Background

The “Advertising:Who Cares” movement seeks to improve the practice of advertising, reversing recent trends that have led to distrust, dissatisfaction, diminishing pride in creativity, and a decline in the appeal of the industry to new graduates.   

One dimension of improvement is with the measurement of advertising’s effect on sales (or other KPIs of fundamental importance to the health of the brand and business)

Deficiencies in these methods has blinded advertisers to the consequences of their decisions leading to abuses such as digital ad fraud.  Advertising is too often viewed with suspicion or even outright hostility in some C-suites in part because the evidence of contribution is either non-existent or couched in terms that are meaningless to CFO’s or CEO’s. Short-termism leads some brands to under-invest and miss market opportunities.

Measurement and accountability go hand in hand. For advertisers to earn the trust of their C suite colleagues they must be able to both measure their contribution and employ methods that help them reliably and consistently improve that contribution.  Stronger measurement methods will set a solid foundation from which marketers can better contribute to the development of business strategy.

We seek to set a high standard for the conduct of practitioners and help business decision makers recognize and reward the value created as a result.

For companies seeking better business performance, one sure path goes through excellence in the practice of measurement.

Setting a high standard for the measurement of advertising effect.

1. C Suite Goals
   1. Begin by aligning the measurement methods with the goals pursued by the C suite; sales, profit margins, new customers acquired or other metrics considered vital to brand and business health.
   1. Use optimization and simulation technology to connect models of these KPIs to prescriptive analytics.
   1. Measure both the opportunity and risk of plans being considered; use these measures to build C suite consensus around strategic choices.
2. Hold a Holistic view of cause and effect; develop models and source data accordingly.
   1. Recognize that once a goal is chosen, we need to explain what drives cause and effect for that goal measure.
   1. Include all forms of advertising and marketing communications efforts including PR, DM, Social (including consumer-generated media), Direct, Digital, Promotions, Sponsorships.  Owned and earned as well as paid media.
   1. Models should use audience measurement data of high quality.  Reference the Who Cares Measurement and Accountability manifesto covering this topic in depth.  (link)
   1. The measurement of business lift due to the creative used should be quantified. At the same time, it may be that creative executions used in market do not show sufficient variation to be measured in these kinds of models; in which case other techniques designed specifically for the evaluation of creative need to be used in parallel.
   1. Consider both short and long term effects of advertising; avoid over-focusing on the short term.  Recognize the asset value media spend and creative together generate.
   1. Enable balance points to be weighed through prescriptive analytics:
      1. Short vs long term; varying planning windows
      1. Offline vs online
      1. Brand vs promotion
      1. Channel mix design
   1. Certainly, advertising will contribute, but so will factors that are not under the control of the advertisers but that can affect outcomes. 
   1. All, or at least the most consequential, of these factors should be taken into account.  These could include, beyond advertising itself:
      1. The economy; local, national and international
      1. Category dynamics such as technology developments
      1. Competitor moves e.g., new product launches, pricing dynamics
      1. Distribution decisions; including the type and quality of sales outlets
      1. Operations decisions such as credit, manufacturing capacity and supply chain management
      1. Weather, where the category is sensitive to variation
      1. Other, as relevant to each brand
3. Build models to a high standard of accuracy
   1. To measure business effect, models should be built to a high standard of accuracy. 
   1. Aim to explain at least 90% of the variation in outcomes over the calibration period
   1. Test the model against hold-out periods and again when implemented, to predict outcomes over a time period relevant to decision making while maintaining high levels of accuracy
4. Models should be designed to be as actionable as possible
   1. grade proposed solutions by the ability and ease of translation into buying
   1. avoid overly-theoretical models that are impossible or at least difficult to translate into buying guidance
   1. if required, link a multi-channel model to channel-specific models to improve predictive accuracy, prescriptive relevance and buying support
5. Ensure a measurement solution can be tested independently of the developer.
   1. while testing should involve the developer, the results should be transparent to all
   1. build testing over time to prove the model can be trusted and prescriptions derived from the model achieve effects along the order of those predicted
6. Models should be judged on the basis of the leverage they bring to decision-making that effects business outcomes.  What lift in incremental business can the model help us create?  At what risk levels?  Think of the model as an asset, and its use as analogous to that of a lever being used to increase the force applied to an object.  As Archimedes famously said “give me a place to stand and a lever long enough and I will move the world”. Properly applied through simulation and optimization, models can be the lever that can move business dynamics in the right direction and at scale.

Moving from Advertising to Marketing

Our group endorses an ambitious role for the measurement of effect, evolving from a focus on advertising/marketing communications alone to the broader topic of the management of marketing. 

Measurements and then prescriptive analytics can support:

• pricing decisions
• distribution dynamics
• creative evolution
• product development
• customer experience design

Advanced solutions should allow the expression of a conceptual model of ad or marketing effectiveness to be translated into intermediate KPIs and then linked to the C suite outcome KPIs discussed above.  Avoid solutions that end with intermediate KPI action plans only.

As with advertising solutions, we recommend vendors and developers be graded for accuracy, actionability and support of sound decision-making that in turn delivers consistent positive outcomes.

Many companies have no systematic capture of their external environment and activities (economy, category, competition, distribution, operations) and marketing investments (resources including but beyond paid media). Marketing should partner with Financial Accounting standards to develop processes. Companies could then organize their internal systems to feed the modeling and reporting data in near real-time.

The industry could exert its wits and collective intelligence to build a standardized non-profit tracker of consumer brand perceptions available to all. Brand metrics can be tied to financial valuation, another outcome of marketing/accounting collaboration. An ambition we hold is to see the development of an ISO standard for ad effectiveness measurement, similar to ISO 10668 standard for brand valuation.

Marketing and advertising effectiveness should in 2020’s be measured by their contribution to the triple bottom line of ‘people’ and ‘planet’ as well as ‘profit’.

Stewart Pearson

David Beaton