Data Entry Test for Employment: Test Accuracy, Not Just Speed

A data entry test for employment should check accuracy and instruction-following, not just speed. Sample tasks, what to look for, and a supervised screen.

What the Admin Skills Screen is: a supervised, recorded ten-minute exercise of real admin tasks, with single-use codes, that gives your team a score and a plain description, never a verdict. What the screen covers and how it works · All guides

A useful data entry test for employment checks whether the candidate enters the right value, in the right place, following the instruction as written, and only then asks how fast they do it. Speed tests are easy to find and easy to run, but a fast typist who transposes a date or skips a step in the procedure creates work rather than finishing it. The most revealing check is a short supervised work sample of exactness tasks, with a separate speed sample only if your volume genuinely requires one.

This page covers what a data entry test should measure, sample tasks, how to read the results, how to run the ScreenYourAdmin screen, and what to add for high-volume roles.

What data entry work actually requires

The title suggests typing. The job, in most businesses, is a set of small accuracy decisions made over and over:

• Reading the source as it is, not as expected. The invoice says 14th; the brain, used to month-end invoices, enters 31st.
• Following the procedure in order, including the exception. "Enter these as usual, but international orders go in the second tab and need the currency code."
• Applying the field rules. Dates in one format, names surname-first, amounts without the currency symbol. Simple rules, applied to every record without being reminded.
• Holding a value between screens. Read a reference number on one window, type it into another a few seconds later.
• Noticing and fixing an error rather than copying it through.

A candidate who does those things at a modest pace produces clean data. A candidate who does them inconsistently at a high pace produces a cleanup project.

Why speed tests alone mislead

Words-per-minute and keystrokes-per-hour tests measure one thing well: how fast the candidate's fingers move on the test's text. They have three weaknesses as a hiring tool for most data entry roles.

They reward the wrong error profile. A speed test that subtracts a small penalty for errors can still rank a fast, careless typist above a slower, accurate one. Your downstream systems do not apply that discount.

They test copying, not following. Retyping a paragraph does not show whether the candidate reads and applies an instruction, which is where most real data entry errors begin.

They are usually unsupervised. A speed test taken at home tells you what someone produced at home. You do not know who typed it. The case against unsupervised online tests applies here as much as anywhere.

Speed matters in some roles. It is just the second question, not the first.

Sample accuracy tasks

The Admin Skills Screen is a ten-minute work sample built from everyday admin tasks, and several of them map directly onto data entry habits. Every attempt draws its own words, numbers, days and names, so the examples here are illustrations, not live items.

Correct the stated error. "This sentence has an error: 'The order ships Friday, not Wednesday.' Retype it, replacing 'Friday' with 'Wednesday'." Candidates still leave the wrong day in, or add the right day without removing the wrong one, or retype the sentence unchanged. Each of those is a data entry error in miniature: the correction was instructed and did not happen.

Follow the stated order. "Do the following in the order 1, 3, 2, 4: (1) type 'north', (2) type 'south', (3) type 'east', (4) type the total number of words you have typed so far." Data entry procedures have an order for a reason. This item shows whether the order is read or assumed.

The conditional step. A four-step instruction in which step three applies only if something has not already been typed, and step four inserts a word in front of the first one. It is the field that is filled only when another field is blank.

Apply a field rule. "Type your first name. If it has more than five letters, mark it 'underline'; if five or fewer, mark it 'circle'." The rule is easy. The finding is whether it is applied at all.

Sort and carry a value forward. "Here are three numbers: 47, 12, 31. Write them from smallest to largest, then give the middle one." Several items later: "Which number did you identify as the middle value?" Holding a reference between screens uses the same habit.

Recall a short list in order. Three items appear for fifteen seconds and vanish; later the candidate writes all three in the original order and names the second.

Each item has its own time limit, generous for the task. Speed is not scored. The full screen takes about ten minutes.

What to look for in the results

You receive a score out of 17 as a percentage, a plain band (Strong, Borderline or Below benchmark), an item-by-item breakdown with a short description of each item, and recordings of the candidate's camera and screen. For data entry, focus on:

• The correction item. It is the closest single item to the job. A partial fix is a miss, and it is the pattern that produces bad records.
• The ordered and conditional instructions. Missing these means procedures will be applied as the candidate assumes them, not as written.
• The list and carried-forward items. In order, complete, correct? Transposed items in a recall task often show up later as transposed digits.
• The rule item. Rarely missed, and telling when it is.

