Design Leadership
Leading Design at Estonia's Largest Media Company
How I run a team of six designers across product, marketing, and advertising at Delfi Meedia, and what I've learned about leading when resources are fixed and scope is not.
The house of brands I manage under Delfi.
This isn't a case study. There is no process diagram. This is an account of how I lead and slices of my experiences. For a design leader, the way the team is run is the actual product.
OUTCOMES: A successful team of 6 across 20+ portals and 5 products; full brand integration shipped in 2.5 months; team eNPS among the highest in the company.
What I walked into
When I took over, the design team had just been merged from two departments. Six designers, two cultures, no shared process. A design system existed on paper and was ignored in practice. Designers found out about projects after the decisions were made, then executed from other people's briefs. The weekly meeting was a task status review. Nobody showed their design work to anyone.
None of this was anyone's fault. It is what happens when an organization chart changes faster than working habits do.
Structure before inspiration
I didn't start with vision statements. I started with mechanics.
The weekly meeting became a design critique. Work gets shown, questioned, and defended every Tuesday, which did more for quality and team cohesion than any offsite could. I introduced 1-on-1s, which sounds basic, but nobody had them before. The design system went from a document to an enforced standard. Briefs from marketing, editorial, HR, and business development now arrive through templated workflows in our task management software, instead of shoulder taps, and the team's knowledge lives in a wiki instead of in people's heads. Also we went through modernizing our tools, like moving from Instapage to Framer.
The harder structural fight was getting designers into project kickoffs. The organization was used to involving design at the end, as a rendering service. That changed meeting by meeting, and it's still changing.
Growing people with no budget
I have no promotions to hand out and no training budget worth mentioning. So I built Design Bootcamp - an internal training programme I run myself, covering design and UX fundamentals, our most used design tools, and AI tooling. It costs nothing and it has expanded what every single person on the team can do.
The clearest example: we were denied a UX/UI hire, so I trained our senior designer into the role, then found another designer on the team who wanted the same path and started mentoring her too. The capability we couldn't buy, we grew.
Engineering around constraints
Running lean is not a phase, it's a permanent condition here. I've stopped treating it as a blocker and started treating it as a design problem.
When the team is at capacity and a stakeholder has budget, I bring in people from a freelancer network I've built over the years, so "we're full" never has to mean "no." For routine production work, we've vibe-coded internal tools using Claude Code, ChatGPT Codex or Cursor so non-designers can self-serve: a social media banner generator, an image compressor, a podcast article image creator, several Adobe and Figma plugins. Every banner a marketer makes themselves is design time returned to solving design problems.
Our custom Figma plugin for easy artboard creation.
The biggest build is now running. When Asana (our previous task-management software) significantly upped their price, it became too expensive and out of our budget. We had to find another way. We looked at competitors like ClickUp and really didn't find anything cheaper or perfect for our work environment, so we decided to build our own. I saw this as an opportunity to improve our workflow, so I analysed our tasks, tried to find the patterns in how design work really gets requested, named, and filed. That analysis became a clean taxonomy, four fields where each one answers exactly one question, with purpose and output format finally separated instead of combined the way the old system left them. The task management tool is the implementation of that research. My senior designer built it with AI tooling under my direction, we've imported the old tasks, and the team is using it now in place of Asana. I treated our own operations as a UX problem, did the research, and shipped the fix. The licence we didn't renew is almost beside the point.
Capacity itself is managed in the open. The team flags load at the Tuesday weekly, work gets rebalanced, and freelancers fill genuine overflow. It's informal, and it works better than the formal systems I've seen fail elsewhere.
oSano - our task management system we tailored especially for our needs.
The stress test
Delfi acquired Geenius, a technology media brand some time ago, and my team owned the full design integration: the portal, four sub-channels, the subscription landing page and order flow, every paywall, newsletters, checkout, billing templates, marketing merch, and a community forum. The brand constraint made it genuinely difficult: the result had to feel like Geenius, fresh but recognizable, and deliberately not like Delfi, because subscribers are asked to pay for it separately.
Deadline: 2.5 months. No added headcount. It shipped on time and it's live: geenius.ee
I include this because culture work is easy to claim and hard to prove. A merged-then-rebuilt team working together and absorbing an entire acquired brand on that timeline is the proof.
Geenius before and after the integration to the Delfi ecosystem.
Measured, not just felt
An independent workplace wellbeing survey runs across the whole group every year. The design team's results rank among the strongest in the company: team satisfaction and management scores well above company averages, and an employee NPS among the highest at Delfi.
The lowest-scoring area was workload. That's the honest cost of running lean, and it's the constraint I manage most actively, through the freelancer network (if there's extra budget), the tooling, and a standing case for headcount backed by exactly this kind of data.
What I know now
Design leadership is mostly not about design. It's about growing people, building systems that keep working when you're not in the room, and finding a workaround every time the answer is no. I still carry a real IC workload alongside the management role, and I've stopped apologizing for that. Player-coach is a legitimate operating model when the systems around it work.
My team recently told me this is the golden time for them at Delfi. There were no raises behind that, and no promotions. What they got was ownership, new skills, a seat at the table, and trust. It turns out that's most of the job.
