Why "closing" is the right word, not "dying"
When people talk about the decline of a professional field, they reach for dramatic language: dead careers, extinct jobs, industries wiped out. The reality is almost never that clean. Fields close the way a store closes at the end of the night: the doors are still open, some customers are still inside, but the lights in the back are already off, the new stock has stopped arriving, and the people who work there are quietly asking their manager whether they should look for something else.
That is the pattern you want to learn to recognize in 2026, because it starts years before the mainstream press writes an obituary. A field that is closing is one where the total number of durable, well-compensated positions is shrinking, where entry points for juniors are narrowing faster than senior attrition, and where the tools of production are being consolidated into fewer hands, fewer platforms, or fewer geographies. It is still hiring. It still has conferences. It still has a subreddit. But the trend lines under all that surface activity have already turned.
Understanding this distinction matters because the mistake most students and career switchers make is binary. They ask "is this field dead or alive?" and get reassured by anyone still employed in it. The better question is "is this field opening or closing, and at what velocity?" Opening fields forgive mediocrity because they need bodies. Closing fields punish everyone below the top quartile, because the survivors are the specialists who accumulated leverage before the contraction.
This article walks through the signals, metrics, and structural indicators that let you assess a field honestly, before you commit two to six years to training for it. It is written for people who are picking a first career, considering a pivot, or advising someone else who is. If you are working with a career orientation mentor, these are the questions to bring to the conversation. If you are self-navigating, these are the checks to run before you enroll in a program, sign a training contract, or move cities for a role.
The framework has seven components: job posting dynamics, entry-level compression, wage stagnation adjusted for inflation, tool consolidation, geographic concentration, credential inflation, and the pipeline of new entrants. Any one signal in isolation is noise. Three or more together, sustained over 24 to 36 months, is a closing field. We will go through each in order, then discuss what to do when you find yourself already inside one.
Signal one: the shape of the job posting curve
Job postings are the most public, most current, and most misinterpreted signal in career analysis. Everyone looks at the headline number: "There are 40,000 open roles for X in 2026." That number tells you almost nothing on its own. What matters is the shape of the curve over time, the ratio of new postings to reposts, and the composition of the roles being advertised.
Start with the trailing 36-month curve for a specific role title, using a source that indexes actual postings rather than employer surveys. Tools like the U.S. Bureau of Labor Statistics Occupational Employment and Wage Statistics, or comparable European sources such as Eurostat's labour force survey, publish this data. The pattern you are looking for is not a single dip, which could be macro-cyclical, but a sustained decline in unique postings that persists across a full business cycle. If postings are down 30 percent since 2023 and the broader labour market is up, the field is contracting, not the economy.
Next, look at the ratio of new postings to reposts. Employers repost roles when they cannot fill them, when the requirements are unrealistic, or when the role is a placeholder they never intend to fill. A healthy field has a high proportion of first-time postings and short time-to-fill windows. A closing field accumulates ghost roles: the same title reposted every 45 days for 18 months, often with slowly increasing requirements as the employer tries to justify hiring nobody.
The third check is composition. Look at the seniority distribution of open roles. In a growing field, the pyramid is wide at the base: many junior and mid-level roles, fewer senior ones. In a closing field, the pyramid inverts. Almost every posting is senior, principal, or staff, because the employer wants somebody who can produce immediately without ramp time, because they have no capacity to train, and because the junior tier has already been automated, outsourced, or eliminated. If you see a field where 80 percent of open roles require ten or more years of experience, the entry gate is welded shut and the field is quietly deciding not to reproduce itself.
Finally, look at role descriptions themselves. Closing fields tend to bundle: a single job posting will list responsibilities that were previously three or four distinct roles. This is not because employers want polymaths, it is because they are consolidating headcount and expecting one survivor to absorb the work of departed colleagues. When you see "we need someone who can do X, Y, Z, and also handle P and Q," the field is telling you it can no longer afford specialists.
Signal two: entry-level compression and the disappearing junior
The single most reliable indicator that a field is closing is what happens to its junior tier. Fields communicate their long-term intentions through their treatment of newcomers. A field that plans to exist in ten years hires and trains juniors even in downturns, because it knows it needs a pipeline. A field that is closing stops hiring juniors first, and the reasons are always dressed up as temporary: budget freeze, hiring pause, focus on senior talent, restructuring.
