The non-technical skills that move an IT career forward
Once you can do the technical work, what moves you on is whether people can rely on you around it: to write clearly, explain a risk to someone who has to decide, keep an incident call in order, get a change through review and give an estimate people can plan around. These skills can be learned on purpose, and SFIA writes several of them into its levels of responsibility.
Working time and rest break rules here are as GOV.UK and Acas set them out for workers aged 18 and over. Your contract may give you more.
Once you can do the technical work, what moves you on is whether other people can rely on you around it: a ticket the next shift can act on, an incident report a manager can read, a change request a reviewer can approve, a risk explained so a non-technical manager can decide, an incident call kept in order, an estimate that holds and a "no" that does not start a feud. SFIA, a framework organisations use to describe IT roles and career paths, writes these into its levels: at level 4 someone communicates with technical and non-technical audiences and, as required, "takes the lead in explaining complex concepts to support decision making". None of it needs a course. All of it improves with deliberate practice.
Writing that other people act on
Most of what colleagues see of your work is written: tickets, knowledge base articles, incident reports and change requests. Each one is read by someone who was not there, often under pressure.
GOV.UK's guidance on using clear language was written for government pages, but it fits technical writing well. It says people with higher levels of literacy also prefer plain English, because it lets them understand the information as quickly as possible, and that specialist terms are fine as long as you explain them the first time you use them. It suggests paragraphs of no more than five sentences, splitting sentences over 25 words and using the active voice.
Tickets
Write the ticket for the person who picks it up at 3am with no context. Say what the user reported, what you checked and found, what you changed (with the exact setting or command), and what is still unknown. "Rebooted, seems OK" tells the next person nothing. "Print spooler on the file server stopped at 09:12, restarted at 09:20, queue cleared, cause not yet known" lets them carry on.
Knowledge base articles
A knowledge base article is a ticket written for everyone who will hit the same problem. It needs a title that matches what users and the service desk search for, the symptom in the words people report it in, the fix in numbered steps, and a way to check it worked. KBY's Graduate Track lesson on publishing a knowledge base article walks through drafting, review, approval and rollback, and KBY's runbooks show the same structure for incidents: symptom, detection, fix and verification.
Incident reports and post-incident reviews
Google's SRE book describes a postmortem as a written record of an incident, its impact, the actions taken to mitigate or resolve it, the root causes and the follow-up actions to stop it happening again. It argues for writing them blamelessly: identify the contributing causes without indicting any individual or team, and assume everyone involved had good intentions and did the right thing with the information they had. Its reason is practical: where people are blamed, they stop bringing problems to light. It is also blunt that an unreviewed postmortem "might as well never have existed", so get yours read.
The habits that make a review useful start during the incident. KBY's incident handover template asks for absolute timestamps and for observed facts to be kept separate from hypotheses and proposed actions. If your notes do that, the review almost writes itself.
Change requests
SFIA's change control skill describes the lifecycle of a change request as registering, assessing, authorising, planning and deploying, and the core of it as assessing and reducing the risks a change poses to availability, performance, security and compliance. A change request that gets approved without a meeting answers the reviewer's questions before they ask: what changes, why now, what could break and for whom, how it was tested, how you will know it worked, how you will roll it back, and who has the authority to do that. KBY's templates include change checklists that show the level of detail: the Terraform pull request checklist asks for every replacement or deletion to be explained and for a named person to apply and verify the change, and the Conditional Access deployment checklist asks for the rollback trigger and authority to be agreed before enforcement.
Explaining risk to people who decide
A non-technical manager does not need to understand the fault. They need to make a decision, and your job is to make it a fair one. Lead with the decision, give the options with what each costs and what it risks, put the risk in their terms (users affected, money, deadlines, reputation), say what you recommend and say plainly what you do not know.
A note might read: "The certificate on the payroll portal expires on 3 December. Renewing it means about five minutes of downtime, which we can schedule for a weekday evening. If we do nothing, staff cannot reach payroll from the 3rd. I recommend we book the evening slot this week." It names the risk, the cost of avoiding it and a recommendation, and nobody needs to know what a certificate chain is.
At level 5, SFIA expects someone to communicate "clearly with impact, articulating complex information and ideas to broad audiences with different viewpoints". Practise on small decisions so it is familiar when a big one comes.
Running and joining incident calls
The SRE book's chapter on managing incidents is worth reading even if you never work at that scale. Its model gives each person on the call one role. The incident commander holds the overall picture, assigns work and holds any role they have not handed out. The operations lead and their team apply the fixes, and should be the only people changing the system during the incident. A communications lead sends regular updates to the team and stakeholders, and a planning role handles the longer-term work such as handoffs and tracking what has been changed so it can be put back. The commander keeps a living incident document, and when command passes to someone else, the handover is explicit and acknowledged, so everyone knows who is in charge.
It also recommends declaring an incident early and closing it if the fix turns out to be simple, rather than starting the process hours into a growing problem. And it asks responders to notice their own state: if you start to feel panicky or overwhelmed, ask for more support.
If you are taking part rather than leading:
- say what you are about to do before you do it, and wait for the go-ahead;
- report what you found as facts with times, not as theories;
- do not make changes outside the people assigned to make them;
- if you have nothing to add, keep quiet and keep reading the incident document.
SFIA's incident management skill shows how the role grows: at level 2 someone advises relevant people of actions taken and gives users and stakeholders updates on the incident's status, and by level 5 they manage incident communications and lead the review of major incidents. Volunteering to keep the record or send the updates is the easiest way to start. KBY's Daily Triage exercises are practice at reasoning from symptom to cause before a real call.
