IT Hardware Refresh: Keep, Repair, Reallocate or Replace

70% of organizations still treat IT hardware refresh as a calendar event — or wait until a user complains. Both approaches are wrong.

Calendar-based refresh replaces perfectly healthy machines and misses the ones already degraded. The result: wasted budget on one side, cascading support tickets on the other. Two methods, two blind spots, zero data-driven governance.

This article introduces a 2-axis, 4-decision framework: Keep, Repair, Reallocate, or Replace. Every device gets scored on technical health and financial value. The calendar drops out. Data steps in.

TL;DR: Fewer than 30% of IT leaders use analytical criteria to drive hardware refresh decisions (Gartner, 2024). The Keep, Repair, Reallocate, Replace framework crosses technical health with financial value to score every device. Result: 20-30% savings on hardware budgets with zero unnecessary replacements.

Why doesn't the 3-year IT hardware refresh cycle work anymore?

Fewer than 30% of IT leaders rely on analytical criteria to decide when to refresh their fleet (Gartner, 2024). The remaining 70% run on calendars or user complaints — two methods that burn budget and let preventable failures slip through.

< 30%of IT leaders use analytical criteria for hardware refresh

The 3-year myth comes from manufacturer warranties. Dell, Lenovo, and HP offer 3-year base coverage on their business lines. When the warranty expires, the IT hardware refresh reflex kicks in. But an expired warranty doesn't mean a dead device. It's an accounting date, not a technical diagnosis.

Consumer thinking leaks into B2B decisions too. "My laptop is 4 years old — should I replace it?" For a single device, fair question. Across 500 four-year-old laptops, the answer looks very different: 200 are running fine, 250 show mild degradation, and 50 are already clinically dead. Treating all three groups the same way throws money away.

Two laptops purchased the same day, same model, same config — and yet 30 to 40 points apart on a health score after 3 years. Why? Usage diverges. A developer compiling code all day doesn't wear out an SSD like an accountant running spreadsheets. Heat, battery cycles, installed software — everything creates variance.

Calendar-based refresh produces two symmetric errors. It replaces machines that are still healthy — budget waste — and keeps degraded machines in production — operational risk. No other management approach generates this double loss.

The 70% of organizations still running calendar-based refresh face a growing competitive risk: as analytical methods mature, early adopters are reallocating savings into AI-ready hardware and talent. Gartner (2024) reports fewer than 30% use data-driven criteria today — but that share is accelerating fast, and laggards will pay the spread in both cost and capability by 2027.

Real computer lifespan by brand — the actual numbers

What is the Keep, Repair, Reallocate, Replace framework?

The framework crosses two axes — technical health and financial value — to produce four possible decisions per device. Gartner estimates that 76% of end-of-cycle devices are refurbishable rather than replaceable (Gartner, 2024). This model captures that potential in a structured way.

76%of end-of-cycle devices are refurbishable

The principle is straightforward. Every device receives a score on two axes. The intersection determines the decision.

Axis 1 — Technical health

The first axis measures the device's actual condition. Not its age, not its purchase date — its functional state right now. Eight dimensions are evaluated: age, system stability, battery, storage, RAM, disk health, boot configuration, and warranty coverage. These eight dimensions consolidate into three scores: Health, Performance, and Security.

Axis 2 — Financial value

The second axis calculates what the device is still worth — and what it costs to keep. Annualized TCO, residual book value, repair cost versus replacement cost. A device can be technically healthy but financially irrational to repair. The reverse is also true.

The four decisions that follow:

• Keep — High health, low maintenance cost. Don't touch it. The device stays in production as-is.
• Repair — Health degraded on a specific point, but the rest is solid. Replace the battery, add RAM, swap in an SSD. Repair cost is well below replacement.
• Reallocate — The device no longer meets its current user's needs, but it's still functional. A developer who needs 32 GB of RAM hands their 16 GB laptop to an administrative role. The device gains 2-3 years of useful life with zero capital spend.
• Replace — Low health, repair cost exceeding 50% of replacement. The device exits the fleet.

Reallocation is the most underused lever. Most frameworks stop at keep, repair, or replace. Yet in a 500-device fleet, 15-20% of machines are "not enough for their current user" but perfect for a less demanding role. That's the best ROI in the entire model.

The Keep, Repair, Reallocate, Replace framework crosses technical health with financial value to score each device into four categories. Gartner (2024) estimates that 76% of end-of-cycle devices can be refurbished rather than replaced, making this decision model exceptionally cost-effective.

Explore the sobrii fleet scoring platform

How do you assess a device's technical health?

