Effectiveness researcher Les Binet and Kantar's global chief strategy officer Dom Boyd presented a study at the IPA Effectiveness Conference on October 7, 2026, finding that three in four UK consumer goods marketers involved in price promotions could not estimate what share of those promotions made money over the previous 12 months - even though promotions now absorb roughly twice as much of marketing and trade budgets as paid media advertising.

In Short

Brands spend more money cutting prices on supermarket shelves than they spend on advertising, and a new survey found most of the people running those discounts do not know whether they make or lose money. The researchers say the discounts often go to shoppers who would have bought anyway, and that repeating them trains people to wait for the next deal. Their answer is not to stop discounting entirely but to measure each promotion properly and move money from the loss-making ones into brand advertising.

A survey of the people who sign off the discounts

The research, titled "Kicking the promotion habit: Why companies are addicted to price promotions. And how to kick the habit to build healthier brand growth," is a 33-slide deck carrying Kantar copyright and the logos of Kantar and Ravensbourne University London, where Binet holds a visiting professorship. The Institute of Practitioners in Advertising published a summary on its website on October 7, 2026, describing the findings as "to be unveiled later today" at the conference.

According to the deck, fieldwork was conducted by Censuswide between August 18 and August 26, 2026. The sample was 250 senior leaders from UK consumer goods businesses, split 60% marketing and 40% insight. Seniority skewed heavily upward: 71% were C-suite, 18% vice presidents, 9% directors and 2% heads of department. By sector, 55% worked in food and drink, 25% in durables and 20% in other consumer goods. The IPA summary adds that all respondents were "directly involved in promotional decision-making, planning or evaluation."

The deck frames the study around three questions: what effect promotions have, why businesses really run them, and what would change their behaviour. Charts drawn from the survey carry the source line "Kantar x Binet Consulting; Censuswide." Neither document publishes a margin of error, a weighting method or a description of how respondents were recruited, which matters because several of the headline figures are drawn from subsets of the 250.

Marketers are not bystanders in this process. According to the survey, 97% of respondents said they had "a great deal or some influence" on promotion level, timing or design, and 94% had planned, approved, evaluated or seen the results of a significant manufacturer-supported UK price promotion in the past 12 months. Asked which function had the greatest influence over the total annual price-promotion budget, marketing came first, ahead of finance, sales and commercial, category and shopper, supply chain, revenue growth management and the chief executive. The chart shows only small gaps between the top functions, and the deck does not print the underlying percentages.

Promotions now outspend advertising

The study opens with a chart titled "Who's been eating our lunch?" It puts price promotions at 36% of marketing and trade expenditure, against 17% for paid media advertising. The IPA summary describes promotions as "the largest single component of marketing spend and around twice the size of paid media advertising," and says spending on them "has grown significantly over the past four decades."

The 36% figure is a composite. Its source line lists Cadent (2020 and 2024), IAB Europe and IAB UK (2025), the Advertising Association and WARC, EMARKETER's US consumer packaged goods data for 2024, Anderson and Fox (2019), Nielsen and POI. Several of those are US sources, while the survey that follows covers UK companies only. The deck does not show how the figures were combined.

For an advertising industry that spends a great deal of energy arguing over the split between brand and performance media, the comparison shifts the frame. Trade promotion budgets typically sit outside the media plan and are often negotiated between sales teams and retailers. Few of the debates PPC Land has covered about how much brand spending is undervalued by traditional measurement have touched on that larger pool.

Where the extra volume really comes from

The deck's second content slide decomposes the sales uplift from a typical promotion. Using an index where normal sales equal 100, a promotion lifts unit sales to 150. Of the extra 50 units, the deck classifies only 22 as genuinely incremental. Fifteen are time shifted, meaning shoppers bought earlier than they otherwise would; nine are cannibalised from the same company's other products; and four are relocated from other stores or channels. The original 100 units, the slide notes, "would have sold anyway, but at higher price."

