If you commission a bespoke VR training module from an agency, published timelines sit at 8 to 10 weeks for a straightforward build and 12 to 20 weeks for a custom programme, with six months or more for complex work. If your own team authors the same scenario in a no-code platform, a first working version takes an afternoon and a reviewed, published one takes two to three weeks. The difference is not the technology. It is who has to be in the room, and how many times.
Key takeaway: Build time is dominated by decisions, filming and sign-off, not by software. Any route that puts a subject-matter expert in direct control of the content removes the slowest steps, which is why the same scenario can be a ten-week project or a fortnight’s work.
What the agencies publish
These are the numbers a buyer finds first, and they are worth reading properly rather than dismissing.
| Source | Timeline | Cost |
|---|---|---|
| Roundtable Learning (updated May 2026) | “The average development time for a VR training program is 8-10 weeks or more”, “as quick as 4 weeks or as long as 6 months or more” | Not stated on that page |
| The Prime VR (June 2026) | “Roughly 12 to 20 weeks from discovery to deployment" | "$35,000 to $250,000 or more” |
| VR Vision Group (updated Sept 2026) | Simple projects 8 to 10 weeks, complex 6 to 9 months | 360° video $8,000 to $15,000 per finished minute; “typical first project lands at $130K-$200K all-in” |
| Frame Sixty (Aug 2026) | “Six weeks to nine months” | $10,000 to $200,000, and “initial development is only 50-70% of total spend” |
One caveat on that first row, because it keeps getting recycled as independent corroboration: the widely cited eLearning Industry article carrying the same 8 to 10 week figure is written by Roundtable Learning. It is one source with several faces, not four studies agreeing.
Roundtable’s phase breakdown is the genuinely useful part: instructional design 1 to 8 weeks, asset creation 4 weeks to 4 months, development and programming 1 to 8 weeks. Notice that only the last of those is software work.
Where the weeks actually go
Strip out the production-studio overhead and every immersive build has the same five jobs. Only one of them is affected by your choice of tool.
| Job | Typical duration | Who owns it | Does the tool change it? |
|---|---|---|---|
| Deciding what the scenario teaches and how it is assessed | 2 to 5 days | Your SME | No |
| Capturing or sourcing the environment | Half a day to 2 days per site | You or a videographer | Slightly |
| Building scenes, hotspots, branching and questions | 20 minutes to 2 weeks | Agency, or your SME | Yes, enormously |
| Review, correction and sign-off | 1 to 3 weeks, in rounds | SME, compliance, brand | Yes, indirectly |
| Publishing, LMS wiring and rollout | 1 to 3 days | You | Slightly |
The expensive variable is the fourth row. In an agency model each correction is a ticket: you watch a build, write notes, wait for a sprint, review again. Three rounds of that is a month of calendar time and almost no working time. When the person who knows the content can open the editor and fix the wording themselves, those weeks disappear, which is the whole argument for no-code authoring and the reason it is worth being specific about it rather than claiming the technology is faster.
The counter-claim, and how far to trust it
Authoring vendors have been quoting minutes for years. ELB Learning published “How I Built a Virtual Reality Course in 30 Minutes” back in 2020. Our own figure is that most first-time creators publish a working scenario within 20 minutes.
Both numbers are real and both measure the same narrow thing: time in the editor, for one scene, with media already to hand. Neither includes deciding what to teach, getting a clinician or assessor to approve it, or filming the ward. A vendor who quotes the editor time as the project time is doing the same thing as an agency quoting the sprint count without the sign-off rounds, just in the opposite direction.
The honest framing is a range with the slow parts named:
- One afternoon to a usable draft scenario with placeholder media, built by one person.
- One to two weeks to a reviewed, assessed, branded scenario including a capture session.
- Two to four weeks to that scenario live to a cohort, with LMS integration and reporting in place.
Merthyr Tydfil College’s Learning and Digital Resource Coordinator described their construction safety build as taking them “from concept to live for students in under three weeks”, including filming and tutor review. That is the realistic shape of a first project when the college owns the content.
The comparison nobody makes
Immersive build time gets judged against nothing, which is why “8 to 10 weeks” sounds either fast or slow depending on the day. The classic benchmark for conventional eLearning is the Chapman Alliance study of 249 organisations: 79 development hours per finished hour of basic eLearning, 184 hours for moderately interactive, and 490 hours for highly interactive. That research is from 2010 and should be treated as a period reference rather than a current figure, but nothing has replaced it at that scale.
Against 490 hours for a highly interactive conventional module, an eight-week agency VR build is not an outlier at all. It is the same industry working at the same pace. The only route that genuinely breaks the pattern is removing the production handoff.
What slows real projects down
From the builds we see, in order of damage done:
- No decision about assessment. Teams design scenes first and work out scoring later, then rebuild. Write the marking scheme before the first scene.
- Filming once. A single capture session always misses an angle. Budget a second visit or accept a reshoot.
- Sign-off with no named owner. Compliance review is not slow because compliance is slow, it is slow because nobody is assigned.
- Scope by enthusiasm. The first scenario should be one route through one task, not the whole curriculum. The second is three times faster because the pattern exists.
- Waiting on hardware. Not needed. Scenarios run in a browser on the devices your learners already have, so a rollout does not have to queue behind a procurement cycle.
The short answer
How long does it take to create VR training? Commissioned from an agency, expect 8 to 10 weeks for a straightforward module and 12 to 20 weeks with a six-figure budget for a custom programme. Authored by your own team in a no-code platform, expect an afternoon for a draft and two to three weeks for a reviewed, published scenario that a cohort can use. The decisive factor is whether your subject-matter expert can edit the content directly, because review rounds, not rendering, are what consume the calendar.