How Business Analysis Canada led requirements, process modelling and acceptance criteria for LELO's global e-commerce, then guided the build and support.
The challenge
Selling a premium product in more than 160 countries means one storefront has to follow many different sets of rules. LELO needed those rules agreed and written down before committing development budget.
Our business analysis approach
The aim of the first phase was one set of requirements that the business had validated and signed off, so development could start from facts rather than assumptions.
Requirements
Rules scattered across teams
A market rules catalogue and requirements for market-level settings: languages, currencies, payment methods, tax display and shipping options
Restricted products visible in some markets
Availability rules applied consistently at catalogue, cart and checkout
Market-specific legal requirements
An age confirmation step for markets that require one, with the rule controlled per market
Conflicting promotions and unclear precedence
Price lists by market, promotion precedence and stacking limits, and an approval step before any promotion goes live
Inconsistent handling across touchpoints
Discretion rules for packaging options, card statement descriptors, email and SMS content, and account pages
Claims handled by email
An online registration and claim flow with proof-of-purchase upload, a product authenticity check, eligibility rules and routing to regional teams
Different return handling by region
Return eligibility rules by market, including hygiene exceptions for unsealed products, a return authorization flow and refund rules
Manual handoffs between systems
Data flows between storefronts, order management, the warehouse and customer care tools, with an agreed owner for each data element
No shared definitions
Agreed definitions for return rates, claim reasons and promotion performance by product and market
Acceptance criteria
Acceptance criteria were the contract between the business and the build. If a feature met its criteria, it was done. If someone wanted something different, it went through change control as a change request. Three examples from the baseline:
Delivery and support
Development and support used the same requirements, so the business validated a requirement once and then saw it carried through to the live site.
Epics and user stories were derived from the BRD, and each story was linked to its requirement, process model and acceptance criteria.
Our analysts joined backlog refinement and sprint planning, answered questions on business rules, and clarified edge cases before development started on them.
New ideas were assessed against the baseline. Each one either became a change request, with its impact on effort, timeline and dependent stories recorded for the product owner to decide, or was parked for a later release.
Test scenarios came straight from the acceptance criteria. Business owners ran user acceptance testing market by market, and a story was accepted only when all of its criteria were met.
After launch, every issue was triaged against the acceptance criteria. If the build did not meet a criterion, the issue was a defect and was fixed. If the build met the criteria but the business wanted different behaviour, it was logged as an enhancement for the product owner to prioritize.
The baseline, decision tables and process models were updated with every approved change, so support teams always had current rules to work from.
Privacy
In this category, privacy affects how customers experience the brand, not just compliance. Discretion requirements were written for every customer touchpoint and tested like any other requirement.
Results
“
Lessons learned
Talk to a senior business analyst about setting a validated requirements baseline before you build, then keeping delivery and support tied to it.
