Remote hiring data: which roles get hired and which ones stay

Most global hiring reports are built on surveys. Someone asks a few hundred HR leaders what they plan to do next year, and the answers become a trend. Plans are cheap. Contracts are not.

This is remote hiring data of a different kind. It is not a survey and not a projection. It comes from live contracts and live payments across roughly 18,000 monthly active workers, more than 1,200 companies, and over 150 countries, covering the last 24 months on the Ontop platform. Nobody was asked what they intended to do. The numbers show what they actually did, who they hired, and how long those people stayed. Ontop is the Global Workforce Engine: financial infrastructure that lets companies hire, contract, and pay teams in more than 150 countries. That position means the platform sees the full arc of a global hire, from signed contract to the payment that lands months or years later, which is where most hiring data goes quiet.

Three numbers that frame everything below

Before the detail, the shape of the finding. These three figures set up every section that follows, and the third one is the one most hiring plans get wrong.

Key takeaways

• Volume and durability are different questions. The category that fills a seat fastest is the one least likely to keep it filled, so a plan built on placement speed pays for it later.
• A fast placement and a lasting one are not the same purchase. Optimizing for time to fill and optimizing for sustained output lead to different hiring plans, and the retention table is where that gap becomes a number.
• Consulting is both the fastest growing category and the most durable, at +30.9% year over year and 42.5% retention. That combination does not appear anywhere else in the data.
• Technology and specialized consulting are 2.5 times more durable than sales, at 41.5% and 42.5% retention against 16.8%.
• Every category spikes in Q1. Read this year over year, never quarter over quarter, or the seasonal pattern will invent a trend for you.
• The order you build in matters more than any single choice. No team builds every tier at once, and what you start with determines what you are still paying for in two years.

Where are the remote jobs actually concentrated?

Technology accounts for 45% of active roles on the platform, more than double the next largest group. Sales follows at 19%, customer service and support at 14%, consulting at 13%, and virtual assistant or administrative roles at 9%. Five categories cover the entire active population, measured across roughly 18,000 monthly active workers.

Here is the full distribution, with what each category actually contains:

The concentration matters more than the ranking. When nearly half of all active global roles sit in a single function, the operational problems of global hiring are mostly tech hiring problems: competing for engineers against local markets that pay in strong currencies, onboarding people who have options, and paying them reliably enough that they do not drift to the next offer.

The second thing worth noticing is what is not on this list. There is no long tail of exotic functions. Companies hiring across borders are not distributing every function globally, they are globalizing a specific and fairly narrow set of roles where the work is genuinely location independent and the talent pool is genuinely deeper abroad.

Why are the categories grouped this way?

These groupings come from raw job titles that were consolidated to remove duplicates and synonyms. That step matters because job titles in global hiring are notoriously inconsistent: the same role can be posted as five different titles across five countries, and counting them raw would scatter one function across the whole distribution.

Where is demand accelerating fastest?

Consulting grew 30.9% year over year from H1 2025 to H1 2026, the fastest of the five categories. Sales grew 21.0%, technology 14.0%, virtual assistant and administrative roles 12.8%, and customer service 4.9%. Four of the five main categories are growing at double digits, so this is broad expansion rather than one category pulling the average up.

Consulting growing fastest is the least obvious result here, and it is the one worth sitting with. Specialized advisory work is the kind of role companies historically kept close, both for confidentiality and because the work is judgment heavy. Its growth rate suggests that reservation is weakening, and that companies are getting comfortable buying senior judgment across borders rather than only buying execution capacity.

Customer service at 4.9% is the outlier in the other direction. It is the third largest category by volume but close to flat in growth, which usually means a category that globalized early and has reached its natural ceiling rather than one that is being abandoned.

Why should you read this year over year instead of quarter over quarter?

Every category on the platform shows a hiring spike in Q1. Budgets reset, headcount plans unlock, and the same seasonal pattern repeats across functions and regions. A quarter over quarter read will therefore show growth in Q1 and a decline in Q2 regardless of the underlying trend.

If you are building a hiring forecast off any global dataset, including this one, compare the same period across years. The seasonal pattern is strong enough to manufacture a trend that is not there.

Which remote roles last, and which ones churn?

