The person reading your CV may not know what the job is

You send the application, and it lands in front of a recruiter holding a list of technical terms somebody else wrote down for them. They cannot tell you what those terms mean. So they compare the characters on your CV with the characters on their list, and if enough of them line up, you go through. We have sat on the hiring side of that process often enough to say it plainly: the first filter has very little to do with whether you can do the work.

That cuts in your favour more often than people expect. During 2018, 58% of EU enterprises that recruited or tried to recruit ICT specialists had problems filling those vacancies, Eurostat found. The figure measures employers’ recruitment difficulty in that year, not an applicant’s odds of an offer, but with demand like that it’s hard not to get hired. Two qualifications come with that. How much experience you have matters, and so does how you market yourself. You also have to qualify somewhat: a laptop at home and enthusiasm won’t win the senior security DevOps role in the ad. Clear that bar and getting hired somewhere is not the hard part. Landing somewhere that teaches you something, pays you properly, and doesn’t cost you two years of your career is the hard part, and that is a different set of decisions.

This is the hiring conversation turned around and seen from the candidate’s chair. Straight out of university, the question is which first employer to pick when you have no idea what any of these jobs involve. A few years in, the question is how to stop applying and get approached instead. The two problems have more in common than they look like they do.

Startup or big company decides what you learn

Before you polish anything on the CV, decide what kind of place you want it to land in, because the size of the employer sets the shape of what you will know in five years.

Should your first job be at a startup or a large company?

A startup gives you breadth and mess; a large company gives you depth in one thing, and that depth comes with a catch.

At a startup you do everything. There won’t be many processes, it will be chaotic, and you will learn a lot quickly. If you thrive in chaos, that life suits you. A big organisation doesn’t need somebody who knows a lot of things. It needs somebody specialised in one thing that fixes a particular problem, and being a cog in that machine is precisely what’s being bought. Both are real options, and there is a whole spectrum of companies in between where you adapt.

The catch is portability. One of us worked at Ericsson while other teams built network stacks on proprietary hardware with proprietary protocols. Some of that knowledge travels to the next employer. Much of it doesn’t, because nobody else runs that stack.

Try things first, then specialise

So the shape of the first job matters. The next decision is how early to narrow down.

Does specialising early help or hurt?

It helps, once you know enough about the field to pick the thing worth betting on.

Specialising hard pays. One of us bet on Google Cloud, went full on it, and passed on jobs that didn’t involve it. Highly specific people get hired more easily, because fewer candidates compete for that work and people trust a specialist with their technology. Picking the bet is the part that needs information you don’t have straight out of university. You are clueless then, and not in an insulting way: you don’t know what the jobs are, what you would be doing in them, or which companies are worth your time. So find a place that lets you touch as many things as possible. After a year or two, some of it will appeal to you and some won’t, and that is when you can go deep with a direction you chose on evidence.

What to look for in the meantime is the same either way: a company that will invest in you rather than one that wants you to move paper from A to B. Large corporations run young talent programmes precisely to find the next generation, and those are worth applying to. If your circumstances allow it, don’t take a job purely for the money at that stage. One of us got lucky with a first employer that spent real time and money on them, and luck is not a plan.

When the pool is five people, stand out

Picking a target changes the odds, because the odds at a small company are much better than they feel from the outside.

How do you compete against candidates with more experience?

Stop trying to be the strongest engineer in the pool and be the one who did the homework.

Nobody has heard of most small companies. The job ad goes out and maybe three people come to talk to them, which means they may well need you more than you need them, and they are choosing from a pool of five or ten. Your job is not to be the best at C. Your job is to stand out, and research is the cheapest way to do it. Read what the company publishes: the open source, the blog, the podcast if there is one. Then come in with questions about what you found.

We interviewed a candidate three weeks before recording who had listened to every episode of this show to prepare and had been through everything in our GitHub. The opening was along the lines of: in this repository, I think you did it wrong, how would you do it differently? Not a single “yes, you’re doing everything right”. All the right questions.

That works because small companies hire like families. They care whether you understand their values and fit in, they don’t expect somebody fresh from university to be strong technically, and they will coach the hard skills. Big companies run the other way round: the technical bar is the prerequisite and the culture interview is a sorting criterion on top. As far as we can tell, Spotify puts real weight on culture interviews, and the hard skills still come first.

Ask what a good hire looks like in six months

Homework gets you into the room. The questions you ask in the room decide whether you can tell a good job from a bad one before you accept it.

What should you ask the people interviewing you?

Start with the mission and the values, then ask what success in the role would look like.

Ask about mission and values and watch what happens. If the people in front of you can articulate them, leadership is probably strong and that’s a place worth being. If nobody has a clue what they are, you may be in the wrong room. Then ask the question that does the most work in the shortest time.

The answers come back concrete: we have fewer incidents, this problem is fixed. You get the expectation right from the start, which tells you what the job is and what the outcome of the job is supposed to be. The conversation lands somewhere much better once you both know what is expected.