Working with change management
Change management can feel like paperwork between you and the fix. Treat it instead as a free review by people who will catch what you missed. SFIA's description of the skill includes developing processes for standard, normal and emergency changes. If you find yourself raising the same low-risk change every week, ask the change manager whether it could be handled as a standard change in your process. If an emergency change was really a planning failure, say so in the review, so the next one is planned. And when a reviewer asks a question you cannot answer, that is a gap in the change, not an obstacle to it.
Estimating, and saying no
Estimates go wrong when they are a single number given on the spot. Give a range, state what you are assuming, break the work into pieces small enough to see, and say what would change the answer. Then compare what you estimated with what it took, so the next estimate is better. SFIA expects someone at level 4 to plan, schedule and monitor work to meet personal or team objectives, which starts with knowing how long your own work takes.
Saying no is mostly saying "not now, and here is what would have to move". When a new request arrives, make the trade visible to whoever sets priorities: this can happen this week if the migration slips a week, or it can wait until the migration is done. At level 4, SFIA expects someone to escalate when scope is affected, rather than absorb it quietly. The guide on moving from senior engineer to team lead covers what changes when protecting the team's time becomes your job.
On-call without wearing yourself out
On-call is where the rules on working time matter, so know them. GOV.UK says you cannot work more than 48 hours a week on average, normally averaged over 17 weeks, unless you choose to opt out. Its guidance on calculating working hours counts time spent on call at the workplace as working time, but not time on call away from the workplace. Acas's guidance on being on call adds that on-call time away from the workplace does not usually count if you can spend it as you choose. The time you spend actually working on a call-out does count towards your maximum hours and rest breaks, and the more control your employer has over what you do, where and when, the more likely on-call time is to count. Whether you are paid for on-call time at home should be written in your contract.
Workers over 18 usually have three kinds of rest: an uninterrupted 20-minute break if the working day is more than six hours, 11 hours' rest between working days, and either 24 hours without work each week or 48 hours each fortnight. GOV.UK lists cases where a worker may instead be entitled to compensatory rest of the same length as the rest missed, including shift workers who cannot take daily or weekly rest between shifts and jobs that need round-the-clock staffing. Your contract may give you more. Read it, and the on-call policy, before you join a rota.
The Health and Safety Executive defines stress as "the adverse reaction people have to excessive pressures or other types of demand placed on them". Its guidance on work-related stress names six areas of work that affect stress levels (demands, control, support, relationships, role and change), and says employers have a legal duty to protect workers from stress at work by doing a risk assessment and acting on it. Its advice for workers lists signs to notice in yourself, including feeling negative or isolated, being unable to concentrate and difficulty sleeping, and says to talk to someone, for example your manager, as soon as possible. If the pressure comes from your line manager, it suggests finding out what policies cover that and, if there are none, talking to a trade union or worker representative, HR, a worker assistance or counselling service, or your GP. It also says stress is not a weakness and can happen to anyone.
Some habits help as well:
- Hand over properly at the end of a shift or an incident, so you can stop thinking about it.
- Take the time back after a night call-out, if your policy allows it, and use it.
- Fix the alert that woke you for nothing, or raise it, rather than learning to sleep through it.
- Tell your lead early when a rota is unsustainable. In HSE's terms that is a problem of demands and control, and it is easier to fix before someone burns out.
Mentoring
Teaching someone else is how you find out what you actually know, and it is part of seniority. At level 4, SFIA expects someone to take responsibility for driving their own and their team members' development, and its performance management skill includes coaching colleagues in the skills they are working towards.
Start small. Pair with a newer colleague on a real ticket and let them drive. Review their write-ups the way you would want yours reviewed. Point them at the knowledge base article rather than answering from memory, and if the article is wrong, fix it together. Over a year, a junior who can take your on-call shifts is evidence of your seniority that no certificate gives you, and it belongs in your case for promotion.
Choosing what to work on first
Pick one of these for the next three months, not all of them. If your tickets get bounced back with questions, start with writing. If you leave meetings with managers feeling unheard, practise the risk note. If incidents feel chaotic, volunteer to keep the record on the next one. Ask your manager or a senior colleague which one they would choose for you: they see your work from the side you cannot.
Common questions
- Does time on call count towards the 48-hour week?
- GOV.UK's guidance counts time spent on call at the workplace as working time. Time on call away from the workplace does not usually count if you are free to spend it as you choose, but Acas says the time you spend actually working on a call-out does count towards your maximum hours and rest breaks.
- What should a post-incident review include?
- Google's SRE book lists the incident, its impact, the actions taken to mitigate or resolve it, the root causes and the follow-up actions to prevent a recurrence. It recommends writing it blamelessly, so it focuses on causes rather than on who to blame.
Sources
Every figure and rule in this guide comes from these pages, read on 17 November 2026. Pay rates, exams and schemes change: check the source before you rely on one.
- Level 4 - Enable (SFIA Foundation)
- Level 5 - Ensure, advise (SFIA Foundation)
- Change control (CHMG) (SFIA Foundation)
- Incident management (USUP) (SFIA Foundation)
- Performance management (PEMT) (SFIA Foundation)
- Use clear language (GOV.UK)
- Postmortem Culture: Learning from Failure (Google (Site Reliability Engineering book))
- Managing Incidents (Google (Site Reliability Engineering book))
- Maximum weekly working hours (GOV.UK)
- Maximum weekly working hours: calculating your working hours (GOV.UK)
- Being on call: working time rules (Acas)
- Rest breaks at work (GOV.UK)
- Rest breaks at work: compensatory rest (GOV.UK)
- Work-related stress and how to manage it: overview (Health and Safety Executive)
- Help for workers on stress at work (Health and Safety Executive)
Spotted something wrong or out of date? Tell us and we will correct it.