Running Agile in a Traditional Automotive Company

The Setup

Doğuş Automotive is one of Turkey’s largest automotive distributors. When I joined as a Product Owner in 2010, the company ran everything on waterfall — long planning cycles, thick specification documents, and change requests that took weeks to approve.

My mandate was to modernize the digital product portfolio. The unspoken challenge? Do it without breaking the organizational machinery that kept a multi-billion-dollar operation running smoothly.

Starting Small: The Pilot Team

The biggest mistake in Agile transformation is going big. I’ve seen organizations announce “we’re going Agile” company-wide and watch as 200 people suddenly attend standups without understanding why.

Instead, we started with a single cross-functional team of six: two developers, a designer, a business analyst, a QA engineer, and me as Product Owner. We ran our first sprint on a dealer management feature that had been stuck in requirements limbo for four months.

Two weeks later, we had a working prototype. The feature shipped in Sprint 3. The business analyst — initially the most skeptical — became our biggest advocate.

What Worked

Visual Management

We installed a physical Kanban board near the coffee machine. This wasn’t a productivity hack — it was a communication strategy. When executives walked past and saw cards moving from “In Progress” to “Done,” they started asking different questions. Instead of “When will it be finished?” they asked “What’s being worked on right now?”

That shift in questioning changed everything.

Two-Week Sprints

In an automotive context where planning horizons are measured in quarters and fiscal years, two-week sprints felt radical. But they created a rhythm the team could trust. Every two weeks, we demonstrated working software. Stakeholders got used to seeing progress, not presentations.

Definition of Done

This was non-negotiable from Day 1. Every story needed to be coded, tested, reviewed, and deployable before we called it done. In the waterfall world, “done” often meant “the specification has been written.” Redefining done was one of the most impactful cultural shifts.

What Didn’t Work

Sprint Planning With Executives

Early on, I invited senior leaders to sprint planning. Fatal error. A session designed for detailed task breakdown became a strategy debate. We learned quickly that executives belong in sprint reviews, not sprint planning.

Pure Scrum in a Matrix Organization

Doğuş Automotive has a traditional matrix structure. Team members reported to functional managers who had their own priorities. Pure Scrum assumes dedicated teams — we didn’t have that luxury.

The compromise: we negotiated 80% allocation for sprint work, with 20% reserved for operational duties. Not ideal, but pragmatic.

Ignoring the PMO

The Project Management Office initially viewed our Agile experiment as a threat to their governance framework. Instead of fighting, we adapted our reporting. We mapped story points to earned value metrics and produced status reports in the format the PMO expected — just populated with data from our sprints.

The Outcome

Over 18 months, we expanded from one Agile team to four. Delivery cycle time dropped from an average of 6 months to 6 weeks. More importantly, the relationship between IT and business stakeholders transformed from adversarial to collaborative.

Key Takeaways

1. Start with a pilot, not a mandate. Let results do the convincing.
2. Adapt the methodology to the organization. Purist approaches fail in traditional environments.
3. Make work visible. Transparency builds trust faster than any presentation.
4. Respect existing governance. Translate your metrics into the language the organization already speaks.
5. Patience is a strategy. Cultural change takes longer than process change. Plan for 18 months, not 8 weeks.

Agile in a traditional company isn’t about converting the organization. It’s about demonstrating that there’s a better way to work — and letting people choose it for themselves.