I've spent more than five years acquiring customers through paid media, mostly for SaaS, e-commerce and services. In most companies that work is split in two: one person runs the campaigns, and someone else builds the models that say where the money should go. I do both. I run the campaigns, and I build the models that decide where the next unit of budget goes.
At Blocks, a product-led SaaS with a marketplace and programmatic SEO, I led growth for about five years. I managed paid media against gross margin, not against CPA or ROAS in isolation, and used logistic regression to find out which channels actually produced recurring revenue. Before that I was a paid-media analyst at the Atacama and BG9 agencies, and spent about three years with Axiom, a technical training company.
Today I run ABADE, my marketing analytics and paid-media consultancy. I build my own campaign-optimisation and data tools on LLM APIs in Google Cloud. My background is a B.Sc. in Information Systems from UFS, where I also did machine-learning research.
Looking for Marketing Data Scientist, Growth/CRM or Acquisition roles where data and media spend answer directly to revenue.
Everything starts from data that can be trusted, moves through four stages that each answer one question about a marketing dollar, and ends in frameworks that turn the method into routine work. Click a block to open its main case.
| Block | Projects |
|---|---|
| Dados | growth-analytics-warehouse · tracking-attribution-lab · dados-sinteticos-ecommerce-tecidos · dados-sinteticos-streaming-cinelume |
| Validate | marketing-mix-modeling: perfect channel ranking against a known truth · ab-testing-toolkit: CUPED cuts variance 75% |
| Scale | unit-economics-olist: paid search eats 71% of first-order margin · lead-scoring-api: 1.54x lift once two leaks are removed |
| Retention | clv-cohort-prediction: top 10% of customers hold 51% of future value · churn-cost-sensitive: £10,285 recovered by the decision rule alone |
| Results | paid-media-budget-optimizer: headroom, not efficiency alone, decides where budget grows · marketing-analytics-portfolio: from campaign report to decision |
| Frameworks | Diagnosis, RevOps funnel, paid-media audit, creative matrix, A/B cadence, unit economics, UTM governance, monthly media report, RFM, growth loops |
Every case opens with the decision it supports and ends with what it cannot show. A hidden assumption is the most expensive way to get an investment decision wrong.
Research and products, outside marketing
- carbon: exon/intron classification in human DNA with a Bi-LSTM (99.80% test accuracy). The paper was accepted at an international bioinformatics conference.
- visppy-cv: computer vision for physical spaces. It was among the 47 projects approved in the preliminary Phase 2 result of Centelha Sergipe III.
- mandacaru: intelligent document processing built at STI/UFS.
- echo-womens-health-research-analytics: survey-based evaluation of a remote women's health training programme.
Acquisition: Google Ads · Meta Ads · GA4 · conversion funnels · attribution and channel evaluation
Revenue modelling: unit economics (CAC, LTV, payback, margin) · MMM · CLV · churn · experimentation
Stack: SQL · Python · DuckDB · dbt · scikit-learn · LightGBM · PyMC-Marketing · FastAPI · Docker
Automation: LLMs via API on Google Cloud for campaign and data tooling
Sou cientista de dados com mais de 5 anos em aquisição de clientes via mídia paga, principalmente em SaaS, e-commerce e serviços. Meu diferencial é juntar duas áreas que costumam ficar separadas: eu opero as campanhas e também construo os modelos que dizem onde investir.
Na Blocks, um SaaS PLG com marketplace e SEO programático, liderei growth por cerca de 5 anos, gerindo mídia paga com foco em margem bruta e usando regressão logística para avaliar quais canais de fato geravam receita recorrente. Hoje toco a ABADE, minha consultoria de marketing analytics e mídia paga.
Busco posições de Marketing Data Scientist, Growth/CRM ou Aquisição em que dados e investimento em mídia estejam diretamente ligados a receita.
ABADE · Strategy · Data · Growth