When Two Teams Want Opposite Things
Aligning stakeholders whose success is measured differently, and turning the conflict into rules both sides accept
Two organizations depended on my team, and they measured success in opposite ways. Growth launches campaigns on fixed dates and wanted speed and numbers. Design wanted quality, because every page we release carries the company’s identity.
I lead five designers who sit in the design organization but work day to day with Growth. This is how I kept both sides moving together, and how EVT-X, a platform I proposed, put it to the test.
- Role
- Design leadershipStakeholder managementCross-team alignment
- Scope
- Cross-organization
Why it was hard
Each side was right about its own job, and each had a reasonable fear of what the other would do. That is what made it hard.
- Different yardsticks. Growth was judged on numbers and launch dates. Design was judged on the quality of what shipped. A win for one looked like a loss for the other.
- More volume than the team could absorb. Campaigns kept coming, and five designers could not build every page by hand.
- A cost that moved out of sight. Once an AI platform made pages cheap to build, the work of handling more requests, coordinating them and checking quality moved to design.
Starting from shared goals
I could not settle this by picking a side, so I started from what the two sides already shared. When I set each quarter’s OKRs, I first find out what Growth plans to do. Then I look for the business lines or processes the design organization already cares about, and tie the two into one goal.
For example, when Growth plans a batch of campaigns, I use that volume to argue for extra resources. Those go into improving a process the design organization cares about, or into a proposal for a redesign. Growth gets its campaigns, and the design organization gets work it wants to do.
There is no shortcut here. I talk to both sides often, so each knows what the other is racing against.
The test: EVT-X
Campaign pages were many and the schedules tight. Together with an engineer, I proposed EVT-X, a platform that generates campaign landing pages with AI. It had two jobs to do: the earlier block-based platform looked too uniform, and it did not follow the design guidelines well.
Here the two sides pulled apart. Growth only asked for speed and did not care much about the quality of what was generated. Design’s position was that each released page represents the company, so quality could not be traded for speed.
I did not pick a side. I proposed sorting requests by value and complexity, because if every campaign used the same amount of design effort, the team would be overwhelmed. We settled on three tiers:
- Full customization
- Partial component changes
- Existing components only
How much design and the design system team take part depends on the tier. To make the third tier trustworthy, we first built the base design system and a first set of static page templates, as the standard option for urgent requests.
Each side kept what it cared about most. Growth kept its speed on routine pages, and design kept its review on the pages that needed it. Both sides’ demands became one rule.
What it delivered
- Pages launched in Q1 202619 more than the previous quarter (+172.7%)
- 30
- Page update actionsIn the same quarter
- 3,511
- People’s worth of work moved onto the platformAd campaigns, agency product pages and the Japan–Taiwan regional campaign hub
- 2
From my Q1 2026 design OKR review. The freed capacity went into higher-value projects such as UEX, TradFi and VIP, which are now live.
What I took from it
It is easier to get resources for a goal both sides already want, so I begin an OKR with what each side finds most painful right now.
When one team’s cost goes down, the organization’s cost does not necessarily go down with it. It can move somewhere else. The tiers were my way of putting that on the table. EVT-X is now the team’s main fast lane for campaign landing pages.