That chart is labelled an "illustrative central case" built from author calculations and a synthesis of academic work including Van Heerde, Gupta and Wittink (2003), Van Heerde, Leeflang and Wittink (2004), Dawes (2012), Ailawadi and others (2007) and Kantar Worldpanel data. It is a model of a typical promotion, not a measurement from the survey. On those numbers, 22 of 150 units sold during the promotion would not have been sold without it - roughly 15%.

This is, in effect, the incrementality problem that digital advertisers wrestle with, applied to the shelf price. A promotion that shows a 50% sales spike looks successful in a weekly report. Whether it is profitable depends on whether the 22 genuinely new units cover the margin given away on the other 128.

The longer horizon

A following slide summarises the academic literature on longer-term effects across eight headings. Most deal buyers are not new customers. Deal buyers tend to be less loyal. Long-term sales volume gains "are rare and small," the long-term revenue effect "can be negative," and "around two-thirds of price promos are unprofitable." Frequent, predictable discounts, according to the deck, "shift sales forward, reduce margins & increase price sensitivity," while easily copied deals "can fuel promo wars, making the whole category more price-driven."

Sources for that slide run from Neslin and Shoemaker (1989) through Mela and colleagues (1997, 1998), Pauwels and others (2002), Srinivasan and others (2004), Steenkamp and others (2005), Dawes (2018), Guyt and Gijsbrechts (2020) and Gielens and Steenkamp (2026). The IPA summary renders the profitability point as "nearly two-thirds of price promotions lose money," attributing it to "the wider body of evidence reviewed in the research" rather than to the new survey.

The profit blind spot

The survey's central finding is a gap between what marketers want promotions to do and what they know about the results.

Asked for the primary commercial objective of their most recent significant promotion, 33% chose manufacturer profit and revenue, split into separate profit and revenue segments on the chart. Clearance of excess stock followed at 17%, retailer relationships, facings and distribution at 16%, volume sales at 13%, share or competitor defence at 11%, and penetration or trial at 9.4%. The IPA summary says profit and revenue were "chosen at roughly twice the rate of any other objective."

Then the confidence. According to the survey, 82% judged their own last promotion profitable - 28% clearly and 55% marginally. Those two components add to 83%, a one-point difference the deck does not explain and that is most likely rounding. The IPA summary adds that "more than 60% of those assessments" were based on "personal judgement or partial analysis rather than formal event-level profit-and-loss measurement." That breakdown does not appear on any slide in the deck.

Against that, 75% could not say what share of their own significant promotions over the past 12 months had been profitable. And 74% did not know the industry-wide rate at which promotions generate positive manufacturer profit over 13 weeks.

There is a base issue worth noting. The IPA summary says "75% of marketers cannot estimate what proportion of their promotions are profitable." The deck's footnote for that slide says the questions were asked "among those who helped to plan, approve or evaluate a promotion" in the past 12 months, a subgroup the deck elsewhere puts at 94% of respondents. The sample also includes insight leaders alongside marketers. The 75% therefore describes senior marketing and insight staff with recent promotional involvement, not marketers in general.

Boyd's assessment, according to the IPA: "Brands are using promotions because they want to drive profit and revenue growth. But what we see is that when it comes to actually delivering that profit and revenue, most brands have a blind spot. Many are measuring promotions on gut feel or partial analysis, not really knowing what the impact on profit is, and then repeating the cycle."

Why the cycle repeats

According to the survey, 76% of respondents said their organisation repeats a similar promotion often or sometimes even when its profit was never assessed. The deck's horizontal bar chart shows the "often" segment at roughly a quarter of respondents, with "sometimes" making up the rest of the 76%.

Measurement windows are short. Asked for the longest period routinely used in post-promotion evaluation, 2% said they look only at the promotion period itself, 10% at four weeks, 27% at eight weeks, 42% at 13 weeks and 19% at 26 weeks or more. The deck labels the first four groups "81% evaluate < 13 weeks." The IPA summary repeats that as "81% of marketers evaluate promotions over periods of less than 13 weeks."

