Three weeks from now your executive team will sit down with the mid-year reforecast. If the cycle has worked, they will spend the meeting choosing between a few defensible options, each one traceable back to a leader who stands behind it. If it has not, they will spend it trying to find strategy inside a spreadsheet that was never built to carry any.
Which of those meetings they get is decided right now, three weeks before it happens, in the inbox of every function head you are about to send a reforecast template to. Every function head is making one decision this week: am I building a point of view, or am I filling in a model?
For most function heads in most companies, the honest answer is the second one. They were given a template, not a question. They did what the template asked.
Where strategy actually lives
Real strategy lives in the calls each function leader is making about where to push, where to hold, and where to admit the original January plan was wrong. The reforecast is where those calls get documented, defended, and rolled up into a coherent picture of the year. The model is how the choices get expressed, not where they get made.
When the cycle works, the executive team, in three weeks, is choosing between options that other people have already thought hard about. When it does not work, finance arrives at the meeting holding a stack of plausible-looking submissions and quietly reverse-engineering judgment that should have been there from the start. The math is correct. The pre-read writes itself. Nobody has decided anything.
This is what most reforecasts miss. The numbers come in plausible. The thinking does not come in at all, because the exercise never asked for it. A function head who is handed a model and a deadline fills in the model and hits the deadline. They will not bring a point of view to the exercise unless the exercise asks for one.
You can test for this in any function's first draft. Sit with the leader behind it and ask one question: Why did you produce this forecast? If the answer is a real argument, something like they are pausing two of the four initiatives in the original H2 plan and concentrating budget on the one where the payoff math has shifted since January, then the thinking is happening. If the answer is a version of last year's plan with this year's numbers stapled on, it is not.
Most function heads, asked honestly, will fail this test. The capability is there. The exercise just did not call for it.
What the tooling shift made possible
The reason it is worth taking this seriously now, rather than next cycle, is that the mechanical part of the work has changed.
What used to take an FP&A team two weeks of spreadsheet work takes an afternoon. Auto-forecasting handles the baseline. Scenario tooling handles the variations. The workload that used to absorb the entire cycle has moved off the critical path. Most companies have used the freed-up time to push more frequent cycles or to add more granularity to the same kind of model. The teams getting real value out of the shift are using it differently. They are buying time for the function heads, not for finance. They are giving the people closest to the work the room to actually choose, then asking finance to play the role of pressure-tester and synthesizer at the top.
That is the opportunity in this cycle. Not finishing a week earlier. Giving the function heads more room to think before they submit.
Three changes to how you run the cycle
None of them expensive. All of them counterintuitive once you have run reforecasts the old way for a few years.
Ask each team for three priorities before you give them a target. The instinct is to lead with the number. Here is what we need from your function, now build to it. The moment you do that, you have traded thinking for compliance. The leader's job becomes hitting the figure, not deciding what the function should do.
Instead, ask each leader to come back with three things that, if they get them right, define a good second half for their function. Just three. With the reasoning behind each. No top-down steer. What you get back is the most valuable signal in the cycle. You find out where each leader's real conviction sits. You find out which of them have a point of view at all. You find out where their view of the business diverges from yours, which is precisely the conversation worth having, and precisely the one a top-down target buries before it can happen. The act of naming three priorities is small, but it forces a choice. A choice is what the cycle has to manufacture.Make the model genuinely easy to play with. Strategic thinking dies the moment every “what if” costs two days of model surgery. If a leader has to rebuild half a tab to see what happens when they move an assumption, they will move it once, if at all, and then defend whatever came out. That kind of friction is fatal to the thinking. People model around it, not into it.
This is where the tooling shift pays off in practice. Auto-forecasting handles the mechanical part, which means a leader can spend their time on the judgment instead of the production. Change a conversion rate, shift a hiring ramp, pull a launch forward a quarter, see the full-year effect in seconds. When the cost of exploring drops to near zero, people explore. The strategic thinking you are asking for is realistic because the plumbing finally supports it.
Coach the thinking, do not grade the number. Once the drafts are in, the role of finance changes from checking that the math foots to pressure-testing the reasoning behind the math. Sit with each leader. Ask why, and then ask why again. Push on the assumption that is carrying the year and watch whether it holds up under a real question. The output you want from that conversation is a sharper argument, not a rounder number.
And do not split the difference to keep the peace. If sales believes the pipeline converts at 22% and finance modeled 18%, landing at 20 so everyone can move on is the worst available outcome. It averages out the one real disagreement in the room and teaches everyone in it that conviction does not matter. Pick a number, name the observable thing in the next 60 days that would prove it wrong, and revisit it. Coaching the thinking means helping the leader hold a defensible position, not negotiating them toward a comfortable one.
What you get when this works
When the drafts come back this way, with three priorities, defensible assumptions, and a couple of distinct scenarios per function, the job at the top of the cycle stops being invention and becomes synthesis. You are no longer reverse-engineering strategy out of a spreadsheet; you are assembling judgment that already exists into a few company-level options that hang together, and handing those to the executive team.
That changes what the meeting actually is. The executive team is choosing between reasoned paths, each one traceable back to a leader who stands behind it. They are doing the job only they can do, which is choosing, instead of the job that should have been done three levels below them.
The compounding part matters more than the first cycle does. Once every function leader has been through a reforecast that was actually run this way, the first draft of the next one already arrives with judgment in it. Each cycle starts further ahead than the last, because the muscle (naming priorities, building real scenarios, defending assumptions under pressure) is built and does not have to be rebuilt every six months. The mindset shift is expensive exactly once. The output gets sharper every cycle thereafter.
The trade is honest. This way of running the cycle takes more thinking than putting numbers into a spreadsheet and marking the task done. The tools have given you the time to make that trade. The only real question is what the team does with the time that freed up. They can use it to think. Or they can use it to plug numbers faster.
You have about three weeks before the executive room. The work this Sunday is the work that decides whether they have anything to actually choose between when the meeting starts. Everything else is downstream of that.
