The Nearest Job Is Not Always the Cheapest Move
Similarity can reduce training time while preserving the same exposure or a wage penalty. Build a move table that scores distance, demand, risk, and pay together.
Adjacent-occupations supplies candidate edges in the occupation network; transferable-skills tests whether the evidence carried across an edge is recognizable and payable by an employer.
The closest role can be declining too, and real capability can remain unreadable when a credential, licence, portfolio, or familiar title is the hiring signal.
Ten listings are open in your browser.
Each one sounds almost exactly like the job you used to do.
The nouns are familiar. The software is familiar. The duties have been rearranged into a slightly different order, and the salary field is missing.
Maybe your role was removed. Maybe the work is still there and the company wants a cheaper version of it. Maybe your search is voluntary, and the old work became intolerable before it became impossible.
Do not apply to all ten.
Rank the moves.
The nearest title is often the fastest way to reuse your experience.
It can also be the fastest way to carry the same decline, the same hiring barrier, and the same wage loss into a new employer.
Similarity is not a destination.
It is one column.
Not the Inventory
Start with your old role, not its label.
Write the work that recurred.
Tasks.
Decisions.
Tools.
Outputs.
Stakeholders.
Constraints.
Results.
Do not write handled customers.
Write whether you answered scripted queries, resolved a billing exception inside policy, retained an account, or redesigned an escalation process.
Those are different capabilities.
Now take each item and ask the harder question.
What did the work require you to notice?
What decision could you make without permission?
What happened when the ordinary process failed?
What did you produce that someone else could inspect?
What result can you document?
You are making a requirement map: your prior evidence in one column, the destination requirement in the other.
The map is not a resume yet.
It is an inventory of things an unfamiliar employer might be able to buy.
Your title is not your inventory.
The title tells a familiar employer a quick story.
The inventory lets an unfamiliar employer test the story.
Edges, Not Answers
Write down five occupations adjacent to yours.
Do not choose them because a search engine puts them beside your title.
Choose them because they reuse a visible part of your inventory: the same client problem, equipment, workflow, compliance decision, technical system, handoff, or output.
Then keep the candidate relationship conditional.
A published similarity score is a hypothesis about retained requirements.
It is not a placement probability, and it does not know the task bundle at a particular firm.
The comparison sets behind those scores cover a few hundred occupational categories.
They do not contain every job title a manager can invent.
Begin in the neighbourhood.
Do not mistake the map for the street.
The Move Table
Take out a sheet of paper.
Make five rows, one for each adjacent occupation.
Make five columns.
Task overlap. Does the destination reuse the concrete tasks, decisions, tools, outputs, and constraints you have already documented?
Employer-recognized evidence. Can you translate your inventory into a work sample, artefact, reference, authorization, portfolio, or precise claim that meets a named requirement?
Local vacancies. Are there actual openings and employers inside the geography, schedule, commute, and contract conditions you can accept?
Wage floor. Does the destination clear the minimum pay that makes the move viable after entry level, hours, benefits, and travel are counted?
Lower exposure. Does the destination move away from the task bundle that is shrinking or being automated, rather than preserve it under a new title?
Give each column a score from 0 to 2.
Use 0 for absent or contradicted evidence.
Use 1 for partial evidence, an unresolved condition, or a real gap.
Use 2 when the evidence is concrete enough to act on.
Total the score.
8–10: Primary search. Apply, contact employers, and tailor your evidence map to the essential requirements.
5–7: Test application. Send a small number of deliberate applications, or seek a conversation that tests the missing condition before you pay for training.
0–4: Do not retrain toward it yet. The move may become real later. Right now it is a story with too many missing columns.
The bands are not validated.
They stop a familiar title from doing all your thinking.
Do not hide the columns inside a single ranking.
An employer, a vacancy, a credential, a wage, and an exposure pattern can each kill an otherwise elegant adjacency.
The table has to make the failure visible early, before you spend money on the gap.
The Cheapest Gap
Training changes the map.
It does not dissolve the constraints.
A longer course lets a model close more of a shortage.
It does not supply tuition, income, child care, transportation, time, a local employer, or a hiring manager willing to recognize the bridge.
The shortages themselves are not all the same.
OECD separates cognitive from task-related requirements.
It finds higher average cognitive-skill distance among low-skilled occupations, and in moves from middle- to high-skilled occupations, while task-related distance is higher within high-skilled occupations than within low-skilled ones.
That is why learn digital skills is not a plan.
