The cut that never happened.

The challenge

An established consumer health and wellness brand faced a familiar brand vs performance marketing question at planning time. Its marketing mix model said television, online video and out of home returned about a dollar in revenue for every dollar spent, while paid search returned more than two. Brand already held only 30% of the $50M media budget, and the draft annual plan followed the numbers: move $5M, a third of all brand spend, into performance. But the marketing team suspected the model was missing something. Brand campaigns seemed to lift sales long after they ended, while the brand’s loyal base did not look like it sustained itself. The brand asked MASS Analytics to find out what the short term view could not see before it signed off the budget.

How the previous model scored brand vs performance marketing

Before the rebuild, the previous model credited each campaign only with the sales that arrived within a few weeks of it airing. The model treated anything later as base demand, the sales a brand is assumed to earn on its own. For performance channels that is a fair view, because their effect is almost immediate. For brand media, however, it is not. Television and video build awareness and consideration that turn into purchases over months and years, yet the model filed all of that under base. On that basis, brand media looked like the weakest investment on the plan.

Revenue returned per dollar of media

Short term view vs full term view, by channel

Short term ROI (previous model)Full term ROI (rebuilt model)

Paid Search

$2.10

$2.21

Display

$1.30

$1.37

Paid Social

$1.45

$1.60

OOH

$0.85

$1.53

Online Video

$0.95

$2.09

TV

$1.05

$2.60

Four things the short term view could not see

• Brand media was judged on a few weeks of results. Television and video returned about $1.00 per dollar in that window, so every planning cycle pulled money toward channels that looked better in the short term.
• Brand equity was hidden in the base. Demand built by years of advertising sat alongside genuinely organic sales, so the model could not say how much of the base the brand had paid for, or how fast it would fade.
• Performance channels took credit for demand they did not create. Paid search captured shoppers that brand campaigns had already persuaded, yet it got the whole sale.
• The cut could not be tested. With no link between brand media and base sales, the model had no way to show what reducing brand investment would cost two or three years later.

The rebuild then closed each gap by measuring advertising’s full path to sales, in three linked stages summarized in Table 1.

Table 1: Advertising’s long term path entered the model as three linked stages. Price, promotions, seasonality and performance media kept their standard treatment.

The solution

So MASS Analytics rebuilt the marketing mix model around the way brand advertising works over time. Advertising builds consideration, consideration accumulates into brand equity, and brand equity keeps customers buying long after a campaign ends. Rather than asking one model to see all of that at once, two linked models followed the chain, an approach known as nested modeling: one explained what drives consideration, the other explained sales. The model used four years of weekly data, and its short term returns also sat in line with the MASS Analytics benchmark database.

Following advertising from first impression to repeat purchase

First, consideration came from the brand’s independent tracker, measured weekly and matched to every campaign. The model let brand channels work at their natural pace, with effects that build and fade over months rather than days. In fact, more than half of the movement in consideration (56%) traced back to media, led by TV and online video. That media built consideration was then carried into base sales as a measure of brand equity, separating the demand the brand had paid for from demand that would exist without advertising. Finally, every channel’s return was split into what arrives within the first 13 weeks and what arrives afterwards, the same cut Profit Ability 2 uses, so the team could see both horizons side by side.

Built to survive a finance review

A long term number only changes a budget if the CFO trusts it. The team checked every stage of the chain before it fed the next, so a weak link would show in the final result rather than hide inside it. Short term returns were checked against the MASS Analytics benchmark database, and the long term uplift against Profit Ability 2, the UK’s largest study of advertising payback, before any figure was reported. The model was then turned into a scenario planner, so the team could compare the draft plan, the alternative and a no brand media scenario on one basis: dollars of revenue.

Results and Impact

As a result, the full term view reversed the draft plan. Brand media returned $2.30 in revenue for every dollar spent, not $1.00, and the $5M earmarked to leave brand stayed where it was, earning an estimated $4.0M more revenue over three years than the draft plan would have.

Brand vs performance marketing: $2.30, not $1.00

The previous model saw only the first $1.00: the sales brand media drives within a few weeks. The rebuild then added the second part, the demand it builds through consideration and brand equity in the months and years that follow, and brought the full term return to $2.30. That 2.3x multiplier sits just above the 2.2x average Profit Ability 2 reports across UK advertisers, but it was far from uniform. The full term view lifted the return on television 2.5 times and on online video 2.2 times, while paid search, social and display barely moved because almost all of their effect is immediate. The base also told the same story: 48% of it was demand built by past advertising. In contrast, stopping brand media altogether would have saved $15M a year but cost 16% of base sales within three years as that equity faded.

Brand vs performance marketing at flat total spend

This was a reallocation question, not a request for more budget. Paid search still posted the strongest short term return, but it was close to saturation: the next dollar there returned about $1.10, against about $1.90 for the next dollar in brand media. Kept in brand, the $5M returns about $9.5M over three years, while moved into search it returns about $5.5M. That $4.0M difference, therefore, is the value of the decision. Total media spend stayed flat, and the brand now plans on marginal ROI, so it judges brand and performance on the same horizon.