Suppose a four-person IT team has six weeks to migrate a set of services. The engineers know how long the migration tasks should take, but two of them also cover the help desk, a third handles escalations, and all four have work on projects already underway. The migration can miss its deadline even if the task estimates turn out to be accurate. The schedule assumed engineers would be available when commitments elsewhere already had a claim on their time.
IT capacity planning starts with a simple question: how many hours can the people with the right skills actually spend on this project, in the weeks when the work needs to happen? You can answer it with a weekly worksheet, a reasonable allowance for support, and a short review when priorities change. The goal is a commitment the team can explain and revise when the workload shifts.
Start IT capacity planning with the work the team already does
List everything that draws from the same people as the new project. Include tickets, incidents, on-call follow-up, vendor escalations, patching, access reviews, maintenance windows, meetings, and existing project commitments. Planned leave and training also affect availability. Work can consume capacity even when it never appears in the ticket queue.
Group those commitments in a way that makes sense to the team. For example, you can track incident recovery and its follow-up under one category or record them separately. Use the same convention each week. If the support history already includes scheduled maintenance, avoid deducting those hours again when you build the plan. The same applies if the staffing calendar already excludes leave.
Some teams bring in outside help for operational work such as backlog coordination or procedure documentation. Estimate how much internal time the arrangement actually frees. If a contractor can take over eight hours of documentation and the team spends two hours briefing and reviewing their work, six hours is a better starting estimate. Keep work that still requires internal access, approval, or accountability with the responsible person in the worksheet.
Check the coverage each service requires. A support queue with a response-time commitment may need a named engineer even on a quiet day. Leave space for the person who can step in if the first responder needs help. A rota can show who is on call, but the project plan also needs to show what that assignment does to the engineer’s availability.
Measure support effort and reserve time for demand
Start with roughly a month of records when you have them. Check tickets, incident timelines, recurring maintenance, and calendar entries. Use effort as well as ticket counts. Ten routine requests and ten complex escalations can occupy very different amounts of time. If your records are incomplete, ask engineers to sample a few representative weeks and include coordination and follow-up.
Time tracking built for development teams is one option for linking effort to work items. A ticket export and a shared worksheet can serve the same planning purpose if they capture the work you need to measure. Keep entries short: a ticket identifier, a category, and an effort estimate are enough for many teams. Explain who can see those records and use weekly totals to discuss workload.
Say the team logged 48, 56, 52, and 76 hours of support in four comparable weeks. The average is 58 hours. The 76-hour week deserves a closer look. If scheduled patching drove the jump and the project overlaps another patch cycle, its reserve should reflect that known demand. If a rare outage caused the jump, test what you would move or pause if another outage occurred. The average helps establish a baseline, while the pattern tells you when to depart from it.
Use a support allowance for each week, and adjust it for events such as a rollout, a holiday cover period, or a vendor migration that is likely to generate escalations. Keep the reserve visible in the plan. When support effort exceeds it, update the remaining forecast before taking on another commitment. Avoid treating the same estimated incident hours as both support and contingency.
Calculate the hours available for a six-week project
Start with scheduled hours and subtract the work that takes precedence or has already been promised. In this illustrative example, four engineers working 40 hours a week for six weeks have 960 gross hours. The team expects support and incident response to take 40% of gross hours. The other categories in the example are tracked separately.
| Six-week capacity calculation | Hours |
|---|---|
| Gross scheduled hours | 960 |
| Support and incident response | −384 |
| Planned leave | −40 |
| Recurring meetings | −48 |
| Scheduled maintenance | −36 |
| Existing project commitments | −52 |
| Available for the new project | 400 |
Check the denominator when working with percentages. If you measured support as a share of time after leave, apply that percentage to a comparable base. In this example the percentage is defined against gross hours, so the support reserve is 384 hours. Then the separate deductions leave 400 hours. A plan that assumes 450 hours of new project work needs a change to staffing, scope, or timing.
