A long-term ROI multiplier is the factor by which an advertising campaign's short-term return on investment (ROI) has to be multiplied to reach its full return, once the sales and profit that arrive months or years later are counted. A multiplier of 2.0 means the campaign eventually earns twice what it appeared to earn in its first weeks. The figure exists because most of the measurement marketers consult daily, from platform attribution to lift tests, looks at a window of days or weeks, while econometric evidence suggests that a large share of what advertising does arrives later. No standards body defines the term, so its exact meaning depends on who calculates it.

How the ratio is calculated

The most explicit published definition comes from Profit Ability 2, a 2024 study for Thinkbox, the UK commercial television marketing body. It defines the multiplier as sustained plus short-term return, divided by short-term return. Short-term covers the first 13 weeks after advertising runs; sustained covers week 14 to roughly 24 months. BERA.ai, a brand analytics firm, gave a simpler version in WARC's report The Multiplier Effect: total return on ad spend (ROAS) divided by short-term ROAS. The two differ in where the windows are drawn and whether the numerator is revenue or profit.

In a hypothetical example, a retailer spends 1 million pounds on television. A marketing mix model (MMM) credits it with 1.8 million pounds of incremental gross profit within 13 weeks, a short-term profit ROI of 1.8. A long-term analysis adds 2.7 million pounds between week 14 and week 104, for a full ROI of 4.5 and a multiplier of 2.5. The same retailer spends 1 million pounds on generic search ads, which return 2.3 million pounds in 13 weeks and a further 1.2 million pounds later, for a full ROI of 3.5 and a multiplier of about 1.5. On a 13-week view search wins, 2.3 to 1.8. Over two years television wins, 4.5 to 3.5. The multiplier reverses the ranking.

The ratio says nothing about profitability: a short-term return of 0.4 with a multiplier of 3.0 ends at 1.2, possibly still below break-even.

Where the long-term number comes from

Four sources supply the numerator. The first is carryover, modelled through adstock, the decay function the British media researcher Simon Broadbent named in 1979. Adstock rarely produces a long-term multiplier by itself. With a weekly retention rate of 0.8, about 95% of an ad's total modelled effect has arrived within 13 weeks, by PPC Land's calculation. Google's open-source Meridian model also cuts effects off at a maximum lag, eight periods by default. Its documentation concedes that "media can potentially have a long effect that goes beyond max_lag" and that, "for practical reasons, we truncate the effect."

The second source is the baseline, the sales a model attributes to neither media nor promotions. Documentation for Meta's open-source Robyn package says the project avoids the terms short-term and long-term "without any agreed definition", and argues that the long-term effect "should describe media impact on baseline sales." Modellers following that logic let the baseline drift, then relate its movements to cumulative advertising or brand metrics such as awareness.

The third is experiments with long follow-up. Leonard Lodish and colleagues summarised 55 in-market television tests for established packaged goods in Marketing Science in 1995. Where an increase in TV weight lifted sales significantly in the first year, the effect roughly doubled over the following two years. Where it did not, no later effect appeared.

The fourth is benchmarks. Profit Ability 2 calculated sustained effects using multipliers supplied by EssenceMediacom, Gain Theory, Mindshare and Wavemaker UK. In practice, many multipliers in circulation are norms drawn from agency databases and applied to a brand, not estimates made from that brand's own data.

Origin and evolution

Lodish's 1995 paper gave the ratio its first experimental footing. Les Binet and Peter Field of the Institute of Practitioners in Advertising (IPA) built the commercial case with Marketing in the Era of Accountability in 2007 and The Long and the Short of It in 2013. The second book analysed 996 campaigns for 700 brands in 83 categories from the IPA Effectiveness Databank and proposed that roughly 60% of budget go to brand building and 40% to sales activation.