That description does not match the chart. The 81% total includes the 42% who use 13 weeks as their longest window. Strictly read, 39% evaluate over periods shorter than 13 weeks, and 81% evaluate over 13 weeks or less. The underlying argument, that standard post-promotion windows end before longer-term effects such as rising price sensitivity become visible, is unaffected.

A second panel on the same slide is revealing. Across all respondents, 54% believe promotions are profitable after 13 weeks. Among a group the deck labels "High knowledge 'Elite'," that figure falls to 36%. The deck does not define how respondents qualified as high knowledge or how many were in the group.

Thirteen reasons

Respondents who repeated promotions despite negative or unassessed profit could pick up to three reasons. Consumer expectations led at 28%, followed by short-term volume and revenue targets at 22% and limited marketing involvement in promotion decisions at 21%. Clearance and supply came next at 19%. Market-share defence, retailer demand, facings and distribution, and annual trade commitments each drew 18%. Senior or quarterly targets, organisational habit and lack of reliable evidence each drew 17%, competitor promotions 15%, and sales-team incentives 14%.

No single reason dominates. That spread is itself a finding: the habit is held in place by retailers, finance calendars, sales incentives and shopper expectations at once, which makes it hard for any one function to break.

The deck turns those pressures into a diagram it calls "the dependency spiral." Short-term sales targets lead to an artificial volume high, continued discounting without proper evaluation, lower price signals and less differentiation, excess inventory, deeper discounts, cuts to brand investment and research and development, reduced demand and pricing power, increased retailer pressure, and consumers who seek value - which brings the loop back to more promotions.

Binet put it more bluntly, according to the IPA: "Price promotions are like class A drugs: expensive, dangerous and addictive. Yes, they give you an immediate volume high. But the high wears off quickly. Sales slump as soon as the offer ends and soon you need another fix. The opportunity is to identify which promotions lose money, cut them out, and reinvest the money in brand advertising."

Marketers already suspect the problem

The survey suggests the people running promotions are not short of doubts. According to the deck, 68% would prefer to run fewer promotions if they could do so without losing retailer support, distribution or competitive position. Agreement with three statements was high: 80% agreed that many promotional buyers would have bought anyway, 79% that the uplift overstates long-term impact, and 75% that repeated promotions raise price sensitivity.

The deck marks each of those beliefs "CORRECT!" and pairs the price sensitivity point with a separate figure: brands that promote repeatedly are "1.2x more price elastic." That number comes not from the survey but from Kantar Worldpanel and GfK data, modelling each brand's pricing power from Kantar BrandZ between 2022 and 2024 against price elasticity derived from real sales data, according to the slide's source note.

The most striking number in the deck may be the one about what would change behaviour. According to the survey, 91% said it would be easy or very easy to shift funds from price promotions to brand advertising if analysis showed brand advertising was more profitable in the long run. The slide headline reads: "But profit evidence is critical for budgets to shift."

That 91% is a hypothetical, and stated willingness in a survey is a weak predictor of what finance departments and retail buyers will accept. Rory Sutherland, in a talk covered by PPC Land in August, cited a figure that 80% of managers in very large organisations would cancel projects they expected to be profitable in order to hit quarterly forecasts. The survey's own reasons list, with quarterly and short-term targets near the top, points the same way.

Brand strength and pricing power

The deck's final research section argues that marketers undervalue the two levers that reduce promotion dependency: brand strength and pricing power.

Respondents were asked which single option was most effective at driving revenue growth and, separately, profit growth. Lower base prices scored 28% for revenue and 25% for profit. Higher base prices scored 23% and 26%. Strengthening the brand scored 20% on both. Price promotions scored 18% for revenue and 17% for profit, and increasing pricing power came last at 11% and 12%.

The IPA summary turns this into two ratios: brand strength rated "1.3 times less effective than lowering prices" and pricing power "1.5 times less effective than promotions." The ratios hold on the revenue figures (28 against 20, and 18 against 11) and come out lower on profit (25 against 20 is 1.25, and 17 against 12 is about 1.4).

