Case Study · Independent · Own-Data Analytics

Affiliate eCommerce Operations Dashboard

Sole operator of an affiliate eCommerce business — and the analyst who instrumented it. One live Looker Studio dashboard became the operation's single source of truth, and the numbers it surfaced were not the ones anyone hoped for. That's exactly why it's here.

The dashboard's biggest finding was a ~50% cancellation rate. Honest measurement first — then decisions.
Looker StudioSQL on WooCommerceeCommerce Ops Sole OperatorLive Dashboard
case scorecard — real, unpolished numberslive
156
orders in the retained export
~50%
cancellation rate the dashboard surfaced
1
operator running the entire workflow
<1min
to read the operation's daily status

01The Setup

An affiliate marketing operation selling physical products in a cash-on-delivery market: I marketed the products, took the orders, and followed them through delivery — as the only operator on the business. No team, no handoffs, and initially, no instrumentation: order status lived in scattered messages and platform screens.

You can't improve an operation you read from screenshots and memory. Step one was measurement.

02What I Built

Live Looker Studio Dashboard

The operation's single source of truth: order lifecycle by status, delivery performance, and recorded commissions — checked before every daily decision, and readable in under a minute by a non-analyst.

SQL Directly on the Store Database

Queries written directly against the WooCommerce database — the same discipline used on production marketplace tables, applied to a small business: define the metrics, own the data model, then visualize.

End-to-End Ownership

Data model, metric definitions, and the dashboard itself — one person owned the pipeline from raw order rows to the decision screen.

Built for an Operator, Not an Analyst

The design target was reading speed: today's orders, their statuses, and what needs chasing — before the first coffee, not after an export.

03What the Numbers Said

The dashboard's headline discovery: of the 156 orders in the retained export, exactly half cancelled (78 of 156) before completing delivery. In a cash-on-delivery affiliate model, that isn't a footnote — it's the difference between a business and an expensive hobby. Instead of burying the number, the dashboard put it front and center: cancellation became a tracked, decomposed, daily-visible metric.

Honest instrumentation beats optimistic reporting: the operation's worst number got the most screen space — because it was the one that decided everything else.
Delivery performance became measurable: statuses tracked through the courier chain showed where orders stalled, and commissions were recorded against completed deliveries only.
Every daily decision started from the same screen: which orders to chase, which products to keep promoting, and whether the funnel deserved more spend.

04Why This Case Is in the Portfolio

This is the only portfolio piece running on my own operational data, live, with real — and imperfect — numbers. It demonstrates the full loop most analyst portfolios only describe: run an operation, instrument it honestly, read the uncomfortable result, and let the data drive what happens next.

Open the live dashboard ↗
SCREENSHOT SLOT
Looker Studio overview page
(or embed via iframe)assets/affiliate-dashboard-overview.png

05Skills Demonstrated

Looker StudioSQL on production DB Metric definitioneCommerce operations Dashboard UX for operatorsHonest analytics