Do this calculation for each week before presenting the total. A six-week total can be useful for sizing the work, but a deployment window can still fail if its engineer has no capacity in the relevant week. Mark dependencies in the schedule too. A vendor approval that takes three days may need only 30 minutes of staff effort. It still affects when the next task can start.
Build a weekly worksheet by person and role
Use one row per engineer per week, or one row per interchangeable group when people can truly cover the same assignments. List scheduled hours, leave, meetings and maintenance, support reserve, existing commitments, and hours available for the new project. Then record the work assigned to that person and when it needs to happen.
| Person or role | Gross hours | Leave | Meetings and maintenance | Support reserve | Other project work | New project hours |
|---|---|---|---|---|---|---|
| Network engineer | 40 | 0 | 5 | 12 | 7 | 16 |
| Systems engineer | 40 | 8 | 4 | 8 | 0 | 20 |
| Week total | 80 | 8 | 9 | 20 | 7 | 36 |
The combined 36 hours may look sufficient. If the cutover requires 18 hours from the network engineer, that role is still two hours short. Move eligible preparation to an earlier week, assign it to a qualified backup, or change the cutover date. Check which tasks can be moved before assuming a different arrangement will solve the shortfall.
Use the same worksheet to make handoffs visible. Note who approves the change, who can perform it, who validates results, and who is available if the team needs to roll it back. NetworkUstad’s change and configuration management lesson explains how approvals and rollback planning fit into a controlled change. Put those activities on the calendar while the plan is still flexible.
Protect the cutover when support demand rises

Decide which work can be interrupted and which work needs a clear window. A review of migration documentation may be easy to pause. A live cutover with customers waiting may require uninterrupted access to the primary engineer and a backup. If the same person covers urgent support, give someone else the first response during the window and agree on when to escalate.
Treat after-hours changes as real commitments. Confirm who can cover the window and how that work affects their availability the following day. If the backup also handles support, check that both roles can be covered at the same time. Review the worksheet before the change freeze takes effect, while there is still room to reschedule or move a preparatory task.
If the week’s support reserve starts running out early, review the next milestone immediately. Reassign triage if there is qualified coverage, defer a lower-priority task, or move a milestone that depends on someone who is busy with an incident. If an external approval is late, bring forward work that can genuinely start without it. A single forecast update should show the new task order, responsible people, and date.
Record scope and schedule decisions alongside the work they affect. Connect each change to the relevant ticket, maintenance window, or incident record, so another engineer can reconstruct what happened. NetworkUstad’s documentation and lifecycle management lesson covers the operational records that help teams maintain that context. Keep documentation and handover effort in the schedule.
Pick a tracking method and review the plan weekly
Use a tracking method engineers can maintain during a busy week. Start with a shortlist if you need to compare options. This comparison of time tracking tools for developers and IT teams is published by actiTIME. Use it to identify options to test with your own team. Test a candidate with a few real tickets before choosing it. Check how easily engineers can enter time, fix mistakes, and export weekly totals. The quality of the planning information matters more than how much detail the tool can collect.
Give the weekly review a decision to make. Confirm upcoming leave and cutover windows, compare actual support effort with the reserve, and check whether named owners can still reach the next milestone. If last week was quiet, releasing reserved hours makes sense only when a task is ready and its owner is free. If the team repeatedly exceeds the reserve, adjust future plans and examine which recurring tasks could be simplified or shared.
State the trade-off clearly when the plan changes. For example, an incident has taken eight hours from the network engineer. The team can still cut over on Friday if a qualified backup takes the validation checks and the lab cleanup moves to next week. Put that choice in front of the people who own the date and the deferred work. Once approved, update the shared plan so both teams work from the same commitment.
At the end of the migration, compare estimated and actual effort by category. Separate project tasks that took longer from support surges, approval delays, and changes in who was available. Use that review when forecasting the next migration. A small, current worksheet can make delivery dates easier to defend because it shows what the team has already promised and what would have to move if demand rises again.
Planning disclaimer.
All hours, percentages, and schedules in this guide are illustrative. Use your team’s actual workload, coverage, and staff availability before committing to a delivery date.



