Econometrics: Detailed data zeroes in on advertising impact

Marketers are constantly looking to determine how to optimise their media mix and econometrics is one of the most important elements in their toolkit, allowing them to quantify the role of marketing in driving consumer purchase behaviour.

For the uninitiated, econometrics is the science of matching sales patterns with current or near-contemporary patterns in other data series. Relying on economic and marketing theory and statistical tests, econometrics can be used show whether individual inputs possibly explain sales, whether two separate inputs may be doing the same job and whether the model as a whole validly represents the sales patterns.

Use of econometrics in media agencies is more widespread in Europe and the US, but there’s a growing trend in its adoption across Asia. Gillian Harrison, managing partner at Oystercatchers, believes that over the last decade the practice has evolved from “being a new discipline for agencies to consider to a more mainstream part of the media offer” and which is becoming “fully integrated into agency planning models”.

Harpreet Kaintel, chief informat ion officer at ZenithOptimedia, attributes this growth to “the push by global clients, tightening of marketing budgets in big markets like China and India, and continuing growth of new marketing channels”.

John Perella, managing partner, Data2Decisions, APAC, puts the change down to the fact that businesses are always looking for new sources of growth.

“Econometrics provides a better way to evaluate marketing in the context of all business drivers,” he says.

Such global businesses are keen to establish their advertising return-on-investment and many econometrics studies have shown that the short-term return on advertising is often less than the investment itself.

However, these studies show that the long-term benefit of advertising does justify the investment; though, to fully appreciate how, researchers and brands must be careful not to over-simplify the issue and should use accurate modelling.

Experts claim econometric modelling is poorly used if it is just a post-campaign ROI justification tool. It is more effective if used to identify diminishing returns and provide the basis for future-facing planning, budget optimisation and forecasting at the strategic planning and the campaign planning stages.

At present in Asia, the demand for econometrics is based on two tiers, says Perella, with “most of the well-established multinationals insisting on econometric modelling”, with the more localisedbusinesses less likely to request it — “although this is now changing”.

According to Harrison, media agencies that offer econometric modelling services can demonstrate both a rounded understanding of contemporary business tools and a compelling case for the media plans they present to clients.

“However, a properly staffed and structured econometric offer does not come cheap so agency profitability will be under pressure unless the service is offered with the help of wider holding company resources,” she adds.

Kaintel believes it can help agencies “build long-lasting client relationships by taking clients on a revenue journey”. By demonstrating both the returns that their recommendations have generated for clients and their dedication to increasing them, “agencies prove themselves to be valuable business partners rather than simply service providers,” he states.

Econometrics MediaCom carried out for P&G in Japan proves this point. In 2012, P&G’s hair care brands were facing mounting competition from new, niche brands, offering non-silicon hair products targeted at women.

Following ongoing econometrics work, MediaCom shifted focus to the profitability of each brand, as opposed to using absolute sales.

The agency considered the ROI derived from each media, examining the historical data for each brand and selected key touchpoints based on its target audience of affluent Japanese women. This helped create a nonlinear optimisation system, which reallocated budgets from lower response activities to drive overall category growth.

As a result, P&G lowered its TV spend from 76 per cent
to 72 per cent of the budget, and diverted this toward digital and magazines. The company’s online video spend shot up by 200 per cent while it dropped OOH altogether as the data showed this was not an effective means of raising awareness. 

How to offer such econometric services can depend on the scale of the agency. Perella says that these services should ideally be offered in the form of a specialist sub-agency (Data2Decisions is part of the Dentsu network). 

“Clients want a model where the analytics are independent and objective, yet completely integrated with the media planning,” he explains. “A culture of collaboration is essential in making this work.” 

That’s a view Harrison shares. “Our experience is that the leading agencies will consider econometrics as a discipline or function that demands resourcing in its own right like, for example, social media metrics or SEO,” she says.

Whether that system could be provided by all agencies remains to be seen, but Kaintel is clear in terms of how media agencies should approach this topic, through embracing a “data and analytics culture”. 

Econometrics can help predicate media buying on the real outcome in precise business value terms rather than just judging output in terms of ratings and impressions — enabling agencies to bring their real value to life in boardroom discussions.