| Employee efficiency is how much useful output a person produces for the time and effort they put in, so anything that consumes paid time without adding to the result reduces it. Technology is one of the largest drains and the easiest to measure, because it costs the same few minutes to everybody who runs into it, every day. |
Ask an owner where the team’s time goes and you will hear about meetings, interruptions and email. Ask the team and you hear something else.
The laptop that takes six minutes to become usable. Four separate logins before morning tea.
An order typed into the job system, then typed again into the accounting package. A file that will not open over the office connection, so someone gives up and comes in on Saturday.
None of it is dramatic. Each costs a few minutes, so nobody raises it, and none of it reaches any report you see.
Added up across a year and a payroll, it is one of the largest uncosted expenses in a small business.
Employee efficiency is managed as a people question. A large share of it is a systems question, and systems are easier to change than habits.
You can find most of it in a week of paying attention. Some of the fixes take days and some take months, and none of them are interesting.
What Is Employee Efficiency?
Employee efficiency is the useful output a person produces for each hour they work. It rises when the same person completes more of the work the business is paid for, and it falls when hours disappear into activity that adds nothing to the result. Not all unbilled work is waste. Training, maintenance, security checks and internal coordination are necessary. Waiting on a system is not.
Employee efficiency carries a commercial edge that owners feel long before they can name it. Two firms of the same size, in the same trade, quoting the same job, can sit twenty percent apart on margin because one of them loses less of every paid hour to friction.
The framing also moves the conversation away from effort. Nobody in a small New Zealand business is sitting idle, and the share of each paid hour reaching real work varies far more between firms than effort does.
What is the difference between efficiency and productivity?
Productivity relates output to the inputs used to produce it, and labour productivity is usually measured as output per hour worked. Efficiency asks how well time and other resources were used to reach the required result, including quality. Total output on its own is neither, and the two measures can move in opposite directions for a year before anyone notices.
A team can lift total output by working longer while its output per hour, and its efficiency, both fall. Employee productivity and efficiency only move together when the extra output comes out of the same hours.
For an owner the practical difference shows up in staff turnover. Output bought with longer days is borrowed, and the loan gets repaid in burnout and resignations.
Why does efficiency get blamed on people first?
Because it is the only part of the problem you can see. A staff member who seems slow gets noticed in a week. A login process that costs ninety seconds, repeated eight times a day by everybody in the building, never gets mentioned at all.
Anything everyone experiences stops registering as a problem. It becomes the way things are here, and it stays that way until somebody counts it and puts the total in front of the person paying for the hours.
That is why the first useful step is measurement. A number turns the problem into a list of jobs with owners and dates against them.
Where Does Employee Time Go in a Working Day?
Most of a desk worker’s day goes into communication and coordination, and what is left arrives in pieces too small for concentrated work. Microsoft measured this across its own platform in 2025.
Its study, Breaking down the infinite workday, draws on two separate sources. Anonymised Microsoft 365 telemetry shows the average worker receiving 117 emails and 153 Teams messages every weekday, most of them skimmed in under sixty seconds. For the busiest fifth of users by message volume, meetings, emails and chats arrive at a combined rate of one every two minutes during core hours.
Alongside the telemetry sits a separate survey of 31,000 knowledge workers across thirty-one markets, New Zealand among them, all of them people who work at a desk. Neither source is limited to large organisations, and neither is specific to New Zealand small businesses.
Nobody in a ten-person firm has a job title that includes protecting anyone else’s calendar. The person losing an hour to a slow system is the same person who quotes the work, invoices it and answers the phone.
What does a single interruption cost?
Considerably more than the interruption itself. The cost sits in what researchers call context switching, the mental reset needed to pick a complex task back up. How long that reset takes varies with the task and with the interruption, which is why a short question can cost a good deal more than the time it took to ask.
This is why an afternoon that felt busy can produce almost nothing, with the time spread across a dozen ten-minute pieces.
Which of these are IT problems and which are habits?
Meeting load, message volume and interruption culture are usually management habits. Slow devices, repeated logins, systems that do not talk to each other and unreliable connections are usually IT problems. Some sit in both, such as a meeting that starts late because the room equipment fails.
The line is worth drawing early, because the two need different owners. Habits change by agreement inside the team. Systems need a budget and somebody’s name against them.
This post stays on the second list, because it is the half an IT provider can put a number against and fix on a timeline.
How Much Does Lost Time Cost a Business?
At New Zealand median wages, one hour a week per person costs about $1,656 a year for each employee before overheads. Across a team of twenty that single hour is roughly $33,000.
