Work Simulation Assessment 2026: What The Firms Publish

Two of six exercises are work simulations

SHL's own list is the only published taxonomy of centre exercises we found. Sorting it by what you have to produce makes the two simulations obvious.

Two of six, which is 33 percent, ask you to act rather than to analyze. The publisher's definition of the second is worth reading in full: "Candidates manage a simulated inbox filled with tasks, messages, and competing priorities. They must decide what needs immediate attention, delegate effectively, and handle disruptions in a predetermined timeframe."

Four demands sit in that sentence: triage, delegation, absorbing disruption, and a fixed clock. Two of the four appear nowhere else in the list of six. Delegation and disruption are not tested by a written report, a group discussion or a presentation.

An inbox in the software, absent from the firms

SHL's candidate support page says what its virtual assessment centre contains, and it is a device instruction rather than an exercise description: "Use a laptop or desktop computer - the virtual Assessment Centre Exercises include an immersive email inbox and document hub."

So the inbox exists in the platform. What no page in our corpus does is name it on a consulting firm's own site. We searched the full text of 371 pages that answered on 30 September 2026 for inbox, in-tray and e-tray, and found nothing at any consulting firm.

Two readings of that are honest and one is not. The honest readings are that the firms using this platform simply do not itemize its exercises, or that they rent the platform without the inbox module. The reading that does not hold is that the exercise never appears in consulting recruiting, because a firm's silence is a fact about its page.

SHL does publish one thing you can use. Its free practice area opens the platform: "Try out SHL's Virtual Assessment Center Exercise platform and complete a short assessment center exercise." The publisher is equally clear about the limits, saying the exercise "is not scored, there are no feedback reports and there is no live interaction involved." That is the only free hands-on route to this format we found anywhere, and it sits with the vendor.

The one firm that publishes a work simulation

Deloitte's UK early careers page publishes a work simulation under its own name, as step three of five.

The firm describes it as designed "to give you a real taste of what it's like to work at Deloitte, focusing on your behavioural skills." The scenarios are work-based, and the firm bullets four answer formats: "Ranking response options in order of preference", "Selecting the most appropriate response(s) from a given set", "Providing written responses", and "Recording video responses".

Two of those four are closed and two are open, which is the single most useful fact on the page for preparation. A candidate ready for multiple choice and ambushed by a video recording has prepared for half the exercise.

The firm also settles the AI question in advance: "Some scenarios may refer to workplace tools like AI. Please know that you do not need prior AI experience, nor will you be asked to use an AI tool during the assessment." And it provides practice inside the assessment itself: "Use the in-assessment practice questions to help you get familiar with the technology and format."

The word inbox does not appear on that page. Neither does a duration, a scenario count, or a number of video answers.

Three published work simulation stages, one unscored

Three large firms recruiting graduates in the United Kingdom publish a stage built out of work scenarios, and they disagree on whether it counts.

EY UK's application process page, checked 6 October 2026, puts its preview inside the assessed stage and names what travels with it. The first online stage "involves different tasks including a Realistic Job Preview - a first stage of online assessment including teamwork and learning capability. There's also a behavioural skills test measuring whether you have the required skills for the role."

KPMG UK attaches the opposite label to a stage that looks the same from outside. Its process page writes: "We've designed a job preview quiz using everyday scenarios that you could face in your chosen role - this step is not assessed but it should help you decide whether the role you've chosen is the right one for you."

So a work scenario stage can be a filter or a brochure. The two are indistinguishable in an invitation email, and the only place that says which one you opened is the firm's own process page. Read it before you spend an evening on the exercise.

EY UK also prices a failure at that stage. The same page says you "can only apply for one programme every six months", and that a rejection "from any stage in the process" carries a six month wait before another graduate or undergraduate application. A preview sitting inside an assessed stage therefore spends half a recruiting year when it goes wrong.

Four work simulation answer formats, and what each reveals

Nothing about the four formats is guesswork; the firm named them. What each one exposes follows from the format itself.

The firm's own recruiter gives the instruction for the first of these, and it cuts against most published advice: "When rank ordering your preferred responses to a situation, go with your gut instinct, don't try to second guess what it is we might be looking for."

SHL says the same thing independently on its candidate page: "Don't make assumptions about how you should respond. If you try to guess what the assessors are looking for you may be wrong." Two unrelated sources, one instruction, and it is the opposite of what a prepared candidate instinctively does.