A foundation gap, a tool gap, a workflow gap, and a regulated-practice gap carry different costs and different evidence requirements.
Keep shortage and excess in separate columns.
A shortage identifies something that needs learning, assessment, or accommodation.
An excess identifies capability the destination will not use.
For high-skilled origins, OECD finds small shortages but large cognitive and task-related excesses on average when moving to other work.
That can mean a job with easy entry and a serious loss of pay, status, or future development.
For professional occupations, the number with no acceptable transition falls from 19 under a short-training scenario to 5 under a moderate one, and to none under the longest.
Your choice to train is not only about whether you can reach the next job.
It is about what the bridge consumes and what the destination throws away.
OECD estimates a moderate-scenario total cost of USD 24,020 per person for managers, against USD 15,940 for professionals.
That is not your invoice.
It is a warning that a high-status move can carry substantial foregone earnings even when the target looks close on paper.
Evidence, Not Resemblance
Similarity becomes useful only when an employer can recognize it.
The employer is not reading a similarity score.
It is looking for a familiar title, a degree, a licence, a portfolio, a work sample, or a reference that lowers the cost of deciding.
Do not answer that with adjectives.
Write a claim that carries scope.
Triaged recurring cases against published criteria, documented exceptions, and escalated outside authority.
Then attach the evidence: a sanitized work product where permitted, a supervisor who can verify the scope, a work sample, an assessment, a credential, or a portfolio.
Separate non-negotiables from proxies.
A licence, safety authorization, legal right to practise, or demonstrable operation can be mandatory.
A specific degree, title, or same-industry tenure may be a proxy an employer can test with a structured interview or work sample.
Do not pretend they are interchangeable.
Where degree requirements have been reset, that is an opening, not an abolition of credentials.
The route only exists if selection changes too.
A skills-first advertisement that still screens for the same employer pedigree has changed its language, not its gate.
So test the gate before you build a plan on the language.
Ask which requirement is mandatory and which is preferred.
Ask what evidence has been accepted in place of the preferred one before.
Experience alone cannot waive a standard, and a standard nobody will name is not a standard.
The Money Test
Do not let a national growth figure choose a local job.
Count the vacancies.
Name the employers.
Read the essential requirements.
Check the commute, schedule, contract, accessibility, and credential condition.
Wage tables exist at national, state, metropolitan, and industry levels.
Use the narrowest one available rather than attaching a national wage to your local decision.
For each destination, compare the wage benchmark with your origin benchmark from the same source and geography.
Write the dollar difference and the percentage difference in the table.
Then call them what they are: occupational statistics.
They do not promise you an offer.
They do not tell you the entry step.
They do not restore your lost tenure.
A higher destination benchmark is a useful screen against obvious downgrading.
It cannot guarantee wage protection on the first day, because a benchmark describes settled workers and you are not one yet.
Now run the exposure column with the same discipline.
Do not call a job safer because it contains more fashionable software.
Name the work the new title requires.
Name the task bundle being automated in your old work.
Check whether the destination preserves it, changes it, or replaces it with accountable, exception, customer, or technical work.
Exposure is not displacement risk.
Displacement risk is not employment impact.
Keep the columns separate.
After the Sort
The table reduces ten vague listings to a primary search, a set of tests, and work that does not justify retraining yet.
Do not convert the test applications into a week of indiscriminate sending.
Use them to test the uncertain column.
Ask a local employer whether your named evidence meets the requirement.
Ask whether a work sample substitutes for familiar tenure.
Ask the training provider what standard it actually assesses, and whether employers recognize it.
Ask what the entry wage and schedule are.
Ask what work remains after the routine task is removed.
The answers may be no.
That is useful information before tuition, a commute, or a lower wage turns the move into a trap.
The move table does not manufacture time, money, or freedom.
It stops you paying for a story that assumes you have them.
Ten open listings feel like ten options.
They are one option repeated ten times until you sort them.
Minimum Viable Action
Before you sleep tonight, write down five occupations adjacent to yours.
Give each a first score for task overlap.
Do not look for a course.
Do not apply to all ten listings.
Make the first column honest.
What you will be able to say afterwards
- Build a five-destination adjacency shortlist from tasks rather than title words.
- Score task overlap, employer-recognized evidence, local vacancies, wage floor, and lower exposure separately.
- Use task, tool, output, scope, and result evidence to test whether a destination employer can recognize prior experience.
- Treat vacancies and local wage benchmarks as a separate local opportunity layer.
