Case Study
Embat is a Spanish fintech that automates treasury management for mid-market and enterprise companies. When the program started they had 5 product and engineering squads working in parallel, active investment and an ambitious roadmap.
The team was good. The problem wasn't the people: it was the absence of a shared system. Each squad worked its own way. No comparable metrics between teams, no way to know if a week had gone well or badly beyond each person's perception.
This is what makes the case relevant: Embat wasn't in crisis — and still found enormous room to improve. If a team in full growth mode had this much space, so does yours.
The symptoms were visible. The root causes were much harder to see from the inside.
Discussions about speed or quality were pure perception. "Seems like we're going slow" has no solution. You need a number, a trend, an objective signal of where the bottleneck is.
Each team had developed its own rules, Jira states and rituals. Moving from one squad to another was like switching companies. Cross-functional collaboration was nearly impossible.
Teams started tasks without finishing the previous ones. Things in QA didn't count as active work. Real WIP was double what was perceived — and nobody could see the full picture.
Before vs. after, weekly averages. Same people. Different system.
CT P85 · SPIDERBANKS
SD CYCLE TIME · SPIDERBANKS
THROUGHPUT DONE · SPIDERBANKS
Not a 400-page framework. A small set of practices that can be taught in one go — and that teams keep when we leave.
Clear task typology (bug/incident/story/spike), mandatory impact traceability to enter the plan. The backlog stops being a wish list and becomes a contract.
Explicit WIP limits and WIP age control. QA counts as In Progress. Fewer things open means more things actually finished.
Review · Retro with 5 Whys · Refinement · Planning. A fixed cadence that creates rhythm, not bureaucracy. Demos show only what's done.
Incidents have their own lane with SLAs. The plan doesn't break when a fire arrives. You manage flow, not chaos.
A task without acceptance criteria doesn't enter the sprint. A task without a Definition of Ready is a rework bomb. The gates protect the team.
CT/LT percentiles, SD, throughput, PVD, fails/100, fix rate. The dashboard stops being a management report and becomes an operational mirror for the team.
Lifting the rugs has definitely added value to Embat. It opened our eyes. We knew we had problems, but you helped us understand how big the problem actually was. If I could go back, we should have done this earlier at Embat.
I loved the entire metrics side, the control we gained, even the product cadence we're now setting. I think that's been good for the team as a whole.
I think you achieved it. You had an impact on what you did and you laid some very strong foundations.
We needed an external wake-up call for leadership to see the real problem. You gave us very clear and realistic guidelines.
My #1 takeaway is the weekly measurement: you see real work, bottlenecks, and give executive visibility. Less gut-feel discussions.
When something is in QA it's still In Progress. It counts for WIP and we avoid opening more. Fewer things open, more gets finished.
We went from subjective to measurable. Capacity and predictability became day-to-day practices, not aspirations.
The dev-led daily freed us. More ownership and focus on closing, not reporting.
Before it was chaos. With controlled WIP, slicing and ceremonies, focus and quality went up at the same time.
Impact mapping gave context: now we know why we do what we do and how it impacts the client. That changes every conversation.
The most valuable thing was making the problem tangible to management: lower WIP, reserve space to think, and coordinate better.
A team in full growth mode found enormous room to ship faster and more predictably — in just 10 weeks. One call is all it takes to see what that looks like for yours.
See What Fixed Looks Like →We take on a limited number of new clients each quarter.