Business owners hear the pitch constantly now: buy managed services, automate the rest, and the operations line item disappears from the payroll. The pitch has a name — NoOps — and like most pitches that promise to remove a department, it is roughly one third true, one third misdirection, and one third a bill you have not seen yet. Owners in aviation-adjacent businesses are prime targets for it, because they already understand outsourced operations better than most industries: hangar leases, contracted maintenance, third-party handling. That experience is an advantage, because the questions worth asking are the same ones you would ask before outsourcing an aircraft’s maintenance to the cheapest tender. This article answers the five questions we hear most often from owners, in the order they usually ask them, with the arithmetic laid bare.
What Does NoOps Actually Mean for an Owner?
Stripped of jargon, it means arranging your technology so that nobody on your payroll performs operational work: the servers, the patching, the 3 a.m. failures all become somebody else’s contract. Two honesty notes before the numbers. First, the fully automatic version — software running everything, no people anywhere — remains a talking point; even the firms building autonomous agents default to a human approving whatever touches production. Second, the practical version does not eliminate the work; it relocates it into contracts, code, and the judgement of whoever signs the cheques. The five questions below are all about where it relocates to, and at what price.
Will Our Costs Actually Fall?
Often yes, and the savings are real — but the fully-loaded number deserves attention before you celebrate. Practitioners who track this carefully compute a total unit cost per million requests, and the instructive part is the composition, not the total: in one published worked example the figure came to 177.6 units, and about two thirds of it was people time — the monitoring, the reviews, the vendor emails, the decision-making — with the machine bill a minority share. Automating the machines is the easy half. If nobody models the human half before the switch, the savings projection is fiction with a spreadsheet attached.
The owner’s takeaway: demand that any automation proposal price the people side explicitly. Which hours of whose month does each purchased service remove, and which hours does it merely move to a different desk? A proposal that cannot answer has not been costed; it has been hoped.
A property parallel makes the composition point cleanly. Moving a building to a managed servicing company removes your maintenance crew’s wages from your books, but the building still needs someone to choose the servicer, read their reports, question their invoices, and notice when the roof report says one thing and the damp patch says another. The wages move off one line and onto another, usually smaller — smaller, not zero. Technology estates behave identically, and the businesses that model both lines in advance are the ones whose savings survive their first supplier dispute.
Does Headcount Go to Zero?
Almost never, and the reason is arithmetic rather than sentiment. Even a service needing only a third of one engineer’s month of actual work cannot be covered by a third of a person, because failures ignore office hours. Guaranteeing that a competent human is reachable whenever something breaks requires a rotation, and rotations have floor sizes once you cap how often any individual may be disturbed. Around-the-clock coverage from one site is commonly estimated at roughly eight people before holidays and illness. A small business solves this differently — by shrinking the hours that need coverage, buying provider-side guarantees, and accepting some risk — but the coverage trap applies at every size: work is a fraction, coverage is a rota, and confusing the two is how “we saved a salary” becomes “we answer the phone ourselves”.
What actually happens in well-run small companies is a role merge, not a role deletion. The duties fold into the founder’s week, or into an office manager’s, or into a contractor’s monthly visit — and that is a legitimate outcome provided somebody wrote the duties down first. The failure mode is the silent version, where the duties fold into nobody and surface as surprises at the worst seasonal moment.
What Changes in January 2027?
From 12 January 2027, the European Union’s Data Act outlaws the leaving-tax: the switching and egress fees that providers have charged customers for taking their data elsewhere disappear. For owners, this changes the calculus of departure — moving your data stops being a line item on the exit invoice. The engineering effort of a move, however, is untouched by legislation. Somebody still has to re-point the systems, re-test the workflows, and re-verify the data on arrival, and that effort was always the expensive part. So the Act improves your negotiating position — the threat of leaving becomes credible — without making leaving free. Owners should note the date in vendor-review calendars and use it in renewals that fall after it.
When Is the NoOps Approach a Bad Fit?
Three business profiles should keep real operational capability, whatever the pitch says. If your workload is steady and heavy around the clock — a booking platform running hot every day — reserving capacity in advance beats paying per-use premiums, and that trade rewards hands-on management. If the business has unusual technical demands — bespoke hardware, hard response-time floors — the managed platforms price the average case kindly and the edge case badly. And if your operation would not survive the retirement of a product you have built on, no savings justify that concentration of risk; exit capability must exist before it is needed. Owners of steady, demanding, or deeply committed businesses tend to land on a lean internal function plus purchased services — not on the elimination the word implies.
A quick self-test covers the rest. Pick the system your busiest month depends on, and ask two people independently what happens if it fails on a Sunday night. If the answers match, and the recovery steps are written somewhere current, your operation can lean further toward purchased services. If the answers differ, or the document is a year old, there is operational work in your building that belongs to somebody specific — and finding that person is worth more than any tool the pitch is selling.
How Do We Start Without Betting the Business?
Small, reversible, and written down. Four moves, in order:
- Open a register of every recurring operational duty and its current owner — the unowned entries are where risk lives.
- Automate the most repetitive duty on the list and measure the hours actually saved, not the hours promised.
- Add one more managed service only after reading its exit terms in full.
- Review the whole picture quarterly, the way you would review a lease portfolio.
The measurement habit transfers straight from the property world, as the walkthrough on simplifying complex property decisions shows: inventory, evidence, review. For teams that need to upskill first, this practical learning roadmap is the sensible starting syllabus, and the background philosophy of deliberate outsourcing runs through our overall approach to complex assets.
One final line for the spreadsheet: whichever way you lean, keep a named human against every duty the vendors do not absorb, and re-price the arrangement once a year. The full framework behind these answers — the metrics, the maturity ladder, the documented failures — sits in this NoOps guide at the Ops Atlas, and it is the reference to hand any adviser who recommends the leap.
