Employee Upskilling: Cost, Funding, and Where to Start
How to upskill employees in a small business: choosing the right skills, the public funding most owners never claim, low-cost tactics, and the real cost.
Almost every guide on this topic opens with the same three recycled statistics, explains the difference between upskilling and reskilling, gives you nine strategies, and recommends a learning platform. I run a small company myself, and the first thing I would tell another owner is the one thing none of them mention: federal money exists to pay for training your existing employees, and the cost-sharing tiers are most generous to the smallest businesses.
Under federal workforce legislation, local boards may reserve a portion of their funding for incumbent worker training, which means training people who already work for you rather than jobseekers. Employers contribute a share based on headcount, and for an employer with 50 or fewer employees that share is at least 10 percent. The minimum rises with size. The program is administered locally, the terms vary, and it is genuinely underused because almost nobody writing about upskilling mentions it.
That omission is characteristic of the whole category. The published advice assumes a learning and development function, a budget line, and a manager layer, and then hands you a framework. This guide is written for the other case: an owner or operations lead deciding whether to train someone or hire someone, with no training department and a real constraint on both money and hours. I build the records and training tooling for businesses in exactly that position at FirstHR.
What follows covers the third option nobody names alongside upskilling and reskilling, how to choose skills in an afternoon, where the public funding sits, the low-cost tactics that actually work, and the cost line every business case leaves out.
What Upskilling Is, and the Third Option
Employee upskilling is training existing staff to perform their current role at a higher level. It is usually discussed alongside reskilling, which moves someone into a different role. Both get covered everywhere. The third move, cross-training, gets almost no coverage and is the one a small business usually needs first.
The reason cross-training deserves top billing at small scale is arithmetic. In a small company, several critical tasks are performed by exactly one person, and the cost of that is invisible until they are sick, on vacation, or leaving. Upskilling that person makes the dependency deeper. Cross-training someone else makes it disappear. Both are training; only one addresses the risk that is actually on your balance sheet.
The distinction also matters for how you describe it to the person. Being told you are being upskilled reads as investment. Being told you are being cross-trained can read as being asked to absorb someone else's work, which is sometimes exactly what it is.
Tell the person what they get out of it, whether that is a defensible case for a raise or simply the ability to take a vacation without anxiety. Naming the benefit is the difference between a willing participant and a resentful one, and it belongs in the conversation about their responsibilities rather than in an announcement.
None of the three is the right answer on its own. The useful discipline is naming which of the three a given piece of training is, because the three have different measures of success. Upskilling succeeds when someone handles harder work. Reskilling succeeds when someone is productive in a new role. Cross-training succeeds when a second person can do the thing unsupervised, which is a much sharper test.
Upskilling Examples
The clearest examples are the ones where you can point at an invoice, a bottleneck, or a person who is currently the only one able to do something. Below are six that come up repeatedly in small businesses, each written as a move from what the person does today to what they would own afterward.
| Role | What they do today | What the upskilling adds | How it gets built | What it replaces |
|---|---|---|---|---|
| Bookkeeper | Prepares the month end and hands it to an outside accountant | Runs the close, and produces the figures the owner actually reads | Two hours a week with the accountant, then one close run alone with a review | Part of the monthly accounting invoice |
| Support rep | Answers tickets and escalates anything technical | Resolves the top three technical categories, and writes the answers down for everyone else | Four sessions beside an engineer, then a queue of real tickets with a review | An engineer losing an hour a day to escalations |
| Office manager | Runs the office and processes the paperwork for new hires | Owns onboarding from offer to day thirty, including the first week schedule | One run alongside the owner, then two alone | Owner hours in every hiring cycle |
| Warehouse lead | Picks, packs, and runs the floor on a normal day | Builds the schedule to a labor target and handles the supplier calls | Shadowing two schedule builds, then owning it with a check before it is published | A manager approving every schedule and every order |
| Marketing coordinator | Schedules posts and updates the site with copy somebody else writes | Builds and reads the reporting, so spending decisions have numbers under them | A free analytics course plus one real monthly report with feedback on it | Part of an outside marketing retainer |
| Field technician | Completes assigned jobs and calls in when something is unusual | Diagnoses the awkward case and quotes the repair on the spot | Riding with the senior technician on the difficult jobs, then quoting under review | A second visit on a share of jobs |
Read the same list a second way and it becomes a cross-training list. When the goal is coverage rather than depth, the person you train is not the one in the left column but the colleague sitting next to them, and the test changes from harder work to the same work performed unsupervised.
