Network Modernization Roadmap: 12-Month Plan From Assessment to Cutover
8 min read ยท 2026-10-25
A network modernization roadmap is a phased plan to replace aging network hardware, circuits and manual operations with a design that fits how your company works now: you assess what you have, set a target architecture, prove it at a pilot site, roll it out in waves and retire the old gear and contracts. For a company with dozens of sites, a common plan runs about a year, with most of the time spent in site waves.
The technology choice is only part of the job. What decides success is an accurate inventory, contract end dates that line up with your waves, change windows the business accepts and a cutover runbook each site team can follow. The roadmap below covers all of that, with a filled-in example, a RACI, the metrics to report and how to present the plan to leadership.
The roadmap at a glance
Goal: All sites run on the target network design with central management, legacy circuits and end-of-life hardware are retired, and the network team reports agreed performance and availability metrics every month. Duration: 12 months
Assess the current network (Weeks 1-6)
Know exactly what you run, where it breaks and what the business needs from it.
- Build an inventory of every site, circuit, router, switch, firewall and wireless controller, with model, software version and support end date.
- Collect contract details for each circuit and support agreement: provider, term and notice period.
- Pull performance data for latency, packet loss, outages and helpdesk tickets by site.
- Interview business and application owners about cloud apps, video, remote work and growth plans.
Milestone: Inventory and assessment report reviewed with the CIO, listing end-of-life gear, problem sites and contract end dates. Metric: share of sites with a complete inventory record.
Target architecture and business case (Weeks 5-12)
Decide what the new network looks like and why it is worth doing.
- Define design principles, such as cloud-first traffic, zero trust access and central management.
- Choose the building blocks, for example SD-WAN for branch connectivity, SASE or cloud security for internet traffic, and standard switch and Wi-Fi designs per site size.
- Create two or three site templates (small office, large office, warehouse or plant) with standard bills of material.
- Write the business case with costs, risks retired and benefits, and agree it with finance.
Milestone: Target architecture and site templates approved by the architecture board, and business case signed off. Metric: sites mapped to a standard template.
Vendor selection and lab (Weeks 10-18)
Pick partners and prove the design works before touching a live site.
- Run a request for proposal with clear requirements and a scoring sheet.
- Build a lab with the chosen gear and test routing, failover, security policies and monitoring.
- Write the standard configuration and the automation scripts or templates for each site type.
- Draft the site cutover runbook with steps, owners, timings and rollback points.
Milestone: Vendor contracts signed and lab tests passed against the test plan. Metric: test cases passed, including failover and rollback.
Pilot sites (Weeks 16-22)
Prove the design and the runbook on real sites with real users.
- Pick two or three pilot sites of different types, with local contacts who will report issues.
- Cut over each pilot during an agreed change window, following the runbook.
- Run old and new links in parallel for a short period where possible, then switch fully.
- Collect user feedback and performance data, then update the templates and runbook.
Milestone: Pilot sites live on the new design for four weeks with no major incidents, runbook updated. Metric: performance at pilot sites against their baseline.
Rollout in waves (Weeks 22-46)
Move the remaining sites in planned, repeatable waves.
- Group sites into waves by region, type and circuit contract end dates.
- Order circuits early, since lead times for new circuits can be long in some locations.
- Run each wave with a pre-check, a cutover window, a post-check and a sign-off by the site contact.
- Hold a short review after each wave and feed lessons into the next one.
Milestone: All waves completed and signed off by site contacts. Metric: sites cut over against plan, and incidents per cutover.
Decommission and operate (Weeks 44-52)
Retire the old network and hand the new one to steady-state operations.
- Cancel legacy circuits within their notice periods and confirm final bills.
- Remove and dispose of old hardware following your asset and data disposal policy.
- Update network diagrams, the configuration database and support runbooks.
- Set monthly reporting on availability, performance, tickets and configuration compliance.
Milestone: Legacy circuits cancelled, old hardware retired and operations handover signed. Metric: legacy circuits and devices still active (target: zero).
Example: a roadmap for a 40-site retailer
Picture a retailer with 36 stores, two distribution centers and a head office, all on an MPLS network with firewalls nearing end of support. Store staff complain about slow cloud apps and payment terminal outages. Here is how the roadmap could look with owners.
The circuit contracts drive the order of the waves in this example. The network manager groups stores whose MPLS terms end in the same quarter, so the company does not pay twice for long.
- Months 1-2, assess: network manager builds the inventory, finance pulls circuit contracts. Milestone: report shows 20 firewalls past end of support.
- Months 2-3, design: network architect sets three templates (store, distribution center, head office). Milestone: architecture board approval.
- Months 3-5, vendor and lab: procurement runs the RFP, network engineers test failover for payment traffic. Milestone: lab sign-off.
- Months 4-6, pilot: two stores and one distribution center cut over at night. Milestone: four weeks stable.
