You have decided to move. Not upward inside the job you have, but sideways out of it, into something near enough that the years behind you still count for something. Every piece of advice you find says the same thing. Lead with your transferable skills.
Which ones. Transferring to what.
The advice names a property and then stops. It sounds like the part of a career change that needs a long walk and some honest self-examination, and for a great deal of the working world it has already been calculated, published, and left sitting on a government website almost nobody opens.
Transferable skills is a category, not a destination
The phrase describes an attribute of a skill. It contains no direction. You can write out stakeholder management, budget ownership and running a delivery schedule, read the list back, and be no closer to knowing what to type into a search box, because what you need is not an inventory of your properties. It is a list of job titles where those properties are the job.
That is a different question, and unlike the first one it has an answer.
The distance between two jobs is a measured quantity
In a 2007 working paper, later published in the Journal of Labor Economics, Christina Gathmann and Uta Schönberg proposed treating occupations as points in a space of tasks.1 They took a German survey that asks people which of 19 tasks they actually perform at work, placed each of 64 occupations at its own position according to the answers, and defined the distance between any two as a single number on a scale from zero to one. It "is zero for occupations that use identical skill sets", and reaches one where two occupations have nothing in common at all.1
That makes it possible to say things that are otherwise only ever said loosely. The mean distance between occupations in their data is 0.24, with a standard deviation of 0.22. The closest pair in the set is paper and pulp processing and a printer or typesetter, at 0.002. The most distant is a banker and an unskilled construction worker.1
When people leave an occupation, do they land near it?
The authors compared where movers actually went against a null in which the destination is chosen with no regard for skill at all, so that the chance of landing in any occupation depends only on how many people it employs. Observed moves came out markedly shorter than that. The distance also shrinks as a career goes on: they find that "individuals move to occupations with similar task requirements and that the distance of moves declines with time in the labor market".1
The part with a price on it arrives later in the paper. Estimating what workers lose when they are displaced, they conclude that "wage losses due to displacement are partially avoided because workers are able to find employment in occupations that require similar skills", and that for high-skilled workers those losses "would be more than twice as large if they were randomly allocated to occupations after displacement".1
The limits are severe. The wage data is a two percent sample of German social security records covering 1975 to 2001. The task data comes from four waves of a survey run from 1979 onwards. The sample is restricted to men employed in West Germany, and there are 64 occupations in it.1 None of that describes your labour market, and none of it is recent. What it does establish is narrower and still useful: the distance between two occupations is a real quantity, it predicts where people actually go, and crossing a large one costs money.
So the useful question is not which of your skills are transferable. It is what sits next to you.
Somebody has already published the list for your occupation
O*NET is the occupational database maintained for the US Department of Labor. Among the files it publishes is one called Related Occupations, and it is close to exactly what someone changing career wants: for each occupation, the ones nearest it, ranked, in three bands. Primary-Short holds the "Five most strongly related occupations after expert review". Primary-Long covers the "6th to 10th". Supplemental runs out to the "11th to 20th". The current release carries 18,460 rows of it.2
Nearness is not left to the imagination. The documentation says the file is built from "three important contributors to occupational similarity: what people in the occupations do, what they know, and what they are called".2
That third contributor is the odd one. Two jobs being called similar things is not evidence that they involve similar work, so a measure that leans partly on naming will pull together occupations that share a vocabulary and little else. Read as a description of the work, that is a weakness. Read as a description of the market you are about to apply into, it is uncomfortably accurate, because the words in the advert are how the vacancy gets indexed and how a reader decides at a glance whether your CV is on topic.
The file also has a history that says who it was built for. Until database release 26.3, ONET shipped two others made for this exact purpose: a Career Changers Matrix, for people moving into an occupation that matched the skills, education and experience they already had, and a Career Starters Matrix for people entering the labour market, both the output of a 2012 research project by HumRRO and the National Center for ONET Development.4 The release notes record that "the newly added Related Occupations file replaces the 'Career Changers Matrix' and 'Career Starters Matrix' files, which have been removed from the database".3 The career-changer question did not go away. It was folded into a better-built file.
You do not have to download anything to start. Every occupation page on O*NET OnLine carries a Related Occupations section, with five titles showing and five more behind a control. On the page for Human Resources Specialists, checked on 14 August 2026, it lists ten in total.5 That is ten of a possible twenty: the Supplemental band, the 11th to 20th most strongly related, sits in the downloadable data file rather than on the page. The browser gives you the near half of the neighbourhood for free, and the file gives you the rest.
