E-commerce category analyst: GMV, assortment and the sale calendar

Breaking GMV into parts you can act on

Gross merchandise value (GMV) is the headline number, but nobody can act on "GMV fell". The analyst's first habit is to split it:

GMV = sessions × conversion rate × average order value (AOV)

Each factor points to a different team. Sessions point to marketing and search placement; conversion points to pricing, availability, product content and delivery promise; AOV points to assortment mix, bundles and discounts.

Worked example: a month that grew for the wrong reasons

Hypothetical numbers for a small-appliances category:

GMV grew by ₹5,84,000. Attribute the change one factor at a time, in the order traffic, then conversion, then AOV:

• Traffic effect: (4,40,000 − 4,00,000) × 2.5% × ₹1,200 = +₹12,00,000
• Conversion effect: 4,40,000 × (2.2% − 2.5%) × ₹1,200 = −₹15,84,000
• AOV effect: 4,40,000 × 2.2% × (₹1,300 − ₹1,200) = +₹9,68,000

The three sum to +₹5,84,000. The headline says growth; the bridge says conversion dropped enough to wipe out all the extra traffic, and a higher basket covered for it. That is the story the category manager needs, and the next question writes itself: why did conversion fall?

One caution to state in any such bridge: the split depends on the order in which factors are changed. Keep the order fixed month to month so comparisons stay fair, and say which order you used.

Where conversion usually goes

When conversion falls, the analyst checks a short list before theorising:

1. 1Availability. Were the top-selling SKUs out of stock or showing long delivery dates for part of the month?
2. 2Price position. Did a key product's price rise relative to comparable listings?
3. 3Traffic mix. Did new traffic come from a low-intent source, such as a broad campaign, that converts poorly by nature?
4. 4Content and ratings. Did a top product's rating fall after a batch of complaints?
5. 5Pincode serviceability. Did delivery coverage shrink for some cities?

In the example, a quick SKU-level cut might show that two high-volume mixer grinders were out of stock for nine days, and that the traffic increase came mostly from a discount campaign. Both are testable, and both are fixable.

The sale calendar shapes the year

Indian e-commerce runs on events: the festive season around Navratri and Diwali, end-of-season sales, brand days and category-specific events. Before each one the analyst builds a plan with the category manager: expected demand by SKU, stock cover, deal pricing and the margin impact. During the event, the work is hourly dashboards and quick calls on which deals to extend. Afterwards, it is the post-mortem: did the event add incremental sales, or pull forward demand that would have come anyway and then dip?

A sample week outside the sale season

• Monday: weekly category review deck with the GMV bridge, top movers and in-stock rate.
• Tuesday: assortment gap analysis. Which price bands have strong search demand but few listings?
• Wednesday: return-rate investigation for one brand whose returns doubled; a size chart problem is suspected.
• Thursday: seller performance scorecard for the account management team.
• Friday: prepare the event forecast for next month's brand day.

Explained movements and forecast accuracy for events

The category's own targets, such as GMV, margin and in-stock rate, are the manager's. The analyst is judged on how quickly and correctly movements are explained, how accurate event forecasts turned out to be, and whether recommendations were adopted.

Category managers, sellers, pricing and warehousing

You report to a category head or a central business analytics lead. You work with category managers, pricing, marketing, supply chain and warehousing, seller or brand account managers, and product teams who own search and listings.

Confused with product analyst and pricing analyst

A product analyst studies how users move through the app or site: funnels, feature adoption and experiments. A category analyst studies what is sold. A pricing analyst goes deeper on price and discount decisions, which the category analyst touches but does not own.

Assortment, selection gaps and BAU-versus-sale signals

• "Own category performance reporting" — recurring reviews, bridges and dashboards.
• "Assortment planning, selection gaps" — demand versus listings analysis.
• "Event planning, BAU and sale" — heavy seasonal peaks.
• "Seller/brand metrics" — marketplace-side work with account managers.
• "Strong Excel, SQL a plus" — expect spreadsheet-heavy work; SQL may be limited.

Rebuild the weekly review, then own a sale event

Month one: rebuild the weekly review yourself so you know every number's source. Month two: add a GMV bridge and an in-stock view. Month three: lead the analysis for one sale event and write its post-mortem. Next roles include senior category analyst, category manager, business finance or pricing roles. Metric bridges like the GMV split above are a recurring exercise in the Data Analyst programme, built from SQL on synthetic order data.