I took an internal pricing product from 0 to 1, from discovery to a phased regional rollout: a single source of truth for commercial conditions, a self-serve app to consult them and build quotes, and an approval workflow with a full audit trail that turns approved changes into ERP-ready records.
A regional user reviews their conditions and proposes a change; margin guardrails and the audit trail are visible before anything reaches the master data.
RegionRegion A
Customer groupGroup 2
DestinationChange event
| Family | Change | Guardrail | Record |
|---|---|---|---|
| Family A | Discount revised | Passed | Queued |
| Family B | Discount raised | Review | Held |
| Family C | Comment added | Passed | Queued |
Problem statement and product bet
Commercial conditions sat in data silos (regional matrices, local records and the ERP) with no single source of truth, inconsistent margin rules, margin-leakage risk and no audit trail of who changed what, or why.
The bet: fix data quality before automating anything, then separate read from write: self-serve consultation for everyone, one maker–checker workflow for every change.
01 · Problem space
It was a mix of data, business rules, exceptions and tribal knowledge, split across regional matrices and the ERP and maintained by different people. Consulting, updating, reconciling or auditing a condition depended on knowing where to look and whom to ask: a textbook key-person dependency.
Reconstructed problem model
Pain point 01
Pain point 02
Pain point 03
Jobs to be done
Governance requirements
02 · Discovery
I pitched the product vision: replace the regional matrices with one governed architecture, keyed by region, customer group and commercial criteria. I then ran requirements elicitation, translating commercial needs into the entities, fields and relationships the data and ERP teams could extract. The result was a data-requirements spec, not a request for a download.
Reconstructed discovery model · findings paraphrased
Problem, target users, value proposition and strategic direction in one proposal: fragmented maintenance; Marketing, Pricing, commercial and regional teams; centralised control and faster updates; one governed architecture.
Sales office, region, pricing family, product hierarchy, price list, customer group, discount value, and material-level identifiers where they applied.
An entity model linking organisation, sales offices, pricing families, tariff types, margin criteria, materials, customer hierarchies, discounts, theoretical net margin and usage by operational centre.
The discrepancies were documented; their financial magnitude was not. Confirmed findings stay separate from unquantified impact.
Riskiest assumption 01
Riskiest assumption 02
Riskiest assumption 03
03 · Data quality
Before automating how the data was used, I made it reliable. I shipped the clean-up as five incremental audits, each de-risking one dimension of data quality, and ran them as federated data stewardship: every accountable owner received their filtered records, a keep / correct / remove decision field, instructions and a deadline. Each agreed decision became a structured change record.
Reconstructed audit sequence
Conditions present in the ERP but missing from the commercial records, and the reverse.
Records held in both places with different percentages, reconciled side by side with the variance calculated.
Material-level discounts reviewed one by one, then retained, corrected or deprecated.
Margin criteria compared across regions, and discounts that took the theoretical net margin below the floor flagged, including strategically priced items that needed their own margin rule.
Commercial comments added and sales-family descriptions rewritten, so the data no longer depended on tribal knowledge.
Standardised: the core
Existence, discount values, margin criteria and descriptions now follow one standard.
Configured: the edge
Where a location had a legitimate commercial requirement and no regional consensus existed, the exception stayed, documented rather than hidden.
Deprecated
Discounts that were no longer commercially justified were retired through the same structured change records.
04 · Solution
With a clean source, I designed the solution as three decoupled layers with a clear separation of concerns, built low-code on Power Apps and Power Automate. Business users self-serve complex pricing rules without ever touching the data model underneath.
Reconstructed product architecture
Data layer
The golden record for matrices, regions, customer groups, families, materials, discounts and margin criteria: validated, synced with the apps and decoupled from the UI.
Experience layer
Conditions sliced by region, office, centre, customer group, family, material, price list and margin criterion; material-level exceptions nested under their family; and a guided, CPQ-style flow to a structured commercial proposal.
Workflow layer
Propose, justify, review, approve: a human-in-the-loop, four-eyes flow. Changed values are flagged automatically, and approved changes become change events for the ERP with requester, timestamp and decision.
Decision 01
Decision 02
Decision 03
05 · Go-to-market
Stakeholder demos closed the feedback loop before launch. A single-region pilot then tested the riskiest question (could a regional user review their own matrix and propose changes through the controlled process, unaided?) before a phased rollout.
Reconstructed rollout model
Formal stakeholder demos, with the feedback applied before release.
Access current data, select the region, review discounts and margin criteria, propose changes, see modified records flagged automatically, and submit for validation: user acceptance in real conditions.
Access provisioning, enablement through user guides and demos, controlled deployment in the organisation’s internal environment, stakeholder comms and hypercare through first-line support.
Regional matrices compared and consolidated into common structures, exceptions reviewed, customer classifications revised, then tested end to end in operations.
A regional user reviews their pricing matrix and proposes changes through the controlled workflow, without write access to the master data.
As a regional user, I want to find and understand my conditions and propose a justified change, so that it reaches review as a structured, traceable record.
Acceptance criteria: all five
Lesson learned. Testing in one region surfaced a customer-group allocation defect outside the application. After the fix, a retest confirmed the discounts applied correctly and the consolidation was signed off. In-app guardrails reduce risk; only end-to-end testing (approved change, ERP load, customer classification, applied condition) proves a change has really shipped.
06 · Outcomes
In operational use
Regional pricing moved from data silos to a single source of truth that people consult, review and change through one audit-trailed process.
Verified outcome
Measurement boundary
Reconstructed measurement model
Share of pricing changes that flow through the governed workflow rather than around it.
Lead time from a proposed change to its approval and ERP load.
Discrepancy rate between the source of truth and the ERP, per audit cycle.
Changes held by margin or validation checks before approval.
Tech debt and risks
Roadmap
Role in practice. End-to-end ownership: alongside the product-owner remit I covered business analysis, the functional data model, solution and workflow design, data-quality coordination, QA, the pilot, internal launch, documentation and first-line support, aligning Marketing, Purchasing, Pricing, Data and PIM, IT, market managers, regional teams and commercial centres. “Product owner” describes the work, not a formal job title.
Attribution boundary. The product operates within a broader commercial process. Wider commercial outcomes require verified measurement and cannot be attributed exclusively to it.
Standardisation was never the goal. A reliable source, justified exceptions and traceable change were.
Professional approach
Explore seven decision frameworks across lifecycle, commercial, launch, availability, product data, adoption and supplier questions.
View decision approachAcademic mobility portfolio
Follow SOLAR MOTION from individual brand and business design through export operations and a collaborative Chile strategy.
View case study