I remember the first long conversation with the Fintera team in Warsaw. They wanted to rebuild their customer portal on Salesforce.
Fintera is a Polish financial aggregator. About 50 people on staff. Customers come to Fintera to compare financial products from a long list of vendors, pick one, and move forward with an application. The business had been running for years on a proprietary web platform the Fintera team had built themselves. It worked. It had also accumulated the kind of quiet complexity that every long-lived in-house system does.
The ask was straightforward on paper. Move the portal onto Salesforce Experience Cloud. Keep every product line, every vendor integration, every journey the proprietary platform already supported. Bring in a design vendor called Scada to own the visual layer.
On paper was the problem.
Why we cut the scope before we wrote any code
A full replatforming, with every vendor and every product line live on day one, was going to run for well over a year. Longer, if any of the third party integrations turned out to be thinner than the documentation suggested. We had seen this pattern before. Long replatforming projects go quiet for months at a time, and the business they serve has to wait for a working portal for the entire duration.
We asked Fintera leadership for a week to look at the product catalogue. The goal was narrow. Find the smallest slice of the business that would still be commercially meaningful if it shipped on its own.
Mortgages were the answer.
The mortgage catalogue had a manageable list of vendors. The journeys were well understood. The customers who came through mortgages were the most valuable ones in the funnel. If we could ship a working mortgage comparison portal in a few months, Fintera would have real traffic, real vendor feedback, and a live platform to hang the rest of the product lines off over time.
The build
We framed the project as an MVP. One product line. A small vendor set. A single customer journey from comparison to application. Experience Cloud for the portal itself. Sales Cloud underneath for the data and the case handling.
The interesting piece was the vendor integration layer. Every mortgage vendor in the Fintera catalogue had a slightly different way of sharing rates, terms, and application endpoints. The proprietary platform had absorbed those differences through custom code. We wanted Fintera’s team to be able to onboard a new vendor without calling an engineer.
We built a configurable integration hub. A vendor template with fields for the pieces that always varied: the rate feed, the authentication, the document handoff. Adding a new vendor became a form to fill in, a set of mappings to confirm, and a test run. Fintera staff could do it themselves.
We worked with Scada on the handoff so the static designs translated cleanly into Experience Cloud components. The portal looked like Fintera, carried the brand, and still fit on the Salesforce platform underneath.
What shipped
The MVP launched in a few months. A working mortgage portal on Salesforce, with a handful of vendors live, a customer journey from landing page to application, and a configurable hub the Fintera team could use to add more vendors as new deals closed.
An advanced version followed later. More vendors. More document workflows. A tidier onboarding flow for new customers. Conversion improved on the onboarding journey, and retention improved among the customers who came through the new portal.
The part we tell other clients
Fintera’s original brief was to rebuild the whole thing. A full platform, every product, every vendor, in one long release. Any consultancy would have taken that budget and delivered a roadmap that matched it.
The better shape, once we looked at the catalogue, was to ship a small thing that worked and grow from there. The MVP gave the business a live product in months. The integration hub gave the team a way to expand the vendor list without a new project every time. The Scada handoff gave Fintera design ownership without a handoff problem.
Long replatforming projects are easy to sell. They are hard to finish. A small MVP with a clear growth path gets more working product to the business inside a year.
– Jeff.