The test is quantitative. Look at the ratio of junior postings (0 to 2 years of experience) to total postings in the field, and track it over 36 months. In a healthy field, this ratio sits between 20 and 35 percent. In a compressing field, it drops below 15 percent and stays there. Once it goes below 10 percent for more than a year, the field has effectively closed its front door. New entrants can still get in through side channels (referrals, internal transfers, adjacent-field pivots), but the public labour market for beginners has stopped functioning.
Why does this matter for someone deciding whether to enter a field? Because the health of the junior tier is what determines your five-year trajectory. If you enter a field with a healthy junior pipeline, you get colleagues, peer learning, structured onboarding, defined progression paths, and a labour market you can move within. If you enter a field where you are one of six juniors in the entire country that year, you have no peers, no benchmark, and no easy second employer if your first one lays you off. You are effectively locked in to whoever hired you, at their terms, and your outside options degrade every year you stay.
A related pattern is the internship-to-hire collapse. In healthy fields, internships and apprenticeships convert to full-time offers at rates above 60 percent, because employers use them as trial hires. In closing fields, conversion rates drop below 30 percent, because employers use internships as cheap labour with no intent to retain. If you are considering a field, ask specifically about internship conversion rates over the past three years. A refusal to answer, or an answer under 40 percent, is a strong closing signal.
The third check is the credentials of who is being hired at junior levels. In an opening field, employers will accept nontraditional backgrounds, bootcamp graduates, self-taught practitioners, career switchers. They need bodies, and they are willing to bet on capability. In a closing field, the same junior role now requires a master's degree, three named certifications, and a portfolio that would have been considered mid-level five years ago. This is credential inflation as a form of gatekeeping, and it means the field is no longer willing to invest in training its own newcomers.
When you talk to a career orientation tutor or advisor, the junior pipeline question should be your first one. It cuts through the marketing and gets to the honest state of the field.
Signal three: real wages, not nominal wages
Wage data is where most career analysis goes wrong, because people look at nominal wages and pattern-match to "still paying well." The honest measure is real wages, adjusted for inflation, for the same seniority level, over a rolling ten-year window. A field where the median mid-level salary in 2026 buys less than the median mid-level salary of 2016 is a field where labour has lost pricing power, which is the economic definition of a closing market.
Run the calculation like this. Take the median salary for a defined seniority (say, mid-level with five to seven years of experience) in 2016. Apply cumulative inflation from your national statistics office. Compare to the 2026 median for the same seniority. In fields that are growing, the 2026 figure is meaningfully above the inflation-adjusted 2016 figure, because the field is bidding up for talent. In fields that are stable, the two figures are within five percent of each other. In fields that are closing, the 2026 real wage is below the 2016 real wage by 10 percent or more, sometimes 20 or 30 percent for occupations that have been badly hit by automation or offshoring.
This calculation is boring. It is also brutally clarifying. Fields that have lost real wages for a decade are fields where every year of experience makes you less valuable in constant currency, not more. That is not a career, it is a slow depreciation.
A secondary metric is the wage compression ratio: the ratio of senior wages to junior wages. In a healthy field, senior practitioners earn 2.5 to 4 times what juniors earn, because experience compounds into productivity. In a closing field, this ratio collapses toward 1.5 or lower, because employers are unwilling to pay for experience they no longer value, and because the field's technical moat has been commoditized. When a 15-year veteran and a two-year junior are paid within 40 percent of each other, the field has stopped rewarding expertise, which means it has stopped valuing it.
The third check is the top-quartile trend. In every field, the top 25 percent of practitioners earn substantially more than the median. In healthy fields, the top quartile is growing faster than the median, because outstanding practitioners get bid up. In closing fields, the top quartile grows slowly or not at all, because even excellent practitioners face a ceiling imposed by the shrinking budget available to the field overall. If you see a field where the median is stagnant and the top quartile is only marginally better, there is no upside for effort. This point is developed in more detail in our piece on choosing a field when money decides the path, which walks through how to think about compensation trajectories over a working lifetime.
Signal four: tool consolidation and the shrinking surface area
Every field has tools: the software, methodologies, platforms, and frameworks practitioners use to produce output. The health of a field can often be read directly from the diversity and evolution of its tooling ecosystem. Opening fields have proliferating, competing tools. Closing fields have consolidating, dominant tools.
The mechanism is straightforward. When a field is growing, many vendors and open source projects compete to serve it, because there is money and attention to capture. Practitioners have to keep up with multiple options, integration patterns proliferate, and job postings mention a wide range of technologies. When a field is contracting or being absorbed, the number of viable tools shrinks. Vendors exit or get acquired. Open source projects become unmaintained. The field consolidates around one or two dominant platforms, often owned by hyperscalers.