One timing rule on the preceding step is worth flagging because it contradicts itself on the page. Deloitte writes "There's no overall time limit for completing the assessment" and then, two lines later, "Some questions are timed, meaning the speed of your response will contribute to your score." Both sentences are the firm's and both apply.

Six questions, one avatar, a human score

KPMG UK publishes the most fully specified recorded interview in the corpus, and it names its interviewer.

The format first: "You will be presented with 6 pre-recorded questions asked by our AI avatar Leah. You will then be given a set time to prepare and then record your responses." Then the reassurance, which the firm evidently felt was needed: "Although your interviewer is AI generated, we'd like to reassure you that your recorded answers will be reviewed and scored by a member of our Student Recruitment team."

On the clock: "As this is a recorded interview, you will be able to complete this stage at a time that suits you and you should allow 40 minutes to complete." Dividing 40 by 6 gives 6.67, so roughly six and two thirds minutes per question including preparation and recording. Not a per-question limit, and the only arithmetic available.

The firm's integrity page carries the other half of the bargain. Candidates confirm they are completing assessments "independently and without the assistance of others, or any artificial intelligence or other resources." Dishonesty leads to withdrawal, and the firm adds that "We will also be re-testing elements of the online assessments." That retest is the only published after-the-fact verification in our corpus.

There is an unremarked tension in those two pages. Candidates may not use AI. The firm's questions are asked by an AI avatar. Both statements are published and the firm does not connect them.

Two platform rules with teeth

Most instructions on these pages are advice. Two are consequences, and both belong to SHL rather than to any firm.

The first: "Preparation activities are timed and can only be accessed once." Opening the reading material to have a look is the same act as starting the clock, and there is no second attempt.

The second: "You are not allowed to open any other Browser / Switch between the Test Screen and other widows/ Minimize the Test Window. The Test will get locked permanently if a candidate is found switching through windows while taking the Test." The spelling of widows is the publisher's own. No appeal route appears next to it.

Alongside those, the publisher asks you to run your own schedule: "Be prepared to manage your own time. Follow your calendar (and/or any individual schedule provided) to ensure that you enter and exit all activities at the correct time." And it gives a fallback when hardware fails: a camera or microphone problem still leaves chat and the whiteboard available.

One more source adds a legal clock. WTW names its vendor, HireVue Assessments, and publishes a notice period: New York City law requires "at least 10 days' notice to applicants that may be requested to complete our video assessments."

Five moves for a timed simulation

Each move below rests on a published instruction, with the source named. None of it is invented.

1. Read the invitation for the vendor name. SHL, Sova and HireVue all appear by name somewhere in this corpus. The rules that can end a session live on the vendor's candidate pages and on no firm's careers page.
2. Do not open anything until you intend to start. Preparation activities open once. A curious click costs the attempt.
3. Answer as yourself on ranking questions. Two independent sources say guessing the framework is likely to be wrong, and one of them wrote the exercise.
4. Expect to be asked to type and to speak. Deloitte's simulation has four formats and half of them are open-ended. Rehearse a written answer and a spoken one under a clock.
5. Track what remains rather than what you have done. The only first-hand failure we found on a timed screen came from misjudging the tail: a graduate who thought one question was left and found it had several parts, in her words, and who notes that others ran out of time.

Under all four Deloitte formats sits the same question of whether you put the answer first and the reasons after. Our synthesis rubric weights Answer first 35 out of 100 and Next step 20, which is close to what a triage decision has to carry: the call, then the reason, then what happens next. Those weights are ours, read from our own code, and no vendor or firm publishes a comparable scheme. The reasoning is on how an answer is assessed and the drills run free and without an account from the tools page. None of them simulates an inbox.

What nobody publishes about the work simulation assessment

Six gaps, and each one is a question worth asking your recruiter.

Nobody publishes whether the inbox simulation is used in consulting. The publisher sells it, the platform contains it, and no consulting firm names it.

Nobody publishes the length of a work simulation. Deloitte gives no duration for its Job Simulation, no scenario count and no number of video answers.

Nobody publishes which questions are timed. Deloitte promises to flag them inside the assessment and does not flag them in advance.

Nobody publishes a per-question limit on a recorded interview. KPMG UK gives 40 minutes for six questions and no breakdown, and does not say how many takes are allowed.

Nobody publishes which elements get re-tested. KPMG UK announces the retest without naming its scope or timing.

And nobody publishes a scoring scheme. SHL points at the employer's competency framework, and no employer in this corpus publishes one. For the group exercise the same publisher sells beside this one see group case interviews, for the whole assessed day see the assessment center written case, and for the platform rules that govern a live video round see remote case interviews.