Nobody boards a houseboat in the backwaters without a chart. The channels braid, the deep water shifts with the season, and the boat that feels captained by the tiller alone is in fact following a route somebody drew long before the passengers arrived. Operations in an automated company braids the same way. The tools make it feel captained by itself, and the chart — the honest map of who does each piece of work — is exactly what most teams are missing when they declare NoOps. This walkthrough draws that chart, in six steps, gently and in order.
A word on what we are mapping. NoOps is the ambition to buy and automate operational work so completely that no dedicated team is needed — an aspiration, its own reference works insist, rather than a finished state. The ambition is reasonable. The mapping is how you find out how much of the voyage is genuinely automated and how much is muscle memory wearing a dashboard.
The NoOps Tracing Walkthrough, Step by Step
The route, before the detail:
- List every duty that recurs for one service.
- Trace each duty to its present holder.
- Weigh the work hours against the coverage hours.
- Audit the robots and their keys.
- Rehearse one failure on purpose.
- Price the exits while the sun is out.
Step one: list every duty that recurs
Take one service — one boat, not the whole backwater — and write down everything that happens for it on a repeating basis: renewals, checks, deploys, updates, password rotations, invoice reconciliations, the Friday glance at the error dashboard. Include the invisible duties, the ones that only become visible when they are skipped. Keep the language plain: “confirm the nightly backup actually finished” will serve you better than any process vocabulary.
Step two: trace each duty to its present holder
Now the honest pass. Against every row, write the name — a person, a vendor, a script, a provider’s support desk — currently doing the work. This is the step that produces the gasps. Duties turn out to be carried by a colleague who never agreed to it, by a former employee’s forgotten automation, or by luck. None of these holders is shameful; uncharted waters never are. The purpose of the pass is simply to bring every row into daylight, because a duty with a name can be negotiated, budgeted, or handed over — while a duty with no name negotiates nothing and simply collects. Nothing is judged in this step; the only rule is that no row may say “everyone” or “it just happens.” Those answers mean the holder is unknown, which is itself the finding.
Step three: weigh the work against the coverage
Two different quantities live in your map, and confusing them is the classic error. The first is work: how many hours the duties actually consume in a month — often surprisingly few, sometimes under forty. The second is coverage: the hours somebody must be reachable in case the duty fails at midnight, at festival season, at the worst moment in the calendar. Coverage is what exhausts small teams, and it is what managed vendors are genuinely selling. Write down both numbers for your service and look at the gap. The gap is the true price of the independence you were considering giving up.
Step four: audit the robots and their keys
Every duty traced to a script or an agent gets a second question: who owns the robot? Automation code needs tending, and the newer arrivals — the incident agents that reached general release through 2026 — hold write access to production, which means their permissions need a name against them and their errors need an owner too. The guidance coming from the largest platforms is refreshingly old-fashioned on this point: let the machine propose, let a person approve, at least for anything in production. A robot with unclaimed keys is not automation. It is a passenger at the tiller.
Step five: rehearse one failure on purpose
Pick the duty whose failure would ruin a week — the restore of the booking data, say — and rehearse it calmly, in daylight, with the phone on the desk rather than in the hand. Write down how long the recovery took and where the procedure surprised you. The date of that rehearsal is the most honest number on your chart, and reliability teams suggest no more than ninety days should ever separate one rehearsal from the next. A backup never restored is a folk tale; the rehearsal is how it becomes a fact.
Step six: price the exits while the sun is out
The final entry on the map is forward-looking. For each vendor in the story, note what leaving would cost — transfer fees, and separately the weeks of human effort — and name the alternative that would receive the work. The calendar is on your side here: a new European data rule retires switching and egress charges from January 2027, which shaves the fee. The effort remains yours, so write it down. Teams that price exits in calm weather negotiate from strength; teams that price them during an outage negotiate from the ceiling.
What Your NoOps Map Tells You
Finished steps one through six, you are holding the artefact that separates serious teams from confident ones. Practitioners grade the species on a ladder whose lowest rung answers to NoOps by assertion, the memo that outruns its evidence, and whose topmost rung, pointedly, is low-ops by design: every duty owned, rehearsed, and priced, with nothing left pretending to be automated that is actually adored. If your map shows rows without names, coverage with no budget, robots with unclaimed keys, rehearsals that never ran, or exits without prices, you have simply located the next quarter’s work. That is not failure. That is a chart doing what charts are for.
The reference collects the ladder’s full detail, the term’s long history, and the arithmetic of work versus coverage at this detailed NoOps reference — worth a slow read with the map open beside it.
Keeping the Chart Alive
A chart drawn once and laminated is a souvenir. The working version gets touched: a small ritual each quarter, an hour with the register open, names refreshed, rehearsals dated, prices revisited after any vendor announcement. Teams that keep this rhythm report a sensation worth describing — not the absence of operations, but the presence of knowledge about their operations, which turns out to be the actual luxury. Ease, it turns out, is not the absence of effort; it is effort that has been mapped, priced, and rehearsed by somebody who knew where the deep water lay. The tiller can be automated. The knowing cannot.
And because the way you do one thing is the way you do everything: the same charting instinct makes travel better. Kerala beyond its famous sights is our own chart of the quieter channels, our guide to planning a first trip to Kerala sequences the essentials so nothing important is left to luck, and this tour of events and local experiences across the state fills the evenings a good chart always leaves room for.
Draw the map before you need it. Then enjoy the water — you have earned the ease, and now you know precisely where it comes from.