The assessment covers 8 dimensions. A concrete example: battery replacement costs 80-150 euros, versus 800-1,500 euros for a new device — a 10:1 ROI (Dell / Lenovo, 2025). But you need to know when a battery actually warrants replacement.

80–150 €battery replacement cost vs. 800–1,500 € for a new device

Here are the 8 dimensions, with alert thresholds for each:

1. Age. Years since deployment. Not a sufficient criterion alone, but a weighting factor. A 5-year-old device in excellent health is still a Keep.

2. System stability. Number of crashes, BSODs, and forced restarts over the last 90 days. More than 3 BSODs per month? The device shifts to Repair or Replace depending on root cause.

3. Battery. Remaining capacity as a percentage of original. Below 60%, it's a Repair trigger. The trap: always-docked laptops hide their degradation. The battery still degrades — throttling stays silent until the user unplugs.

4. Storage. SSD or HDD? An HDD in 2026 is an instant bottleneck. An SSD upgrade at 50-80 euros transforms perceived performance.

5. RAM. Average utilization above 90% means a bottleneck. Windows 11 consumes 4 GB at idle. 8 GB is the functional minimum. 16 GB is the comfortable standard.

6. Disk health. S.M.A.R.T. data: reallocated sectors, read errors, temperature. An SSD with climbing error rates is a Replace candidate, not a Repair.

7. Boot config. TPM 2.0, Secure Boot enabled, UEFI. Without these, no Windows 11 — and a growing security risk.

8. Warranty. Active or expired coverage. A device out of warranty with a motherboard issue goes to Replace. Under warranty, it's a free Repair.

These 8 dimensions consolidate into three scores: Health (stability + disk + age), Performance (RAM + storage + battery), and Security (boot config + warranty + OS). And at fleet scale? Same model, same age, 30-40 points apart between the best and worst device. That's exactly why the calendar doesn't work.

Fleets that skip battery-level diagnostics leave the cheapest ROI on the table: at 80-150 euros per swap versus 800-1,500 euros for a new device, every missed repair is a 10:1 return forfeited (Dell / Lenovo, 2025). By 2027, predictive battery scoring alone is projected to defer 15-20% of full replacements in enterprise fleets — turning a minor maintenance line into a strategic budget lever.

Complete battery health diagnostic for enterprise fleets

How do you calculate residual financial value?

A laptop purchased at 1,200 euros depreciates over 5 years — leaving 240 euros of book value in year 4. But book value doesn't reflect functional value. With $6.08 trillion in global IT spending projected for 2026 (Gartner, 2025), every euro matters.

$6.08Tglobal IT spending projected for 2026

Book depreciation and functional value often diverge. A device depreciated to zero can still run 2-3 more years. Conversely, a device with 500 euros of book value might be unusable if the SSD fails and repair costs 400 euros.

Annualized TCO clarifies the decision. The formula:

Annualized TCO = (Purchase price + cumulative maintenance + support costs) / years in service

A 1,200-euro device used for 3 years with 200 euros in support = 467 euros/year. The same device used for 5 years with 400 euros cumulative support = 320 euros/year. Extending lifespan mechanically reduces annualized TCO — as long as maintenance costs don't spiral.

And that's where the 50% rule kicks in. If repair cost exceeds 50% of an equivalent new device, you replace. Battery at 120 euros on a 1,200-euro laptop? Repair — that's 10%. Motherboard at 650 euros? Replace — that's 54%.

Reallocation crushes every other option on cost. A reallocated device costs only reconfiguration time — no purchase, no parts. That's 50 euros per year versus 320 euros for a replacement. Across 100 devices reallocated instead of replaced, savings exceed 27,000 euros annually.

The annualized TCO of a 1,200-euro device drops from 467 euros over 3 years to 320 euros over 5 years. The 50% rule simplifies the call: if repair exceeds half the cost of a new device, replace. Reallocation remains the cheapest decision at 50 euros per year per device.

Calculate real per-device TCO with sobrii

How do you shift from reactive to predictive refresh?

By 2028, more than 70% of large enterprises will have adopted analytical Device Lifecycle Management (Gartner, 2025). Continuous monitoring combined with automated alerts lets you plan replacements 3-6 months ahead — instead of absorbing breakdowns.

70%+of enterprises will adopt analytical DLM by 2028

The reactive model works like this: a user calls the helpdesk, a technician investigates, the purchase order goes out, the new device arrives 2-3 weeks later. Meanwhile, degraded productivity, workaround solutions, frustration. The same scenario repeats every month across different users.

The predictive model reverses the sequence. K/R/R/R scoring updates monthly for every device. A 4-tier policy drives actions:

What does this actually change? No more surprise breakdowns. IT hardware refresh budget gets planned quarterly instead of managed in crisis mode. Users don't endure months of silent degradation before someone files a ticket.

