As AI collapses digital IT costs, physical becomes the budget.
A research-grounded view of how onsite IT support — the one layer AI cannot reach — becomes both the most expensive and the slowest service in enterprise IT between now and 2032. Modelled for Fortune 500 enterprises. Cited from Gartner, MetricNet, Reworked, and IDC.
Physical IT tickets cost 3× more than digital ones.
Per MetricNet's 2024 IT Service Desk Benchmarks, the average cost to resolve a Tier 1 digital ticket — password reset, software issue, generic service desk query — is around $22. The cost to resolve an onsite, desktop, or field-support ticket — the broken laptop, the new-starter kit-out, the peripheral exchange — is around $70. The physical ticket costs roughly three times the digital one.
That gap is the unsurprising one. Onsite resolution carries engineer time, dispatch overhead, travel, parts logistics, and coordination cost. None of that is news to anyone running an IT services budget. The 3× ratio is the floor most CIOs and CFOs have known about for a decade.
The surprising part is what the ratio does next.
As AI matures, the cost gap compounds.
Drag the slider to advance the year. The physical cost-per-ticket stays the same — AI cannot reach through a screen. The digital cost shrinks. The gap widens from 3× today to 14×+ by 2032.
Unit costs are the symptom. Budget share is what your CFO sees.
The cost-per-ticket ratio is the operational view — useful for service-delivery managers, MSPs and engineering leadership. It explains why the gap matters. But the budget review happens one layer above. What the CFO sees on the spreadsheet is the line item: "Onsite IT support." And what makes that line item the standout problem is not its absolute size, but its share of operational IT cost — and the trajectory of that share over time.
The chart below models that trajectory for a representative Fortune 500 ICP.
Three Gartner forecasts compound across the rest of the stack.
The trajectory above isn't a projection on its own. It's the arithmetic of three independent Gartner forecasts, each touching a different layer of IT operations, all bypassing onsite physical support.
A caveat worth naming: Gartner itself (June 2025) forecasts that more than 40% of agentic AI projects will be cancelled by the end of 2027 due to escalating costs, unclear business value, and inadequate risk controls. The technology is real — but procurement discipline matters. The trajectory above assumes realised savings, not vendor claims. Even at a 50% adoption haircut, the share trajectory still bends sharply upward.
And it's not just expensive. It's slow.
Every employee in your business experiences instant gratification in their personal life — same-day delivery, on-demand food, AI answers in seconds. Then they walk into work and wait 2.9 days for a replacement laptop. As AI accelerates every other service inside enterprise IT, slow physical IT becomes the standout bad experience inside the modern workplace — and the line item IT leaders increasingly get fired for delivering.
Your employees ordered groceries in 30 minutes this morning.
Everyone in the building lives in a world calibrated to instant. The Spotify queue advances in milliseconds. The Amazon Prime parcel arrives the next day. ChatGPT answers questions in two seconds. Microsoft's State of Global Customer Service research consistently finds that 55% of customers expect a higher level of customer service year-on-year — and "customer" here means anyone receiving a service, including the employee receiving the IT one.
Then employees raise an IT ticket for a broken laptop or a new peripheral, and wait. Two days. Five days. Sometimes twenty. They don't know what a P4 SLA is. They don't care that the engineer is overworked. They know that everything else in their life is instant, and IT is the slowest service they touch. This is precisely why the industry is now moving away from SLAs as the primary IT metric and toward Experience Level Agreements (XLAs) — which measure perceived employee experience, not just compliance against a clock the employee never agreed to.
Every day of waiting costs the business measurable productivity.
The productivity cost is measurable and consistent across the research. StrongDM's 2025 employee onboarding survey found that 43% of new starters were still waiting for basic workstation tools after their first week, and 18% were still missing essential tools after two months — every wait day is a day of paid salary delivering reduced productivity. InsightGlobal's workforce research separately finds that 78% of employees say they are missing at least one of the tools they need to do their jobs effectively. And on the new-starter end specifically, industry-consensus benchmarks (SHRM, Aberdeen) place time-to-full-productivity at 6 to 8 months — a window that begins on Day 1, with the laptop handover.
The economic damage is direct. The reputational damage is bigger. Every wait shortens the half-life of "IT is a partner" and lengthens the half-life of "IT is what's slowing me down" — and once an IT function lives in the second category, it doesn't move back.
Your CFO sees the cost. Your employees feel the speed. The IT leader who can fix both wins the seat at the table. The one who can't, loses it.
The line doesn't have to keep rising. The wait doesn't have to keep stretching.
Smart Collect® automates the physical handover itself — the device exchange, the peripheral swap, the new-starter kit-out, the broken-laptop return. The ticket closes inside ServiceNow without dispatching an engineer. The cost-per-ticket drops toward digital parity. The 2.9-day wait drops to minutes. Both lines bend.
Smart Collect® has already delivered 500%+ throughput uplift in pharma, 60% onsite ticket reduction in nuclear energy, and 35% engineer travel reduction in aerospace — all before AI was driving the workflow. As Now Assist matures and the digital ticket closes itself, the physical-handover platform delivering these results today multiplies its impact across every workflow that needs to reach the workplace.
Every IT leader is making this decision now.
Every MSP CFO and every Fortune 500 CIO is currently writing a 24-month plan for how AI lowers cost-to-serve. By the end of that window, the digital layers of IT services will be substantially automated. The physical layer will not have moved on its own.
The decision being made today — and the decision their clients, their boards, and their competitors are watching them make — is whether to address the physical layer now, or watch a competitor do it first.
Smart Collect® is the only ServiceNow-native answer to that question. The MSPs and enterprise IT teams that close the physical layer in the next 18-24 months capture the services-margin advantage. The ones that wait become the ones being compared against in 2029.
Common questions on cost, speed, and what changes.
What does your IT budget look like in 2029?
A 60-minute discovery workshop with a Velocity ServiceNow architect maps your specific environment, calibrates this model to your numbers, and identifies the use cases where Smart Collect® bends the trajectory.