Aiming a budget cut

A reduction arrives late in the planning cycle. The group has committed to a three year investment envelope, EUR 48.0M across the remaining two years, spread over four business units and eight workstreams. Finance is now told to find twenty percent of it. That is EUR 9.6M to remove from a plan that has already been approved, staffed and in places signed into contracts.

The instinct in the room is always to spread it. Twenty percent off every line is fast, and nobody can accuse you of favouritism. It also requires no analysis. In most portfolios I have worked on, it is the most expensive way to absorb a cut.

The figures and business units below are invented, but the method is the one I use.

A flat percentage assumes every euro in the portfolio is equally removable. It never is. Some lines are signed contracts where cancellation costs more than delivery. Some are the compliance work that keeps a licence to operate, or the maintenance that keeps a plant running. Others are genuinely discretionary. An even cut treats all of them the same way, so it lands hardest wherever the money was least free to move.

Before deciding anything, look at what each unit's budget is made of.

The four units are not comparable. Industrial Operations is dominated by contractual and business critical work, so a flat cut there is not a reduction in ambition. It is a breach negotiation. Commercial sits at the other end, with a third of its envelope in value adding work that can be deferred without anything breaking this year. Applying the same percentage to both is arbitrary, and it quietly moves pain from the unit with slack to the unit without any.

The mechanism is unglamorous. Every budget line gets a tier, and each tier has one rule a budget holder can apply without calling a meeting. The rule has to test consequence rather than enthusiasm. Ask people how important their work is and you get a portfolio where everything is tier one.

Applied, it looks like this. Below is one workstream as it sits in the file, every line carrying a tier alongside its cost element, its spend by year and its share of the workstream. Retag a line and the rollup underneath moves, which is the entire mechanism at the scale of a single screen.

The split between tier 1 and tier 1.a matters more than it looks. Both are contractual, but only one is immovable. A commitment that can be pushed into a later year is not a saving in the strict sense, since the money is still owed, but it relieves the current envelope. Collapsing it into tier 1 throws that flexibility away, and most of the negotiating room in a portfolio sits there.

One discipline holds this together: assign the tier before anyone knows the size of the cut. Tag the portfolio while the number is still unknown and the classification is honest. Tag it afterwards and the tiers reverse engineer themselves into whatever protects the current plan.

With the portfolio tagged, absorbing the reduction stops being a negotiation and becomes arithmetic. Work up the ladder from the bottom. Exhaust the deferrable tier, take what is still needed from the value adding one, and stop as soon as the target is met.

The whole EUR 9.6M comes out of tiers 4 and 3. Tier 4 gives up everything it has and tier 3 covers the remainder, keeping about sixty percent of its budget intact. Contractual and business critical work never moves, so the group carries no penalty exposure and no compliance gap, and the resulting EUR 38.4M plan can be defended line by line.

The cut is still not painless. EUR 9.6M of real work disappears and the deferred workstreams have owners who will feel it. What changes is that the pain follows a stated rule instead of following whoever argued hardest in the room. That is a much easier conversation to have twice.

The waterfall says how much. It does not say who. Rolling the same tagged lines up by workstream shows where the portfolio has give and where it has none.

Two shapes matter here. Regulatory and compliance is almost entirely tiers 1 and 2 with nothing at all in tier 4, so there is no discretionary money to find and pressing it produces excuses rather than savings. Commercial expansion and Facilities and workplace sit at the other end, with more than two thirds of each in the deferrable half of the ladder. Those two absorb a disproportionate share of the cut, which is the correct outcome rather than a failure of fairness.

This is also the view that survives an executive committee. Telling someone their workstream is being cut invites a fight. Showing them that it is seventy percent deferrable while the compliance workstream is four percent moves the discussion onto ground where the numbers do the arguing.

A model that cannot be overruled gets ignored the first time it is wrong, so the tool carries two deliberate escape hatches. The first is the retagging in the ledger above, available to a budget holder without a rebuild, and marked as changed the moment they use it. The second is a qualitative adjustment column, a free hand override applied to a rollup rather than a line, with a reason attached.

Both exist to keep the override visible. Without them the same overrides still happen, silently, in the version of the file someone edits the night before the meeting, and by the third revision nobody can reconstruct which numbers were computed and which were negotiated. Making the adjustment a column rather than an edit keeps the arithmetic reproducible and the judgement attributable.