Lifetime Mission As Service Assessment

Setting Up a Lifetime Mission As Service Assessment

I ran into this properly back in 2022 when a client wanted to shift their industrial equipment division toward a outcomes-based pricing model. They kept asking for a "lifetime mission as a service" framework, and nobody on the team actually knew how to price it. So we built one. Here is what I learned doing it three times since then. It is a structured evaluation that estimates the total cost, reliability, and performance degradation of delivering a mission-critical outcome over the full expected lifespan of the asset, rather than just selling the asset itself. You are pricing the result, not the hardware. That means you need failure data, spare part lead times, energy consumption curves, and technician labor rates across at least a 10 to 15 year window. Anything shorter and you are just guessing with extra steps. The first step most people skip is gathering historical mean time between failures for the specific equipment in the actual operating environment. Factory floor MTBF numbers are useless here. I found that on a packaging line project in Rotterdam, the published MTBF was 2,400 hours and the real world number was 680 hours because the ambient humidity and product abrasiveness doubled the wear. Using the published number would have underpriced the service contract by about 31 percent. We ended up using field service logs from three comparable sites instead.

Once you have reliable failure data, you build a degradation curve. This is not a linear decline. Most mechanical and electrical systems follow a bathtub curve with an early burn-in period, a stable operational phase, and then accelerated wear. You model each phase separately. I use a simple Weibull distribution fitted to the failure data, then layer in seasonal variations if the asset operates outdoors or in uncontrolled environments. Next comes the cost layering. You need part replacement costs at year 3, year 7, and year 12, not just current pricing. I keep a rolling 12 month log of actual part invoices because supplier price changes eat margins faster than failures do. On one contract I forgot to index for a motor supplier raising prices by 18 percent in a single quarter. That cost us roughly 9 percent of the projected margin on a four year deal. After the cost model, you calculate the service delivery overhead. This includes scheduled maintenance visits, remote monitoring subscriptions, spare parts inventory carrying costs, and the probability-weighted cost of emergency callouts. The callout part is where people lose money. A breakdown at 2 AM on a Sunday costs about 2.3 times a standard weekday visit once you factor in overtime rates and rush shipping on parts.

Lifetime Mission As Service Assessment in Practice

Here is the section that does not get written about anywhere. The customer acceptance criteria. When you sell a mission-as-a-service contract, the customer will define "mission accomplished" differently than you will. I worked on a compressed air system deal where the customer guaranteed minimum output pressure of 7 bar at the point of use. My model assumed 8 bar at the compressor outlet accounting for pipeline losses. The customer had undersized piping on two of their three production lines, so the actual delivered pressure was 6.8 bar during peak demand. The contract had a penalty clause tied to that 7 bar threshold, and we breached it for about 14 percent of operating hours in the first six months. We fixed it by adding a small buffer tank at the critical usage points, which cost 4,200 euros and eliminated the penalty exposure. Always validate the customer end-to-end system configuration before you sign. Their infrastructure assumptions are not your problem until the contract says they are. Over-relying on OEM reliability data. Manufacturers publish optimistic figures under ideal conditions. Cross-reference with independent field data whenever possible. If you cannot get field data, apply a conservative derating factor of at least 0.65 to the published MTBF. That is a rough heuristic but it keeps you from being embarrassed. Ignoring supply chain lead time on replacement parts. A bearing that takes 6 weeks to arrive from a specialized supplier in Germany is not the same risk as one you can pull off the shelf. Downtime cost multiplied by lead time is a real line item in your assessment. I started tracking supplier lead time variability as a standard input after a 2023 chip shortage made a $200 control board take 14 weeks to replace a $2,000 drive unit that would have been available in 3 days.

Underestimating the administrative overhead of long-term contracts. A five year service agreement requires quarterly business reviews, annual part price renegotiations, and change order management. I budget about 40 hours per year per contract for administrative work that never shows up in the technical model. On a ten contract portfolio that was roughly 400 hours annually, or about 10 percent of what a dedicated account manager's time costs.

When This Approach Fails

Lifetime Mission As Service Assessment does not work well for assets with highly variable usage patterns. If the customer runs their equipment 24 hours a day in year one and 8 hours a day in year two because of market conditions, your degradation model is wrong and your pricing is wrong with it. I recommend switching to a usage-based billing model instead when utilization variability exceeds plus or minus 40 percent from the baseline. The math gets simpler and the risk stays with the customer where it belongs in that scenario. It also breaks down for technologies that are on the edge of obsolescence. If a new generation of equipment is likely within three to five years, locking into a long service contract gives you downside risk without upside potential. The customer will want to upgrade and you will be stuck maintaining aging hardware under the old contract terms. In those cases, shorter assessment windows with renewal options are safer.

Quick Reference

Gather field-level failure data, not factory specs. Model degradation in phases using Weibull or equivalent. Index part costs annually. Include callout cost multipliers for off-hours service. Validate the customer end-use configuration. Track supplier lead time variability. Budget administrative overhead separately. Switch to usage-based pricing when utilization varies more than 40 percent. Avoid long contracts on near-obsolescence technology. I keep a spreadsheet template that covers all of this and it takes about two hours to fill out for a standard asset after the first few assessments. The data collection is the time sink, not the modeling. If your client can provide SCADA logs or maintenance histories, you are mostly just cleaning and organizing. If they cannot, you are back to estimating and nobody likes that part.