Of everyone hired into these categories in the last 24 months, consulting retains best at 42.5% still working, followed by technology at 41.5%, customer service at 34.1%, virtual assistant and administrative at 30.9%, and sales at 16.8%. Sales generates the most new placements of any category and keeps the fewest of them.

Getting hired is not the same as building a career, and this is the table where that distinction becomes a number:

Read the table across rather than down. Sales is the only category that is simultaneously large, fast growing, and poorly retained. Consulting is the only one that is fast growing and well retained. Technology is the only one that is large and well retained. Those are three different strategic situations that a single ranked list would have flattened into one.

The sales result is the finding most likely to change a decision. A 16.8% retention rate means five out of six sales hires are gone. If your global hiring plan is weighted toward sales because those roles fill quickly and the pipeline pressure is immediate, you are buying placements rather than a team, and you will pay the onboarding cost repeatedly.

Technology and specialized consulting retaining at roughly 2.5 times the sales rate reframes what those roles cost. A tech hire that takes longer to close and costs more per month is not necessarily the expensive option once you price in the replacement cycle on the alternative.

Does high churn mean the hire was wrong?

Not always. Some categories carry structurally shorter engagements, and a consulting contract that ends on delivery is a completed engagement rather than a failure. The distinction that matters is whether the churn is planned or unplanned, because only one of those has a cost you can design around.

Sales is the harder case, because sales churn is rarely planned. When five of six placements end inside 24 months in a function where ramp time is measured in quarters, the arithmetic gets uncomfortable: a meaningful share of those hires never reached the point of producing what they were hired to produce.

Why does sales generate the most hires and the least retention?

Sales combines the highest placement velocity with the lowest durability because it is the easiest category to hire into and the hardest to succeed in. The barrier to entry is lower than tech, the evaluation cycle is faster, and performance is measured in months. Those same properties that make sales quick to fill make it quick to empty.

There is a second dynamic specific to cross-border sales roles. A sales hire in a different market is often being asked to sell into a territory, a language, or a buyer profile that the hiring company understands less well than it thinks. When the results do not arrive, the diagnosis usually lands on the individual rather than on the territory assumption, and the seat gets refilled with the same assumption intact.

The practical implication is not to stop hiring sales globally. It is to stop treating placement speed as the success metric. If you are staffing a global sales function, plan the cost model around a replacement cycle you can actually predict, and treat the first two quarters as a validation of the territory rather than only of the person.

Paying people reliably is a smaller lever than compensation design but it is not a trivial one, particularly in markets where local payment infrastructure is slow or where currency volatility eats a paycheck between the transfer and the withdrawal. The Ontop Global Account exists for that specific problem: workers receive and hold funds in a currency they control rather than absorbing an involuntary conversion every cycle.

Where should you invest first when building a global team?

Start with software engineering and technology, which carries the highest volume, solid growth, and the best durability of any large category. Use sales and customer service as fast entry paths where speed matters, paired with training routes toward more durable roles. Treat design, data, legal, medical interpretation, and project management as specialized niches, best entered where adjacent skills already exist.

One clarification on tier 3, because those five names do not appear anywhere in the distribution above. They are not a sixth category sitting beside the other five. They are smaller specializations that live inside or just below them, each too small to surface as its own line at this sample size. That is precisely what makes them a targeted entry rather than a base to build on: you go there when your team already has the adjacent skill, not to establish a beachhead.

This is not about picking a single skill. It is about sequencing what to invest in first, because no team builds all three tiers at once and the order determines what you are still paying for in two years.

What does this change about how you budget?

It moves the unit of budgeting from cost per hire to cost per retained year. A category with a 16.8% retention rate and a low cost per hire can easily be more expensive over two years than a category with a 41.5% retention rate and a higher monthly rate, once you count the recruiting, onboarding, and ramp time you spend twice.

Getting the classification and the contract right at the start is part of that arithmetic, because a contract that has to be restructured mid engagement is its own form of churn. If you are hiring contractors across borders, the risks around contractor misclassification are assessed per worker and per country, so a small mistake repeated across a growing team compounds quickly.

How does AI change which roles are worth hiring?

AI will transform parts of these roles over the next few years, and the exposure is uneven. The most exposed work is entry level QA, content moderation, and narrowly scoped junior development tasks. The least exposed is work that requires judgment, context, and accountability, which is concentrated in the consulting and senior technology roles that already retain best.

