The final panel on Day Two of Advertising Research Foundation’s (ARF) “Attribution & Analytics Accelerator” conference Tuesday echoed insights reflected during a series of outstanding presentations throughout the day, as well as tackling some of the issues raised during the first day’s sessions on various marketing and attribution modeling issues.
Thanks to modeling, Janelle Bowman, senior director of advanced data & integrated analytics at Kellogg Co., better understands the value of loyalty metrics and the impact of targeting across different brand category sub-groups. However, the difficulty of obtaining a complete 360-degree views of target audience-based (persons!) sales across all outlets remains a problem -- although represents a potential opportunity for “data clean rooms.”
Data clean rooms are environments where walled gardens like Google, Facebook and Amazon share aggregated -- as opposed to individual consumer-level -- data with advertisers, while exerting strict controls.
Thibaut Munier, co-founder and co-CEO at Numberly, stated clean rooms have a long way to go. He noted the data is awfully expensive, the data quality is not harmonized, and brands should put pressure on retailers for richer data as they compete directly with each other.
“Can attribution models really measure long-terms effects?,” asked Mick Hedberg, principal at Foresight Associates, adding, “And how critical is this with the continuing focus on the lower funnel and short-term sales response?” He also underlined the importance of data-cleaning for any modeling exercise.
Pernod Ricard proprietary marketing mix model (MMM), Matrix, has connected long-term and short-term aspects into a single model to help balance ad investments by recognizing brand equity and advertising’s contribution as part of ongoing advertising touch points. Maxime Lamagat, marketing effectiveness lead expert shared that for its category, “it is better not to invest than under-invest.”
Jim Spaeth, partner at Sequent Partners, which co-hosted the conference with the ARF, reminded attendees of the fundamental importance of understanding “net present value” (NPV) of long-term marketing effects.
The pull between short-term lower funnel-marketing and long-term brand-building was stressed by Tina Wilson, executive vice president marketing analytics at Nielsen, who co-presented Michele Thomas, senior manager of communication planning insights at Nestlé Purina PetCare.
They have used MMM to underline the balance required in any brand marketing strategy and the importance of “awareness” and “consideration” -- both longer term measures -- which impact marketing efficiency. Are CMO’s ready to embrace the duality of marketing objectives?
Thomas cited the importance and opportunities for creative messaging that can really resonate in the pet care category.
On cue during the panel, Leslie Wood, Chief Research Officer of NCSolutions restated her well-supported conclusion that the key to outcomes remains creative, when crafted to impact distinct brand targets.
Woods’ presentation with NCSolutions’ Director of Research & Development Andrew Bernier and Kellogg’s Bowman, was the highlight of the day. It was based on four years work with 52 brands.
Critically, it referenced how loyalty metrics affect brand growth and the importance of three long-term keys to understanding success: net annual brand churn; repertoire of brands consumed; and the depth of repeat-buying within a category.
Depth of repeat is central to long-term brand health although difficult to achieve in high-churn categories.
Churn and depth of repeat are heavily influence by the category. Consequently, it is difficult to maintain the long-term effects of advertising unless you understand where the brand sits in the category and the nature of the category.
Three key takeaways:
Focus on retention and depth of repeat to build brands.
Ad responsiveness tends to be most closely aligned to churn.
And a surprising observation, targeting appears unrelated to lift.
Put this all together, NCSolutions believes advertisers do have the opportunity to optimize campaigns.
Clearly modeling is helping us recognize the contributions of advertising and its effects despite its shortcomings. Stay tuned for my commentary on Day 3.