Why I built explōrātiō
In 18 years in advertising, I worked at the four houses that produce the most communication in the world — Dentsu, WPP, Omnicom, Publicis. I built data hubs from scratch for some of the brands below. I saw what fits inside each budget and each decision.
At every one of those houses, I saw the same scene repeat. A meeting called at the last minute to decide where to cut budget. A director asking for "the number that justifies it" with no one knowing where that number came from. A creative scaled because "this one works" — and discontinued because "this one did not". Price moved on the sales team's gut feel. Media reallocated on the planner's intuition.
I also saw the opposite. Desitin Roxo grew 29% after a three-week elasticity analysis. adidas cleared winter inventory in July because an Opportunity Map anticipated the curve. The adidas e-commerce revenue during the 2014 World Cup was forecast with 99% accuracy by a regression model.
The difference between the two sides, in my reading, was never access to data. A company with a large agency has data. What separates a good decision from a bad one is method: named model, declared formula, visible variables. A real possibility of auditing the calculation.
explōrātiō exists to serve that difference. I take the analytical products that repeatedly deliver clarity at large brands — MMM, elasticity, basket analysis, attribution models, creative analysis, demand anticipation — and deliver them ready for companies, agencies, and marketing teams of any size. No black box. No 80-page PDF. No promise of certainty.
We sell clarity. We sell the path. And we sell to whoever arrives in three ways: the company that does not yet have a data function, the agency that needs analytical capacity on a client project, and the marketing team with structured BI that needs the study that does not fit the internal queue. In all three, the product is the same. What changes is where we fit in your workflow.
Impact by brand and method.
From spreadsheet to corporate hub in 12 months.
At Performics, I led the build of Ambev's data hub: AWS datalake architecture, metrics and dimensions mapping, ETL process, and dashboards in Looker Studio, Power BI, and Datorama. In 90 days, taxonomic adherence went from 27% to 78%. After 12 months, it reached 98% — above the operation's target. The model was replicated for 8 other clients, totaling 12 datalakes and more than 40 million rows operated under LGPD standards.
+29% sales with elasticity, without compromising margin.
At Global Shopper, we led a price and demand elasticity study for the RDO line. The analysis indicated where and how much to move — and where not to move. Result: 29% increase in Desitin Roxo sales with no negative impact on margin or competitiveness.
Winter inventory cleared in July, with 60-day anticipation.
At iProspect, I applied the Opportunity Map (proprietary algorithm on Google Trends and search data) to anticipate adidas winter product demand at least 60 days ahead. E-commerce cleared inventory in July. For the 2014 World Cup, a multivariate regression model forecast total e-commerce revenue during the event with 99% accuracy.
Air conditioning in the Midwest: 2.9x above expectation.
Also at iProspect, the Opportunity Map anticipated air-conditioning search seasonality in Brazil's Midwest, allowing media reallocation before the peak. Sales in the period were 2.9 times higher than expected. Attribution models applied to Brastemp and Consul e-commerce delivered +38% ROAS over the previous period.
Stories from the founder's work at agencies in the Dentsu, WPP, Omnicom, and Publicis groups. They do not represent explōrātiō clients.
Want the same method on your data?
A 30-minute conversation on WhatsApp, no obligation. We look at a sample of your data and say plainly: there is an explōrātiō product that fits, or it is not the right moment yet. If it fits, we align timeline, format (direct, with your agency, or alongside your BI team), and investment in the same call.