The breakdown is worth more than the total. A candidate weak on the memory items but perfect on the correction and order items may do very well with source documents open side by side. A candidate who misses the correction and both twist items is the one to discuss carefully before an offer, however fast they say they type. Our guide to testing attention to detail goes further on reading these items.

A worked example: two candidates, one role

Consider two candidates for a remote order-entry role who both land in the Borderline band.

The first gets every correction and ordered-instruction item right, including the conditional skip, and loses points on the memory items: two of three list items recalled, and the carried-forward middle number left blank. Her profile is exact with what is in front of her and weaker at holding values in her head. For a role where the source document is always on screen, that is a manageable gap, and a habit of copying values directly rather than from memory addresses most of it.

The second gets the memory and arithmetic items right and misses the correction (the wrong day is left in alongside the right one) and both twist items. His profile is good recall and a tendency to do what he expects rather than what is written. For a role governed by a written procedure with exceptions, that is the profile that produces records that look right and are not.

The band is the same. The hires are not. That is why the breakdown matters more than the number, and why the follow-up questions in the interview should be different for each.

If your role needs a speed check too

For high-volume roles, add a short speed sample of your own after the accuracy screen. Keep it honest:

• Use a realistic sample of your own records, anonymized, with a written procedure.
• Give every candidate the same sample, the same instructions and the same time.
• Score accuracy first: count records with any error, not just characters.
• Run it supervised, on the same kind of video call, so you know who did the work.

Decide in advance how you will weigh speed against accuracy, and apply the same rule to everyone. The ScreenYourAdmin screen does not include a typing-speed or keystroke measure; it is deliberately about exactness.

How to run the screen with ScreenYourAdmin

1. Create a code. Sign in, type the candidate's name and create a single-use code. For a candidate who has asked for an accommodation, set 1.5x or 2x time on the code, which lengthens every reveal and answer window.
2. Send the setup note. Laptop or desktop, Chrome, and the video-call link in a second tab. The candidate needs no account and installs nothing.
3. Supervise the call. The candidate shares their entire screen and you confirm they are alone. Read the code aloud once they are ready. They sign the consent, enter the code and start. Stay on the call throughout.
4. Read the result. Score, band, breakdown and both recordings appear on your dashboard as soon as the candidate finishes. The candidate is told the exercise is complete and never sees a score.
5. Decide. Your team reads the breakdown next to the interview and any speed sample. The software never accepts or rejects anyone.

Remote data entry hires

Many data entry roles are now remote or offshore, which removes the early signals an office would provide: you no longer see the first day's work as it happens. A supervised work sample puts one of those signals back before the offer. Our guide to how remote hiring removes the signals explains why that matters. The screen runs identically wherever the candidate sits; schedule the call at a reasonable local time for them.

Fair and consistent use

Give every candidate for the role the same test, at the same stage, under the same conditions. Tell them beforehand what it is: a ten-minute exercise of everyday admin tasks on a supervised, recorded video call, one attempt, nothing to study. Offer an accommodation to anyone who asks, without penalty. Never let one score be the whole decision. See our overview of pre-employment testing legal basics for the general principles.

Get started

Take the screen yourself first; it takes ten minutes and the values you see will never appear for a candidate. Credits come in packs of 5, 20 and 50, with no subscription, and one credit is one supervised attempt. See the credit packs on the pricing page, or read how the screen works on the home page.

Frequently asked questions

Does the ScreenYourAdmin screen measure typing speed? No. Each item has a generous time limit and speed is not scored. It checks accuracy habits: correcting a stated error, following an instruction in order, applying a rule, recalling a list and carrying a value forward.

How long does the data entry screen take? About ten minutes, plus a few minutes of setup on the video call.

Can candidates use a spreadsheet or notes during the test? No. The candidate confirms they are working alone, without help, notes or an AI tool, and the supervised call and recording are there to see that.

Can I use the same test for every data entry candidate? Yes, and you should: same test, same conditions, same stage of the process. Every attempt still draws fresh values, so no candidate can pass the answers to the next.

What device does the candidate need? A laptop or desktop computer with Chrome, a camera and microphone, and the video-call link in a second tab. Phones are not supported.