Ask the rest of it too. Will they let you work on the things you want to work on? How will they take care of you once you’re in? Do they have regular meetings to coach you toward where you want to go? If the answer is no, maybe that’s not the company for you. Almost nobody asks any of this. People arrive scared and stressed and treat the interview as a chance to show off their skills, when it’s closer to a date. You want to know as much as you can about these people, because you’ll spend more waking hours with them than with your family.

Practice on interviews you don’t want

Knowing what to ask is not the same as being able to say it out loud under pressure, and the second part only comes from repetition.

How do you get good at interviews?

Go to interviews for jobs you don’t want, because the answers have to be polished and you have no time to prepare them on the spot.

You get maybe 45 minutes. Somebody asks a question and you answer it there and then. You can’t say you’ll get back to them on that one. Talking about yourself, your strengths and your weaknesses is a skill like any other, and one way to get the reps in is to take interviews you have no intention of accepting.

Reading the market is the other half. Go through the job ads on LinkedIn for the kind of work you want, see which technologies keep appearing, learn one of them and put a project using it on your CV. That gets you invited far faster than a general claim to be a quick learner.

Boot camps and the skill underneath them

Job ads tell you what to learn. They don’t tell you how, and one of the routes on offer asks for three months of your own time and money up front.

Are coding boot camps worth it?

They can jump-start a career, and the price is a few months of unpaid time plus the risk that you find out this field isn’t for you.

The good side is real. Boot camps are short and intensive, and they usually help with placement afterwards: support, introductions, a job at the end. Course Report’s 2020 alumni report puts average tuition at $14,214 and 78% of alumni in full-time employment, counting respondents across all graduation years. Narrow it to people who graduated in 2020 and it reads $15,423 and 52%. Read those numbers knowing what they are: a survey of alumni, so the people who walked away halfway through are mostly not in it. They exist. Of the people we know who enrolled, plenty quit before the end or straight after, having realised it wasn’t for them.

The complaint we have is the money. Those months are usually unpaid, it’s your own time and your own savings, and not everyone can plan for that. If you expect easy money on the other side, the landscape is tougher than the brochure.

Underneath all of it sits one skill. In IT you have to keep up, so learning becomes the main thing you do. If you’re not ready to learn continuously, we’d honestly suggest finding another field. Boot camps are also a minority route: in the 2021 Stack Overflow Developer Survey, 10.24% of respondents had learned to code at a boot camp, against 59.53% from online resources and 51.53% from books.

You may also be asking a question you can’t answer yet. One of us started in banking software writing COBOL and lasted four months before it got soul-crushing. That job still paid off: a stable nine-to-five that showed how the world works and gave something to compare everything else against. Sometimes you don’t even know the question yet. What you enjoy is personal and nobody can answer it for you, so try things. Worst case, you find another job.

Portable skills, unstable jobs

Where you learn something decides how long it stays worth money, and that trade-off is sharper than it looks.

Is it safer to specialise in your company’s stack or in the market’s?

Neither is safe, they just fail differently.

Go deep into one big company’s own technology and you can switch off when you leave the office. What you know is unrelated to the market. That is the one case where you may have less need to follow changes in mainstream tooling. Getting to that point takes a lot of learning, so it isn’t the easy option. The exposure is that the more specialised you are in what one company does, the less stable your position is. Departments get outsourced and whole functions move.

Go the other way, into what everyone uses, and the skills are portable. Infrastructure engineering, SRE, cloud ops, whatever we’re calling it this year, is highly reusable: our customers all use AWS in much the same ways, which makes what we sell a commodity service. Others can do it too. That means more competition, and it means the ground moves faster, because everyone is pushing the tooling forward at once.

The job stability part of this needs care about what the numbers measure. The feeling is that almost nobody stays anywhere more than two years. The US Bureau of Labor Statistics put median tenure at 4.1 years in January 2020, and 2.8 years for workers aged 25 to 34 against 9.9 years for those aged 55 to 64. Tenure there is time already served with a current employer, counted mid-job, so somebody two years into a ten-year stay is recorded at two years. It is not the length of finished jobs, and it says nothing about when anyone should move. It’s US data too, and it doesn’t break out our corner of the industry. What it settles is that a couple of years in one place is ordinary for younger workers. Keeping your skills portable is our advice, from watching departments get outsourced, not a conclusion the data reaches.

Make the job look for you

All of the above assumes you’re the one applying. The position worth getting to is the opposite one.

How do you get offers without applying?

Publish enough that people can form an opinion of you without ever talking to you.

A conference talk somebody can watch. An open source project they might already depend on. A blog. A podcast. One of us gets a steady flow of job offers because of this show while not looking for work at all. That is the point: not looking can give you a much stronger negotiating position, because you are not the one asking.

