Predictive planning and scenario planning share a spine. Scenario planning came out of Shell in the early 1970s, and the academic method Thomas Chermack later wrote down is the version most planners learned. Predictive planning is what that method turns into when the inputs go live, the cost of a scenario collapses, and the operating environment starts moving faster than the calendar that was supposed to track it. Same logic. Different cadence. If you have run a scenario project before, you are not starting over. You are a quarter of the way there.
What scenario planning was built to do
The original job was rehearsal. Pierre Wack’s team at Shell did not predict the 1973 oil shock. They wrote a world in which it happened, walked the leadership through what they would do inside that world, and when the shock arrived Shell moved while its peers were still arguing about whether it was real. That is the whole value of a scenario: a decision you have already rehearsed is a decision you can make quickly.
The method that grew out of that work has a recognizable shape. Identify the driving forces. Pick the two or three uncertainties that matter most. Build a small set of plausible, distinct futures on those cleavages. Name them. Rehearse the strategy against each. Chermack’s contribution was to make that repeatable inside an organization, with phases, a team, a facilitator, and a deliverable. It is rigorous and it works.
It was also built around two constraints that no longer hold. It was expensive, so it ran as a project every few years rather than as an operating rhythm. And it was expert-dependent, so the scenarios lived with the people who knew how to build them and sat on a shelf between uses. Neither constraint was a flaw. They were the honest response to a world where a scenario took six weeks and a strategic assumption held for years.
Three things changed underneath the method
The first is cost. What used to take a strategy team six weeks, gathering data, modeling assumptions, pressure-testing, writing up, now takes a senior operator with the right tools an afternoon. Not a publishable afternoon. A useful one. When the unit cost of a scenario drops by an order of magnitude, the frequency at which scenarios are worth running changes with it.
The second is that the signals went live. Fed decisions, mortgage rates, jobless claims, credit spreads, all of it publishes daily or weekly and is free. Prediction markets price the events in between. Chapter 3 of the book makes the case that market-implied probability from Kalshi and Polymarket is a checkable input to a scenario set, not a curiosity. In 1973 the hardest part of scenario work was finding the signal. In 2026 the hardest part is deciding which signals to ignore.
The third is velocity. The half-life of a strategic assumption has compressed across every dimension that matters. The Fed raised its target ceiling on September 17 while most 2027 plans were being drafted, and by November the market had a December hike priced near certainty. A scenario set built in 2024 and reviewed in 2027 would miss that whole move. The episodic cadence was a response to cost and scarcity. Remove the constraints and the cadence has to go too.
What predictive planning keeps
All of the scenario logic. A small set of named worlds, three to five, built on real cleavages, not good-base-bad. Indicator sets for each one, so a scenario has consequences instead of just a narrative. The rehearsal, so the decision is pre-made before the pressure arrives. The refusal to forecast a point when a range is the honest answer. None of that is being thrown out, and anyone who tells you predictive planning makes scenario planning obsolete has not read either carefully.
The lineage is on the record. The man who wrote the standard academic text on scenario planning wrote the foreword to the book that argues for its successor. That is not a coincidence. It is the method acknowledging its own next version.
What predictive planning changes
The discipline runs as a four-phase loop, Woodring’s Loop: Scan, Story, Stake, Steer. Scenario planning is Story. Predictive planning puts a continuous Scan in front of it and a Stake and a Steer behind it, and then refuses to let the loop close. Scan runs weekly. Story is read monthly. Stake is committed quarterly. Steer corrects every couple of weeks. The scenarios are the same kind of artifact. What is new is everything around them.
The most important addition is the forced commitment. Classic scenario work ends with scenarios. The organization now “has” them, and a surprising share of the time it does nothing with them. Predictive planning ends every revolution with a sized stake, using the Stake Sizing Matrix to match the size of a commitment to the cost of being wrong, and then puts every active scenario in front of the people with budget authority once a month at a Convergence Review, where each one gets sorted into converging, holding, or collapsing and the stakes move accordingly.
The second change is ownership. Scenario planning belonged to the facilitator and the strategy function. Predictive planning has a chair. The Predictionist runs the Loop on behalf of the leadership team, and each phase has a named owner with a cadence. The discipline survives a leadership change because the roles do, not because the last consultant left good slides.
The third change is the input layer. Each scenario carries public signals that would confirm or kill it, and where a prediction market exists for the event, the market’s price gets attached as the crowd’s current read. A scenario that assumes the Fed holds in December is a scenario the market currently gives about a 4% chance. You can still hold it. You cannot hold it without knowing that.
The honest side-by-side
- Cadence. Scenario planning: a project every few years. Predictive planning: a loop that never closes, with weekly, monthly, and quarterly beats.
- Trigger. Scenario planning starts when leadership decides it is time. Predictive planning is always running; a signal moving is the trigger.
- Owner. A facilitator and a project team versus a standing role with phase owners.
- Output. A set of scenarios versus a sized commitment and a steering cadence.
- Inputs. Expert interviews and research versus a live signal layer plus the same expert judgment plus market-implied probability.
- Between cycles. Nothing versus a monthly Convergence Review.
Why successor and not replacement
If your organization ran a scenario project in the last three years, the worst thing you could do is throw the worlds out. They are probably still structurally right, because good cleavages age well even when the numbers inside them do not. What they are missing is indicator sets, attached signals, a stake sized to each, and a standing meeting where someone has to say which one is converging. Add those four things and you have not replaced scenario planning. You have put it on the cadence the world now runs at.
What to do Monday
Pull the last scenario set your organization built. For each world, write three indicators you would expect to see in the next sixty days if it were converging, and attach a public series to each one from the Signal Library. Where a prediction market prices one of your cleavages, note the current price from the market layer. Book a ninety-minute Convergence Review for the first week of December with the people who control budget. That is one revolution of the Loop built on work you already paid for.
If you want to know how far your current process is from running continuously, the Maturity Read takes about six minutes. The full method, including how the Loop retires the annual cycle it replaces, is in the book.