Turning unused payment data into a service that merchants want to buy.

The service to invent: turning payment terminal transactions and the bank's anonymised client profiles into a tool for retailers, without creating any new data. The input material existed: 9 B2B personas and an internal study. Research first sized reality: the market exists, but usage of the bank's digital tools remains low; the solution would have to be simpler than what exists.

Segmentation eliminated two entire segments as not interested: multi-shop retailers already equipped with data support, and B2B businesses. Single shops remain a secondary target. The chosen target: multi-shop retailers without a data team. The benchmark confirmed the space: one bank in the world offered this service, BBVA in Spain, with a catalogue of 120 metrics; the other reference, Proximus's data intelligence offer, sells a visualisation. Neither sells decision support.

User research with SMEs delivered a clear balance of forces: a single driver, easy access to massive data no other player holds, against 6 barriers phrased in the retailers' own words. Those 6 barriers became the criteria for deciding what to build, and what to leave aside.

The decisive insight lies elsewhere: these retailers have never worked with statistics and have no marketing knowledge, yet they understand that data can help their business. The barrier is not conviction, it is access. The value proposition wrote itself in response: an easy-to-read dashboard that prompts insight, carried by a bank as a trusted data partner.

Marketing had commissioned a Spotfire data discovery POC. It did its job: proving that bank accounts and card transactions on terminals could be extracted and cross-referenced. On a real pilot retail chain of 8 points of sale, it processed 63,477 transaction lines, filtered by card type, and profiled customers by age, gender, language and income.

It was never tested with retailers, and no one was convinced of its value for the user. Research had just shown why: a free exploration tool is unusable for a target that has never worked with statistics. The product would have to guide the analysis. To settle that question, I organised the data analysis workshop.

The data analysis workshop aggregated and categorised the real needs of the chosen segment: 18 candidate objectives on the wall, 10 kept in the MVP scope, grouped into 4 named goals: Performance Analysis, Business Expansion, Evaluating Actions, and Market Research, deferred to phase 2.

The principle behind this choice is written in the Concept Report: retailers come with precise goals, want information fast and effortlessly, and will not explore more data than they can process. So there would be no exploratory metrics system, but goal-oriented reports where every screen delivers learnings and actions.

The MVP targets a Triple Win written in black and white: value for the bank, for the retailer and for the cardholder. Its constraints are deliberate: data available for only a few hundred clients, access solely via Easy Banking Business where security and roles are already solved, daily refresh, and the interactive dashboard as the only channel, with no raw data export and no advisory layer.

4 client requests were explicitly refused from the MVP by the Lean Team, and documented as such. What the product does not hard-code, the settings absorb: high and low buyer thresholds, the definition of a lost customer, age brackets, and up to 6 configurable comparison sectors, the real answer to the barrier about competitor relevance.

The product architecture fits in one loop: control the state of the business with Performance Analysis, learn and act with Business Expansion, observe the impact of actions with Evaluating Actions. The third goal requires no new data: it reuses the same data as the first two, in the same order.

Each goal was mocked up screen by screen on a fictional demonstration business: the home with its 3 indicators, the shop comparison table, customers gained, maintained and lost, retention, and customer origin across Brussels' 12 municipalities.

Under every chart, a “What's important” block writes the takeaway for the user, in a shopkeeper's language: “You possibly close 2 hours too early.” The gap between two periods is calculated and written out by default, sparing any mental math; precision lives in the roll-over.

The same discipline applies to time: no free date picker, but 3 canonical comparison bases. And on first use, a 4-step guided tour, optional and refusable, because blocking tutorials kill momentum and get skipped.

Service design work also happens off-screen. The existing access journeys were mapped: a non-client retailer went through 15 steps, 1 to 3 days of postal delay and a branch visit for the smart card and identity check. The itsme option was investigated then discarded on evidence: the bank card remained required and the identity check maintained.

The target journeys reuse Easy Banking Business authentication: subscription reserved for the contract manager, terms and contract signed online in a single session, and if someone else tries to sign up, the system notifies the manager by email instead of losing the intent. Deep-link login opens the report in 2 actions.

The 86-page Concept Report handed to the sponsor sequences the work into 4 deliverables: the research, the value proposition, the concept, then the wireframes and access journeys. Phase 2 is already framed: the Market Research goal, whose tab is reserved in the navigation, and user profiles, deferred until after the MVP.

The service serves a quantified business goal: the acquiring offer opened by Corporate Banking in Q2 2018, acquirer status targeted for 2020, and a payment terminal market share raised from 7% to 20%. The investment committee gave its green light, backlog ready, the 4 teams aligned on the same scope.