Your pilot worked. So why isn't the program scaling?

XR gives learners something traditional training can't: a safe place to fail. A veterinary student can get kicked by a horse in VR instead of the real thing. A police officer can practice de-escalation and taser use before ever carrying one on shift. Whether XR works isn't really in question anymore.

The harder question is why so many programs still get stuck in pilot instead of scaling past it.

At AWE 2026, ArborXR CEO and co-founder Brad Scoggin gave a keynote on exactly that problem. After watching thousands of enterprise XR deployments live or die, drawing on ArborXR's work with over 2,000 companies and roughly 60 Fortune 500 organizations, he's identified the patterns behind why programs stall and what separates the ones that survive.

Here are the 5 obstacles most XR pilots will face, and what it takes to overcome them.

1. Selecting the right champion

Most programs live or die with one person. A champion needs to genuinely believe in the problem being solved, since real momentum only comes from real conviction. Someone has to own the initiative end to end, and success needs to be defined, with timelines and measurable outcomes, before the program ever launches.

Nicklaus Children's Hospital in Miami is the clearest example of this working. One committed champion, Dr. Christina Potter, turned an early idea into a hospital-wide training program almost entirely on her own initiative. She identified VR as a fit for both staff training and patient care, secured the budget, and has run the program herself for years. Staff trained on CPR and other life-saving techniques in VR now show an 80% retention rate a year later, against roughly 15% for traditional training. The hospital also uses VR to calm children before procedures, which it credits with an 80% reduction in pre-operative anxiety and faster recovery times, across about 8,000 patients a year.

Great technology with a weak champion usually fails. A strong champion can overcome imperfect technology.

2. Picking the right use case and ISV

The second failure point shows up even when a champion is in place: choosing XR because it's exciting rather than starting with a real business problem. The technology should match the problem, not the other way around, and the right ISV partner is the one who will honestly assess where XR actually creates value instead of just selling a use case.

Delta Air Lines built its de-icing training program around exactly that discipline. Training someone to de-ice a plane traditionally requires a plane, a trainer, a trainee, hoses and cables, and one ingredient nobody controls: weather. In a location like Oklahoma, that's roughly six usable weeks a year. By matching the right long, high-stakes process to the right ISV partner, Delta moved the training into VR and can now train a thousand people a day, regardless of season or location.

The wrong use case makes great content fail. The right one makes adoption almost inevitable.

3. Building for results and ROI

Proving ROI isn't the hard part. The hard part is defining measurable outcomes before deployment instead of trying to reconstruct value after the fact. Time saved and faster onboarding matter. So do fewer errors and better safety. Programs that skip this step are exposed the moment a champion leaves without documented results to point to, since undocumented value doesn't survive a change in leadership.

Mars Wrigley's approach is the model here. Rather than deploying broadly, the team tied a pilot to business metrics from day one: one factory, one job (the person loading the gum-packing machine at an Extra facility), one six-month window. They tracked time to proficiency, employee turnover, waste, and output against traditional training. The result was almost $20 million in savings across those four metrics in six months, a number that made the value of the program impossible to question.

Without proof of value, a pilot stays an experiment. With ROI, it becomes a program.

4. Avoiding vendor lock-in

Deployment cycles change. Hardware, software providers, and internal priorities all shift over time, and limited flexibility means slower innovation and single-provider dependence. The fix is choosing platforms that support multiple devices from the start, and owning content and data in a way that holds up over the long term rather than optimizing for what's easiest on day one.

Walmart, which serves more than 10,500 stores in 19 countries, is one of the most advanced XR deployments in the world, and its own journey out of a single-vendor walled garden is a well-documented example of what that flexibility looks like in practice: an enterprise built to evolve as hardware, software, and priorities changed, rather than one locked into a single provider's roadmap.

Successful XR programs are built for flexibility, not dependency.

5. Driving internal adoption

The last failure point shows up after everything else has worked: the champion is in place, the use case is proven, the ROI is documented, but people still aren't consistently engaging with the training. XR belongs in an organization's broader learning strategy, connected to existing programs and involving L&D, operations, safety, and business stakeholders, not treated as a side initiative.

The clearest version of this mistake mirrors what happened with early web-based training in the 2000s: a genuine step forward that didn't reach mainstream adoption until it could sync directly with a company's existing LMS. Before that, it was fully siloed, employees finished their regular coursework, then went somewhere completely separate to do the additional training, and the results never made it back anywhere useful.

ArborXR built Insights, which syncs with 500+ LMS platforms, to close that same gap for XR: a trainee launches training from inside the LMS they already use, confirms identity and content with a PIN entered inside the headset, and results sync back automatically. No separate room, no separate reporting step, no silo.

The goal isn't XR adoption. It's business adoption, enabled by XR.

Where this leaves the next pilot

None of these five are technology problems. They're planning and ownership problems, and they show up the same way whether the deployment is VR headsets in a hospital, a de-icing simulator at an airline, or a fleet of devices across a Fortune 500 rollout. The programs that make it past pilot are the ones that assign a real owner, pick a use case with a fast and measurable win, define success before deployment, and build for flexibility over dependency.

The best teams don't ask what headset to buy. They ask how they'll create sustainable business value.

Watch the full keynote from AWE 2026.