When product, CX, and technology need one roadmap
Parallel plans look orderly on org charts. Customers experience them as inconsistency and broken promises.
Planning season brings three separate AI roadmaps. Product optimizes features, CX optimizes NPS, technology optimizes platform migration. The customer sees one brand. Merging roadmaps is a governance choice, not a workshop.
One star, not three OKRs on the same customer
Pick one customer-visible outcome and derive team OKRs from it. Pilots belong on the shared map with exit criteria; see 90-day pilot design, not as side quests.
| Function | Owns in prioritization | Brings to the room |
|---|---|---|
| CX | Customer truth, friction evidence | What frontline can promise |
| Product | Sequencing, tradeoffs | Journey ownership |
| Technology | Feasibility, debt, risk | Staffing and dependency reality |
| Executive | Stop-doing courage | Capacity truth |
Publish what you stop
Every AI program needs explicit stops: duplicate pipelines, zombie dashboards, pilot connectors that will not scale. Leadership should ask what we ceased, not only what we shipped.
| Question | Product | CX | Technology |
|---|---|---|---|
| Same north star? | ✓ | ✓ | ✓ |
| Shared release calendar? | ✓ | ✓ | ✓ |
| Stop-doing list published? | ? | ? | ? |
If the third row is blank, capacity is imaginary.
Technology leaders should bring debt and staffing to the same table; see the paired memo on team strength vs vendor delivery.
Frequently asked questions
Who should run the merged roadmap session?
Usually product or a program office facilitates, but the session only works if CX field reality and honest engineering feasibility are on the table. Without both, you get alignment on paper only.
What if departments refuse to merge roadmaps?
Show the customer journey collision with concrete examples: the same user getting two conflicting promises from two teams. That is a CX defect, and escalating with evidence is fair game.
How do pilots fit on the shared roadmap?
Every pilot appears on the map with exit criteria and feeds the portfolio when it ends. A pilot not on the map is informal experimentation; it rarely transfers learning into the organization.
For, CTO and technology leaders
Bring staffing and debt visibility. Prevent CX from promising what the roadmap cannot carry.
Paired perspective: When your AI program should strengthen teams instead of adding a delivery vendorFurther reading
- McKinsey: rewiring to capture AI value, 78% use AI; fewer rewire
- Google: Article structured data, One clear story for crawlers and teams
- Gartner: AI project production rate, 48% reach production
Running a program on this topic? Describe your context, technology, experience, or both.
Discuss an engagement