Business opportunities don't wait. This was the catalyst for the entire company: we want to deliver something very quickly, without the usual time spent on design and long discussions. And we felt motivated to do it.

This article provides a detailed look at replacing the core system during Direct's transformation.

‍

Step 1: Customer journey workshop (1 day, 50 people)

The foundation of our success was a one-day workshop based on a specific customer journey consisting of about seven stages. With 50 people, we mapped out and broke down against that journey everything that needs to happenfor the product to go live in three months.

The workshop included everyone who needed to be there: developers, designers, frontline staff, key stakeholders, and process mappers. Everyone brought value and their own perspective. No presentations—just pure collaborative work.

The key: you don't start with "what the stakeholder wants," but with the customer journey. Only then does it become clear what needs to be built.

‍

Step 2: Initial backlog refinement (approx. one week)

The workshop was followed by roughly a week and a few days of design, breakdown, and initial backlog population. This allowed us to map out all product requirements and set one clear goal: by the turn of the year, when the product launches, we must be able to underwrite.

‍

Step 3: Ruthless scope cutting based on a single criterion

A clear goal made it possible to prioritize at a high level — and, most importantly, to say no firmly. There was only one criterion: the primary goal is to be able to sell the product. Everything else was set aside:

• We don't need a polished client portal if we can't sell the product.
• It is enough to be able to handle a claim if one arises.
• From administrative processes, only those required by law — often adopted from the operations of a traditional insurance company.

This is the hardest and most important discipline of the entire story: scope is not cut based on what looks nice, but based on a single business goal.

‍

Step 4: One owner, cross-company collaboration

The crisis taught us a lesson about ownership. That is why one Product Owner held the main backlog for the entire product and collaborated with other POs on the parts that their segments needed to cover.

The product was put in the charge of a product team B2C (it was primarily a client product and onboarding). At the same time, the team digitizing claims processing had to manage their part. And because it was a company-wide priority project, the collaboration worked.

And here is the real shift: on this specific product, we transformed the classic siloed organization. Previously, requirements, designs, and development were tossed between departments. Now, we brought business people into one place, practically into one team — without any organizational change. They met frequently and refined the backlog; designers knew what to focus on to eliminate "wait time" between design and development (= time when work is sitting idle). Handovers happened overnight. The cooperation was strong, natural, and human.

The result: we accomplished the same amount of work as with the old products but we minimized waiting time and work that wasn't immediately necessary.

‍

Results and broader impact

We delivered the product in three months. And it created a blueprint for the entire company — product team: representatives from all relevant departments who plan, build, test, review, and launch together.

Key systemic shift: previously, priorities were assigned to departments in parallel — now they go to a product team that has all the roles and competencies in one place.

‍

What we would do differently next time (an honest retrospective)

To ensure this isn't just a success story: we lacked continuous discovery. The product had to be adjusted on the fly — designs and copy changed frequently, and we had to be careful not to damage the brand. Because we were cutting hard and moving fast, some things were only fine-tuned after launch. Next time, we would invest more in continuous validation during those three months. It wouldn't have extended the timeline. And we would have reduced the amount of rework.

‍

Key takeaways from this module

1. Start with the customer journey, not a feature list. The customer journey reveals what is necessary — and what isn't.
2. A single clear goal is a tool for cutting scope. "Knowing how to sell it" is a tough filter that saves months.
3. One backlog owner + cross-functional collaboration beats tossing tasks between departments — and doesn't require a reorganization.
4. Reduce handoffs to days, not weeks. Most time lost in product development is spent waiting, not working.
5. Speed comes at a price: discovery. When moving fast, plan for continuous validation so you aren't debugging after launch.

‍