Search this question and most of what comes back is about cloud computing: site reliability engineering for software teams, not for a plant that breaks. For the person actually holding the decision, a maintenance manager with a resignation letter on the desk, a role that has been open for five months, or a budget round that will fund exactly one answer, the heavy-industry version deserves a straight treatment.
Here it is, from a firm that sits on all three sides of the decision. We hire reliability engineers, we run consulting campaigns, and we operate a managed reliability service. Each option is genuinely the right answer somewhere. Each fails somewhere specific. The expensive mistake is buying one when your situation calls for another.
Quick Links
The short answer
Hire in-house when the reliability work is continuous, the site can fill the seat, and the role will be protected from breakdown duty. Buy a consulting campaign when the work is a defined transformation (a strategy build, an FMECA program, an operational readiness scope) with a start and an end. Choose a managed service when the work is continuous but the seat will not fill, will not stay filled, or cannot be justified full-time. Mature sites usually end up combining them.
The three options at a glance
| Dimension | In-house engineer | Consulting campaign | Managed service |
|---|---|---|---|
| Best for | Continuous reliability work on a site that can protect the role | Defined transformations with a start and an end | Continuous work where the seat will not fill or cannot be justified |
| Time to capability | Months to recruit, then ramp-up | Weeks to scope, then the campaign runs | Weeks to stand up |
| Cost shape | Salary plus on-costs, fixed | Project-priced, finite | Subscription, scales with scope |
| Budget it comes from | Headcount, which needs an establishment change | Project or capital budget | Maintenance services, with no establishment change |
| Depth on tap | One person’s specialisms | A team, deep, for the project’s length | A bench across the toolkit, on a cadence |
| When it ends | Resignation takes the knowledge with it | You keep the report; the capability leaves at handover | The capability is the contract; context stays |
| Fails when | The role drifts into breakdown duty | Nobody owns the follow-through after handover | The site wants a person in the room at 6am |
Where the in-house engineer wins
In-house wins on presence and authority, and nothing else matches it. A good resident reliability engineer is in the pre-start meeting, hears the pump that sounds wrong, and carries site relationships no outsider can duplicate. When the role is set up properly, with the analytical core protected and the boundaries against planner and firefighter drift held, it is the strongest long-term answer for a large site with continuous reliability work.
The honest caveats are structural, not personal. Everything rides on one person: their specialisms are the site’s specialisms, their resignation restarts the whole build, and Australian sites routinely wait months to fill the seat, with remote operations waiting longest of all. The cost question is also rarely run honestly. Most operations cost the salary and never the vacancy, the ramp-up, or the fraction of paid hours that actually go to reliability engineering once the breakdowns start pulling. We have published the full hourly-cost comparison on the Reliability Operating Centre page; run it against your own numbers before deciding anything.
Verdict: the right answer for continuous work, if you can fill the seat and protect it.
Where the consulting campaign wins
Consulting wins when the work is a project, not a function. A maintenance strategy build across an asset fleet, an FMECA program, a readiness scope before first ore, a reliability assessment with a hard deliverable. These have a start, an end and a defined output, and they deserve a team rather than one person: facilitators, modellers and data builders working in parallel, at a depth no single hire carries. That is what reliability engineering consultants are for.
The failure mode is just as specific: the campaign ends. If nobody on site owns the follow-through, the strategy decays, the register drifts, and three years later the next consultant is quoting the same scope again. A consulting project hands you the report and the system. It cannot hold the standing responsibility for what happens next. Buy campaigns for transformations. Do not buy them as a slow, expensive substitute for a function.
Verdict: unbeatable for defined scopes, wrong shape for continuous work.
Where the managed service wins
The managed service wins on continuity. It contracts out the reliability engineering function rather than a set of projects. The analysis runs on a cadence, the improvement register has an owner, and the failure data gets reviewed every month whether or not something broke. There is no vacancy risk, no ramp-up repurchase after each resignation, and the depth is a bench rather than one person’s toolkit. What a managed reliability service is, and what sits out of scope, has its own plain-language guide.