What pricing power is worth

The case for pricing power rests on data from "The Effectiveness Equation," a Kantar and Google analysis. The relevant slide credits "Kantar x Google 2026 The Effectiveness Equation." It matches Kantar BrandZ brands with S&P Capital IQ company financials and splits them into the top and bottom 10% by pricing power. The top tier averaged a net income margin of 15% against 10% for the bottom tier - "1.5x higher." During what the slide calls an "inflationary period," the gap widened to 1.78 times.

The IPA summary reports only the 1.5 figure, as margins "around 1.5 times higher." There is also a date inconsistency in the source lines: one slide cites "Kantar x Google 2026 The Effectiveness Equation," while another cites "The effectiveness equation, Google and Kantar 2025." Google's report of that name, developed with Kantar, Nielsen and others, was covered by PPC Land in March 2025, which suggests the 2026 label refers to an update or a later analysis built on the same work.

A separate Kantar, Google and Marketing Week study cited in the deck, "The Language of Effectiveness 2026," found that 91% of marketers agree strong brands can charge higher prices, but only 42% say they can quantify the link between brand strength and price elasticity or pricing power. The deck calls this a "capability gap."

A worked example and a disputed percentage

To show why elasticity matters, the deck models the same 14% price rise for two brands. With what it labels stronger price inelasticity (0.6), volume falls 7% and revenue rises 7%. With weaker inelasticity (0.7), volume falls 10% and revenue rises 2%. The example echoes a UK skincare brand case from the original Effectiveness Equation, in which a 14% price rise cost about 7% of volume.

The slide headlines the difference as "+350% revenue impact." The arithmetic does not support that label. A 7% revenue gain is 3.5 times a 2% gain, which is 250% more, not 350%. The label appears to confuse a multiple with a percentage increase.

Salience against difference

Another Kantar chart splits the drivers of demand power and pricing power. Salience - how readily a brand comes to mind - accounts for 42% of demand power against 20% for difference, according to the slide. For pricing power the order reverses: difference accounts for 49% and salience 6%, which the slide headlines as "Difference is 8x more important than Salience." The footnote on that slide reproduces the survey question wording from an earlier chart about promotional influence, which appears to be a copying error; the underlying source for the salience data is not stated.

The deck also reproduces an IPA Effectiveness Databank analysis of 559 awards cases from 2000 to 2023. Cases reporting four or more very large branding effects reported very large reductions in price sensitivity 11.7% of the time, against 3.8% for cases reporting none and 3.5% for those reporting one. Cases with two or three such effects both came in at 10.3%. That slide carries a second source line for Kantar BrandZ Global Top 100, attached to a chart showing the BrandZ strong brands portfolio up 504% between 2006 and 2026, against 451% for the S&P 500 and 234% for the MSCI World Index.

Six steps, and when to promote

The deck closes with a "six steps to getting clean" programme. Marketers are told to get trained, on the basis that "trained marketers get better results"; to learn best practice from the literature; to use econometrics to "measure & optimise promos, price & ads with a unified financial model"; to model short and long-term effects, because "13 weeks is too short"; to measure pricing power; and to reinvest by cutting loss-making promotions and putting the money into brand.

The recommendation for a unified model is notable because many marketing mix models treat price and promotion as control variables rather than as budget lines to optimise alongside media. PPC Land has reported on the limits of standard mix models, including research showing a standard model overstated paid search returns by 2.5 times on synthetic data, and on a Winterberry Group survey in which only 9% of US marketing executives said they performed unified measurement extremely well.

The deck does not argue for abolishing promotions. A "selective not reflexive" slide lists when promotions can work: when they recruit genuinely incremental buyers, generate full-price repeat buyers, expand consumption or create new occasions, support a valuable launch, secure quantifiable distribution value, unlock a new pack or channel format, solve a defined clearance problem, or reinforce a brand's "meaningful difference and value positioning." They fail, according to the same slide, when they subsidise existing buyers, cannibalise the portfolio, bring future purchases forward, are repeated with no profit assessment beyond three months, erode base price and elasticity, or divert investment from building the brand. The IPA summary describes the target as promotions "that subsidise demand the brand already owns, are repeated without proper assessment and/or that divert investment away from long-term brand growth."

