Halden Industrial: servitisation, risk appetite and a three-horizon portfolio

Viability rests on capability, not intent. An outcome-based contract transfers availability risk to Halden. Halden can only price and carry that risk if it can (a) sense asset condition, (b) predict failure with enough lead time to act, and (c) dispatch service at a cost it controls. Today the evidence is thin: remote monitoring covers 400 pumps out of an installed base that is presumably in the tens of thousands, and the prediction work is a proof of concept with no owner.

The hardest constraint is measurement, not technology. The service division is measured on contract margin. A subscription model is measured on availability and lifetime value. If the internal scorecard rewards the opposite of the strategy, the strategy will not execute - managers will rationally protect margin by avoiding risk-bearing contracts.

What must be true:

1. Installed-base data is good enough to underwrite availability - else the contract is an unhedged bet.
2. Pricing capability exists for deferred cash and risk transfer (or a partner/insurer absorbs part of it).
3. Service delivery can meet 99.2% at a defined cost, demonstrated on a bounded cohort first.
4. Capital allocation is willing to fund deferred cash flow through the transition - this is a multi-year earnings drag, not a cost saving.

Judgement: the direction is defensible because the market is moving and equipment margin is exposed. The pace is the risk. A 30%-by-year-four target set before any cohort has run under an availability contract is a target without a baseline; the recoverable position is to commit to the direction, prove it on a segment, and re-set the percentage against evidence.

First, resolve the capacity question honestly. The engineering function is firefighting a safety-relevant recall. Any portfolio that assumes that capacity is available is fiction. The recall should be recognised as a protected, funded workstream that displaces discretionary change for its duration, with an explicit date at which capacity returns.

Second, stop treating six pilots as six initiatives. They are fragments of one capability: sense asset condition → predict failure → act. Consolidate into a single programme with one accountable owner and one benefit hypothesis (e.g., 'reduce unplanned outage hours on the monitored cohort by X%, enabling an availability guarantee'). Two abandoned pilots are evidence of orphaned investment, not of failure - kill or fold them formally so the portfolio reflects reality.

Third, segment the portfolio by horizon and govern each differently:

• Run (recall, existing service contracts): risk and compliance metrics, weekly.
• Grow (extend monitoring across the installed base, portal): standard business case, progress against adoption and cost metrics.
• Transform (outcome-based contracts): staged funding, decision gates tied to evidence - a first cohort of contracts with measured availability, priced risk and a validated cost-to-serve. Do not fund the 30% target as a single commitment.

Fourth, fix the scorecard. Until the service division is measured on availability and contract lifetime value alongside margin, the portfolio and the incentive system point in opposite directions and the portfolio will lose quietly.

A useful appetite statement is specific enough to make a decision reversible. It should contain four parts.

1. Boundary - what risk we will not take. Halden will not write an availability guarantee on an asset class where condition data coverage is below a defined threshold, nor exceed a stated maximum exposure to penalty payments in any one contract. This converts the risk conversation into a pre-agreed limit rather than a case-by-case argument.

2. Quantum and horizon of acceptable loss. The board accepts a defined cumulative earnings drag and balance-sheet exposure over the transition period in exchange for a defined subscription revenue position. Naming the acceptable loss is what makes deferred cash a decision rather than a surprise.

3. Evidence required before scaling. Approval to move beyond a pilot cohort requires demonstrated availability performance, cost-to-serve within tolerance, and a priced risk model validated against actual claims/penalties. In other words, appetite is graduated by evidence, not granted once.

4. Accountability and tolerance triggers. A single executive owns the outcome-based P&L; the board reviews availability against guarantee, penalty exposure, and cash deferral quarterly. Pre-agreed triggers - e.g., availability shortfall on the cohort, or penalty exposure above the stated cap - force escalation or rollback.

Why this shape: it separates the strategic risk the board is deliberately taking (deferred cash, transferred performance risk) from uncontrolled risk (writing guarantees on unmonitored assets). It also gives the finance director a lever: the strategy can proceed on a small scale while the risk framework matures, which is normally the only way such a proposal gets approved at all.