Money comes last, and deliberately so. Early in the process they don’t know who you are, so saying no to your number is cheap. Once you’ve shown what you can do, the number attaches to something, and the value you bring justifies the price you put on it. For a senior role, the expectation runs the other way: you’ve been around, you know the market, and both sides can assume the compensation is in a sane range or you wouldn’t still be talking. Where you are matters here. Colorado’s Equal Pay for Equal Work Act took effect on 1 January 2021 and requires employers to announce openings together with the pay range, so in some markets the number is on the table before you are. That range is a starting point rather than a boundary: the rules behind it let an employer “ultimately pay more or less than the posted range” if it was a good-faith estimate when the job went up.

The cheap version of all this is available to anyone. Tools are free, most of them are open source, and the only thing you spend is your own time. Learn one, write up what you learned, post it on your blog or on LinkedIn. Someone curious, investing their own hours, willing to share what they found: that’s the candidate we’d want to interview.

What this means for you

Most of this is a reallocation of effort rather than extra work. Less time polishing a CV nobody reads closely, more time on the things that make a small company’s five-person shortlist easy to win, and on being visible enough that the next role finds you.

The one decision worth making consciously is the first employer. A place that invests in you and a place that wants a cog both pay a salary. Choose on the investment: the coaching conversations, the young talent programme, somebody whose job it is to get you where you want to go.

Four things to do this week:

• Go through twenty job ads for the work you want and write down which technologies appear most often. Build one small project with the top one and put it on your CV.
• Book an interview at a company you wouldn’t join, purely to practise answering out loud with no preparation time.
• Before your next real interview, read the company’s public code, blog or podcast and bring two questions about something specific you found there.
• Ask what would make this a good hire six months in, and ask what the coaching and growth conversations look like once you’ve started.

Common questions, answered

When should I bring up salary in a job interview?

As late as the process allows. Early on, the people interviewing you don’t know what you’re worth yet, so rejecting your number costs them nothing. Once you’ve demonstrated what you can do, the number has something to attach to and they can justify it against the value you bring. Where a range is published with the ad, as in Colorado since January 2021, you have a reference point to open from. A posted range is the employer’s good-faith estimate rather than a ceiling on the offer, so it informs the negotiation instead of settling it.

What do I put on my CV when I have no professional experience?

A project using the technology that shows up most often in the job ads you want to answer. Read twenty ads in your target area, find the tool that keeps recurring, learn it on your own time and build something small with it. The tools are free and mostly open source, so the cost is hours rather than money. Write up what you learned as a blog post or a LinkedIn post as well, because showing curiosity and a willingness to share counts for a lot with the people reading applications.

Should I accept the first offer I get?

Only if you’ve checked that the company will invest in you. Demand for people in this field is high enough that you have choices, so be intentional instead of grabbing whatever arrives first. Ask what success in the role looks like six months in, ask who will coach you and how often, and see whether the people interviewing you can explain their own mission and values. An offer you turn down is still useful: it’s an option, and having options is what makes the next conversation go your way.

Related episodes

• #31, Hiring an infrastructure automation person: the same process from the other chair, including the take-home assignment we set and what watching someone work tells us.
• #7, How Do We Learn: our own answers on continuous learning, which feeds we read and how we keep track of what we find.
• #22, Who are Mattias, Julien and Andrey?: the three career paths behind the advice above, and how each of us ended up doing infrastructure work.

Resources

• Hard-to-fill ICT vacancies: an increasing challenge: Eurostat’s figure that, during 2018, 58% of EU enterprises recruiting ICT specialists had problems filling the vacancies. It measures employers’ recruitment difficulty, not how easily an applicant finds work.
• Stack Overflow Developer Survey 2021: how people actually learned to code, with online resources at 59.53%, books at 51.53% and boot camps at 10.24%.
• 2020 Coding Bootcamp Alumni Outcomes and Demographics Report: average tuition of $14,214 and 78% in full-time employment across respondents from all graduation years, and $15,423 and 52% for the 2020 cohort. It surveys alumni, so it cannot tell you about the people who dropped out.
• Employee Tenure in 2020: the US median tenure numbers, 4.1 years overall and 2.8 years for workers aged 25 to 34. It measures years with the current employer, so it does not say how long completed jobs last.
• Colorado Senate Bill 19-085, Equal Pay for Equal Work Act: effective 1 January 2021, requires employers to announce job openings along with the pay range.
• Equal Pay Transparency Rules, 7 CCR 1103-13, Statement of Basis and Purpose: adopted 10 November 2020, it settles that a posted range is a good-faith estimate and that an employer may pay more or less than it.

Since we recorded

Alphabet announced about 12,000 job cuts in January 2023, as filed with the SEC. That is one employer rather than the whole market, but it is enough to say that “it’s hard not to get hired” was a claim about 2021 and not a standing condition. In Europe, the EU pay transparency directive gives applicants a right to pay information before they are hired, and member states must transpose it by 7 June 2026. We came back to what is worth learning now in #112, DevOps Careers: What Should You Learn Now?, and to where juniors get their reps when agents write the code in #113, Books, Juniors and Mattias’s Dark Factories.