BlueShop AI Engineering Lab

BlueShop BI

One shared BI plane for five business questions. Dashboards connect customer activity, commercial performance and operations without treating route counts as the story.

PerformanceGMV, orders, margin and demand trends in one shared view
CustomersValue, retention, geography and RFM segments
ConversionCheckout, promotions, affinity and recommendation CTR
OperationsStock pressure, fulfillment and support SLA

1. The Business In One Glance

BlueShop BI begins with three practical questions:

The Executive Overview brings GMV, orders, average order value, margin, customers, promotion activity and stock health together. From there, stakeholders can move from the overall position into the customer journey, commercial actions or operational outcomes without losing the business context.

In the 14 July 2026 Business Morning capture, the Executive Overview shows CHF 2.1M GMV, 12,634 orders, CHF 610K gross margin and 4,425 customers. These are recorded development-data figures, not current production totals.

2. Five Questions BlueShop BI Answers

Business question What stakeholders can explore
How is the business performing? GMV, orders, AOV, margin, revenue trends, categories and demand periods
Who are our customers? Customer value, geography, loyalty, retention and RFM segments
What drives conversion? Checkout progression, promotions, basket affinity and recommendation CTR
Can we fulfill demand? Stock pressure, turnover, delivery performance and support SLA
Are customers satisfied? Ratings, review velocity and moderation outcomes

Decision map diagram

Recommendation architecture diagram 1

Text alternative: BlueShop BI brings business performance, customer value, conversion and engagement, operations health and product quality into shared decisions.

3. Dashboard Capability Map

The twenty dashboard routes documented on 14 July 2026 are grouped by the decision they support rather than by the service that emits the source event.

Business performance

Customer value

Conversion and engagement

Operations health

Product quality

4. Sales And Commercial Performance

The sales views answer both the immediate and longer-term questions:

Money remains tied to the accepted order-time facts. Historical sales are not recalculated from a product price or cost that changed later.

5. Customer Value And Retention

Customer BI moves beyond a customer count. It combines:

The analytical customer dimension does not store names, email addresses, phone numbers or street addresses. It keeps only the business attributes needed for aggregate views.

6. Conversion And Engagement

BlueShop follows the customer journey from intent to outcome:

Cart created
→ item added
→ checkout started
→ order placed

The checkout funnel shows progression and abandonment. Promotion views connect offers to GMV and margin. Basket affinity shows which products appear together, while Recommendation Performance connects recommendations shown to recommendations chosen.

This gives commercial teams a shared vocabulary for customer value, conversion and engagement without reducing every decision to a single metric.

7. Stock, Fulfillment And Support

Operations BI follows three connected responsibilities:

  1. Serve demand: stock pressure, shortages, recovery and rejected quantity.
  2. Deliver reliably: picked, shipped and delivered lead times with median and p90.
  3. Respond when needed: support backlog, first response, resolution and agent workload.

Together, these views help teams discuss availability, delivery and customer service as one operating flow rather than isolated departmental reports.

8. Product Quality

The Product Quality & Reviews dashboard completes the journey after delivery. It brings together:

Only approved, non-deleted reviews contribute to product rating averages.

9. How Business Activity Reaches BI

The architecture stays out of the stakeholder's way, but it explains why the business views remain connected to operational activity.

Event-to-dashboard architecture diagram

Recommendation architecture diagram 2

Text alternative: Operational services publish transactional outbox events as Kafka CloudEvents. Analytics Service converts them into ClickHouse facts and dimensions. Metabase dashboards are embedded in the Keycloak-protected BI application for stakeholder decisions.

Operational services keep ownership of their business state. The BI plane consumes events, builds analytical facts in ClickHouse and exposes signed Metabase dashboards inside the role-protected Angular application.

10. How To Read The Numbers

Data-window note (14 July 2026). Cart and Fulfillment facts begin on 2 July 2026; Recommendation facts begin on 11 July 2026. Support and Review facts are forward-only from their respective implementation windows. Dashboard families therefore do not cover historically comparable periods, and no backfill is implied.