Why It Matters, Without the Recycled Statistics
Upskilling matters for three practical reasons: growing a skill inside the business tends to cost less than hiring it in, the roles on your payroll are shifting anyway, and in a small business a second person who can do the work is what keeps it moving when someone is out.
The standard argument for upskilling relies on a handful of survey numbers that appear on every page in this space, usually stripped of their original year and source. The reasons above are more durable, and none of them requires those numbers.
Start with the build-versus-buy comparison, which is the decision actually in front of a small business. Academic research examining internal and external moves into the same jobs at a US investment bank found that external hires were paid substantially more than people promoted from within to do the same work, and were considerably more likely to be terminated.
The study is Paying More to Get Less, by Matthew Bidwell in Administrative Science Quarterly, 2011. It put the pay premium at around 18 percent and the involuntary exit rate around 61 percent higher, both measured against workers promoted into the same jobs. That is a peer-reviewed finding rather than a vendor survey, and it points the same way everyone selling training claims.
The second argument is defensive rather than aspirational. Jobs change whether or not you plan for it, and federal employment projections show which occupations are expected to grow and shrink over the coming decade, occupation by occupation, at no cost (Bureau of Labor Statistics).
Looking up the roles on your own payroll takes ten minutes and gives you a checkable answer to where the ground is moving, rather than a survey figure about skills changing by some future date.
There is a third argument that applies only at small scale and is rarely made. In a small company, a person who can do two jobs is not a nice-to-have but the reason the business can absorb an illness, a resignation, or an unexpectedly good month.
Capability depth is an operational resilience question long before it is a talent-development one, which is why so much of what gets filed under small business HR is really just risk management with a friendlier name.
The Benefits of Upskilling and Reskilling
The three moves pay back in different currencies, which is the practical reason for naming which one you are running. Upskilling buys capability you would otherwise hire or send out, reskilling buys a person you keep instead of replacing, and cross-training buys the ability to survive an absence. Each also pays the employee something, which is why the training gets used when nobody is watching.
| What the business gets | Which move delivers it | What it looks like | What the employee gets |
|---|---|---|---|
| Capability you stop buying in | Upskilling | The monthly invoice to a contractor, an agency, or an outside specialist shrinks or stops, because the work now happens inside the business | Harder work with more judgment in it, and a defensible case for a raise |
| Cover when somebody is out | Cross-training | A second person runs the thing that used to stop when one person was sick or away, so a week of leave gets planned rather than negotiated | Time off without the phone ringing, and a wider view of how the business fits together |
| A person kept rather than replaced | Reskilling | A shrinking role ends without a separation, and the context that person carries about your customers and systems stays in the building | A future here instead of a layoff notice |
| Speed and cost against hiring | Upskilling and reskilling | No agency fee, no vacancy, and no first quarter spent learning how this particular business works | A route upward that does not require leaving to take it |
| Retention, because the next step is visible | All three | People with something to learn next stop treating the job as a stepping stone to somewhere else | Skills that outlast this job and transfer to the next one |
Upskilling sits inside a wider training and development plan rather than replacing it, and it is usually the line with the clearest payback. None of these benefits arrive on their own. Each depends on the work actually moving to the person once they are able to do it.
Nick Anisimov, FirstHR founder
Deciding Which Skills to Build
Every guide recommends a formal skills-gap analysis and almost no small business ever completes one. Five questions get you the same answer in an afternoon, and they are ordered by how obvious the return is.
The first question is the one to act on. A recurring invoice for work you send out is a business case that requires no modeling: you know the annual cost, you know the frequency, and you know exactly what capability would replace it.
The work that most often shows up on those invoices is bookkeeping, basic design, simple contract review, routine maintenance, and payroll administration, and at least one of them is usually a bad fit for outsourcing at your scale.
