Measuring What Matters
The measurement shift that unlocked 15% sales growth for Estee Lauder
+26%
retailer revenue
+15%
total sales growth
+132%
share of voice
About Company
Estée Lauder Companies is a leading player in the UK prestige beauty market, manufacturing and marketing skincare, makeup, fragrance and hair care products across a portfolio of more than 25 luxury and prestige brands.
Company size
Enterprise
Industry
Beauty
Brainlabs featured services
Data Strategy
See what Brainlabs can do for your team
Speak to a BrainlabberThe challenge
ELC was losing market share in a highly competitive market, and their media setup was part of the problem.
Their strategy was Search-dominated, efficient at volume but not at building brand value. The fix required a significant pivot toward upper-funnel investment, but their measurement framework made that pivot impossible to justify. Built around last-click thinking and short-term digital metrics, it couldn't unify media's impact across ELC's complex retail and direct sales network. Upper-funnel activity was effectively invisible, leaving it critically underfunded and unable to drive sustained revenue.
Key friction points:
- A last-click measurement framework that couldn't capture upper-funnel contribution
- Search-heavy media mix building volume but eroding brand equity
- No way to connect brand activity to total business sales across a complex retail network
- A necessary strategic pivot that couldn't be made without the data to back it
Our approach
1. Building 20 bespoke Bayesian econometric models
Our data science team built 20 bespoke Bayesian econometric models across five portfolio brands, each separating baseline from incremental sales across every route to purchase. Using over 75 custom priors, the models holistically assessed every input from media spend to pricing and competitor activity. We validated the models by feeding in past in-platform and geo-based incrementality tests, achieving a MAPE score of 5% and an R-Squared score of 0.95.
2. Proving the value of full-funnel investment
The modelling gave us what last-click attribution never could: statistical proof of upper-funnel media's contribution to business outcomes. This allowed us to make the case for a strategic shift in how ELC allocated budget, moving investment toward brand-building activity with the data to justify it.
3. Turning the models into an always-on optimisation tool
Rather than deliver a one-time static report, we built a tool that transforms the analysis into continuous action. It incorporates every response curve from the modelling, prescribing the most efficient spend across every channel and stage of the funnel and outputting an optimised media plan calibrated to ELC's sales and revenue targets. To keep the tool accurate, the model undergoes a six-monthly Bayesian refresh, ensuring every strategic decision is based on the most current data available.
Our data science team built 20 bespoke Bayesian econometric models across five portfolio brands, each separating baseline from incremental sales across every route to purchase. Using over 75 custom priors, the models holistically assessed every input from media spend to pricing and competitor activity. We validated the models by feeding in past in-platform and geo-based incrementality tests, achieving a MAPE score of 5% and an R-Squared score of 0.95.
2. Proving the value of full-funnel investment
The modelling gave us what last-click attribution never could: statistical proof of upper-funnel media's contribution to business outcomes. This allowed us to make the case for a strategic shift in how ELC allocated budget, moving investment toward brand-building activity with the data to justify it.
3. Turning the models into an always-on optimisation tool
Rather than deliver a one-time static report, we built a tool that transforms the analysis into continuous action. It incorporates every response curve from the modelling, prescribing the most efficient spend across every channel and stage of the funnel and outputting an optimised media plan calibrated to ELC's sales and revenue targets. To keep the tool accurate, the model undergoes a six-monthly Bayesian refresh, ensuring every strategic decision is based on the most current data available.
The results
The modelling validated a 91% strategic shift toward upper-funnel media investment across the portfolio, with overall media contributing 15% toward total sales.
Since working with Brainlabs:
Since working with Brainlabs:
- 26% increase in retailer revenue, driven by a validated full-funnel media strategy
- 15% total sales growth, with media directly contributing to overall business performance
- 132% increase in share of voice, reflecting the impact of reinvesting in brand-building activity
+26%
retailer revenue
+15%
total sales growth
+132%
share of voice
