By Claire Fenner, national CEO, Atomic 212°
With the advertising market the way that it is, it’s very tempting for marketers and businesses to fall into a cost-cutting mindset. Budgets are tightened and every dollar is scrutinised for maximum efficiency. While reducing wasteful spending is important, simply slashing budgets without careful optimisation risks leaving significant revenue on the table.
In this environment, data-driven approaches like media mix modelling – which provide valuable insights into past campaign performance – are all too often misapplied to solely justify reductions rather than to maximise effectiveness.
Media mix modelling analyses vast amounts of past campaign data and contextual inputs to understand what combination of channels and tactics drive the best results, as well as the impact of other internal and external factors. By looking at the contribution of various factors, including media investment, price, economic variables, and competitive activity, the modelling can predict future outcomes based on the existing mix, as well as different scenarios or combinations of these inputs. The risk occurs when brands simply cut the channels with the lowest predicted performance to maximise efficiency.
By focusing only on cutting underperforming channels rather than testing ways to improve them, we’ve seen clients leave up to 10% in missed growth opportunities on the table. Beyond this, there is a huge opportunity growth opportunity from increasing investment in the strongest performing channels.
In tough markets, marketers need to work smarter, not simply spend less. A more holistic approach to media mix modelling is required to fully leverage its capability to deliver incremental growth with refined spending. This means avoiding the trap of sub-optimal cost-cutting and instead using testing and optimisation to ensure no stone is left unturned.
To avoid leaving growth on the table, marketers should take a test-and-learn approach to media mix modelling. With smaller budgets, the focus should be on refining the channel mix toward what works best historically, then maintaining a testing budget to gradually expand the channel mix while mitigating risk, to prove to the business the opportunity to scale the investment.
More generous budgets give the opportunity to enhance lower performing channels through experimentation. Testing new tactics within a lower-performing channel can improve overall performance and is a strategy worth considering before removing a channel entirely, where budget permits. There are so many factors that will impact a channel’s performance, including reach, frequency, dayparts, formats, durations, audience targeting, and creative. It’s not unheard of for us to conduct A/B tests in lower performing channels and find unexpected wins that boost results for clients above the baseline prediction.
While these suggestions will enhance the application of modelling to inform growth rather than cost cutting, all too often, lack of buy-in from other business stakeholders prevents marketers from fully leveraging this opportunity in their media mix modelling. People outside marketing may question the validity of modelling outputs or the methodology and fail to see it as a growth tool. Additionally, years of budget cuts can ingrain a “cost-cutting mental model” that is difficult to shift.
A critical first step in establishing buy-in from key stakeholders is engaging key decision-makers – including the CFO or representatives from finance – early in the model-building process to establish confidence and take the opportunity to address any concerns. Getting others onboard requires both simplifying complex methodologies and demonstrating accuracy through “hold out” testing. With buy-in addressed, marketers are freed to leverage modelling not just as a cost savings exercise but a way to refine spending and test optimisations that can drive significant incremental gains for the business.
It’s also critical for marketers to avoid falling into the trap of only considering the short-term role of media, focussing purely on the CPA without considering longer-term brand objectives. Relying solely on these predictions risks falling into the “garbage in, garbage out” trap. The modelling can only work with the historical data provided, so unique contextual factors or new innovations will likely not be accounted for correctly. Hypotheses must be formed and tested to continuously optimise for growth.
With rigorous testing frameworks and engagement across departments, marketers can absolutely shift their organisation’s mindset around marketing spend and the value of media mix modelling to unlock real short-term and long-term growth, especially in a challenging market.
See also: Atomic 212° rolls out O’Brien’s first campaign aimed at younger demo
–
Top image: Claire Fenner