The second question produces a different kind of answer, and it deserves the same weight. Single points of failure are not a training aspiration but an operational risk that already exists, and cross-training is the cheapest insurance available against it. The test is blunt: name the person, name the task, and ask what happens on the Tuesday they call in sick.
Digital Upskilling
Digital upskilling is the version most small businesses actually need, and it is narrower than the phrase suggests. It is not a curriculum in data science. It is the specific software the business already pays for, used properly, by more than one person.
Start with your own subscription list. Most teams use a fraction of what they license, and the gap between a competent user and a fumbling one on a system you already own is usually wider than anything a new tool would buy. An hour of structured practice on the accounting package beats a course nobody will apply.
The second layer is judgment about tools that write or summarize: when the output has to be checked, what should never be pasted into a public system, and which tasks are worth automating at all. That is closer to a policy question than a skills one, and AI training for employees works best treated as an ordinary training target rather than a separate category.
Staying future-ready is mostly a matter of keeping the list current. Once a year, look again at the tools you pay for and at the federal outlook for the roles on your payroll, then move one item onto the training list. That is a much smaller exercise than a workforce strategy document, and it is the one that gets done. Where the gap is genuinely technical, price the hours before the course.
There Is Public Funding for This
Federal workforce legislation funds training for people who already have jobs, not only for jobseekers, and the cost-sharing tiers favor small employers. This is the least known fact in the whole subject and the one most likely to change what you can afford.
The mechanism is called incumbent worker training. Local workforce development boards may reserve up to 20 percent of their combined adult and dislocated worker allocations to fund it (20 CFR 680.800), and employers participating are required to pay a non-federal share of the cost, meaning the part the federal funds do not cover.
The Workforce Innovation and Opportunity Act (WIOA), the statute behind the program, sets the floor for that share by headcount: not less than 10 percent for an employer with no more than 50 employees, 25 percent between 51 and 100, and 50 percent above 100 (WIOA section 134(d)(4)(D), 29 U.S.C. 3174).
| 50 or fewer employees | At least 10% of the training cost The most favorable tier, and the one most small businesses sit in |
| 51 to 100 employees | At least 25% of the training cost The share rises with headcount under the statutory tiers |
| More than 100 employees | At least 50% of the training cost Larger employers carry at least half the cost |
The worker has to qualify too, not only the employer. Under the regulation, an incumbent worker needs to be employed, to meet the wage-and-hour requirements for an employer-employee relationship, and to have an established employment history with the employer of six months or more. When training a cohort, not everyone needs six months as long as a majority of those being trained do (20 CFR 680.780).
The employer also makes a commitment beyond the cost share. The regulation describes incumbent worker training as conducted with a commitment by the employer to retain the workers trained or avert their layoff (20 CFR 680.790), so the person you put forward should be someone you intend to keep.
Two practical notes. The employer contribution can often be met in kind rather than only in cash, which matters when what you have is staff time rather than budget. And approval generally has to be in place before training begins, so look into it while you are still planning rather than after you have paid an invoice.
Reskilling: Moving Someone Instead of Replacing Them
Reskilling employees means training someone for a materially different job, usually because the one they hold is shrinking. It is slower and costs more than upskilling, and the honest comparison is not against doing nothing. It is against a separation and a hire.
The trigger is normally visible months ahead. A function is being absorbed by software, a client segment is going away, or a service line is closing. From there you have three options: retrain the person, replace them with a hire, or let the role go. Retraining wins when the person carries context you cannot buy, and in a small business most people do.
Federal workforce policy treats this as a category of its own. Rapid response regulations (the federal rules for responding to closures and mass layoffs) define layoff aversion as strategies to prevent or minimize the duration of unemployment resulting from layoffs. They list developing and funding incumbent worker training and other worker upskilling approaches among the appropriate activities (20 CFR 682.320). A plan that keeps someone employed is what a local board is funded to support.