The term gained currency with the first Profit Ability study, launched on November 16, 2017, by Ebiquity and Gain Theory for Thinkbox. Covering more than 2,000 campaigns in 11 categories, it found average short-term ROI, measured within three to six months, of 1.51 pounds per pound spent, rising to 3.24 pounds over three years. Television's return rose from 1.73 to 4.20 pounds. Marketing Week reported television's long-term multiplier as nearly 2.5 and online display's as one, with no examples of pay-per-click search driving long-term response. Matthew Chappell, a senior partner at Gain Theory, wrote that "there is a long-term multiplier effect that many C-suite executives may not know about."

Profit Ability 2 followed on April 24, 2024, covering 141 brands, 1.8 billion pounds of media spend and 10 channels from 2021 to 2023. Average short-term profit ROI was 1.87 pounds and full ROI 4.11 pounds, an implied multiplier of 2.2. Sustained effects accounted for 58% of advertising-generated profit. Implied multipliers ranged from about 1.5 for generic search (2.29 to 3.52 pounds) to about 3.3 for linear television (1.82 to 5.94 pounds). On January 28, 2025, WARC published The Multiplier Effect with Analytic Partners, BERA.ai, Prophet and System1. It cited Analytic Partners' ROI Genome database as putting long-term multipliers generally between 1.1 and 2 times revenue ROI, varying by channel.

Why it matters for marketers

A plan optimised on 13-week or seven-day returns will consistently favour channels with low multipliers, because their value is front-loaded and visible. Google's "The Effectiveness Equation" report, citing an Ekimetrics meta-analysis of studies from 2017 to 2022, found that marketing returns in months 5 to 24 equal those of the first four months, which implies a multiplier near two.

TransUnion and MMA Global case studies published in October 2025 found long-term conversion lift 1.8 times greater than short-term lift at Ally Bank, 2.5 times at Campbell's and 6.0 times at Kroger. Conventional tools captured 17% of Kroger's online sales effect. Standard sales lift and brand lift studies, which typically run for weeks, cannot see such effects.

The ratio has also become a selling point for media owners. YouTube's creator marketing materials cited a Circana meta-analysis, commissioned by Google, that found 86% higher incremental long-term ROAS than paid social, modelled through brand equity pathways across 10 US packaged-goods brands.

Limitations and disputes

Definitions do not match. The 2017 study measured short-term returns over three to six months and total returns over three years; the 2024 study used 13 weeks and about 24 months. Two multipliers can differ because of the windows rather than the advertising.

The baseline is hard to assign. Long-term effects are usually inferred from baseline movements, which also reflect distribution, pricing, product changes and competitors. Ryan Dew, Nicolas Padilla and Anya Shchetkina found in 2024 that nonlinear and time-varying effects often cannot be separated in standard mix model data, with conflation in up to 99% of high-carryover settings. The same coverage reported a 2026 simulation in which a conventional model put paid search ROAS at 10.61 against a true 4.20.

Point estimates hide uncertainty. Industry studies usually report headline multipliers without intervals, even though the long-term component rests on more extrapolation than the short-term one.

Samples and sponsors. Profit Ability 2's sample consists of large brands that commission econometrics, and Thinkbox exists to promote television. Byron Sharp, director of the Ehrenberg-Bass Institute, said in August 2022 that Binet and Field had used "a very weird data set, which is award submissions," and called the 60:40 rule "really terrible, very misleading," according to Mi3. Platform-commissioned figures carry the obvious conflict, a concern PPC Land raised in its examination of Meta's Robyn.

Ratios are not profit. Binet himself told Google Marketing Live in Sydney in June 2026 that "the main driver of profit is not ROI, it's the budget by a factor of eight or nine to one," according to B&T. A high multiplier on a small budget can be worth less than a modest one on a large budget.

Not the same as

ROAS and iROAS are revenue-to-cost ratios over an attribution window or test period. The long-term multiplier is a ratio of two returns and describes how a return is spread over time.

Customer lifetime value measures what a customer relationship is worth over its life, whatever acquired it. The multiplier concerns advertising effects, including effects on existing customers and on price sensitivity.