Those figures come from Stats NZ. In the June 2026 quarter median hourly earnings from wages and salaries were $35.96. The table below models $36 an hour, five working days and 46 working weeks, which allows for annual leave and public holidays. Those are modelling assumptions rather than measurements, so use your own labour costs wherever you have them.
| Time lost per person | Cost per person, per year | Team of 10 | Team of 25 |
|---|---|---|---|
| 15 minutes a day | $2,070 | $20,700 | $51,800 |
| 30 minutes a day | $4,140 | $41,400 | $103,500 |
| 1 hour a day | $8,280 | $82,800 | $207,000 |
The figures are built on gross pay, so what the business actually carries is higher once KiwiSaver contributions, ACC levies and leave are added. If you substitute a fully loaded hourly rate, do not add those costs again on top. The figures also assume the time you win back goes somewhere useful, which does not happen unless somebody directs it.
Keep the causes separate when you build a case. Take an eight-person engineering consultancy. Suppose four people lose ten minutes a day waiting on ageing machines, and six lose ten minutes a day entering the same job into two systems. At $36 an hour the device delays are worth about $5,500 a year and the re-keying about $8,300. Replacing four laptops is then a decision about the $5,500, because those are the people it affects. The re-keying needs its own fix and its own figure. Both are illustrative, and the same arithmetic runs on your own payroll.
Two terms are worth separating. Downtime is the time a system is unavailable to everybody, and it gets reported. Utilisation is the share of paid hours reaching productive or chargeable work, and it absorbs all of this without ever being itemised.
How do you work out the figure for your own team?
Take your average hourly wage cost, multiply by 46 weeks, then by the staff affected, then by the hours a week you believe are lost. Do not chase precision. An estimate that is roughly right and defensible will win a budget argument that a vague sense of frustration never will.
Which hours do you actually get back?
Not all of them, so do not build a business case on the full figure. Removing a five-minute wait does not hand back five minutes of finished work, because attention takes a moment to return.
Rather than settle on a single recovery rate, run the case at more than one. Model it at a quarter, a half and three-quarters of the time removed, say which figure you are using, then measure what actually happened.
It also helps to be clear about what kind of result you are claiming. Released capacity means staff spend less time waiting or repeating work. A cash saving means overtime, contractor cost or other spending actually falls. Additional output means more work is completed with demand and quality maintained. Those are three different claims and each needs its own evidence.
What Causes Low Employee Efficiency?
A good share of the causes are technical rather than personal, and the same five turn up again and again. Each one is charged against everybody who runs into it, on every day they do.
- Ageing devices. A computer past its useful life costs minutes at every start-up, every application launch and every large file.
- Too many applications. Staff keep a dozen tools in their head and lose time working out which one a given job belongs in.
- Repeated logins. Every separate credential is a small tax collected several times a day from everyone who holds it.
- Manual re-keying. The same information entered into two systems, with a second chance to get it wrong.
- Weak connections and meeting equipment. Nine minutes lost at the start of a call is nine minutes multiplied by everyone waiting in the room.
None of these is worth complaining about on its own. They only become visible when someone multiplies them by headcount and by 46 weeks.
They also compound. A slow device makes each of the other four worse, because every login, every application switch and every file transfer waits on the same processor.
How old is too old for a work computer?
Three to four years for most office roles, and less where the work is heavy. The useful test is whether the machine still keeps pace with the person using it, measured in start-up time and application load time.
A harder deadline sits over part of most fleets. Standard support for Windows 10 version 22H2 ended on 14 October 2025. Eligible devices can be enrolled in Extended Security Updates, and a few editions run on their own lifecycle, so check what each machine is actually entitled to before you plan its replacement.
Those devices are frequently the slow ones as well, which makes the decision easier to sign off. Our guide to Windows 10 end of life covers the upgrade paths and the hardware that cannot take Windows 11.
How many applications is too many?
The total matters less than the overlap. Two tools doing the same job cost time daily, because staff have to work out which one holds the current version before they can start.
Sprawl carries a licence bill nobody is reading as well. SaaS management covers the inventory side, including subscriptions still charging for people who left eighteen months ago.
The question worth asking is narrower than the licence count. For each task a person does weekly, is there one obvious place to do it?
How Do You Measure Employee Efficiency?
Three sources will tell you, and none of them needs new software: ask the staff, read the helpdesk log, and sit with one person for an hour.
What should you ask staff?