| Settle this first | Why it decides the answer | What a no means |
|---|---|---|
| Does the destination role exist here? | Training toward a job you will not create is a course with no landing place | Do not start. Help them find it elsewhere while they are still paid |
| Does the person want it? | A different job is a different career, and consent is not the same as compliance | You have converted a retention plan into a resignation with a delay |
| How much of the current work transfers? | Adjacent moves take weeks, unrelated ones take quarters | Budget the longer number, and say that number out loud |
| Who covers the old job meanwhile? | Nobody holds both, and pretending otherwise is how both fail | Reduce the old workload formally, or delay the start date |
| What happens if it does not take? | Decide it while everyone is calm rather than at month four | Agree on the review date and the fallback in writing, at the start |
The sequencing that works is to start the new work before the old work ends. Give the person a real piece of the destination job while they still hold the current one, at a reduced load, so capability and confidence build against actual output rather than against exercises. Workforce reskilling in a small business is one person at a time, done properly.
Say the reason plainly as well. Tell someone their role is being rebuilt around a change they can already see, and they become an ally in it. Move someone without explanation, and they read it as a demotion and start looking, which turns a retention plan into the turnover you were trying to avoid.
Building the Program
An upskilling program at small scale is a short list of named people, named skills, protected hours, and a date by which each person demonstrates the thing. Everything beyond that is administration.
Those eight steps produce a record, and it is worth keeping as one. Three sheets, one row per training decision, carried from the decision itself through what it cost to whether it worked.
| A | B | C | D | E | F | G | H | I | J | |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Capability we cannot currently cover | Why it matters now | Move: upskill, reskill, cross-train, or hire | Who is being trained | Who supports them | Protected hours per week | Start | Demonstrates it unsupervised by | What they own afterward, and who it comes off | Status |
| 2 | ||||||||||
| 3 | ||||||||||
| 4 | ||||||||||
| 5 | ||||||||||
| 6 | ||||||||||
| 7 | ||||||||||
| 8 | Note | Three live rows at a time. A fourth means none of them has protected hours | ||||||||
| 9 | Note | Name the move before the method. Hire is a legitimate entry and saves you training toward something you should buy | ||||||||
| 10 | Note | A row with an empty handover column is a row that will decay, whatever the demonstration says |
The first sheet holds the decision: the move, the person, the supporter, the protected hours, the demonstration date, and what they own once they can do it. The second holds the cost the business case usually omits, with the learner and supporter hours priced separately from the fees and a column for whether the funding was checked before anything was spent. The third holds the baseline and the two checks against it.
Keep the record beside whatever training matrix you already maintain rather than inside it, because the matrix shows who can do what, while the record shows what you decided to do about it.
Three targets is a deliberate constraint rather than modesty. A plan with ten items is a plan where nothing has protected hours, because the hours available did not change when the ambition did. Finishing three and then choosing three more produces more capability in a year than starting ten, and it has the side effect of making the program visibly work, which is what buys you the hours for the next round.
The last of the eight steps, handing over the work, is the one people skip and the one that determines whether any of this survives. Building a capability and then leaving the work where it was is how you get a trained person doing their old job, which is indistinguishable from not having trained them.
Deciding the handover in advance also makes the training concrete, and it belongs in whatever development plan the person has rather than only in your head.
Low-Cost and No-Cost Tactics
Most of what works in a small business costs nothing in fees and something in hours. That trade is usually the right one, because the alternative is a license you pay for whether or not anyone opens it.
Stretch assignments with a review are the highest-return item on that list, and they are underrated because they do not look like training. Handing someone the next-harder piece of real work, with an explicit agreement that you will review it before it leaves the building, produces capability and a finished deliverable at the same time. The review is not optional. Without it the assignment is a gamble on a live piece of work.
Teach-back sessions come second, for a reason that is easy to miss: the person doing the teaching gains more than the audience. Sending one person to something and requiring them to present it back turns an individual expense into a team one and gives the attendee a reason to pay proper attention.
Structured cross-training deserves the word structured. The difference between cross-training and someone occasionally helping out is a written list of the tasks, a date by which the shadow performs them alone, and a person watching when they do. Without those three things it is goodwill rather than coverage, and goodwill does not survive the Tuesday somebody is genuinely unavailable.
Where a task carries compliance weight, record the sign-off with the rest of your employee records rather than in a chat message.
The Real Cost
The number that sinks upskilling plans is never the course fee. It is paid hours, and it is missing from almost every business case because the fee is the thing with an invoice attached.