Why does this matter for career choice? Because tool consolidation is often the last visible symptom before automation. When a field's work can be performed by two dominant platforms, those platforms have the data and the incentive to automate that work, and they eventually do. The field's practitioners then become operators of the automation rather than producers of the underlying craft, which pushes their compensation toward the operator wage, not the specialist wage.
Look at the tool graph over the past five to ten years. If the field went from 15 competing solutions to 3, if the vendors have been consolidating through mergers, if open source alternatives have died off, and if the surviving platforms increasingly market themselves as "end to end" or "AI powered," the field's surface area is shrinking. There is less craft to master, which sounds like a benefit until you realize it also means there are fewer differentiated things a human practitioner can be uniquely good at.
Contrast this with what employers are actually asking for right now, which we cover in what tools employers actually ask for. Growing fields have complex, sometimes contradictory tool requirements, because the field is still figuring out what wins. Closing fields have a small, boring, consolidated tool list, because the answer has been decided and there is not much left to argue about.
A related check is the direction of investment. Follow the venture capital, corporate R&D budgets, and academic research funding into your candidate field. Money follows opening surface area. If the field's total funding has been flat or declining for five years while adjacent fields have doubled, the market is pricing the field as closed. Investors are not always right, but when they collectively withdraw, they are usually early rather than late.
Signal five: geographic concentration and offshoring pressure
Fields have a geographic distribution, and that distribution changes over time. When a field is opening, it disperses: more cities, more countries, more remote-friendly employers, more distributed teams. When a field is closing, it concentrates: fewer cities, more offshore hubs, and increasing pressure to relocate to whichever geography can offer the work at the lowest cost.
There are two variants of geographic closure, and they matter differently for your career. The first is concentration into elite hubs. Certain fields (elite finance, top-tier research, some legal specialties) concentrate their remaining high-value work into a small number of expensive cities. If you can move to New York, London, Zurich, Singapore, or Palo Alto, and you can compete against the pool that already lives there, you can still have a career. If you cannot, the field is effectively closed to you, regardless of your ability. This is a specific form of closure that is worth naming honestly, because well-meaning advisors often tell young people to "just work hard" when the actual constraint is geography plus network.
The second variant is offshoring. Fields with routine, digital, English-language work migrate to lower-cost geographies over 5 to 15 year cycles. The pattern is predictable: the work moves first for the largest employers, then for mid-market, then for everyone. Wages in the origin country stagnate throughout, then decline in real terms, then jobs disappear. Fields currently in the middle of this transition include large parts of application development, quality assurance, first-line customer support, junior legal review, and some categories of financial analysis.
How do you detect offshoring in progress? Look at where new hires are being made. If the employer's headcount in the origin country is flat or declining while their headcount in India, the Philippines, Poland, Vietnam, or Mexico is growing at 20 to 40 percent per year, the field is offshoring in that employer's segment. Extend this across the top 20 employers in the field and you have a clear picture of where the work is actually going.
A third check is language. When job postings in the origin country start requiring "experience managing offshore teams" or "vendor management skills," the field has already offshored the production work. The remaining local roles are coordination and oversight, which is a smaller, less durable set of jobs than the production work that used to exist. That coordination tier is often the next to be automated, because it is easier to build software that manages a workflow than software that performs the workflow itself.
Signal six: credential inflation and the moving goalposts
When a field is opening, credentials are useful but not gatekeeping. Employers care more about capability than paper. When a field is closing, credentials inflate. What used to require a bachelor's degree now requires a master's. What used to require a master's now expects a doctorate or a portfolio equivalent. Named certifications multiply, become mandatory, and cost more.
The reason is straightforward: when supply of practitioners exceeds demand for practitioners, employers use credentials as a cheap filter. They do not necessarily believe the credentials predict performance, but they use them to reduce their candidate pool from 500 to 50, then screen the 50 on capability. If you are not in the initial 50, your capability never gets evaluated.
The test is longitudinal. What did the job listings for this role require in 2016, and what do they require in 2026? If the requirements have visibly ratcheted, and the work itself has not become materially more complex, credentials are being used as rationing. Rationing means demand is exceeding supply of positions relative to interested candidates, which is closing behavior.
