The Roadmap You Will Not Follow: What a Technology Plan Is Actually For
Most technology roadmaps are approved in the autumn and overtaken by the spring, and everyone involved knows it. The problem is not the planning horizon. It is that a plan listing everything has decided nothing, and left the decision to March.
Most technology roadmaps are approved in the autumn and overtaken by the spring. The people who wrote it know this. So, generally, do the people who signed it.
That is not a planning failure. It is a sign the document is being asked to do something it cannot do.
1. The Annual Plan Is Not Pointless
There is a fashionable view that annual planning is obsolete, and that organisations should work in rolling quarters instead. It underrates what the plan is for.
Funding is released against it. Suppliers are engaged on it. People are recruited to it — and recruitment, notice periods and onboarding together take a good deal longer than a quarter.
No organisation of any size runs on intent alone. Something has to be written down far enough ahead that the money and the people turn up in time to be useful.
2. It Is Built to Be Approved
The difficulty is what happens to any document that has to clear a board. It grows, and it grows in one direction only.
Each function adds the item it cannot be seen to have left out. Very little is refused, because refusing costs more than including — and the cost of including is not paid in that room.
What arrives at the meeting is a complete list of everything anybody wants — arranged by quarter, and approved, because there is nothing in it to object to.
3. Shorter Cycles Do Not Fix It
The usual remedy is to plan in smaller increments. Quarterly, rolling, continuously reviewed.
That improves accuracy and leaves the real problem untouched. A quarterly list of everything is still a list of everything.
Replanning more often mostly means having the same argument four times a year rather than once — and calling the repetition responsiveness.
4. The Currency Is Capacity, Not Money
Money can be found mid-year. It moves between budgets, comes out of contingency, arrives from an underspend somewhere else.
Capacity does not behave that way. Behind every line on the plan sit the same architects, the same testers, the same two people who understand the integration layer — and no amount of approved budget creates a third.
A plan approved on cost has not been tested against the thing that actually constrains it. Funding is elastic. The people are not.
5. Somebody Will Prioritise It Anyway
A list containing everything has not decided anything. It has only postponed the decision, and moved it somewhere less visible.
That decision gets taken in March, when two items collide — usually well down the organisation, under time pressure, by somebody with no sight of what the board actually cared about.
Which is not a failure of nerve on their part. Nobody gave them the order of preference, so they used the only signal available, which is whoever is shouting.
6. Leave Some of It Unallocated
Most plans commit the whole of the available capacity, which guarantees that anything unforeseen will displace something already promised to somebody.
A plan committing perhaps two-thirds, with the remainder named and held back, looks unambitious in the approval meeting — and survives contact with the year considerably better.
The reserve is not slack, and it should not be described that way to a board. It is what absorbs the work nobody could reasonably have known about in October.
7. What a Useful Plan Records
The list of initiatives is the least useful part of the document. Four other things matter more, and none of them needs a tool:
- which few items will be protected when something urgent arrives
- what has been deliberately deferred, and who has been told
- which people the plan depends on, and what else they are already committed to
- how much capacity is held back, and who is allowed to release it
The second is the one that gets skipped, because writing it down means somebody has to be told no in October rather than discovering it in June.
8. Protect a Few Things Properly
An organisation that protects three things and allows the rest to move will finish the three, and will know in advance which three they were.
One that protects twenty finishes whichever happened not to be interrupted, which is a different thing entirely and not something anybody chose.
The plan is not a forecast of what will happen. It is a statement of what will be defended when the year does what years reliably do.
Final Thought
Roadmaps are usually judged by how much of the list was delivered — a measure that rewards the plans which promised least, and tells nobody anything useful about the year.
The better question is what the organisation still managed to protect once the year had rearranged itself. That, rather than the document approved in the autumn, is what it actually chose.