Required job-related training is hours worked. Federal rules treat attendance at training as working time unless four conditions are all met: it takes place outside regular working hours, attendance is voluntary, the training is not directly related to the job, and no other productive work is performed during it (29 CFR 785.27).
Training you asked for and follow up on fails at least two of those four conditions, so the hours count. For non-exempt staff (employees entitled to overtime pay), those hours also count toward overtime under the Fair Labor Standards Act.
| Cost line | How to estimate it | Usually in the plan |
|---|---|---|
| Course, certification, or materials | The invoice | Yes |
| The learner's paid hours | Hours per week multiplied by weeks multiplied by rate | Rarely |
| The supporter's paid hours | Their hours beside the learner multiplied by their rate | Almost never |
| Coverage while they are training | What does not get done, or who covers it | No |
| The productivity dip while applying it | Slower for the first few weeks of real work | No |
| Rework on early attempts | Real and temporary, and worth budgeting once | No |
| Funding offset | Subtract, if the local board approves it | Almost never |
Run the arithmetic once and the picture usually improves rather than worsens, which is the point of doing it. Four hours a week for eight weeks is thirty-two hours, which is under one week of a person's year. Compared against a recurring outsourcing invoice or the cost of hiring, that is frequently a favorable trade, and it is a trade you can only see once the hours are on the page.
Knowing Whether It Worked
Measure the thing you were trying to change, not course completion. Completion tells you somebody sat through something, which is the least interesting fact available.
| What you were trying to do | The measure | When to look |
|---|---|---|
| Stop paying an outsider | Did the invoice stop or shrink | Next billing cycle after handover |
| Remove a single point of failure | Has a second person done it unsupervised | Day 30, then a real absence |
| Clear a permanent backlog | The size of the queue against baseline | Day 90 |
| Prepare for a role change | Is the person doing part of the new role yet | Day 60 |
| Retain someone who was restless | Are they still here and what are they asking for now | Day 180 |
| Reduce errors or rework | Rework volume against the pre-training figure | Day 90 |
Write the baseline down before the training starts, because reconstructing it afterward is impossible and estimating it is self-serving. Then check twice: once at around thirty days for whether the behavior appeared at all, and once at ninety for whether the outcome moved. That two-stage check is the useful core of the standard evaluation frameworks.
Be honest about attribution. If the queue cleared and you also hired someone that quarter, the training is one of two explanations and probably not the larger one. Naming the alternatives yourself is what keeps the finding credible when somebody asks.
Where This Goes Wrong
The failure patterns are consistent and almost all of them involve treating training as an event rather than as a transfer of work.
Training with no handover is first and the most wasteful. The capability gets built, the work stays where it was, and within a few months the skill has decayed and the team has learned to treat training as a formality.
Not protecting the hours is second. Training expected to fit around a full workload does not happen, and the person who fails to complete it is then treated as the problem rather than the schedule.
Deepening a single point of failure is third and is a genuinely counterproductive move. Upskilling the one person who can already do the critical thing makes your dependency on them stronger. If that person is the answer to what stops when they are away, the correct move is cross-training somebody else.
Skipping the funding check is fourth, and it is free to avoid. One phone call to a local workforce board before you spend anything is the entire cost of finding out whether a large share of the bill is covered.
Choosing skills from a catalog is fifth. Starting with what courses exist rather than with what your business cannot currently do produces a plausible-looking program aimed at nothing in particular.
And letting fear of departure decide it is last. Someone might leave after you train them. Someone will certainly leave eventually, and a business where critical skills exist in exactly one head is more exposed to that, not less.
The way to hold on to people you have invested in is a role that grows and a clear picture of what comes next, which belongs in your workforce plan alongside the hiring you are trying to avoid.
Whatever you run, keep the record. Who was trained on what, when, whether they demonstrated it, and what they own now is the institutional memory that stops you paying twice for the same capability. Keeping it beside the roster in a single employee directory is what makes it retrievable when the person who arranged it has moved on.
Frequently Asked Questions
What is employee upskilling?