The catch is jurisdictional. O*NET describes the American occupational structure, in American titles, on American codes. Work in Europe and the shape will still be recognisable and the neighbours still informative, but the words are not the ones in your local adverts.
Europe grades the skill rather than the job
ESCO is the European Commission's counterpart, maintained by its employment directorate. Version 1.2.1, last updated in December 2025, "provides descriptions of 3,039 occupations and 13,939 skills", is "translated into 28 languages (all official EU languages plus Icelandic, Norwegian, Ukrainian, and Arabic)", and "can be consulted free of charge".6
It is built on a different principle: where O*NET hands you a list of neighbouring jobs, ESCO hands you a graded list of skills and lets you work out the neighbours. Two of its distinctions do real work for someone moving occupation.
The first separates essential from optional. Essential skills are "those knowledge, skills and competences that are usually required when working in an occupation, independent of the work context or the employer".7 Optional ones "may be required or occur when working in an occupation depending on the employer, on the working context or on the country".8 When you read a job advert and cannot tell whether a requirement belongs to the profession or to that particular company's version of it, there is a published opinion on the question.
The second answers the transferability question head on. Every skill in the classification is graded by reusability, on four levels. Transversal skills are "relevant to a broad range of occupations and sectors", and the examples given are working in teams, office software, negotiating compromise and communication. Cross-sectoral skills are "relevant to occupations across several economic sectors", the example being animal welfare, which turns up in agriculture, in veterinary work and in amusement and recreation parks. Sector-specific skills are "specific to one sector, but are relevant for more than one occupation" inside it, as monitoring livestock is. Occupation-specific skills are "usually applied only within one occupation and its specialisms", and the example is milking operations.9
That is a published answer to which parts of your experience travel, and how far. It is worth reading in both directions. A skill graded transversal is, by ESCO's own definition, one that applies across a broad range of occupations and sectors, which is exactly what makes it portable and exactly what makes it silent about destination. Communication does not point anywhere in particular.
Be careful what you take from that. The reusability grade measures how widely a skill applies. It does not measure how many other applicants have it, or what employers reward, and ESCO claims neither. Read it as a map of reach rather than of advantage: the transversal parts of your experience are the ones most likely to survive the move, and the sector-specific parts are the ones that say which move you are making.
Reading it as a shortlist, not an oracle
Start from the job you do, not the title on your contract. Titles inflate and deflate by company, and looking up the wrong one produces a neighbourhood that is not yours. Find the occupation whose task description you recognise, even if nobody has ever called you that.
Read past the first five. The top of any related list is mostly the moves you had already thought of and the ones with similar names. The interesting entries sit lower down, where the tasks still overlap but the vocabulary has changed enough that you would never have searched for them.
Mark the essential skills you can evidence, not the ones you could learn. Go through a candidate occupation's essential list and mark only those where you could name a specific episode, with what you did and how it came out. Whatever is left unmarked is your actual gap. It is almost always shorter, and more concrete, than "I would have to retrain".
Then do the arguing yourself. No employer is going to look up your O*NET code. A crosswalk tells you where the overlap is; your application still has to demonstrate it, described in the target occupation's words rather than your current one's.
If you would rather start from a conversation than a database, JobCraftly can suggest directions your experience already supports and then search for real openings in whichever one you pick. It is the same move, made with less reading: start from the work you have actually done rather than from a job title you have already thought of.
What a crosswalk cannot tell you
It measures the work, not the gate in front of it. Two occupations can sit right next to each other in task space and be separated by a licence, a registration or a protected title. A list of related occupations will not warn you about that, and finding out is worth doing before you spend a month.
It says nothing about demand. Relatedness is a property of occupations, not of the vacancies within commuting distance of you this month.
It describes an occupation rather than a person. These lists are built on what people in a job typically do. If your own version of your job has been unusually analytical, or unusually hands-off, then the further your work sits from the average the less the neighbourhood applies to you.
And the finding that people move short distances carries more than one explanation. Gathmann and Schönberg read it as evidence that skills partially transfer, which their wage results support.1 But an observed move records where somebody ended up, which is the product of what they wanted and what they were offered. A labour market in which employers simply refuse to hire across large gaps would produce a similar picture, and the comparison with random mobility cannot separate the two. If the second explanation is doing some of the work, the crosswalk is telling you something slightly different from what it appears to: not only where your skills would function, but where an employer is likely to believe they would.
None of this decides anything for you. What it does is convert a question that does not yield to more thinking, the one that goes what else could I even do, into ten names on a public web page, or twenty if you fetch the file. Most of them will be wrong. Crossing off the ones that are is an afternoon's work, and it is a great deal more than most people changing careers ever get to start from.
