Your invitations are ready, your assessors are booked and your competency framework has been signed off. One decision is still open in the planning document: should the assessment day be virtual or in person?
It can look like an operational detail, but the format affects how candidates first experience your organisation, who is able to take part, how consistently assessors can score and what the process costs in time and budget. It also shapes the evidence you end up with. The more useful question is not which format is better in general. It is which one gives candidates the fairest chance to show you what you need to see.
The two formats are not simply interchangeable. A large 2025 field study of 10,898 candidates found broadly similar assessor reliability in virtual and face-to-face centres, but also differences in how some capabilities were expressed and in the performance ratings candidates received. The researchers supported the use of virtual centres while noting that more work is needed to understand where the differences matter. In practice, that puts the weight on design.
An assessment day is valuable because it lets you see candidates doing things rather than describing what they have done before. That might be working through a group problem, presenting, handling a difficult conversation, analysing information or responding to a role play.
Many of these exercises move online well. Written tasks, analytical exercises, presentations and structured interviews can all be delivered virtually, and the stimulus material is often easier to standardise on screen than in a printed pack. Others need more thought. A group discussion with five or six people on a small screen is harder to run well than the same exercise in a room, so the brief, the group size and the way turns are managed matter more. Role plays that depend on reading subtle interpersonal cues also deserve a careful look.
The test is not whether an exercise can be recreated online, because almost any of them can. It is whether doing so changes the behaviour you are trying to observe. If the role involves working with colleagues remotely and sharing information on screen, a virtual exercise may be more realistic than a room. If the role depends on moving around a physical environment or dealing with people face to face, there is a stronger case for bringing candidates together.
Format changes who can attend. A virtual assessment day removes travel time, accommodation costs and geographical barriers, which matters for national campaigns, for candidates with caring responsibilities and for anyone who would otherwise need to take a full day out of work or study.
Virtual does not automatically mean accessible, though. Candidates still need suitable technology, a reliable connection and somewhere they can take part without interruption. Moving online removes some barriers and creates others. In either format you need a clear, confidential route for requesting adjustments under the Equality Act 2010, and your assessors need to know how approved adjustments affect timing and scoring.
Face to face has its own case. Meeting future colleagues and seeing the workplace can help candidates judge whether the organisation is right for them, and if that is a real part of your recruitment objective, it is worth the journey. If candidates are travelling across the country only because that is how assessment days have always been run, it is worth questioning.
The reliability of an assessment day depends less on where it happens than on what assessors do before it starts. Whichever format you choose, the same principles apply.
Assessors should score against predefined behavioural anchors, not overall impressions. A score of three for communication should mean the candidate explained ideas clearly and listened actively, not that they seemed confident. Assessors record what the candidate said or did, link it to a competency and only then give a score. Building that sequence into the briefing separates observation from judgement and keeps moderation discussions productive.
Calibration matters more for virtual delivery, because assessors often work alone rather than in a shared room. A calibration exercise using a sample response, plus a short recorded briefing they can return to mid-session, helps a great deal. Add a clear approach for technical interruptions, so a candidate is not marked down for a lost connection rather than a lack of capability. In every format, hold a moderation meeting before ratings are confirmed, so decisions reflect patterns of behaviour across the whole day rather than one strong or weak exercise.
One more point on fairness. Lunch and other informal time can help candidates get a feel for your organisation, and in person it gives assessors a view of people in unscored moments. That should not become unrecorded evidence. Only what is observed in scored exercises, against the agreed criteria, should feed the decision.
A candidate who arrives flustered because they could not find the entrance, or who is struggling with a camera setting, is not showing you their capability. Logistics are a data quality issue as well as a courtesy.
For in-person days, the usual problems are scheduling collisions, too few assessors for the number of candidates and transitions that eat into scored time. If a lunch overrun pushes the afternoon back, candidates at the end of the queue can end up with a shortened interview, which makes scores harder to compare. Build in ten-minute buffers between activities, confirm rooms and AV the day before, and send a clear brief covering arrival time, what to bring and how to get into the building.
For virtual days, technical preparation replaces venue logistics. Send a pre-day check covering browser, camera and microphone, explain what happens if a connection drops mid-exercise, and say whether candidates need a quiet space or any materials to hand, such as a notepad or a calculator if permitted.
There are real practical benefits to virtual delivery. Assessors can join from different locations, venue space stops being a constraint, parallel cohorts can run in the same week and every scored activity leaves a digital record that can be reviewed, moderated and retained as evidence. Those are fair considerations when you have a budget, limited assessor time and a deadline.
The risk comes when efficiency becomes the assessment strategy. A cheaper or faster process is not better if it makes it harder to observe the behaviours that matter, and an in-person day is not better simply because everyone is in the same room. Some employers use both: virtual for large early-stage volumes, then an in-person final stage for a smaller group where meeting the team adds something real. The format should support the assessment design, not dictate it.
A general video call gives you a meeting room. It does not give you a structured process. A dedicated assessment platform brings schedules, activities, assessor evidence, scoring and reporting into one place, so assessors work from a single record rather than reconciling spreadsheets at the end of the day. That supports consistency without dictating how you design the assessment itself, and it takes a good deal of the admin out of running a larger campaign.
Whatever you use, test it before candidates are involved. Run a small internal dry run with a handful of colleagues, put them through a shortened version of the programme and check that assessors can use the scoring interface, that the adjustments route works and that the output reflects your competency framework. You will know within an afternoon whether the format suits your scheme.
The choice between virtual and in person is a quality decision more than an operational one. Start with what candidates need to do, what assessors need to observe and which environment gives people the fairest opportunity to show it. Once that is clear, the format follows more easily, and the foundations stay the same either way: a clear competency framework, calibrated assessors, behaviourally anchored scoring and a candidate experience that cuts out unnecessary uncertainty from invitation to feedback.
I explored the wider question in my article for Prospects Luminate. If you are weighing this up for an upcoming campaign and would find it useful to talk it through, we are always happy to have an exploratory conversation.