Employee upskilling means teaching the people you already employ to do their present job better: taking on harder work, making more of the judgment calls, or getting fluent with a tool the role now depends on. The person stays in the same job and grows deeper in it. Reskilling prepares someone for a different job altogether, typically because their current one is shrinking. Cross-training is the third move: one person learns a colleague's job well enough to step in for them. Put simply, the first adds depth, the second points a career somewhere new, and the third protects the business when someone is away.
What is the difference between upskilling and reskilling?
Upskilling deepens capability within the role someone already holds. Reskilling moves a person into a different role. The practical distinction is whether the job title stays the same. If a bookkeeper learns to run the month-end close rather than prepare it, that is upskilling. If the same person moves into customer support because the bookkeeping is being automated, that is reskilling. Upskilling is generally cheaper and faster because it builds on existing context; reskilling costs more and is usually the alternative to a layoff.
How do you decide which skills to upskill?
Five questions will get you most of the way, and they fit in an afternoon. Start with the work you pay outsiders for, because a recurring invoice tells you both what the training is worth and what it has to teach. Next, name the tasks that halt when one particular person is out; those single points of failure are the targets with the highest return. Then find the backlog that never shrinks, since it usually points to a bottleneck at one desk. After that, check the federal government's published job outlook to see which of your roles are shifting. Finally, ask what people requested at their last review, because a skill someone chose gets practiced without anyone standing over them.
Is there funding available to train existing employees?
Yes. Federal workforce law funds a program called incumbent worker training, which pays for training people who already have a job with you, not just people looking for work. Each local workforce development board can set aside as much as 20 percent of the money it receives for adult and dislocated worker programs to pay for it. Your business covers part of the cost, and the statute sets the smallest required share for the smallest employers: with 50 or fewer employees, you must put in at least 10 percent. Local boards run the program, so the terms, caps, and application steps differ from one area to the next.
Who qualifies as an incumbent worker for training funding?
The federal regulation defines an incumbent worker as someone who is currently employed, whose working arrangement qualifies as an employer-employee relationship under federal wage-and-hour law, and who has worked for that employer for at least six months. Group training gets some flexibility. When you put a cohort through together, a majority of the group has to meet the six-month mark, but not every individual does. States and local areas add their own policies on which workers and groups are eligible, so your local workforce board has the final word on the details.
How much does upskilling cost a small business?
The visible cost is course fees or certifications, and it is usually the smaller line. The larger one is paid hours, because required job-related training counts as hours worked for the employee and often for whoever is teaching them. If someone spends four hours a week for eight weeks on a new skill, that is thirty-two paid hours, plus the coverage gap while they are away from their normal work, plus the supporter's time. Public funding can cover much of the direct cost but not the hours, which is why the hours belong in the plan from the start.
Do you have to pay employees for time spent upskilling?
Yes, when the training is required and tied to the job. Under federal hours-worked rules, time in training counts as paid working time unless all four of these hold together: the session falls outside the employee's normal schedule, going is truly optional, the content is not directly related to their job, and they do no productive work while attending. Training a person picks for themselves, on their own time and in a subject unrelated to their current role, can fall outside paid hours. Training you requested, put on the calendar, or check up on afterward will generally count as hours worked.
How do you measure whether upskilling worked?
Judge it by the outcome the training was meant to produce, never by whether someone finished a course. When the aim was to bring outsourced work in-house, see whether the outside invoice ended or got smaller. When it was to end a single point of failure, see whether a second person has now done the task without help. When it was a backlog, compare the queue with where it started. Record that starting figure before training begins, look again around day thirty for the new behavior and around day ninety for the result, and list anything else that changed in the meantime before giving the training the credit.
What if you upskill someone and they leave?
Some people will, but buying the skill from outside instead is the worse bet. A peer-reviewed study that followed people moving into the same positions at one US investment bank, some hired externally and some promoted internally, found the outside hires earned substantially more than their internal counterparts and were considerably more likely to be fired or let go. Turnover happens sooner or later whether you invest in people or not. Worrying that a trained employee might walk out is a weak reason to leave your business relying on skills that sit with a single person.
Make HR the easy part of running your business
Onboarding, e-signatures, training and core HR in one place. Up and running in 5 minutes.
- 7-day free trial
- No credit card required
- Cancel anytime