Planning for performance when the cycle moves in five directions at once

A divisional planning review in an industrial group usually runs business area by business area. Trucks reports. Construction equipment reports. Buses, the engine business, the financial services arm, each in turn. Every set of numbers is defended, every forecast is reasonable on its own terms. And then someone has to add them up into a single group outlook the CFO can stand behind in front of analysts, and that is where the exercise quietly breaks.

It breaks because the divisions are not moving together. They never really were, but in a stable economy the divergence was small enough to absorb. In 2026 it is not.

A recovery that arrives unevenly

The clearest illustration is freight. ACT Research describes an industry that enters 2026 in a stronger position than it held a year ago, though the environment remains defined by transition rather than acceleration, with freight demand still uneven across sectors. The recovery is real, but it is not arriving everywhere at the same time. Class 8 demand has strengthened meaningfully as fleets respond to better freight conditions and replacement needs, while medium-duty markets remain softer and more tied to housing and small-business activity.

For a single-product company that is a forecasting challenge. For a multi-business industrial group it is a structural one. One business area is planning for a ramp while another is still planning for under-absorption. The group cannot pick a single macro assumption and cascade it, because no single assumption is true across the portfolio. And it cannot simply let each division plan in isolation, because the divisions share suppliers, share plants in some regions, share a balance sheet, and answer to one set of analysts.

This is the planning problem that defines industrial groups right now. Not a downturn, and not a recovery. A cycle that has fragmented, moving at different speeds through different parts of the same company, while the group is still expected to produce one coherent number.

Where the group view falls behind

Three things tend to go wrong, and they compound.

The first is consolidation lag. When five business areas plan in their own models, on their own cycles, with their own definitions, assembling a group view is a manual exercise. By the time the consolidation is finished it describes a position that has already moved. A mid-quarter tariff change or a currency swing lands days or weeks before the group model can reflect it, so the board is briefed on a picture that is structurally out of date. Not wrong when it was built, just no longer current.

The second is the flattening reflex. Faced with that lag, groups often respond by imposing uniformity, one set of assumptions, one template, one cadence, applied across every division. It makes consolidation faster. It also makes the output less true, because it strips out exactly the divisional nuance that mattered, the reason trucks and construction equipment needed different assumptions in the first place. The group buys speed by sacrificing accuracy, and then has to caveat the result anyway.

The third is scenario poverty. A group navigating an uneven cycle needs to test interactions, what a freight pickup in one region does to a plant that also serves a softer division, what a tariff shift does to a shared supplier base. Most planning environments can model one variable at a time, in one division at a time. They cannot model the portfolio responding to several moving inputs together, which is the only scenario that actually reflects the year.

What groups that hold the line are doing differently

The industrial groups that produce a credible through-cycle forecast, rather than a defensible one, share a structural characteristic. They plan in a connected layer that sits above the divisional systems rather than replacing them.

Each business area keeps its own planning logic, its own drivers, its own view of its own market, because that local accuracy is the point. But those divisional plans feed a common group model continuously, so consolidation is not a month-end assembly job, it is a live state. When an assumption changes in one division, the group view updates with it, and the variance is visible immediately rather than discovered in the next quarterly pack.

That connected layer is also where genuine scenario work becomes possible. Because the divisions are modelled in one place, the group can test how the portfolio behaves when several inputs move at once, the interaction effects, not just the isolated ones.
Three questions tend to locate where a group actually stands.

1. How long does a full group consolidation take, and is that fast enough to act on the result before it moves?
2. When an assumption changes inside one division, who learns about it, and how long after?
3. Can the group model two business areas responding to the same shock together, or only one at a time?

The answers usually point to the same place: not a strategy problem, a planning infrastructure one.

Where to start

None of this is solved by a system alone, and it is rarely solved all at once. The groups that close the gap tend to start narrow, with one consolidation cycle or one cross-divisional scenario, and prove the connected approach there before extending it.

That first step is worth a conversation. We work with finance and supply chain leaders in industrial groups on exactly this problem, and we are always interested in how different groups are handling an uneven cycle, because no two portfolios diverge in quite the same way.

What to read and do next

If the picture in this piece is recognisable, we would value the exchange as much as you might. You can reach us at info@bedfordconsulting.com, or follow Bedford Consulting on LinkedIn for the next piece in this series on connected planning in automotive and industrial manufacturing.