- Set a wage floor using an origin-relative benchmark before applying.
- Keep skill shortage and unused capability separate when evaluating retraining.
- Use an application score as a structured prompt rather than a placement probability.
Questions this raises
Is the job most like the old one always the best move?
No. Similar requirements can reduce the amount of new learning, but similarity does not establish local hiring, a workable wage, lower automation exposure, or employer recognition. A usable destination combines a close task relationship with vacancies, a wage benchmark, acceptable entry requirements, and evidence that the employer can understand the applicant's prior work.
How should someone compare five possible next occupations?
A worker can score each destination from 0 to 2 on task overlap, employer-recognized evidence, local vacancies, wage floor, and lower exposure. A total of 8 to 10 makes it a primary search; 5 to 7 makes it a test application; 0 to 4 means it is not yet worth retraining toward. The score is a structured prompt, not a probability of placement.
What counts as evidence that skills transfer to another job?
Evidence names the prior task, decision, tool, output, scope, constraint, and result, then links each item to an essential target requirement. A statement such as triaging cases against published criteria, documenting exceptions, and escalating outside authority is more legible than a claim of organizational skill. It shows what level of work was actually performed.
Why is a close skill match not enough to get hired?
Employers still use conventional signals to reduce assessment cost. OECD-reported World Economic Forum survey data finds that about 80% of more than 1,000 employers across 55 economies expect to prioritize work experience in hiring over 2025–30, and more than 40% expect to rely on university degrees. Adjacent experience needs a task-level evidence map or a credible work sample to be readable.
Should a person apply for a job that pays more on a wage website?
A higher destination wage is an occupational benchmark, not an individual offer. The comparison needs the same source and geography for origin and destination, then needs entry level, hours, tenure, benefits, commuting, and the time required to reach the benchmark. United Kingdom evidence from 1993 to 2012 finds roughly a 15% real-wage fall for lower-end job-to-job movers and more than 20% after unemployment.
What does a vacancy count prove about a career change?
A vacancy count proves only that a destination exists in a defined place and period; it does not prove that the openings are accessible to people from the origin occupation. The local opportunity layer also needs employer requirements, location, commuting, schedule, credential rules, and pay. A national growth projection can coexist with no usable opening in the local labour market.
Does more training make every career move possible?
No. In an OECD exercise covering 123 occupations, 57 origins, or 46%, had no acceptable transition under a small-training scenario. When moderate effort was allowed, 87% of origins had at least one acceptable transition. The larger set reflects a model that permits more shortage to be closed; it does not establish that training is affordable, recognized, or sufficient for every worker.
What is the difference between a skill shortage and having too much experience?
A shortage is a target requirement that needs learning, assessment, or accommodation. An excess is capability present in the source profile that the destination does not use. OECD comparisons find that moves from high-skilled origins tend to involve small shortages and larger excesses. Excess can mean underuse of capability, lower pay, weaker progression, or loss of status; it is not a free advantage.
Can a missing degree be ignored if the job posting talks about skills?
Not automatically. Harvard Project on Workforce analysis reports material degree resets between 2017 and 2019 in 46% of middle-skill occupations and 31% of high-skill occupations, with 63% of changing occupations classified as structural. A removed degree requirement only opens a route when selection also accepts task evidence or work samples. Licences, safety authorizations, and legal rights to practise can still be mandatory.
What should a worker do when a transfer looks close but nobody recognizes the experience?
The worker should build a requirement map and test it against actual employers or a work sample. Recognition of prior learning can document work samples, appraisals, references, photographs, and testimonials against a stated standard; an OECD Mauritius case used a facilitator-supported portfolio over three months before interview assessment. The route can establish partial credit, but it cannot waive a requirement without a credible assessment.
Why should exposure be checked after finding a similar job?
A close destination can preserve the same vulnerable task mix. Similarity measures retained requirements and transition effort; it does not establish that the destination has lower exposure or better staffing prospects. Exposure, displacement risk, and employment impact are distinct questions. A destination should therefore be checked for the tasks being automated, local adoption, vacancies, and the work that remains after the routine work changes.
Can an occupation network predict that one person will get a job?
No. Network research finds that skill similarity predicts occupation-to-occupation transition rates in one nationally representative survey and three résumé datasets, but the result is probabilistic. A network describes a pattern of movement, not an offer. Individual feasibility still depends on evidence, licensing, location, vacancies, pay, schedule, training resources, and the employer's actual selection process.