This does not argue against hiring in exposed categories. It argues against hiring for a single tool. The durable skill is full software engineering literacy rather than training on one framework, because base tools change and the people who understand the fundamentals stay relevant when they do.

The same logic applies to how you structure the engagement. Roles most exposed to automation are the ones where a rigid multi year commitment carries the most risk, and where the flexibility to change scope matters more than locking in the lowest rate.

Should you slow down hiring in exposed categories?

No, but you should change what you screen for. In exposed categories, screen for adaptability and fundamentals rather than tool specific fluency, and structure the engagement so that scope can move without renegotiating the entire relationship.

It is worth being precise about where the exposure actually sits, because it is not where you would expect. Entry level QA and bounded junior development live inside technology, which is the second most durable category in this data at 41.5%. The least durable category, sales at 16.8%, is not one of the exposed ones. So automation is not arriving where retention is already weakest. It is landing on the junior end of a category that otherwise holds, which puts pressure on the pipeline into the durable roles rather than on the durable roles themselves. That is a slower problem to notice and a harder one to staff around.

What this data does not tell you

This is platform data, not a census. It reflects the Ontop client base, which skews toward technology and fintech businesses hiring globally, so the concentration in tech roles is real but amplified by who is in the sample. The findings are directionally solid. The precise magnitudes are not universal.

Three specific limits are worth stating plainly. First, the sample is self selected: companies that hire globally through a single platform are not a random draw from all companies that hire globally. Second, retention here measures whether a worker is still active on the platform, which is a close proxy for still working but not identical to it. Third, the role categories are consolidated from raw job titles, and any consolidation makes judgment calls at the boundaries.

None of that undermines the core findings, which are large enough to survive the caveats. A 2.5 times gap in durability between categories is not a rounding error, and a 45% concentration in one function is not a sampling artifact. But a company using this to plan should treat it as a strong directional signal rather than as a benchmark to hit.

Frequently asked questions

What is the most hired remote role globally?

Technology is the most hired category, at 45% of active roles across the Ontop platform in the 24 months through H1 2026. It covers development, QA, DevOps, mobile, and data, and it is more than double the next largest group, sales, at 19%.

Which remote roles have the highest turnover?

Sales, by a wide margin. Of everyone hired into sales roles in the last 24 months, 16.8% are still working, which is 1 in 6. Consulting retains best at 42.5%, followed by technology at 41.5%, customer service at 34.1%, and virtual assistant or administrative roles at 30.9%.

Is remote hiring still growing?

Yes, and broadly. Four of the five main categories grew at double digit rates from H1 2025 to H1 2026: consulting +30.9%, sales +21.0%, technology +14.0%, and virtual assistant or administrative roles +12.8%. Customer service grew more slowly at +4.9%.

How should a company decide which roles to hire globally first?

Start where volume, growth, and durability overlap, which in this data is software engineering and technology. Use sales and customer service when speed to coverage matters, and budget explicitly for their higher churn. Enter specialized niches only where your existing team already has adjacent skills.

Does hiring contractors internationally require an entity in each country?

No. Companies can contract and pay workers across borders without opening a local entity, provided the classification matches the actual working relationship and the contract fits local law. What matters is that the arrangement is genuinely a contractor relationship rather than employment wearing a different label, since regulators assess that per worker and per country.

Closing thoughts

The most useful number in this data is not the one that says where people are hired. It is the one that says who is still there two years later. Volume tells you what is easy. Durability tells you what worked.

Global hiring plans tend to be written from the first number and paid for with the second. A function that fills in three weeks and empties in nine months looks efficient on a recruiting dashboard and expensive everywhere else. The categories that retain best here are the ones that took longer to build, which is an uncomfortable finding for anyone under pressure to show headcount growth this quarter.

If you want a country level breakdown or a deeper cut by occupation for your own team, book a demo and we will walk through it with you. More platform data and reports are published in Ontop guides and reports.

Data window covers the last 24 months of live Ontop platform data on active and historical contracts, across roughly 18,000 monthly active workers, more than 1,200 companies, and over 150 countries. Role categories are consolidated from raw job titles to remove duplicates and synonyms. Growth figures compare H1 2025 to H1 2026.