- Discovery partnered with Incubeta to use advanced Marketing Mix Modelling (MMM) to measure media effectiveness and optimise marketing investment across its business units.
- The modelling combined two years of media spend, sales data, seasonality and external factors to assess the true contribution of each channel.
- MMM revealed diminishing returns in performance channels and opportunities to improve efficiency by reallocating spend.
- Discovery uncovered the underreported impact of TV and radio, including their indirect and delayed effects on conversions and their ability to boost paid search performance.
- MMM has given Discovery an evidence-based framework to justify marketing spend, improve media allocation and demonstrate marketing’s contribution to business growth. *
Discovery was investing heavily across multiple business units and platforms, but couldn’t clearly see what was actually moving the needle, or where money was being wasted.
Partnering with Incubeta, they used advanced Marketing Mix Modelling (MMM) to turn a complex, multi‑channel media mix into a single, evidence‑based roadmap for where to cut, where to double‑down, and how to defend every rand of marketing spend.
“Before MMM, our biggest challenge was understanding true incremental impact across channels. We had strong reporting on delivery and platform-level performance but struggled to disentangle correlation from causation particularly in a complex, multi-channel environment,” explains Karl Vermeulen, business insights head at Discovery.
Vermeulen says the company’s channels such as paid search and digital were relatively easy to attribute, but upper funnel channels (TV, radio, and OOH) were underrepresented in performance narratives.
“This created a structural bias toward short-term, last-click efficient channels, potentially at the expense of long-term growth. We also lacked a unified framework to evaluate diminishing returns, saturation, and cross channel interactions at a portfolio level,” he adds.
Old methodology inhibited a clear view
Like most enterprise businesses, Discovery was allocating their budget based on information from a combination of historical performance, platform reported metrics, and incremental testing where possible.
“We leaned heavily on digital attribution models and cost per acquisition benchmarks, supplemented by campaign level analyses and market intuition. While this worked tactically, it didn’t provide a holistic view of how channels contributed across the full funnel or over time,” Vermeulen explains.
“As a result, optimisation tended to favour channels with clearer attribution signals rather than those with the strongest total business impact.”
Three reasons MMM offered the solution
Vermeulen explains that the Discovery teams had already explored multi touch attribution (MTA), but says it was limited by data fragmentation, privacy constraints, and an over reliance on user level tracking.
He saw three distinct advantages offered by MMM: holistic measurement across all channels; causal, econometric grounding rather than platform reported attribution; and the ability to incorporate external factors like seasonality, pricing, and macroeconomic conditions.
“MMM allowed us to step back and understand how the system behaves, which was ultimately more aligned to how our business operates,” he shares.
Making the data do the heavy lifting
Incubeta’s specialist Marketing Intelligence team, combined at least two years of media investment data (digital and above-the-line) with sales performance, seasonality and relevant external factors.
This data was structured and analysed using advanced modelling techniques, with outputs visualised through Fovio.
“The result is a science-based framework that shows which channels are genuinely driving sales, where there is over- or under-investment, and how reallocating spend can unlock efficiencies,” explains Calvin van Rensburg, client solutions partner at Incubeta.
Some surprises along the way
Vermeulen reveals that the work with Incubeta has helped his team uncover some valuable insights, some of which came as a surprise.
“We now understand the extent of diminishing returns in performance channels and that beyond a certain threshold, incremental efficiency drops off significantly. We had also underestimated the contribution of upper funnel media, and channels like TV and radio showed strong indirect and lagged effects on conversions. Finally, we now have clear evidence that certain channel combinations delivered disproportionately higher returns when run together,” he says.
Interestingly, Vermeulen says MMM has fundamentally reframed how the marketing teams view TV.
“Rather than evaluating it purely on direct response, we now understand its role as a demand generator and amplifier. The model showed that TV drives both immediate and lagged effects, often increasing the efficiency of lower funnel channels like search,” he says.
Insights enable a shift in executive conversations
Vermeulen says the solution has also shifted how the team communicates with decision makers across the organisation.
“By quantifying incremental impact and demonstrating diminishing returns, we can now justify not just how much we spend, but where and why. MMM has given us a more credible and defensible narrative. Instead of relying on fragmented metrics, we can now present a unified view of how marketing investment translates into business outcomes,” he says.
“Increasing market volatility is forcing a structural shift towards greater accountability, agility, and intentional spending. But one of the biggest challenges still facing CMOs is proving their strategies. So many marketing leaders can see platform‑level metrics, but they still can’t answer ‘what’s really working’ across the full funnel or over time,” says Jaco Lintvelt, MD Incubeta SA.
“Advanced MMM elevates performance and strategic planning, including the all-important financial decisions, helping CMOs defend strategies in a language that the board and shareholders understand and appreciate.”
* Summary created by AI