Two situations account for most of the sites that choose it. First, the site that cannot justify or cannot fill the full-time seat. Regional and remote operations feel this worst, because the role is hardest to recruit exactly where it is hardest to do without. Second, the site that has the seat but is not getting the function, because the incumbent has been consumed by daily operations. What it does not give you is a person in the room at 6am with site authority. That is the in-house engineer’s ground, and a fourth option, agency labour hire, exists precisely to cover known gaps fast with one person’s specialisms on hire.
Verdict: the strongest answer when continuity matters and the seat is the problem.
Which should your site choose?
A large site with a working reliability team: stay in-house at the core, and buy consulting campaigns for the lumps: the strategy refresh, the readiness scope, the fleet FMECA program your own team cannot absorb without dropping the day job.
A site where the seat will not fill, or keeps emptying: managed service. Every month the seat sits vacant, the function is not merely paused. The data decays and the PM program drifts unvalidated. Contract the function, and revisit the hire when the market or the roster changes.
A project moment, meaning a new plant, an expansion or a strategy rebuild: consulting campaign, scoped hard, with a named site owner for the handover. The end date is a feature, not a flaw. The shape of that campaign is not the same in every case, because greenfield, brownfield and restart projects each load the readiness work differently.
A site building reliability capability for the first time: counterintuitively, do not start with the lone hire. A first-ever reliability engineer rarely inherits usable data, walks into an unvalidated PM program, and has no protection. It is the hardest version of the job, set up to fail. Start with a campaign or a managed foundation, then hire into a working system. The engineer you eventually recruit will thank you, and stay.
The combinations are the point
The options are presented as rivals; on the ground they sequence and stack. Campaign first, managed sustainment after, so the transformation does not decay. Managed foundation first, in-house hire second, into a system that already works. In-house core with campaign surge for the work that would otherwise wait a year. The question is not which option is best. It is which combination matches the work your site actually has, at the maturity it actually has.
Get a straight answer for your site
We sell two of the three options on this page and hire the first, so we have no interest in steering every site to the same answer. The fit is what makes any of them work. Tell us the shape of your reliability workload and the state of the seat, and we will tell you which option, or which sequence, fits.
Talk to our reliability engineers See the Reliability Operating Centre
Frequently asked questions
When does hiring an in-house reliability engineer make sense?
When the reliability work is continuous rather than a project, the site is large enough to keep a full analytical workload running, the seat is fillable in your market, and management will protect the role from being consumed by breakdown response. Miss any one of those four and the hire underdelivers, not because the engineer is weak but because the structure is.
Should a small site hire a reliability engineer?
Usually not full-time. Below a certain asset base the real analytical workload is part-time, and the role either pads out with planning work or dissolves into firefighting. A small site typically gets more from a managed arrangement or periodic campaigns, the function on a cadence without the seat, and the budget difference funds actual improvements.
Can you combine a consultant with an in-house reliability engineer?
That combination is the normal operating state of mature sites. The in-house engineer owns the site relationships, the daily decisions and the follow-through, while campaigns deliver the concentrated lumps (a fleet strategy build, an FMECA program, a readiness scope) that would otherwise consume the resident engineer for a year. The failure pattern to avoid is the reverse: using serial consultants as a substitute for the function.
What should a reliability consulting scope include to be worth buying?
A defined deliverable, a data handover in your systems rather than a slide deck, named site people in the working sessions so capability transfers, and an explicit owner for the follow-through after the campaign ends. If a proposal cannot say who holds the outcome twelve months after handover, you are buying a report, not a result.
Which option gets reliability capability running fastest?
The contracted routes are measured in weeks. A campaign takes weeks to scope and a managed service weeks to stand up, while recruiting an in-house engineer is measured in months before ramp-up starts. Speed alone should not decide it, but when failures are costing money now, waiting half a year for a hire to start learning the plant is the most expensive patience there is.