One question, asked privately: what part of your job takes longer than it should, and what makes it slow?
Ask it one to one. In a group session it turns into a discussion about meetings, which belongs to somebody else.
Write the answers down without solving any of them yet. The pattern across ten people matters more than any single complaint.
What can the helpdesk log tell you?
It shows which faults keep coming back. A printer that fails every week costs fifty interruptions a year, spread across whoever happened to need it that morning.
Ask your provider for tickets grouped by recurring cause rather than by month. That report is the heart of proactive IT reporting, where the point of the work is to stop the tickets being raised.
A provider who cannot produce that report has told you something useful about how the account is being run.
Is watching someone work for an hour worth it?
Yes, and it is the step owners skip. Sit beside one person for an hour and count the waits: the loading, the logging in, the switching between windows, the re-typing.
You will see things nobody thought to report, because they have done them the same way for two years and stopped noticing.
Is there a quicker version of all this?
A one-week tally gets you most of the way. Give everyone a sheet of paper, ask them to mark it each time they wait on a system, and collect the sheets on Friday.
It is a crude instrument and it produces a number nobody in the room argues with.
Do Employee Monitoring Tools Improve Efficiency?
Not on their own, and not for the reason the category is sold on. Monitoring software records what a person did and for how long, which describes the shape of the problem without reaching its cause.
A dashboard showing four hours in a spreadsheet does not say how much of that went into waiting, re-typing and hunting for the current version of a file. The waiting looks like work.
There is an observer effect as well. Staff who know an activity is counted optimise for the count, and the reporting improves while the output does not.
Monitoring has narrow uses: billing client time, meeting a record-keeping duty, or dispatching field staff. As a way to lift output in an office of twenty, it measures the symptom and leaves the cause running.
What should you spend that budget on instead?
The three sources above, and the fixes they point to. A week of asking, reading tickets and watching produces a shorter list than any dashboard, and every item on it has an owner.
Which Fixes Return the Most Time?
The cheap ones that affect everybody, which is rarely where the budget goes. Match the symptom to its cause before buying anything, because a good deal of efficiency spending goes on tools that solve a problem the business did not have. The table below covers what we see most often.
| What staff say | What to investigate | Where the fix sits |
|---|---|---|
| It takes forever to start up | Device past its useful life | Hardware refresh and lifecycle planning |
| I log in over and over | Sign-on, session settings or identity configuration | Identity and access setup |
| I type the same thing twice | Two systems that do not talk to each other | Integration or workflow automation |
| Files are slow to open | Connection or storage bottleneck | Network and internet review |
| Meetings start late | Room equipment and joining friction | Meeting room technology |
| It has broken again | Recurring fault with no root cause found | Proactive support and ticket analysis |
Investigate the cause before changing anything, and do not weaken multi-factor authentication to cut the number of prompts. Fewer sign-ins should come from single sign-on and sensible session lengths, not from removing a control.
The cheapest items in that table are the ones nobody asks for. A slow laptop feels like a personal failing, so people put up with it and ask for new software instead.
Budgets should follow the same order, with the cheap universal fixes first and the larger projects after them.
Where does AI fit into this?
None of the rows in that table are AI problems. Repeated logins, duplicate entry and recurring faults are configuration and integration work, and rules-based automation handles the predictable parts: moving a record between two systems, raising a notification, routing an approval.
AI earns its place on the variable language work that sits alongside those rows, such as drafting and summarising, where the output still has to be checked and the net time saved still has to be measured.
Adding an assistant on top of a machine that takes six minutes to wake up moves the bottleneck without removing it. AI productivity tools sets out which ones earn their licence fee.
How Do You Improve Employee Efficiency?
Start with the faults that affect everyone every day, and work down in this order:
Fix the worst start-up and load times
Diagnose before you buy, because a slow machine can be storage, software, configuration or a network dependency as readily as age. It is the least interesting line item on the list and the one staff notice within a day.
Reduce the number of logins
Single sign-on across your main applications removes several of those daily interruptions outright.
Retire one duplicate system
Pick the pair doing the same job and switch one off, with a date attached.
Automate the worst re-keying job
Start with the one people complain about by name.
Remove the top three recurring faults
Take them from the ticket log and fix the cause, not the symptom.
Ask the same question again after ninety days
The answers tell you whether the time came back and where the next hour is hiding.
None of that is a transformation programme. It is six pieces of deferred maintenance, and it is usually cheaper than the new tool being proposed alongside it.
