A marketing analytics lead has an ad budget and a room full of managers who read the business through sales volume. This dashboard makes the case for spending it elsewhere, and lets the room test the recommendation live.
Built in Tableau. What it demonstrates: turning analysis into a decision a room can act on: the bridge between the technical work and the business language. Live on Tableau Public · Interactive dashboard
Trap. Budget follows sales volume. Phones sells $330K, Chairs $328K, Tables is #4. So the money goes to the biggest sellers, including one that loses $17,725.
Insight. The discount ceiling. No discount level above 20% makes money: at or below 20% the business earns +$422K; above 20% it loses −$135K. The states confirm it independently: average discount and margin correlate at r = −0.98.
Machine. The ceiling as a testable number. The analysis verifies it two ways: the discount ladder (+$422K at or below 20%, −$135K above) and the state map (r = −0.98). A Discount Ceiling slider prototyped after publication lets a reader perform the cap instead of taking the number on trust. The live workbook ships four interactions: filter, highlight, Profit Target, highlighter.
Situation. As marketing analytics lead, defend an ad budget to management. Management reads the business through sales volume and wants the sub-category picture across the USA. Dataset: 9,994 order lines, 17 sub-categories, 49 states, 2016–2019.
Task. Build an interactive dashboard (linked charts, ≥2 actions) plus a 200–500 word management report with findings and recommendations.
Action. Four linked views (sales bar, sales-vs-profit scatter, discount ladder, state map) driven by four interactions: a filter action, a hover highlight, a Profit Target parameter with reference line, and a type-ahead highlighter. A Discount Ceiling what-if slider was prototyped after publication (see methodology). A five-point Tableau Story walks the argument: what sells → selling ≠ earning → the discount cause → the map proof → Fund / Fix / Stop. Colour rule throughout: blue = profit, orange = loss, readable for red-green colour blindness.
Result. The recommendation: fund Paper (43%), Copiers (37%), Accessories (25%), Labels (44%), Envelopes (42%); fix pricing first on Binders, Storage, Chairs; stop funding Tables, Bookcases, Machines, Supplies. Capping discounts at 20% lifts profit 47% on its own.
dashboard/: the five-point story as a click-through walkthrough (one idea per view, like the Tableau Story).docs/analysis.md: the five story points, one idea each, with the verified numbers.docs/business-report.md: the 407-word management report.docs/methodology.md: parameters, calculated fields, discovery notes, QA log.data/: summary tables (aggregates only; row-level data lives in the workbook extract).
Sample data, disclosed as such. The value is the mechanism, not the dataset.
- Live workbook (Tableau Public): https://public.tableau.com/app/profile/lorisca.cessia.tuuk/viz/A1_EXTRACT_v3/Story1