The forecast was a vendor roadmap

Meta description: The strategy was anchored to a capability timeline that slipped — and no decision survived the slip. The future in the deck could move without a single decision changing. That was the tell.

The strategy deck had a future in it. I have read many of these now: a curve of AI capability, milestones dated by quarter, market projections in a confident band, and a strategy arranged around them — investments timed to the capability’s arrival, commitments justified by the projected market. Then the capability slipped, as capabilities do, and I watched what happened to the strategy: nothing. The deck was updated with a new timeline and the decisions carried on unchanged. A forecast that can slip without altering a single decision was never load-bearing. It was decoration, and the strategy had been built around a decoration.

[AWAITING MIKLOS — scene: a strategy anchored to a capability timeline that slipped, with no decision that survives the slip]

What is the forecast actually made of?

When I trace the forecasts in these decks to their sources, the genealogy is consistent: an analyst projection citing vendor statements, a consultancy’s market sizing built on adoption surveys, a capability timeline that mirrors, closely, a vendor’s roadmap. None of this is fabricated — it is worse than fabricated, it is recycled. The company’s independent-looking forecast is the industry’s promotional material, laundered through two intermediaries. The strategy team experiences it as external validation. It is the vendor’s plan, arriving in the boardroom wearing an analyst’s jacket.

Which decision depends on the timeline holding?

[AWAITING MIKLOS: forecast errors he has seen cost a planning cycle.]

This is the test I apply to every forecast-anchored strategy: name the decision that reverses if the timeline slips by half. In the strategies I am describing, there is none. The investments are justified by the future but not conditioned on it — they would be made anyway, for adjacent reasons; the commitments are directional, survivable in any timeline. This sounds prudent and is actually diagnostic. A real forecast-driven bet has a tripwire: if the future does not arrive, we stop, reverse, or double down elsewhere. A decorative forecast has no tripwire, because nothing was staked on it. The future was in the deck to make the deck feel like strategy.

Why does updating the slide preserve the error?

The false explanation is forecast hygiene: refresh the projections, update the timeline, keep the strategy current. I have watched the slide get refreshed annually while the underlying relationship — strategy resting on an unexamined future — stayed intact. Updating the forecast treats the problem as stale numbers. The problem is structural: the organization has no practice of asking what a forecast is for, which decisions hang from it, and what evidence would change those decisions. Without that practice, every refresh replaces one decoration with a newer decoration, and the strategy continues to be rationale in search of a future.

What pattern marks a bad picture versus ordinary uncertainty?

Ordinary uncertainty is honest: the future is unclear, the strategy holds options, the decisions are reversible where possible. A bad picture is confident in a specific way I have learned to recognize: it is confident exactly where confidence is convenient. The capabilities the strategy needs are projected to arrive on the strategy’s schedule; the adoption the business case needs is projected at the business case’s rate. When the forecast’s confidence aligns perfectly with the plan’s requirements, the forecast was not made; it was commissioned — sometimes by a vendor, more often by the plan itself.

[AWAITING MIKLOS: the timeline scene.]

[AWAITING MIKLOS — scene: second scene]

Miklos Roth

What question separates a bet from a decoration?

Forecast conversations ask whether the projection is right — an unanswerable question that generates reassuring debate. The separating question asks what the projection is doing in the strategy: which decision would change if it were wrong, and when would we know? I have watched that question sort an entire strategy offsite in an afternoon — real bets on one side, suddenly visible as bets, with owners and tripwires; decorations on the other, suddenly quiet. The future did not get more predictable. The strategy got more honest.

[AWAITING MIKLOS: the question, and why it is the hinge]

PR excerpt: [PR — awaiting Miklos material; cut 80–120 words from his paragraphs once available.]

By Miklos Roth — AI marketing consultant, Budapest. More: his profile.

© Copyright sikerdijaskeresooptimalizalas