Two things worth doing this week
Find out the age of the oldest computer still in daily use, and how many separate passwords a new starter is issued in their first week.
Both answers should take one email, and both are usually worse than the person asking expects.
Find Out Where Your Team’s Time Is Going
Exodesk has supported New Zealand businesses since 1989, from offices in Christchurch and Dunedin. We find where technology is costing your team time, put a figure on it, and work through the causes in order of what each returns. It sits inside our ongoing IT support.
Contact us today to discuss how we can help your business or connect with us on LinkedIn to stay updated with more insights.
Frequently Asked Questions
What is a good employee efficiency rate?
There is no published benchmark worth relying on for office work, because the output of a knowledge job resists a single number. Track two figures of your own instead: the hours a week each person reports losing to systems, and whether that number falls after each fix. A rate you calculated yourself and can repeat beats an industry average you cannot check.
What is context switching and why does it matter?
Context switching is the mental reset needed when attention moves from one task to another, and it costs several minutes each time before full concentration returns. It is the reason a day of small interruptions produces so little finished work. Microsoft recorded meetings, emails and chats arriving every two minutes during core hours for the busiest fifth of its users by message volume.
How much does poor employee efficiency cost a business?
At the Stats NZ median hourly earnings figure of $35.96 for the June 2026 quarter, one lost hour per person per week costs about $1,656 a year for each employee, or roughly $33,000 across a team of twenty. Half an hour a day per person takes that same team past $82,000. Both figures are built on gross wages, and the number the business actually carries is higher once leave and employer contributions are counted.
Is employee efficiency an IT problem or a management problem?
Both, and it is worth deciding which one you are looking at before you act. Meeting load, message volume and interruption culture belong to management and change by agreement. Slow devices, duplicate systems, repeated logins and unreliable connections belong to IT and change when someone is funded to fix them. Starting with the second list is easier, because the results are measurable within weeks.
How old should a work computer be before it is replaced?
Three to four years suits most office roles, and heavier work shortens that. What matters more is whether the machine keeps pace with the person using it, measured in start-up and application load time. Machines still running Windows 10 have a separate reason to go, because standard support ended on 14 October 2025. Extended Security Updates can cover eligible devices for a limited period, and some editions have their own lifecycle dates, so confirm what each machine is entitled to rather than assuming the updates have stopped.
How do you raise employee efficiency without upsetting staff?
Fix the systems before you examine the people. Staff rarely object to a faster laptop, fewer passwords or one less place to enter the same data, and the people doing the work are the ones who can tell you where the time goes. Treat what they report as evidence to investigate.
How many business applications should a small team use?
There is no correct number, and duplication matters far more than the count. Where two products cover the same ground, staff pay a decision cost every time they open either one. Aim instead for a single obvious home for every task somebody performs each week.
Will AI tools fix workplace efficiency?
Partly. An assistant can take real time off drafting and summarising, provided somebody checks the output and measures what was actually saved. Moving data between two systems is usually a job for rules-based automation rather than AI. Neither has any effect on a slow laptop, a weak connection or a login process with six steps, so the foundations have to come first. Bought too early, the licence becomes another subscription nobody opens.
How long does it take to see an improvement?
Device replacements and single sign-on show up within days, because staff feel the difference at every start-up and every login. Consolidating systems and automating re-keying take one to three months, since they change how work moves between people. Measure again at ninety days against the answers collected at the start.
What is the difference between employee efficiency and workforce productivity?
Workforce productivity relates a team’s output to the inputs behind it, most often as output per hour worked. Employee efficiency asks how well time and other resources were used to reach the required result, including quality. A team can raise total output by working longer hours while both its output per hour and its efficiency fall.
Does hybrid or remote work reduce employee efficiency?
Not by itself, since remote staff report better focus where there are fewer interruptions and what suffers instead is coordination. What decides it for a hybrid team is whether the connection, the devices and the meeting equipment make joining from anywhere effortless.
How can Exodesk help improve employee efficiency in Christchurch or Dunedin?
Exodesk assesses where technology is costing a team time, puts a figure on it, and works through the causes in the order that returns the most hours. That covers device fleets, logins and identity, duplicated systems, connectivity and recurring faults. Teams in Christchurch and Dunedin support businesses across New Zealand, and Exodesk has operated since 1989.
How many hours a week is your setup costing?
Nobody budgets for the slow login, the second system holding the same data, or the fault that comes back every month. A free IT assessment sets out what you are running and where the time is going, so the fixes can be ranked by what they give back.
Or read more about managed IT services.
