Case study · 2026
Weldaad Business Intelligence Dashboard
Exact held years of sales data, but extracting management insight from it still meant digging through an accounting system.
- Exact Online API
- Shopify App
- React
- Node.js
- Discipline
- Business intelligence · Exact Online · Sales analytics
- Year
- 2026
- Engagement
- Fixed scope
- Status
- Live
01Context
- Build management-level sales and customer analytics.
- Support flexible reporting periods instead of all-time-only metrics.
- Use Exact historical sales rather than only Shopify or active orders.
Weldaad already had Exact Online as its central operational system, but its reporting tools were focused primarily on inventory and accounting.
Management wanted clearer commercial answers: which customers are new or returning, what products are genuinely best sellers, where revenue comes from, and which customers drive disproportionate value.
HeapByte added a BI layer to the existing portal so the same Exact data could become a practical management tool.
02What was in the way
- 01
Historical classification
Identifying new, returning, and reactivated customers requires looking outside the currently selected reporting period.
- 02
Metric definitions
“Bestseller” was defined as quantity sold divided by MOQ, not simply revenue or raw units.
- 03
Data completeness
Active-order caches were insufficient for historical reporting and had to be expanded.
03Architecture
What we built, and what we deliberately left alone.
- 01
Historical Exact dataset
Delivered and active order data are synchronized for reporting.
- 02
Derived customer segments
Customer history is evaluated relative to the requested period.
- 03
Period-aware metrics
Revenue, order count, AOV, bestsellers, and customer behavior are recalculated for the selected range.
Delivery
- Metrics mapping
- Dashboard implementation
- Historical synchronization
- Data validation
- Reporting refinements
04Outcome
- Customer records
- 1,400+
- Refresh interval
- 10 minutes
- Top customers
- Top 20
- Reporting
- Period-based
Customers with historical orders surfaced in the portal
Operational data refreshed automatically
Ranked with cumulative Pareto context
Date filters apply to sales and bestseller analysis
05Engineering notes
- Bestseller formula
- Product performance can use units sold divided by Shopify MOQ to estimate ordering frequency.
- Customer segmentation
- New, returning, and reactivated states are derived from complete customer order history.
- Pareto view
- Top-customer ranking includes cumulative contribution to expose revenue concentration.
- Exact history
- Historical delivered orders were added after early versions showed only the active order backlog.
What we took from it
- 01A dashboard is only as useful as the definitions behind its metrics.
- 02Historical customer segmentation requires data outside the visible reporting window.
- 03ERP data becomes more valuable when presented around business questions.
- 04Raw revenue and product popularity answer different questions.
Disciplines applied
Next case study
Säsongsgruppen Seller QR Platform
We built a Shopify fundraising platform that attributes purchases to sports-club sellers through QR links and manages teams, reporting, payouts, and accounting.
Similar problem?
We will tell you whether this approach fits.
Every store is different. Send us yours and we will say what we would actually do — including when the answer is less work, not more.