AI dashboard

AI dashboard · retention

Customer retention dashboard, built by AI on your Data Model

Whether new visitors come back and buyers order again, for an online store: monthly cohorts compared at the same age, repeat buyers against last year, and retention by channel and customer segment. Claude built it through OWOX MCP, and every number comes from a governed Data Mart, not from the AI's guess.

4 Data MartsBuilt by Claude via OWOX MCPRefreshed every Report runE-commerce data model

The dashboard

Hand this to your AI

Revenue is growing, and you still can't tell whether it comes from customers who stay or from new ones you keep paying to win.

Paste this into your AI assistant. It builds this dashboard on your own data with OWOX Data Marts and tells you whether new visitors come back, how many buyers order again, and whether the channel or the segment makes a difference.

The prompt

build me a customer retention dashboard from my visitors, customers and orders: monthly visitor cohorts and the share still visiting one to six months later, compared with last year at the same age, the share of buyers who ordered again and orders per customer against last year, and retention by acquisition channel and customer segment

What each view answers

Five questions, one Data Model

Every view answers one question an owner asks about growth that has to last: do the people we win come back? Between them they read four Data Marts of the e-commerce Data Model, and each view names the ones it uses, so you can open the definition behind any number.

The E-commerce data model: 12 Data Marts and how they join
E-commerce data model · 12 Data Marts · 15 relationships · open it →

Summary

Do new visitors and buyers come back?

Opens with what the numbers mean: how many of this year's new visitors return against last year's at the same age, how many buyers order again, whether the channel explains it, and what a second order from one-time buyers is worth. Then new visitors, returning visitor rate, registration rate, orders per customer and repeat buyers against the same dates last year, cohort by cohort.

VisitorsCustomersOrdersReturning visitorsReturning customer rate

Cohorts

How long do new visitors keep coming back?

Each month's new visitors, and the share still visiting one to six months later, as a table you read left to right. Then 2026 cohorts against the same months of 2025, at the same age, so time alone never reads as a decline.

Visitors

Buyers

Who orders again?

Buyers by the number of orders they placed this year against last, orders per customer and average order value, then new customers and orders per customer month by month.

OrdersCustomersAverage order valueNumber of orders

Channels

Does the channel decide who comes back?

Every acquisition source and medium side by side: its share of new visitors, the share that returned and the share that registered.

VisitorsTraffic Sources

Segments

Do our best customers reorder more?

Platinum, Gold and Standard buyers: how many ordered, how many orders each placed and how many ordered again, against last year.

CustomersOrders

Metrics

The metrics on this dashboard

Each one is defined once in the Data Model and read the same way in every view. Follow a chip for what the metric means and how to calculate it.

Returning visitorsReturning customer rateAverage order valueNumber of ordersRegistration rateOrders per customerCohort return share

Connected, not a snapshot

An AI dashboard is only as true as what it reads

Most AI dashboards start from an exported file. That works until the second export.

On a CSV

The AI reads time as a trend, every time

One export says 34.2% of last year's new visitors came back, the next says 25.3% of this year's, and the deck reports a collapse. Half of it is only that this year's visitors have had less time to return, and nothing in the file says so. It is out of date the moment it is saved.

Two exports of new visitors giving a return rate of 34.2% and 25.3%

How much of the drop is just time?

On your Data Model

Every visitor carries their cohort

Visitors keeps each visitor's first and last visit, cohort month and acquisition channel, and links the customer they became, so cohorts are compared at the same age and a buyer's orders tie back to the visit that won them.

The e-commerce Data Model: Visitors joined to Customers, Orders and Traffic Sources

One definition, every view.

Connecting the AI straight to the warehouse is not enough either. A warehouse rarely holds the business context, so the AI still guesses what the numbers mean. The Data Model is where that context lives.

How it is built

From your warehouse to a dashboard in five steps

Your analyst does step two once. After that, anyone with access asks their AI for the dashboard they need.

  1. 01Your warehouseBigQuerySnowflakeDatabricksRedshiftAthenaYour data stays where it is. OWOX runs on it and keeps no copy.
  2. 02Data ModelData MartsjoinsmetricsYour analyst defines once what a visitor, a cohort and a returning customer mean.
  3. 03OWOX MCP and APIgovernedpermissionsThe only door the AI gets: the Data Marts it may read, nothing raw.
  4. 04Claude or ChatGPTyour own AIAsk for the dashboard in plain words. It builds it on the Data Model.
  5. 05AI dashboardReportevery runKept as a Report, refreshed every time it runs, like Sheets or Excel.

Build yours

Build this on your own data

Connect your warehouse, describe your business once, and let your AI build the dashboards your team asks for.