A job simulation assessment is an online test that walks you through a fictional working day in the role you applied for. Instead of separate test blocks, you read e-mails from a manager, rank responses to workplace dilemmas, pull numbers from a table, write a short reply and record a video answer, all inside one storyline. Large graduate employers in professional services, banking and consulting use the format, often built by vendors like Cappfinity or HireVue. Length varies by employer and version; allow up to 90 minutes. This guide explains the format, how it is reported to be scored, how to prepare, and walks you through an original eight-task mini simulation with model answers.
Job simulations are employer-specific: each firm commissions its own storyline, and the mix of tasks changes by year, region and programme. What follows describes common patterns reported by vendors and prep sites, not a fixed standard.
Key facts
- Also called
- Immersive assessment, virtual job tryout, day-in-the-life
- Vendors
- Cappfinity, HireVue, SHL and others; often not named
- Format
- SJT, inbox triage, data tasks, written and video answers, work-style
- Time
- Varies by employer and version; allow up to 90 minutes
- Timer
- Often none overall; some items and video answers are timed
- Navigation
- Usually forward only; video sections in one sitting
- Adaptive
- Not reported; content fixed per employer and version
What "job simulation" means (and what it does not)
The search term "job simulation" covers three different things:
- Job simulation assessments (this page): scored selection tests sent after you apply.
- Virtual work experience programmes, such as those on Forage: free, self-paced tasks designed by employers that you do voluntarily to learn about a role. They are not selection tests, although some employers let you mention them.
- Simulation games like VR "job simulator" titles: entertainment, unrelated to hiring.
In hiring, a job simulation is a modern, online version of the work sample test: a small slice of the real job that you perform under observation. Work samples are among the better predictors of job performance. A 2022 re-analysis of selection research by Sackett and colleagues put their average validity at about .33, close to cognitive ability tests (.31) and behind structured interviews (.42). Simulations also double as a realistic job preview: you see what the work is like, and some candidates decide it is not for them, which employers count as a benefit.
What a job simulation assessment looks like
Most simulations follow the same arc, whatever the employer:
- Welcome and integrity statement. A short video or text intro to the fictional company, your role and your manager. Some include practice questions to learn the interface.
- Scenario tasks. E-mails, chat messages, voicemails or short videos set up a situation. You then rank options, choose the most and least effective response, or rate each option on a scale.
- Embedded data tasks. A table, chart or short report appears inside the story, and you answer a numerical or verbal question about it. The material can be mock e-mails, data tables, graphs or short articles.
- Written responses. A short e-mail reply or summary, typed into a box, sometimes with a word or time limit.
- Video responses. A prompt appears, you get preparation time and then record a short answer. Expect fixed time limits, and plan to do the video section in one sitting.
- Work-style questions. Some simulations ask how you prefer to work, on sliders or Likert scales. These are scored against the role profile, so consistency matters.
Two design features make simulations feel different from a test battery:
- The story moves on. Information you get later can change what the best action is. A client who was "happy" in task 2 may complain in task 6.
- Responses are tied to values. Employers write the scenarios around their own values or "strengths". Many publish them, and their scoring keys reward behaviour that matches them.
How job simulations are scored
Employers rarely publish scoring rules. What vendors and prep sites report:
- SJT items are scored against a key, usually built from the judgements of experienced employees or subject-matter experts. Ranking items often give partial credit for near-misses.
- Speed can count. Some simulations time individual questions, so the speed of your response can contribute to your score. Prep sites report that Cappfinity-built reasoning tasks are time recorded rather than time limited.
- Video and written answers are scored by trained assessors against criteria, or by a mix of people and automated tools.
- Feedback is fairly common in the UK: some employers send candidates a short feedback report after the online stage.
Prep sites claim scores from all parts are combined into one profile. That is plausible but not confirmed.
Original mini job simulation: a morning at Northbridge Advisory
Everything below is original: a fictional firm, invented people and numbers. Give yourself 25 minutes for all eight tasks, then check the answers.
Setting. You have just joined Northbridge Advisory as a graduate analyst in a small team that helps family-owned manufacturers. Your manager is Priya, a senior consultant. Today's client is Hallam Tools, a maker of garden tools.
Task 1: Inbox triage (rank 1–4)
It is 8:45. Four messages are waiting. Rank them by what you handle first.
- (A) Priya: "Can you pull Hallam's last three years of sales by channel for my 10:30 call? A table is fine."
- (B) IT: "Your laptop update is due this week. It takes 40 minutes."
- (C) Hallam's finance manager: "The file I sent yesterday had the wrong month in tab 2. Corrected version attached."
- (D) A colleague: "Coffee at 11? Want to hear about the project you'll join next month."
Answer: C, A, D, B (D and B can swap). C first because the corrected file changes the data A depends on; it takes a minute to save and note. A is the urgent deliverable. Reply to D quickly. Schedule B for a quiet slot this week rather than now. The trap is ranking A first and building the table from the wrong file.
Task 2: Most and least effective
While building the table you also check the monthly product-returns tab and notice that March and April returns are exactly the same figure, €48,312, to the euro. Priya's call is in 50 minutes. What is the most and least effective action?
- (A) Leave it; the client provided the data.
- (B) Send the table with a note flagging the identical figure as a possible copy error.
- (C) Replace April with the average of March and May.
- (D) Ask the client's finance manager to confirm, and send Priya the table with the flag in the meantime.
Answer: Most effective: D (you act on the risk and still meet the deadline). Least effective: C (you invent data). B is reasonable but leaves the question open; A ignores a visible risk.