- Months 6-11, waves: five waves of six to eight stores, store managers sign off each cutover. Milestone: all stores live.
- Months 11-12, decommission: MPLS circuits cancelled, old firewalls retired, operations team takes over. Milestone: handover signed.
Who owns what
Network projects involve more teams than people expect: security, procurement, finance, facilities and local site staff. Write a RACI at the start and confirm it with each team lead.
The site contact role is often missing. Without a named person at each site who can grant access and confirm things work, cutovers stall.
- CIO or head of infrastructure: sponsor, approves the business case and the change windows.
- Network architect: target design, site templates and technical decisions.
- Network program manager: plan, wave schedule, RAID log and reporting.
- Security lead: security policies, zero trust and internet security design, and sign-off on each template.
- Procurement and finance: vendor contracts, circuit orders and legacy contract cancellations.
- Site contacts: access, local testing and cutover sign-off.
- Service desk and operations: monitoring, incident handling and steady-state support.
Metrics to track
Set baselines in the assessment phase and report the same metrics every month. Report per site type, because a warehouse and a head office have very different needs.
Keep project metrics and service metrics apart. Project metrics show whether the roadmap is on track. Service metrics show whether users are better off.
- Sites migrated: sites cut over against the plan, by wave.
- Cutover quality: incidents and rollbacks per cutover.
- Availability: uptime per site from your monitoring tool.
- Application performance: latency and packet loss to key cloud apps and data centers.
- Helpdesk tickets: network-related tickets per site per month.
- Legacy retired: old circuits and end-of-life devices still active.
Risks and dependencies
Circuit lead times, contract notice periods and change freezes are the usual causes of delay. Put them in the RAID log from the first week, each with an owner and a fallback.
Plan around the business calendar. A retailer will not accept cutovers in peak season, and a manufacturer may only allow changes during planned shutdowns.
- Circuit lead times: owner is procurement. Fallback: temporary wireless or broadband backup links at sites that are waiting.
- Contract overlap: owner is finance. Fallback: align waves to contract end dates and send notices on time.
- Change freezes: owner is the program manager. Fallback: plan waves around known freeze periods.
- Hardware supply: owner is procurement. Fallback: order for two waves ahead and keep spares.
- Skills gap: owner is the network manager. Fallback: vendor training before the pilot and partner support during early waves.
How to present it to leadership
Fit the plan on one slide: the problem in business terms (outages at stores, slow cloud apps, hardware out of support), the target design in one line, the six phases on a timeline, the business case summary and the decisions you need, such as approval for change windows or the contract strategy.
Leave out device models and protocol detail. Leaders want to know what gets better for users, what risk goes away and what it will disrupt. During rollout, report a simple wave tracker with green, amber and red status and the next three waves.
How to build the roadmap view
Build the plan in a tool your stakeholders already use, such as Smartsheet, Microsoft Project, Excel or Jira. Use one swimlane per workstream: design, procurement, security, site waves and decommissioning. Put each wave as a bar with the sites listed in the details, and mark contract end dates and change freezes as milestones on the same timeline.
Add a site tracker as a second view, with one row per site and columns for wave, template, circuit order date, circuit live date, cutover date, sign-off and legacy cancellation date. Filter it by wave for each weekly review.
Common mistakes to avoid
- Starting design before the inventory is complete leads to surprises at cutover, so finish the site and contract inventory first.
- Ignoring circuit contract end dates means paying for old and new links at the same time, so align waves with notice periods.
- Skipping the pilot leaves the runbook untested, so prove the design at two or three different site types first.
- Treating every site as unique slows the rollout, so standardize on a small set of site templates.
- Leaving security to review at the end creates late redesigns, so make the security lead an approver of the target design and templates.
- Forgetting to decommission leaves old devices and circuits running for months, so give decommissioning its own phase, owner and target.
Frequently asked questions
What is network modernization?
It is the work of replacing aging network hardware, circuits and manual processes with a design suited to cloud apps, remote work and current security needs. It often includes SD-WAN, cloud-delivered security, updated wireless and more automation.
How long does a network modernization project take?
It depends on the number of sites, circuit lead times and change windows. Many mid-size companies plan for about a year, with most of the time spent on site waves. Plan by phases and waves rather than one end date.
Where should a network modernization roadmap start?
Start with a complete inventory of sites, devices, circuits and contracts, plus performance data and interviews with business owners. This shows which sites hurt most and which contracts set the timing.
What is the difference between SD-WAN and SASE?
SD-WAN manages connectivity between sites over several types of links. SASE combines networking and security functions delivered from the cloud. Many roadmaps start with SD-WAN and add cloud security services over time.
How do you minimize downtime during a network cutover?
Use a tested runbook with rollback points and cut over during agreed change windows. Run old and new links in parallel where possible. Have a site contact test key apps, such as payments or the ERP, right after the switch.