A subtle form of this is portfolio inflation. In some fields (design, some engineering specialties, quantitative finance, editorial work), the required portfolio for entry has expanded from "three good pieces" to "a website, active presence on three platforms, published work, side projects, and community contributions." This looks meritocratic but is really just a longer, more expensive credential. Candidates from wealthy backgrounds who can afford unpaid time to build the portfolio pass. Candidates who cannot, do not. The field self-selects for people who could afford to spend two years unpaid demonstrating enthusiasm, which is a narrower and narrower pool.
Certification treadmills are another warning. When a field has three, four, or five active certifications that expire and require renewal, and when employers list several of them as required, the field has monetized its own scarcity through credentials. Practitioners are effectively paying for the right to keep working. This is a wealth transfer from workers to credential vendors, and it happens most aggressively in fields where the underlying demand has plateaued.
Compare this to fields still in expansion, where you can enter with a bootcamp, self-taught background, or one lateral pivot. Growing fields will accept messy resumes because they need bodies. Closing fields want pristine resumes because they can afford to reject.
Signal seven: the pipeline of new entrants
Watch what young people are choosing, and more importantly, what they are leaving. The pipeline of new entrants into a field is a leading indicator that shows up in university enrollments, bootcamp cohorts, apprenticeship applications, and switcher-oriented programs like ours. Fields that are opening see rising interest. Fields that are closing see the exact opposite, often five to ten years before the labour market catches up.
The reason young people are a good early signal is not that they are prescient. It is that they have less sunk cost. A 40-year-old professional in a declining field has 15 years of specific human capital to defend, so they rationalize staying. A 19-year-old choosing a major has nothing to defend, so they follow the honest signal: where do their smartest friends want to go, where does the money seem to be, where does the future feel exciting. Their collective choice, aggregated, is a fairly reliable forecast of which fields will still be relevant in a decade.
How do you read this signal? Look at enrollment trends in the relevant degree programs over 10 years. Look at bootcamp cohort sizes for the field. Look at applications to entry-level roles per opening. If any of these have compressed by half or more, and the compression is not explained by demographics (national birth rates declining, for example), the field's incoming talent pipeline is voting with its feet.
A parallel signal is faculty movement. Where are the productive researchers and senior practitioners choosing to work? When you see multiple respected senior figures leaving a field for an adjacent one, taking their doctoral students with them, the field's intellectual momentum has moved elsewhere. The remaining work will still be done, but the frontier has relocated.
A specific case worth naming: fields that have become synonymous with a single company or platform. When "the future of X" is entirely coupled to one vendor's roadmap, and that vendor is not itself a growth company, the field's fate is linked to a single boardroom. That is not a diversified bet. Compare this to genuinely opening fields where dozens of companies, research groups, and open source projects are all pushing the surface area outward. That distributed pattern is what an opening field looks like from the inside.
A field that combines these seven signals (falling postings, junior compression, real wage decline, tool consolidation, geographic concentration, credential inflation, and pipeline erosion) is not a field to enter in 2026. It may still be a field to work in, if you are already senior in it and have five to ten years of leverage before you retire or pivot. It is not a field to train into.
What the signals miss: the acceleration case and legitimate second acts
Every framework has failure modes, and this one has two important ones you should know about before applying it too rigidly. The first is the acceleration case. Sometimes a field looks like it is closing because it is being disrupted, and the disruption creates a new adjacent field that absorbs most of the value. The people who worked in the old field are best positioned to move into the new one, if they see the shift early enough.
Example: prompt engineering as a distinct role has been visibly closing in 2026, with employers no longer treating it as a standalone job. But the underlying work has not disappeared. It has evolved into context engineering, retrieval design, evaluation, and applied AI engineering. Someone who was strong at prompt work in 2023 is well positioned to move into these adjacent roles, which are opening rapidly. We wrote about this transition specifically in our comparison of context engineering versus prompt engineering, because it is a clean example of a field closing and reopening under a different name.
The lesson: when a field appears to be closing, check whether it is actually being absorbed into a larger, differently named field where the same underlying skills are still needed. If yes, your career can pivot with relatively low friction. If no, the closure is real and you need to plan a more substantial move.
The second failure mode is late-career specialty fields. Some fields shrink to a small number of specialists who serve legacy systems, and these specialists earn very well, because supply has contracted faster than the residual demand. COBOL programmers, mainframe operators, specific medical device technicians, certain regulatory specialists. The field is closed to newcomers but very open to the surviving practitioners, who can name their price.
If you are already senior in a closing field, one legitimate strategy is to double down on becoming one of the last remaining specialists. This works when three conditions hold: the legacy systems have long lifespans (multi-decade), the cost of replacing them is prohibitive, and the number of qualified practitioners is falling faster than the workload. Under those conditions, a closing field can be one of the most lucrative places to spend the last 10 to 15 years of your career.
