Why ROAS Alone Cannot Measure Marketing Success

Industry & Competitive Context

By 2016, digital advertising had become the dominant channel in global marketing spend, with the Association of National Advertisers (ANA) and its chairman, P&G Chief Brand Officer Marc Pritchard, describing digital media as having grown to roughly $200 billion out of a worldwide total marketing spend of $600 billion. This growth coincided with structural problems in the digital media supply chain. An ANA-commissioned 2016 study into media agency contracts had surfaced concerns about non-transparent rebate arrangements between agencies and media owners, a controversy Pritchard publicly referenced as central to his subsequent reform agenda. At the same time, programmatic buying, ad fraud, bot traffic, and opaque viewability standards had become systemic issues across the Google–Facebook-dominated digital ecosystem. Within this environment, Return on Ad Spend (ROAS) and other last-click, platform-reported metrics had become the default currency by which marketers justified digital budgets, even as the underlying measurement infrastructure was, in Pritchard’s own words, “murky at best and fraudulent at worst.”

Brand Situation Prior to the Initiative

Procter & Gamble, the world’s largest advertiser by spend, had been a major participant in this shift toward digital and programmatic buying. However, internal data began to reveal a gap between reported platform performance and actual consumer exposure. Pritchard disclosed that transparent viewability data showed the average view time for an ad on a mobile newsfeed was 1.7 seconds little more than a glance. This finding was significant because it emerged after P&G had already begun acting on brand-safety concerns: in the quarter from April to July 2017, the company had cut digital ad spending by between $100 million and $140 million due to bots and brand safety concerns, a decision disclosed during a company earnings call rather than through a marketing campaign announcement. This indicates that P&G’s digital spend, even where it may have cleared conventional ROAS thresholds on a platform-reported basis, was not reliably reaching or engaging real consumers a distinction ROAS as a formula does not capture, since it measures attributed revenue against spend, not the integrity or quality of the underlying exposure.

Strategic Objective

Pritchard articulated the company’s objective publicly on two major occasions: a speech at the IAB Annual Leadership Meeting on January 29, 2017, and an address at the ANA Media Conference. The stated goal was not to abandon digital media but to force structural transparency across the “media supply chain,” so that spend could be evaluated against verified human reach and brand-safe placement rather than platform-reported performance alone. Pritchard framed this as a responsibility rather than a preference, stating that driving media transparency “should not be optional” and that the industry needed an “antiquated media buying and selling system” rebuilt for the digital era, with savings reinvested into better advertising to drive growth. As ANA chairman, he also pursued a parallel objective of eliminating non-transparent agency compensation structures and reducing the number of agencies and production costs across the system, targeting a combined reduction of roughly $400 million through agency model changes and up to half a billion dollars in savings from reduced agency fees and ad-production costs, with an ambition of squeezing $1 billion or more out of the system.

Campaign Architecture & Execution

The execution was financial and structural rather than creative. Having already cut $100–140 million in the first fiscal quarter affected, P&G extended the reduction through the remainder of calendar 2017, bringing the full-year digital ad spend cut to $200 million, with reductions concentrated among several large digital platforms, cut by between 20% and 50%. The company was explicit that this was not an overall media spending cut: funds were reallocated rather than eliminated, with a company spokeswoman confirming that funds were reinvested to increase media reach, including in areas such as TV, audio and e-commerce media. Context for scale: in 2017 the company spent $7.1 billion on advertising overall, with digital accounting for around a third of that spend. CFO Jon Moeller reinforced the operational logic on the related earnings call, stating that P&G did not “need to be spending money that is seen by a bot and not a person,” directly tying the cuts to measurement integrity rather than to a reduction in overall marketing ambition.

Positioning & Consumer Insight

The underlying insight that Pritchard publicized was that transparent, third-party-verified viewability and reach data told a different story than platform dashboards did. A metric like ROAS, calculated from attributed conversions against spend, can appear efficient even when the underlying “view” barely registers with a human being — as the 1.7-second average mobile newsfeed view time illustrated. Pritchard’s broader positioning, delivered as ANA chairman, was that the industry had been “giving new media a pass” on accountability standards that would never have been tolerated in traditional media, and that marketers needed to reassert control over measurement rather than accept platform-reported outcomes at face value. This reframed the marketing question from “did this impression convert” (a ROAS-style question) to “was this impression real, viewable, brand-safe, and reaching the intended audience” a reach-and-quality question that ROAS alone does not answer.