On a typical 500-device fleet after 3 years of use, the observed distribution is: 35% Keep (175 devices), 20% Repair (100 devices), 15% Reallocate (75 devices), 30% Replace (150 devices). That means 70% of the fleet does not need a brand-new replacement.

Automated scoring also eliminates human bias. A technician who sees a 4-year-old laptop thinks "old." The score might say 78/100 — Keep. Conversely, a 2-year-old device with a failing SSD and 12 BSODs in 30 days deserves immediate Replace, regardless of age.

Gartner (2025) projects 70%+ DLM adoption by 2028 — meaning the remaining 30% will face supplier deprioritization, longer lead times, and inflated spot-buy pricing as vendors shift incentives toward data-mature customers. Organizations that delay predictive scoring by even 12 months risk locking themselves into the costliest procurement tier during the next hardware cycle.

sobrii scores every device automatically — explore the platform

What ROI can you expect from data-driven IT hardware refresh?

Organizations that switch to data-driven refresh see 20-30% savings on their annual hardware budget (Gartner, 2025). On a 500-device fleet, that represents 48,000 euros per year.

20–30%savings on hardware budget with a data-driven approach

First lever: avoided replacements. In a 3-year calendar cycle, 30-40% of devices are still healthy at the time of replacement. Across 500 devices refreshed at 160,000 euros per year, that's 48,000-64,000 euros of machines replaced for nothing.

Second lever: zero emergency downtime. A planned replacement takes 30 minutes of transition. An emergency replacement costs 4-8 hours of lost productivity — not counting stress and temporary workarounds.

Third lever: environmental impact. Manufacturing a laptop generates 300-400 kg of CO2 (ADEME, 2024). Extending average lifespan from 3 to 5 years reduces emissions by 40% across the full lifecycle. That's a measurable CSR argument, not a slogan.

During an audit of a 7,000-device public-sector fleet (Montpellier Metropole), 40% of devices flagged for calendar-based replacement were still functional after scoring. That's 2,800 devices that didn't need replacing — and potential savings of several million euros.

The 20-30% savings from data-driven refresh compound year over year: Gartner (2025) notes that organizations reinvesting those savings into fleet analytics see a further 10-15% cost reduction in year two as scoring accuracy improves. For a 500-device fleet, this turns an initial 48,000-euro saving into a 60,000+ euro annual advantage by the third cycle — widening the gap against calendar-driven competitors.

Measure your fleet's environmental impact with sobrii

FAQ

How often should you refresh your IT fleet?

There's no universal frequency. The enterprise average is 3.7 years for a laptop (Gartner, 2024), but that figure hides enormous variation. The right answer: when technical health data and TCO justify it — not when the calendar says so.

How do you build a business case for hardware refresh?

Three data-backed arguments: annualized TCO per device (shows the real cost of keeping vs. replacing), the cost of NOT replacing (breakdowns, lost productivity, security risks), and the 50% rule that objectifies every decision. Finance teams respond to numbers, not gut feelings.

Should you buy new or refurbished devices?

Gartner estimates that 76% of end-of-cycle devices are refurbishable (Gartner, 2024). Refurbished devices reduce costs by 40-60% compared to new. The key: require a technical health score before purchase and at least a 12-month warranty on refurbished units.

When should you reallocate instead of replace?

When the technical score is moderate (60-75/100) but the current user profile is demanding. A laptop with 16 GB of RAM and a 256 GB SSD no longer cuts it for a developer — but it's perfect for an administrative role. Reallocation costs 50 euros per year versus 320 euros for a replacement.

What software should you use to manage hardware refresh?

sobrii automates K/R/R/R scoring across both axes (technical health + financial value). Partial alternatives exist: GLPI for inventory, Intune for deployment. But none automatically crosses all 8 health dimensions with TCO to produce a per-device recommendation. It all starts with an IT asset inventory that answers 7 questions per device.

Discover sobrii's fleet management features

Conclusion

Calendar-based refresh is an outdated model. Here's what matters:

• Fewer than 30% of IT leaders use analytical criteria to refresh — the rest waste budget or absorb preventable failures.
• The K/R/R/R framework crosses technical health (8 dimensions) with financial value (TCO, 50% rule) to produce 4 objective decisions.
• Reallocation is the most underused lever — €50/year per device versus €320 for a replacement.
• 20-30% savings on hardware budgets when you switch to data-driven refresh.
• 76% of devices at end of cycle are refurbishable, not disposable.

A modern IT hardware refresh strategy replaces the calendar with data. It's that simple.

Request a fleet audit — free scoring on 50 devices Explore the sobrii platform