Analytics for industrial operations and distribution
Spare parts and MRO inventory, long and variable lead times, and project-based demand that defeats standard forecasting.
Demand that does not behave
Industrial operations — fabrication, distribution, equipment, chemicals, building products — share a planning problem that standard methods handle badly. Demand is lumpy, lead times are long, and the cost of a stockout is often downtime rather than a lost sale.
- Spare parts and MRO inventory where most items move a few times a year, making conventional forecasting useless and judgement-based stocking expensive.
- Long and variable lead times on imported or fabricated items, where the decision has to be made months before the requirement is known.
- Project-based demand that arrives in large blocks and does not resemble a time series at all.
- Supplier concentration, where a handful of single-sourced items carry risk out of proportion to their spend.
Where we usually start
-
Criticality-weighted stocking policy for spares and MRO, using methods suited to intermittent demand.
-
Replenishment and expediting recommendations that apply lead time, criticality and supplier performance consistently.
-
A pipeline-based requirement plan for project demand, rather than forecasting it as if it were steady.
-
Resolve the duplicate part numbers and inconsistent units that make an MRO catalogue hard to analyse.
Stocking decisions for critical spares
The question for a critical spare is not what demand will be. It is whether the cost of holding one is less than the expected cost of not having it — which depends on downtime cost, lead time and failure probability rather than on sales history.
That makes it a decision problem, not a forecasting problem, and it needs your maintenance team's knowledge as a direct input. We structure that judgement rather than replacing it, so the policy is consistent across thousands of parts instead of varying with whoever last reviewed them.
Common questions
Most of our spare parts barely move. Can they be forecast?
Not as a point forecast, and anyone promising one is overselling. Intermittent demand needs methods built for it — Croston's method and its variants — and the practical output is a stocking policy rather than a monthly number: whether to hold the part at all, how many, and where.
How do you decide which spares to stock?
By weighing the cost of holding against the cost of not having it, which for a critical spare is downtime rather than a lost sale. That makes criticality the dominant input. We work through it with your maintenance team, because the consequence of a stockout is operational knowledge rather than something in the data.
Our demand is project-driven and lumpy. Does any of this apply?
Yes, but the approach changes. Project demand is better handled as a pipeline than as a time series: known projects with probabilities and dates, feeding a requirement plan. Forecasting it like steady demand produces confident nonsense.
Can you quantify supplier concentration risk?
Yes. Mapping spend and criticality by supplier usually shows that a small number of single-sourced items carry most of the exposure. That list is actionable — qualify a second source, hold more, or accept it knowingly rather than by default.
Find out what your data can already tell you
A free 30-minute data readiness assessment. We look at what you have, tell you what it can support today, and name the first thing worth building. No preparation needed.