Media & Channel Strategy

The documented channel shift moved spend away from narrowly targeted programmatic digital inventory and toward channels with verified broad reach: television, audio, and e-commerce media. This is consistent with the company’s own description of its reallocation strategy, reported independently across multiple outlets, that the $200 million was invested into other areas with stronger “media reach,” including e-commerce, audio and television. Separately, and on a larger structural level, P&G pursued agency consolidation and fee renegotiation as part of the same transparency push, a strategy distinct from channel mix but aimed at the same underlying goal: reducing spend that could not be independently verified or justified.

Business & Brand Outcomes

The most directly attributable outcome reported in the press was reach-based rather than sales-based: multiple outlets, citing P&G and Adweek’s reporting, stated that the $200 million cut in 2017 digital ad spend was associated with a 10% increase in total reach. This is a critical data point for the topic at hand, because it demonstrates a scenario in which reducing spend presumably measured by platforms using ROAS-oriented logic increased a core marketing outcome (reach) a result that a narrow ROAS lens, focused on attributed conversions per channel, would not have predicted or explained.

On business performance, P&G’s own quarterly disclosures are the most reliable record. For the fiscal quarter in which the $140 million reduction occurred, reduced digital ad spend contributed to the company beating earnings expectations, and organic sales growth also beat both analyst predictions and competitor performance, rising 2% despite the reduced advertising support. Looking at the full fiscal year 2017 period covering the described cuts, the company also reported that e-commerce organic online sales rose approximately 30% for the period, significantly outpacing offline sales, with online sales representing more than 5% of total business in fiscal 2017, and that CEO David Taylor stated the company met or exceeded its fiscal 2017 objectives, including accelerating organic sales growth and cost savings.

For the subsequent full fiscal year 2018 (the period following the full $200 million annual reduction), P&G’s reported net sales were $66.8 billion, a three percent increase versus the prior year, with organic sales increasing one percent on a two percent increase in organic volume. Within that year, Beauty segment organic sales increased seven percent, driven by product and packaging innovation and increased investment in consumer communication, while Grooming segment organic sales decreased three percent, attributed to investments in consumer and customer value in North America. CEO David Taylor characterized the year as one in which the company “made important progress,” delivering strong volume and consumption growth and market share improvement, albeit with organic sales slightly below target.

Strategic Implications

The P&G case, read alongside long-standing academic and industry research on marketing effectiveness, illustrates a structural limitation of ROAS as a standalone success measure. ROAS is, by construction, a short-window, attribution-dependent, channel-specific ratio: it credits conversions to the last measurable touchpoint and does not account for reach quality, brand-safety, viewability, or the compounding effect of brand-building investment over time. The P&G case demonstrates this limitation directly, since spend reductions in channels where ROAS-style platform metrics were presumably being reported did not damage, and were associated with improved reach and continued organic growth.

This finding is consistent with the most extensively cited independent research on marketing effectiveness, the Institute of Practitioners in Advertising’s (IPA) Effectiveness Databank, analyzed by Les Binet and Peter Field across 996 campaigns entered into the IPA Effectiveness Awards between 1980 and 2010. Their analysis concluded that short-term, directly attributable “activation” effects and long-term “brand-building” effects behave differently and require different budget allocations, with their headline finding recommending an optimum balance of around 60% of budget on brand-building activity and 40% on sales activation. A follow-up decade of IPA research reaffirmed this architecture, with Peter Field warning that the industry’s growing short-termism, or “tide of activation,” was correlated with declining marketing effectiveness, measured as the number of very large business effects reported in IPA case studies. Binet and Field’s original 2013 report was explicitly framed as a response to the risk that marketing was “sleepwalking towards a precipice” by over-using short-term metrics in ways that would damage brand profitability over the long term a concern structurally identical to the measurement gap P&G’s leadership identified in its own supply chain.

The strategic implication for marketing organizations is that ROAS, while useful for optimizing short-term, lower-funnel, directly attributable spend, cannot on its own evaluate reach quality, brand safety, long-term share growth, or price elasticity all of which require separate, often slower-moving measurement frameworks such as verified viewability audits, econometric modeling, and brand-tracking studies. A marketing organization that optimizes exclusively for ROAS risks, as the P&G data suggests, under-investing in broad-reach channels whose contribution to business outcomes is real but not captured by last-click attribution, while over-crediting narrowly targeted digital inventory whose “performance” may reflect fraudulent or low-quality impressions rather than genuine consumer engagement.