You just got a “yes” from a new client, your calendar looks lovely, and the first invoice is already living in your head. This is the best possible moment to try something slightly awkward: pretend the engagement has already gone wrong. Freelancers who run a quick pre-mortem before signing tend to catch the boring, preventable stuff that otherwise eats weekends. Here are eight ways a freelance DevOps gig commonly falls apart, written as if they have already happened, plus a friendly nudge on what to do about each.
First, what is a pre-mortem, really?
It is a ten-minute exercise. You imagine it is three months from now, the contract ended badly, and you are explaining what happened over coffee. Then you write the reasons. Since nobody has lost anything yet, you can be brutally honest, and the fixes are usually just a sentence added to a scope document.
Eight ways the gig fell apart
1. “The scope was a fog”
You agreed to “improve the deployment process” and the client heard “fix everything that has ever annoyed us”. The fix: write down three concrete outcomes you will deliver and one or two things you explicitly will not touch. A short list beats a long conversation.
2. “I waited a week for access”
The cloud console, the repository, the monitoring tool, and the ticket queue all needed approvals, and each one lived with a different person. Meanwhile your paid days ticked by. The fix: ask for an access list before day one, and make the start date depend on getting it.
3. “Nobody on their side owned me”
Your contact was enthusiastic but had no authority to decide anything. Decisions drifted. The fix: ask who signs off on changes, and get that person’s name in the kickoff message.
4. “I was the only person who understood it”
You built something good, then went on to the next client, and a month later it broke at midnight and nobody could read it. The fix: plan a short handover into the price, with notes someone else can follow. Think of it as the last deliverable, not a favour.
5. “The payment slipped and so did my mood”
The client was late once, then twice, and awkwardly you were still working. The fix: milestone payments, a clear invoice date, and a tidy way to pause work if an invoice is overdue. If tax rules are on your mind too, the guide on money basics for new DevOps freelancers covers the financial side in plain language.
6. “They quietly expected on-call”
You were hired for a migration. Somewhere along the way, a production issue happened on Saturday and everyone looked at you. The fix: state your availability in writing, including what happens out of hours and what it costs.
7. “I said yes to a second job that overlapped”
Two clients, one week, and a deadline that appeared out of nowhere. Something had to give, and it was your sleep. The fix: keep a rough view of your capacity, and leave a gap you do not promise away. If you are balancing this with a salaried role, the article about taking on freelance work responsibly next to a day job has useful guardrails.
8. “The client never really wanted the outcome”
Sometimes a project exists because a manager needs a line in a report. The work is cancelled mid-flight, or quietly ignored. The fix: ask early what happens after the project succeeds, who benefits, and who will use it. If the answers are vague, price and plan accordingly.
One more thing about the list: the stories get better when you write them from your own past. Scroll back through your last two or three clients and ask which of these eight showed up, even in a small way. Most people find that the same two or three keep returning. Those are the ones to hard-wire into your template contract, so that you stop renegotiating them from scratch every time a new client appears in your inbox.
A short stop at a security-minded neighbour
If you like thinking this way, DevSecOps Now is worth a visit. It is a DevSecOps site, and its blog covers securing CI/CD pipelines, which is basically professional pre-mortem thinking applied to delivery tooling. Reading it can sharpen the instincts you bring to client systems.
Turning the list into a one-page habit
You do not need to use all eight each time. Instead, try this light routine before every new engagement:
- Block ten minutes and open a blank page.
- Write “It is three months later and this went badly because…” at the top.
- List as many reasons as come to mind, and do not edit while you write.
- Circle the two that would hurt most.
- Turn each into a sentence you can add to your proposal or kickoff email.
After a few rounds, you will notice your own favourite failure patterns. Some freelancers keep forgetting about access; others undersell handover. Knowing your pattern is half the cure. If you are just starting and the whole thing feels like a lot, the starter questions new DevOps freelancers usually ask are a gentle place to begin.
Will clients find this weird?
Rarely. Most clients are relieved when someone names the awkward parts out loud before the money moves. You are not being negative; you are showing that you have done this before. A calm line such as “I like to list what could go wrong at the start, so we can avoid it” tends to land well and quietly raises your standing.
Wrapping up
Freelancing rewards optimism, because you need to believe in your pitch. A pre-mortem simply gives that optimism a safety net made of paper. Run it before the next contract, keep the page where you can see it, and enjoy having fewer unpleasant surprises to deal with.
