Six product themes survived the triage on a cloud platform. Two — self-service & automation, and cost optimization — took 80% of the firepower. Roughly €6M a year in cost avoided and reduced came out the other end.
By the time I took over Spryker Cloud/PaaS, the backlog wasn't a list. It was an organism. Five different mouths, external roadmap, internal roadmap, security, customer operations, internal ops, all feeding off the same starved team, all convinced their item was the one that mattered this quarter.
I could have run a workshop. Instead I looked at what we actually had capacity for, which was not much, and asked a blunter question than any framework wants you to ask: which of these, if we shipped it, would actually change where this company is headed, not just make it feel like progress?
Six things survived that question. Cost reduction. Automation and self-service. Security. Observability. Cloud native. Closing the gaps we'd been overpromising customers on for a year. Everything else, and there was a lot of everything else, got told no, in writing, with a reason attached.
But six themes is a menu, not a diet. The real decision happened inside the six: self-service and automation first, then cost optimization, took roughly 80% of the firepower between them. The other four got enough to stay alive. In retrospect, that ruthlessness was the whole point — the €6M traces back to those two.
They didn't all earn their place the same way. Automation and cloud native grew the platform; customer gaps kept accounts from walking; security did neither. Nobody outside the team ever saw it work, no customer asked for it by name, and on a growth or engagement scorecard it would have lost every time. It stayed funded anyway, because I put its margin case on the same page as the other five, in front of the same leadership team, instead of letting it lose by default to anything with a demo. That's the part that actually cleared confusion in the room: not the six bets themselves, but showing, side by side, which ones grow the business and which ones just keep it from losing money quietly.
Saying no to a backlog doesn't feel like leadership. It feels like disappointing people in a Slack thread, over and over, while someone's pet initiative slips another quarter because you decided it wasn't worth feeding. I did that a lot.
6
themes survived, out of a backlog with no real ceiling
80%
of the firepower on two: self-service & automation, cost optimization
~€6M/yr
cost avoided + cost reduced, recurring
This isn't a story about the best year the company ever had. I joined in May 2022, a few months into the war in Ukraine — not background noise for Spryker. A meaningful part of our people were based in Ukraine, including some on my own product team, and the war helped trigger a macro selloff that hit the whole sector hard right after the COVID-era party ended. Spryker was already under real financial pressure before any of that. What the six bets actually bought was real: roughly €6M a year in cost avoided and reduced, on a recurring basis, and a sales-support and operations footprint that shrank because the platform finally did the work itself instead of a person doing it by hand.
"Angel operates at a rare intersection of deep technical expertise, strong product judgment, and real business awareness. He doesn't just understand how systems are built, he understands why decisions matter in the context of customers, cost structures, and long-term platform scalability. […] He's the kind of product leader you want owning critical parts of your platform and driving decisions that actually matter."
— Elena Leonova, Co-founder & CEO/CPO at OneRank.io, formerly Angel's CPO at Spryker
Not every call was right on the first attempt. A couple needed a second pass before the numbers actually moved, and Spryker gave me the room to get there instead of pulling the plug at the first miss. That margin matters more than it usually gets credit for. Without it, this is a much smaller story.
I left in October 2025. Three years earlier there was no real platform product function at Spryker to speak of; by the time I left there was automation, profit lines that hadn't existed before, and the foundations and roadmap for self-service already laid down — the console itself shipped a few months after I left, built by the team on the plan I'd set. What remained after that was mostly finishing that build and general maintenance, and neither was the kind of strategic work I'm built for. I wanted to set the next direction, not execute the last one. Spryker's still in the fight today, still building on some of what got fixed while I was there. I'd like to think I left more good bones than bad ones.
Two bets got the firepower. Four themes got kept alive. Everything else got a no, in writing, with a reason — consistent enough to survive contact with a real budget.
