The question of how far a roadmap should commit is really a question about honesty. Commit too far in detail and you are publishing fiction that everyone will hold you to. Commit too little and you have given the organisation no direction to align behind. Both extremes are common, and both erode trust.
The principle I work to is that certainty and distance are inversely related, and a good roadmap shows that openly. The near term should be concrete. Specific outcomes, real dates, named owners. The middle distance should be directional. The outcomes you are aiming for, with the sequence clear but the detail deliberately soft. The far term should be intent, not commitment. Where you are heading, and why, without pretending you can see the path.
The mistake that does the most damage is presenting all three horizons with the same confidence. When a roadmap shows a date eighteen months out in the same firm font as a date next month, it invites people to plan against something that was always a guess. Then, when it moves, as it always will, the roadmap loses credibility and so does the person who owns it.
So I am explicit about the confidence attached to each horizon. It is better to say "this is our direction, not a commitment" out loud than to let a slide imply a promise you cannot keep. Stakeholders can handle uncertainty when it is named. What they cannot forgive is certainty that turns out to have been false.
A roadmap should also have a rhythm for changing. A plan that never moves is being ignored. A plan that moves constantly was never trusted. Somewhere between those is a roadmap that is reviewed on a known cadence, where changes are deliberate and explained, and where the act of updating it is normal rather than a crisis.
For an executive, the signal to watch is not how far the roadmap reaches. It is whether the person presenting it is honest about how firm each part really is. A leader who tells you exactly where the confidence ends is more reliable than one who promises the whole horizon in detail.