Laurence Green, the IPA's director of effectiveness, said according to the institute: "This research reinforces the importance of evaluating marketing activity through the lens of profitability and long-term effectiveness, rather than short-term sales uplift alone, helping marketers make more informed decisions about how pricing, promotions and brand investment work together to support growth."

The Volkswagen case

The deck's single case study is Volkswagen UK. A market share chart running from 1980 shows the brand at roughly 4% to 5% through the 1980s and early 1990s, a period the slide labels "Low SOV, Deep discounts." From the mid-1990s, money was "shifted from discounts to ads," its share of voice doubled, and market share rose to about 9% by 2011. The slide claims the "fastest growth in UK history," "+£2bn revenue in first 8 years alone" and "zero additional spend." Those figures are presented without a source line and come from a well-known effectiveness case rather than from the new research. The final slide reads "Just Say No."

Why the findings land differently for digital marketers

For the performance marketing community, the study sits on familiar ground. The same arguments about short measurement windows, non-incremental volume and the gap between attributed and real return have run through debates about search, social and retail media for years. Earlier this year, Omnicom Media Group UK's Charlie Ebdy argued in an IPA paper that advertising returns face a structural ceiling, a claim that set off a public argument among effectiveness specialists over whether measurement or execution was to blame.

Discounts are also moving into ad platforms. Google said it would begin extracting discount offers from advertisers' websites and attaching them to some Search and Performance Max ads as promotion assets from October 12, 2026, an automatic enrolment PPC Land reported on October 6. That change does not set the size of any discount, but it puts promotional offers in front of more searchers by default.

The pressure from shoppers is real as well. An Adtaxi survey of more than 1,100 US adults reported by PPC Land in September found 40% had switched to lower-cost brands and 49% were comparing prices more often. In Europe, private label reached 38.5% of grocery value across 17 markets in 2024. And the long-term cost of cutting brand investment to fund short-term volume has precedent: PPC Land's account of how Kraft mac and cheese lost market share traced share falling from 45% to 39% after years of zero-based budgeting.

What the Binet and Boyd study adds is a direct measure of how little the people approving promotions know about their returns. Whether the 91% who say budgets would move given better evidence will be tested depends on who builds that evidence - and whether retailers, finance teams and sales incentive plans allow the money to move once it exists.

Timeline

Summary

Who: Les Binet, visiting professor at Ravensbourne University London, and Dom Boyd, global chief strategy officer at Kantar, with fieldwork by Censuswide among 250 senior marketing and insight leaders at UK consumer goods companies. The IPA hosted the presentation and published a summary, with comment from its director of effectiveness, Laurence Green.

What: A survey and evidence review finding that price promotions take 36% of marketing and trade spend against 17% for paid media, that 75% of respondents involved in promotions cannot estimate what share of their promotions were profitable, that 76% repeat promotions whose profit was never assessed, and that 81% evaluate them over 13 weeks or less. The deck proposes six steps for shifting money from loss-making promotions into brand building and pricing power.

When: Fieldwork ran from August 18 to August 26, 2026. The findings were presented at the IPA Effectiveness Conference on October 7, 2026, the same day the IPA published its summary.

Where: The survey covers UK consumer goods manufacturers in food and drink, durables and other categories; the spending comparison draws partly on US data. The presentation took place at the IPA Effectiveness Conference, run by the London-based Institute of Practitioners in Advertising.

Why: Promotions are the largest single line in many consumer goods marketing budgets, yet the survey suggests most are judged on gut feel and short windows. For advertisers, the findings matter because any money moved out of trade promotions is a potential source of new brand and media budget - if companies can produce the profit evidence the respondents say they need.