An e-commerce data architecture

Every data source an online business already has, unified in one lakehouse it owns, powering the outputs the team actually uses. This is a worked example, not a template. Every setup is tailored to the business.

Online shop, marketplaces and paid channels.

Data flows in from 8 sources
Sources flowing into the lakehouse
Insights flow out
Lakehouse flowing into outputs
DATEV

Accounting & finance

Your Business plans & goals

Inventory

Market & competitor data

Sales Channels

· incl. marketplaces

Shopify
Amazon
eBay

+ more

CRM

Marketing Performance

Google Ads
Meta
Instagram
LinkedIn

+ more

GA4 (Google Analytics 4)

Web tracking

Data Lakehouse

Your central source of truth

Raw data transformed into metrics
Raw data transformed into metrics

Investor & fundraising materials

Board decks, data rooms, metrics

Dashboard chart types
Dashboard chart types

Dashboards

Live KPIs across the business

Custom reports, forecasts & more

Ad-hoc analysis, projections

+ 6 more sources: web tracking, CRM, inventory, market data, your business plans, accounting

+ 6 more sources: web tracking, CRM, inventory, market data, your business plans, accounting

How to read this

Left: the sources. Middle: the lakehouse (one place, one definition per number, owned by you). Right: what the business gets out of it. The sources change from client to client; the shape rarely does.

See the hospitality example →

Curious what yours would look like?

I usually sketch the target architecture in the first two weeks, before anything gets built.

Curious what yours would look like?

I usually sketch the target architecture in the first two weeks, before anything gets built.

Curious what yours would look like?

I usually sketch the target architecture in the first two weeks, before anything gets built.