Term Plain-language meaning
GMV Gross value of non-cancelled orders using the accepted order-time amounts
Gross margin Accepted selling value minus the order-time cost snapshot
AOV GMV divided by order count
Attach rate Share of orders carrying the selected promotion or reward
CTR Recommendation clicks divided by recommendation impressions
RFM Customer recency, purchase frequency and monetary value
p90 A duration that 90% of the measured journeys complete within
Retention Customers from a cohort who return and place another order

Each dashboard also exposes its calculation notes directly below the embedded BI view.

11. See BlueShop BI In Action

EXECUTIVE STORY / 1M35S

A Business Morning in BlueShop

Tour the Executive Overview, checkout funnel, product promotions, recommendation engagement, customer RFM, stock pressure, fulfillment, support SLA and product quality. Each dashboard scene carries concise burned-in narration without overlapping the dashboard question.

Open the WebM directly

Scene and narration timeline
  1. 00:00-00:04 — Story question: what happened, why it matters and where attention is needed.
  2. 00:05-00:13 — Executive Overview: GMV, orders, average order value and margin establish the position.
  3. 00:14-00:23 — Checkout Funnel: where carts progress and where momentum is lost.
  4. 00:24-00:33 — Product Promotion Effectiveness: promoted GMV, discount, units and margin.
  5. 00:34-00:43 — Recommendation Performance: products and SKUs connect impressions, clicks and CTR to choices.
  6. 00:44-00:52 — Customer RFM: recency, frequency and monetary value shape the portfolio.
  7. 00:53-00:61 — Stock Pressure: shortages, recovery and rejected demand identify attention.
  8. 00:62-00:70 — Fulfillment: percentiles show how quickly paid orders reach delivery.
  9. 00:71-00:79 — Support SLA: backlog, first response and resolution guide service work.
  10. 00:80-00:88 — Product Quality: ratings, review velocity and moderation complete the journey.
  11. 01:29-01:35 — Closing: one business view connects performance to customer outcome.
EXECUTIVE PERFORMANCE / 48S

Six Orders Move the Business View

This film stays on Executive Overview and shows only the before/after effect of a six-order test window: 2,460 to 2,466 orders and CHF 344,155 to CHF 350,049 GMV. The customer journeys and event processing occur outside the frame.

Open the WebM directly

Scene and narration timeline
  1. 00:00-00:03 — Title: the dashboard effect of a six-order test window.
  2. 00:04-00:11 — Before: CHF 344,155 GMV and 2,460 orders.
  3. 00:12-00:19 — Six customer journeys occur outside the dashboard frame.
  4. 00:20-00:27 — Event processing also occurs outside the film.
  5. 00:28-00:39 — After: CHF 350,049 GMV and 2,466 orders.
  6. 00:40-00:48 — The visible claim is limited to the dashboard before/after effect.
OPERATIONS / 38S

Availability, Delivery and Service

Move through Stock Pressure shortages and recovery, Fulfillment & Delivery lead-time percentiles, then Support Operations backlog, first response and resolution.

Open the WebM directly

Scene and narration timeline
  1. 00:00-00:03 — The film asks whether demand can be served and orders delivered reliably.
  2. 00:04-00:13 — Stock Pressure: shortages, recovery and rejected demand.
  3. 00:14-00:22 — Fulfillment & Delivery: lead-time percentiles and delivery flow.
  4. 00:23-00:31 — Support Operations: backlog, first response and resolution time.
  5. 00:32-00:38 — Closing: availability, delivery and service become one operating conversation.

For deeper commercial and recommendation stories:

12. Result

BlueShop BI turns activity across commerce, customers and operations into a shared business language. Stakeholders can start with the overall position, move into the journey that needs attention and return to the same connected context for the next decision.