Marketing mix modelling is the method that most often produces the multiplier, not the multiplier itself. An MMM can report short-term returns only.

The Multiplier Effect, WARC's 2025 framing, refers to brand and performance activity amplifying each other, an interaction between channels rather than a time profile.

Recent developments

Platforms are stretching their counting windows. At Google Marketing Live on May 20, 2026, Google introduced Qualified Future Conversions, and announced that it plans to feed the signals into Meridian without publishing how circularity would be handled. The metric counts conversions up to 180 days after a click once a qualifying action occurs within seven days.

In January 2026, Charlie Ebdy of Omnicom Media Group UK argued in an IPA essay that distraction and clutter act as a structural drag on returns, prompting an exchange among industry figures over how to measure brand building. As of October 2026, no industry body has standardised the windows, and each published multiplier remains specific to its study.

Timeline

  • 1979: Simon Broadbent introduces adstock, the carryover concept used in mix models.
  • 1995: Lodish and colleagues publish 55 in-market TV tests in Marketing Science, finding first-year effects roughly double over two further years.
  • 2007: Binet and Field publish Marketing in the Era of Accountability through the IPA.
  • 2013: Binet and Field publish The Long and the Short of It, based on 996 IPA Databank campaigns.
  • November 16, 2017: Thinkbox launches Profit Ability, by Ebiquity and Gain Theory, reporting short-term ROI of 1.51 pounds and three-year ROI of 3.24 pounds.
  • August 2022: Byron Sharp criticises the 60:40 rule and the IPA award-entry data.
  • March 2024: Google unveils Meridian.
  • April 24, 2024: Thinkbox launches Profit Ability 2, defining the long-term multiplier and reporting 1.87 pounds short-term and 4.11 pounds full ROI.
  • January 28, 2025: WARC publishes The Multiplier Effect with Analytic Partners, BERA.ai, Prophet and System1.
  • January 29, 2025: Meridian opens to all marketers.
  • March 2025: Google publishes The Effectiveness Equation, citing equal returns in months 5 to 24 and months 0 to 4.
  • October 2, 2025: TransUnion and MMA Global report long-term lift 1.8 to 6.0 times short-term lift.
  • January 2026: Charlie Ebdy publishes his IPA essay on structural limits to effectiveness.
  • May 2026: YouTube cites Circana research claiming 86% higher long-term ROAS than paid social.
  • May 20, 2026: Google introduces Qualified Future Conversions at Google Marketing Live.
  • June 2026: Les Binet tells Google Marketing Live in Sydney that budget outweighs ROI as a driver of profit.
  • August 21, 2026: Niklas Heusch posts research showing a mix model overstating paid search ROAS 2.5 times.

Summary

  • Who: Econometricians, agencies and measurement firms such as Ebiquity, Gain Theory, Analytic Partners and Ekimetrics calculate long-term multipliers; trade bodies including Thinkbox, the IPA and WARC publish them; platforms such as Google and YouTube now cite long-term returns in their own research; researchers including Les Binet, Peter Field and Byron Sharp dispute how the evidence should be read.
  • What: The ratio of advertising's full return, including effects that arrive after the first weeks or months, to its short-term return, commonly calculated as sustained plus short-term return divided by short-term return.
  • When: Experimental evidence dates to Lodish's 1995 study; the IPA work of 2007 and 2013 popularised long-term effects; the term spread through Thinkbox's Profit Ability studies in 2017 and 2024, with published multipliers mostly between about 1.1 and 3.3 depending on channel and study.
  • Where: In marketing mix models, econometric meta-analyses, long-run experiments, media trade body research and, increasingly, platform measurement tools with extended counting windows.
  • Why: Short-term measurement favours channels whose effects arrive quickly; the multiplier attempts to correct for that, though its value depends on the windows chosen, the treatment of baseline sales and who commissions the study.