AI and hybrid work have written new rules for matching a device to a role. Equipping the workforce by persona was once a straightforward exercise: the org chart supplied the groups, a stereotype supplied the requirements, and the model held because most people worked roughly the same way. That no longer holds. AI keeps redrawing what roles need, and hybrid work has scattered where they need it. Together, those two shifts have changed the inputs persona building runs on.

The inputs changed, and the principle held. Smart device fleets still get provisioned by sorting employees into representative types and equipping each accordingly. The older build — a role, a department, a rough sketch — came from a workplace that has since changed shape. The new build starts with the work itself: where it happens, what it demands, what it costs when it stops. The device follows from that, role by role.

Reading each role by the work it does

The new build reads each role against the factors that decide what its device has to do. Where does the work happen — headquarters, home offices, client sites, a hospital floor? What does it demand of hardware at peak — not on an average Tuesday? How much of the day runs through a camera and a headset? What data does the role touch, how fast is it changing, and what does an hour of downtime cost? An hour is not the same hour for a field engineer and a back-office coordinator.

Run that reading across a workforce and the real groupings surface: roles with nothing in common on paper turn out to make the same demands of their devices, and each pattern anchors a standard configuration with a defensible reason behind it. The fleet gets simpler — fewer configurations, better matched, easier to support.

Those questions aren’t a checklist — they’re a sample. The complete reading runs eight dimensions deep.

The eight dimensions behind a well-built persona

1. Mobility & Environment. Where does the work physically happen — and how often does it move? That answer sets connectivity, durability, battery, and form factor before anyone opens a spec sheet.
2. Work Demands. What does the role ask of hardware at peak — and where does its AI actually run, on the device or in the cloud? The moments that decide whether a device fits sit at peak load, and averages hide them.
3. Collaboration Load. How much of the day lives in calls, meetings, and screen shares? For some roles, the camera, audio, and bandwidth are the job, and the device gets chosen around them.
4. Digital Dexterity. How self-sufficient is the role with technology, and how much support does it generate? Some people turn extra capability into output, and others turn it into tickets. Neither answer is a judgment; both are provisioning facts.
5. Data & Access Footprint. What systems and data does the role reach, and from where? Two roles with the same hardware needs can carry very different access risk, and the device has to reflect that.
6. Cost of Disruption. What does an hour of downtime actually cost this role? Reliability spend should follow that math, concentrated where a stoppage hurts most and kept lighter elsewhere.
7. Volatility. How fast is the role itself changing — new tools, new workflows, new AI in the mix? Fast-moving roles need headroom, and stable ones can skip paying for it.
8. Security & Compliance. What does the role’s risk profile require the device to enforce, not merely support? Compliance sets a floor, and raw performance does not substitute for it.

Read a role across all eight, and the right device for it stops being a debate.

Persona work is a lifecycle discipline

Rollout is where a persona starts earning its keep, and the work runs on from there. The match has to hold as roles evolve: watched for drift, corrected when the work outgrows the configuration, and closed out cleanly when devices come back at end of life. That’s lifecycle work, and most provisioning models leave it unplanned. MCPC, a managed IT services provider focused on device lifecycle management, has spent more than twenty years there — managing devices from planning and provisioning through support, refresh, and recovery, so the fit that was right on day one is still right in year three.

The bottom line

Persona-based provisioning didn’t get harder. It got deeper — and the organizations that read the work behind each role will run fleets that cost less, support easier, and fit the people using them. The ones that don’t will go on buying good hardware for the wrong desks.

MCPC’s new guidebook — built with HP, Microsoft, and AMD — puts the argument to work as a method, with a step-by-step process for building personas your fleet can run on. Download the full guidebook for all eight dimensions in depth and the questions to ask of each role here.