Task 3: Data task
Hallam's sales (€ thousands):
| Channel | Year 1 | Year 2 | Year 3 |
|---|---|---|---|
| Garden centres | 4,200 | 4,000 | 3,700 |
| Online | 900 | 1,150 | 1,500 |
| Hardware chains | 2,400 | 2,450 | 2,500 |
Which statement is correct? (A) Total sales fell from year 1 to year 3. (B) Online grew faster from year 1 to year 2 than from year 2 to year 3. (C) Garden centres' share of total sales fell below 50% in year 3. (D) Hardware chains grew by more than 5% over the two years.
Answer: (C). Year 3 total = 3,700 + 1,500 + 2,500 = 7,700; 3,700 ÷ 7,700 = 48.1%. (A) is false: year 1 total is 7,500, year 3 is 7,700. (B) is false: year 1 to 2 is 900 → 1,150 (+27.8%), year 2 to 3 is 1,150 → 1,500 (+30.4%), so growth accelerated. (D) is false: 2,500 ÷ 2,400 = +4.2%. The trap is (B): the absolute increase (+250, then +350) and the percentage both rise, but candidates who compare the wrong pair of years pick it.
Task 4: Chart reading
Priya forwards a bar chart from Hallam's board pack. The vertical axis starts at €3,000k, not zero. The year 3 garden-centre bar looks little more than half the height of the year 1 bar. A board member writes: "Garden-centre sales have halved." Is that justified?
Answer: No. With a truncated axis, bar heights exaggerate change. The real fall is 4,200 → 3,700, about 12%. A good reply points out the axis and gives the percentage.
Task 5: Written reply (5 minutes, about 80 words)
Write to the board member, copying Priya, explaining the garden-centre decline politely.
Example answer: "Good question on the garden-centre chart. Its vertical axis begins at €3.0m, so the bars overstate the decline: sales went from €4.2m to €3.7m over two years, about 12%. Over the same period online grew from €0.9m to €1.5m, so total sales rose slightly to €7.7m. I'm happy to share a version of the chart with the axis starting at zero before Thursday's meeting." Short, factual, no blame, one concrete next step.
Task 6: Video response (30 seconds to prepare, 90 seconds to record)
Prompt: "Hallam's owner is worried that moving towards online will upset the garden centres that built the business. What would you say to her in our first meeting?"
Outline: acknowledge the concern and the relationships; state the facts (online now almost a fifth of sales and growing, garden centres still the largest channel); suggest options that protect partners (exclusive ranges, click-and-collect via garden centres); propose a next step (talk to three key garden-centre buyers). Speak to the camera, not the notes.
Task 7: Work-style ratings (no right answer)
Rate from 1 (not like me) to 5 (very like me): - "I prefer to finish one task before I start another." - "I like changing priorities during the day."
There is no key you can guess reliably. What matters is that your answers are honest and consistent with how you answered the scenario tasks: if you ranked inbox triage confidently, rating yourself a 1 on handling changing priorities sends mixed signals.
Task 8: The story moves on
At 12:00 the finance manager replies: the April returns figure in Task 2 was a copy error; the real figure is €61,200, up about 27% on March. What should you do now?
Answer: Update the returns tab, tell Priya what changed and whether it affects her call (returns are rising, not flat, which matters if online sales drive returns), and thank the client. The point of the task: corrections are routine, and the expected behaviour is to fix, inform and move on.
Traps and how to avoid them
- Treating it as an aptitude test only. Most of the score is behavioural. Read every scenario for what the employer values.
- Rushing the reading. The key fact is often in an attachment, a footnote or an earlier e-mail.
- Escalating everything. Employers want you to own small problems and keep people informed, not to hand every issue upwards.
- Inventing data or guessing silently. Flag, ask and carry on.
- Inconsistent self-ratings. Work-style questions are compared with how you behave in the scenarios.
- Doing the video on the wrong device. Use a laptop or desktop with a webcam, stable internet and a quiet room. Check the employer's browser advice.
- Starting without enough time. Many simulations must be finished in one sitting. Block 90 minutes even if the invitation says 45.
How to prepare for a job simulation
Five days, 45 minutes a day.
| Day | Focus |
|---|---|
| 1 | Read the employer's values or strengths page and its official process page; note which tasks they describe |
| 2 | 20 SJT items in mixed formats (rank, most/least, rate); explain each answer in one line (SJT guide) |
| 3 | Inbox and prioritisation practice (in-tray guide); 10 numerical items built on tables and charts (numerical guide) |
| 4 | Write three short business e-mails in 5 minutes each; record three 90-second video answers and watch them back (video interview guide) |
| 5 | Do the employer's own practice material or preparation hub, if offered, then rest |
On the day: read the integrity statement, close other apps, keep pen and paper for numbers and names, and use the practice questions to learn the interface before the scored part.
Sources
- UK Digital Marketplace, Cappfinity service listing (Simulate: asynchronous video and virtual in-tray; Preparation Plus)
- HireVue, Virtual Job Tryout and Modern Hire acquisition press release (2023): https://hirevue.com/platform/assessment-software/virtual-job-tryout, https://www.hirevue.com/press-release/modernhire
- HumRRO, summary of Sackett, Zhang, Berry and Lievens (2022), Journal of Applied Psychology 107, 2040–2068 (validities: structured interviews .42, work samples .33, cognitive ability .31)
心理学修士、7年の受験歴。
最終更新：Oct 1, 2026