Skip to content
E-commerce analytics

Margin and sales analytics for marketplace sellers

Analytics that show the real margin on every SKU, including commissions, logistics, returns and advertising. Demand forecasting, price optimisation and unit economics, on ClickHouse.

Discuss a project Common questions
Stack and integrations
ClickHouse PostgreSQL FastAPI Marketplace APIs Grafana Metabase
In short
What we build
Real margin analytics across marketplaces: unit economics per SKU, demand forecasting, advertising.
Who for
Sellers at a scale where the platform's reports no longer reflect actual profit.
Stack
ClickHouse with FastAPI and the marketplace APIs, dashboards in Grafana or Metabase, ML forecasts.
Timeline
MVP in five to eight weeks, a full BI system in two to three months.
A good fit when
  • → The platform's own reports do not show your real margin
  • → Commissions, penalties and logistics quietly eat the profit
  • → You need demand forecasting and purchase planning
  • → Advertising on the platforms is a black box on return
What you get
  • → A full export of platform data into ClickHouse
  • → Unit economics per SKU, including commissions, logistics and returns
  • → Dashboards for profit, turnover and margin by channel and cluster
  • → Demand forecasting and purchasing recommendations
  • → Advertising analytics: cost share, effective bids, return on spend
  • → Alerts on anomalies — falling sales, rising returns
Process
  1. 01 An audit of the current reports and data sources
  2. 02 An MVP with a ClickHouse dashboard of the key metrics
  3. 03 Expansion: forecasts, advertising, alerts, accounting integration
Why not an off-the-shelf product

Custom development earns its place where your processes matter.

Ready-made SaaS is good for standard scenarios. But once the business logic depends on specific roles, documents, integrations, security or data, the cost of the workarounds quickly exceeds the cost of a proper architecture.

We start with discovery, separating what genuinely has to be built from what is cheaper to cover with an existing service. That is why the project ends up smaller, clearer and easier to run.

Related cases

Similar problems from the portfolio.

/cases/marketplace-manager
FAQ

Common questions

How is this different from the SaaS analytics tools?
Those give external analytics and estimates about other sellers. We build an internal system on your data: exact cost of goods, integration with your accounting, and forecasting models built for your assortment.

What does it cost?
An MVP with ClickHouse and profit dashboards per SKU is the entry point. A full system with ML forecasts and advertising analytics costs roughly twice that.

How long does it take?
MVP in five to eight weeks. A full system with ML, alerts and accounting integration in two to three months.

Is there a real example?
A marketplace manager for a cosmetics brand: ClickHouse analytics and a 12-point margin gain through automated repricing. The write-up is on the cases page.

How accurate is demand forecasting?
For stable SKUs, typically 10–20% MAPE over a four-week horizon. For seasonal and promotional items, hybrid models with manual correction.

How are returns and damaged goods handled?
Returns arrive from the platform reports with a 30–60 day lag and retroactively recalculate margin. The dashboard shows two versions: operational and adjusted.

Can advertising return be calculated separately?
Yes. A separate module for on-platform advertising: cost share per SKU, effective bids, return on spend per campaign, pulled from the platforms' advertising accounts.

How do we measure the effect of the system itself?
Margin gain (usually 5–15 points, from correct prices and assortment structure), reduction in dead stock, and purchasing accuracy.

Written up in more detail

Next step

Let us go through your problem.

We will show you a possible architecture, the risks, the order of the budget and what an MVP could prove.

Discuss a project