You are the analytics lead for Amazon’s Devices business (think Echo/Ring/Fire TV-style consumer hardware). The product line sells across Amazon.com in the US, UK, and Germany, and through a handful of big-box retail partners. The line does $1.2B annual gross revenue, ships ~18M units/year, and has meaningful downstream impact on subscription attach (e.g., home monitoring) and Alexa engagement.
Over the last two quarters, the GM for the product line is asking for a single, consistent way to track revenue, margin, and market share that can be used in weekly business reviews and in quarterly planning. The immediate trigger: revenue is up +9% QoQ, but gross margin rate fell from 28% to 22%, and a competitor launched a similar device at a lower price. Leadership wants to know whether the margin decline is “healthy” (intentional price investment to gain share) or “unhealthy” (mix shift, returns, channel leakage, or cost regressions).
You have access to internal transaction and cost data, plus syndicated market data (e.g., Circana/IDC-style) with weekly category sales by country and channel.
| Source | What it contains | Grain |
|---|---|---|
orders | order_id, order_ts, sku, units, list_price, paid_price, coupon_amount, currency, country, channel (Amazon/retail), customer_id | order line |
returns | order_id, sku, return_ts, return_reason, refunded_amount, condition (new/open-box) | return line |
cogs | sku, week, standard_cost, landed_cost, freight_cost, duty_cost | sku-week |
promo_calendar | promo_id, start/end, country, channel, promo_type, funded_by (vendor/retailer) | promo |
market_syndicated | week, country, channel, category_units, category_revenue, competitor_units, competitor_revenue | week-country-channel |
fx_rates | date, currency, usd_rate | day-currency |
Assume you can query these tables in a warehouse and publish a dashboard used by execs weekly.