A third exception is fields that are geographically closed but not universally closed. A field can be dying in your country while thriving elsewhere. If you are willing and able to relocate, and if immigration paths exist for your target country, the field may still be viable for you personally even though it is closed in your local labour market. This is a real option for many people and is often underexplored, because career advisors default to assuming their client will stay in their current country.
These exceptions do not invalidate the framework. They refine it. The framework tells you the aggregate direction of the field. The exceptions tell you where individuals can still find opportunity inside an unfavorable trend.
What to do when you find yourself in a closing field
Running the seven-signal check on your own field is uncomfortable, especially if you are five or ten years in. The most common reaction is denial, followed by rationalization, followed by paralysis. None of those help. If your assessment says the field is closing, you have three practical strategies, and the right one depends on how far into the field you are and how transferable your skills are.
Strategy one: pivot early, while you still have adjacency options. This is the right choice if you are in the first three to five years of the field, have not yet built deep specialization, and can identify an adjacent opening field that values 60 to 70 percent of your current skills. The pivot takes 12 to 24 months and costs some compensation short term, but it moves you into a trajectory with upside instead of decay. Backend engineering into infrastructure and platform engineering is a clean example, which is why we maintain a current view of the top backend skills to learn for people in exactly this situation.
Strategy two: become a specialist within the closure. This is the right choice if you are 10 or more years in, have specific irreplaceable knowledge, and can identify a niche within the closing field that is contracting more slowly than the field average. Legacy systems specialists, regulatory experts, and specific technical historians of a field can earn well for 10 to 20 years even as the broader field disappears around them. The requirement is that you are truly one of the best in the niche, not just present in it.
Strategy three: monetize the accumulated expertise through teaching, consulting, or advisory work. If you have deep expertise in a field that is closing, other people (particularly juniors and switchers who did not read this article in time) still need what you know. Teaching and mentoring is a legitimate second career, and it works particularly well as a bridge while you are transitioning to a different primary field. If you are considering this path, you can become an instructor on Refonte Learning and monetize your expertise directly, either as a primary income or alongside a role in an adjacent field. This is also how the Refonte Learning platform sources practical, honest teaching from people who have actually done the work.
Whichever strategy you choose, the worst option is inaction. Fields do not un-close. Once the seven signals are all pointing down, the trajectory is set for a decade or more, and time spent waiting to see if things improve is time subtracted from your pivot window. Every year you stay in a closing field, your adjacency options narrow, your accumulated capital becomes more field-specific, and the pivot becomes harder.
Applying this to your own decision in 2026
The practical exercise is straightforward. Take the field you are in, or the field you are considering entering. Score it against the seven signals using publicly available data. Give each signal a rating from 1 (strongly opening) to 5 (strongly closing). Sum the scores. Anything above 25 out of 35 is a field to leave or avoid. Anything between 18 and 25 is a field to enter with caution and a clear exit plan. Anything below 18 is genuinely healthy.
Do this exercise annually, not once. Fields shift, and the assessment from three years ago is not necessarily still accurate. A field that scored 15 in 2023 may score 22 in 2026, and vice versa. The point of the framework is not to produce a permanent verdict but to give you a repeatable, honest look at where you are.
Bring the results to someone who can pressure-test your reasoning. That could be a mentor, a peer in a different field, or a structured advisor. The Refonte Learning career orientation program is designed for exactly this kind of conversation, and connecting with a mentor there gives you a second set of eyes on your assessment before you make a costly decision.
The hardest part of this work is not the analysis. It is the honesty. Fields close slowly enough that people inside them can convince themselves the closure is temporary, cyclical, or exaggerated. The signals in this article are designed to cut through that self-deception with data that is verifiable and independent of your emotional investment. Use them, and use them early.
About Refonte Learning
Refonte Learning is an EdTech platform operated by Refonte Infini Infiniment Grand (a French SAS, SIREN 949 841 605, verifiable at https://data.inpi.fr/entreprises/949841605), with a UK operational office at 1 Poulton Close, Dover, Kent, United Kingdom, CT17 0HL. Refonte Learning offers practical training and mentorship in AI, data, cloud, DevOps, and software engineering, with a focus on skills that are actually being hired for in 2026. Practitioners who want to teach on the platform can apply to become an instructor and share their field experience with learners who are making the same career decisions you once made.