales lift through to customer lifetime value.

MiQ's Sigma platform as the commercial context

The research was conducted using MiQ's Sigma platform, which the company describes as an AI-powered advertising operating system. According to MiQ, Sigma aggregates data from multiple sources into a single intelligence layer designed to identify audience states, optimize across screens, and measure outcomes across the open web and major closed platforms. In the past 12 months, MiQ ran more than 13,000 campaigns from over 1,200 advertisers through the platform.

MiQ reported that Sigma delivered double-digit improvements in reach, efficiency, and speed across consumer packaged goods, retail, travel, and political campaign verticals. In categories including media and entertainment and automotive, the platform delivered triple-digit conversion performance improvements. According to MiQ, some clients have seen 8x more conversions and over 50% lower cost per acquisition compared to standard programmatic approaches, though these figures represent individual campaign outcomes rather than platform-wide averages.

The report arrives at a moment of significant activity for MiQ as a business. In March 2026, MiQ acquired the Latin America business of Adsmovil, creating what the companies described as the largest independent programmatic provider in that region, with more than 150 employees across 12 markets and access to more than 400 million unique IDs. In April 2026, the company acquired Rocket Lab, a Miami-based mobile app growth company, in a deal that extended Sigma's signal coverage into in-app environments. Earlier, in January 2026, MiQ's Sigma trading agent was deployed within PubMatic's AgenticOS, the first operating system built for agentic advertising, with live campaigns running through the first quarter of that year.

"Couches have become the new storefronts as consumers move from screen to screen, almost in the same breath," said Jordan Bitterman, Global Chief Marketing Officer at MiQ. "Today, a full customer journey can conceivably take place in hours or minutes, instead of weeks or days. To keep up, marketers must shift from rigid funnels to flexible frameworks. This isn't just a planning shift for marketers, but a mandate for the adtech industry to move beyond channel silos toward systems that unify data, intelligence, and activation, so every investment is accountable to both brand and business outcomes."

The phrase "channel silos" has particular resonance given the fragmentation documented across the programmatic measurement landscape. Traditional measurement platforms - search, social, retail media - each report performance within their own ecosystems. When consumers move fluidly between those ecosystems within the same purchase journey, the attribution signals get split across platforms that never communicate with each other.

Why this matters for the programmatic and media buying community

For media buyers and programmatic practitioners, the most operationally significant finding may be the 71% distraction figure. If nearly three quarters of consumers abandon purchase research mid-way and only return when something prompts them, the entire retargeting stack - which is designed to recapture that lapsed intent - becomes structurally dependent on timing and frequency in ways that most campaign setups do not account for. The industry has been grappling with measurement frameworks for commerce media in parallel, with the IAB releasing incrementality guidelines in November 2025 to address precisely the kind of fragmented journey attribution that MiQ's data makes visible.

The scale of the Sigma data set - 700 trillion signals, 53 million households, seven days of continuous observation - puts it in a different methodological category from most industry survey research. Sigma's ability to analyze how many devices an individual consumer used in a single household, how often two devices were used within 30 minutes of each other, and how behavior transitioned at 15-, 30-, and 60-minute intervals produces a granularity that self-reported survey data cannot match. That said, the Sigma data covers a single seven-day window in January 2026 and the geographic scope of the household analysis was not disclosed in the report's published summary. The consumer survey and advertiser survey components are limited to four English-speaking markets.

Timeline

More than 100 years ago: Elias St. Elmo Lewis develops the first marketing funnel model, establishing the awareness-consideration-purchase framework.
November 2023: MiQ acquires Grasp, a French media governance and data quality company, extending its programmatic governance capabilities.
October 6, 2025: YouTube launches its Activation Partners program, naming MiQ Digital as one of four vetted partners for campaign management and audience intelligence.
November 2025: MiQ publishes its first AI confidence curve report, surveying 3,169 marketers worldwide on AI readiness gaps.
December 2025 - January 2026: MiQ and Censuswide field consumer survey (4,000+ respondents) and advertiser survey (600 respondents) across Australia, Canada, the UK, and the US for the From Funnel to Flexibility report.
January 2026: MiQ's Sigma platform processes 700 trillion signals across 53 million distinct households over seven days for the proprietary data component of the report.
January 5, 2026: PubMatic launches AgenticOS with MiQ's Sigma trading agent as an early participant, running live campaigns through Q1 2026.
January 20, 2026: Comscore's State of Programmatic Report finds 82% of marketers consider AI-powered optimization essential, and CTV captures 26% of media budgets.
March 14, 2026: IAB Europe's CTV Working Group documents measurement fragmentation and cross-device attribution challenges in CTV.
March 25, 2026: MiQ signs agreement to acquire Adsmovil's Latin America business, creating the region's largest independent programmatic provider.
April 2, 2026: Deep Sync and MiQ expand their political programmatic partnership for the 2026 US midterm cycle, promising voter data activation in under 48 hours.
April 7, 2026: MiQ acquires Rocket Lab, a Miami-based mobile app growth company, extending Sigma's signal coverage into in-app environments.
April 22, 2026: MiQ releases "From Funnel to Flexibility," a global report based on 53 million households and surveys of 4,000+ consumers and 600 marketers across four countries.

Summary

Who: MiQ, a global programmatic media company founded in 2010 in London and operating from more than 33 offices worldwide, with research conducted in partnership with Censuswide across consumer and advertiser populations in Australia, Canada, the UK, and the US.

What: MiQ released "From Funnel to Flexibility," a global research report combining proprietary Sigma platform data from 53 million households with consumer and marketer surveys. The report finds that 87% of consumers switch digital activities at least once per hour, 42% describe their purchase path as random, 91% use a second device while watching television, and 45% already use AI chatbots when shopping. Only 43% of marketing professionals report confidence in their measurement capabilities.

When: The report was published on April 22, 2026. The underlying survey data was collected in December 2025 and January 2026. The Sigma household data covers a seven-day observation window in January 2026.

Where: The consumer and advertiser surveys covered four markets - Australia, Canada, the United Kingdom, and the United States. The Sigma data analysis covered 53 million households, though the geographic breakdown of that proprietary data set was not disclosed in the published report.

Why: The report argues that the traditional marketing funnel - which assumes consumers move sequentially from awareness through to purchase - no longer maps onto actual consumer behavior. With consumers switching states between watching, browsing, and buying multiple times per hour, and with AI tools and social commerce compressing purchase journeys into minutes, MiQ contends that marketers need state-based measurement models and flexible, cross-screen media strategies rather than the stage-based planning frameworks that have dominated the industry for over a century.