The five-tier roadmap nobody can execute
You have seen this deck. A consultancy delivers it after an eight-week engagement. There are five horizons: “foundational,” “operational,” “transformational,” “competitive moat,” “industry-leading.” Each horizon has six to ten initiatives. The total is sixty initiatives over three years. There is a chart at the back that estimates ROI per initiative to three significant figures.
The deck is beautiful. The deck is also unexecutable. Here is what happens to it within ninety days:
- Engineering disagrees with the sequencing.
- Finance disagrees with the budget assumptions.
- Two initiatives become urgent that aren’t on the map.
- Three initiatives that were “tier 1” turn out to be blocked by data work that wasn’t scoped.
- The CEO loses confidence in the document and asks for “a fresh strategic look.”
This cycle repeats every nine months. The document is the product. The document is also the problem.
The three-column roadmap
The version that ships looks completely different. It is one page (literally one page), maintained as a living document, reviewed monthly, and has exactly three columns: Shipping. Next. Watch.
Shipping (this quarter)
Three to five initiatives that are actively in build. Each has:
- An owner (a person, not a team).
- A one-sentence success definition.
- A kill criterion (“if this isn’t in production by X, we kill it”).
- A rough cost band (5-figure / 6-figure / 7-figure).
That’s it. If you can’t fit an initiative into that template, it doesn’t go in the Shipping column. Either tighten the definition or move it to Next.
Next (next quarter, committed in principle)
Three to five initiatives we’re committed to scoping in the current quarter so they can move to Shipping next quarter. They have a sketch — not a plan — and an owner. They don’t have a budget yet. They might not survive scoping; that’s the point of putting them in Next instead of Shipping.
Watch (under evaluation, not committed)
Open-ended. Ideas, vendor pitches, internal proposals, technology shifts we’re tracking. A line per item. No commitment. The Watch column is where the wishful thinking lives, separated from the executable plan but not deleted. Items get promoted to Next when they earn it.
Why the format works
The three-column roadmap works because it forces three things that the five-tier roadmap actively prevents.
1. It forces honesty about commitment. Most “tier 1, year 1” items on a five-tier roadmap have no owner and no budget. They are aspirational. The three-column format makes you commit to exactly what’s shipping right now. Aspirational items go to Watch, where they belong.
2. It survives reality. Five-tier roadmaps assume the world holds still. The three-column roadmap is designed to be rewritten monthly without losing coherence. Things move from Watch to Next to Shipping. Things get killed and disappear. The document evolves with the program instead of resisting it.
3. It fits on one page. The whole roadmap is visible on one screen. The CFO can read it in two minutes. The board reviews it without a binder. Every monthly leadership meeting opens with it on screen. Compare this to a 40-slide deck that nobody opens between strategic offsites.
What you lose
The format does cost you two things, and both are worth paying.
You lose the illusion of a three-year plan. Five-tier roadmaps imply you know what year three looks like. You don’t. Nobody does. AI is moving too fast for a three-year plan to mean anything more specific than “AI will be important and we will continue to invest.” The three-column format is honest about that and frees you from defending a plan that’s going to be obsolete.
You lose the consulting deliverable shape. A three-column page is hard to charge $400k for. That is the actual reason most AI roadmaps are five-tier decks; the deliverable has to look expensive. The format is a billing artifact. Replace it.
Anatomy of a working three-column roadmap
Below is a redacted version of a real roadmap from a mid-market client, ~$80M revenue, in their second year of structured AI work.
Shipping (this quarter — Q1)
- Sales call summarization (production). Owner: VP Revenue Ops. Success: 90% of sales calls have a structured summary in CRM within 1 hour. Kill: if accuracy doesn’t reach 85% by week 8, pause. Cost: 6-figure annual.
- Customer support draft assistant. Owner: Head of CX. Success: agents accept the AI draft 60%+ of the time. Kill: if acceptance is below 30% by week 6. Cost: 5-figure annual.
- Internal knowledge agent (pilot in engineering). Owner: CTO. Success: at least 30 engineers use it weekly. Kill: if engagement is below 10 weekly users by week 8. Cost: 5-figure.
Next (Q2, committed to scoping now)
- Customer-facing chat with handoff (CX expansion).
- Outbound prospect research agent (sales tooling).
- Multi-language document translation (operations).
Watch
- Voice-based meeting transcription (vendor evaluations in progress).
- Internal forecasting model (data work prerequisite incomplete).
- AI-driven QA for product (capability not yet mature enough).
- Self-service contract review (legal hesitation).
That’s the entire document. It fits on one page. The whole company can read it. Every initiative has an owner, a definition of done, and a kill criterion. Every speculative thing is in Watch and isn’t pretending to be a commitment.
The monthly cadence
The roadmap is reviewed monthly in a 45-minute meeting:
- Shipping — status of each item, blockers, kill candidates.
- Next — what got scoped, what’s ready to promote.
- Watch — anything new, anything to promote, anything to delete.
- Adjustments — what moved between columns.
The first time a company runs this meeting it feels lightweight. By the third one, the team realizes they have more control over the program than they did with the 40-slide version. They also realize they’re killing initiatives earlier, which compounds.
What to do this week if you have a five-tier deck sitting around
Three steps:
- Identify what is actually in build right now. Be honest. Probably 2–4 things. Put them in Shipping.
- Identify what you’ve committed to scoping in the next 90 days. Probably 2–4 more. Put them in Next.
- Take everything else from the deck — and the dozens of ideas not on the deck — and put one-line entries in Watch.
You now have a one-page roadmap. Delete the deck. Schedule the monthly review. Refresh the document live in the meeting.
The deeper point
Strategic documents should be tools, not deliverables. A five-tier deck is a deliverable; it justifies the engagement that produced it. A three-column page is a tool; it helps the executive team make decisions every month for years. The shape of the document is the shape of the program. Choose the shape that makes the program possible.
Building the three-column roadmap is part of how a Fractional CAIO engagement starts. If you have a deck that’s not getting executed, schedule a call — we’ll often rewrite it on the line.
