Planning a zero-downtime cutover to VICIdial

Cutover day is the moment your old dialer goes quiet and VICIdial takes every call. Done well it is invisible to agents. Done poorly it is a frantic hour of missed dials and angry supervisors. The difference is almost entirely in the preparation you do the day before, not the day of.

Pre-cutover checks (the day before)

Start with a checklist that your team signs off on at least 24 hours before go-live. The goal is to make sure VICIdial is not just installed but production-ready. Verify that your SIP trunk registers cleanly - place a live test call in both directions and confirm audio is clear in both ears. Check that every Campaign you plan to cut over has the correct dial prefix, the right Lead list loaded, and realistic call times set for your timezone. Confirm agent logins work for at least a representative sample of your team.

• SIP trunk registers and passes a two-way voice test
• All campaigns have correct lead lists and dial settings
• Agent logins and softphones confirmed for every team member
• DNC lists exported from old system and imported to VICIdial
• Supervisor and admin accounts tested end-to-end

The full migration context, including lead export formats and carrier verification steps, is in our complete VICIdial migration guide.

Phase the cutover by campaign

Cutting all campaigns simultaneously is the highest-risk approach. Instead, phase by Campaign complexity. Start with your lowest-volume, simplest outbound campaign - ideally one with a small Lead list and no complex IVR (interactive voice response) routing. Let it run for 30 minutes with two or three agents watching call quality and Disposition accuracy. If everything looks right, bring in your next tier of campaigns.

Keep the old dialer idle but running - not deallocated - for at least the first full shift. This keeps the rollback path open without any additional provisioning time.

The cutover state diagram

stateDiagram-v2
    [*] --> PreCheck
    PreCheck --> ParallelRun : all checks pass
    PreCheck --> Blocked : check fails
    Blocked --> PreCheck : fix applied
    ParallelRun --> PhasedCutover : pilot campaign stable
    PhasedCutover --> FullCutover : all campaigns green
    PhasedCutover --> Rollback : error rate exceeds threshold
    FullCutover --> Monitor : old dialer idle
    Monitor --> Done : shift complete, no issues
    Monitor --> Rollback : critical failure
    Rollback --> ParallelRun : old dialer restored

Go/no-go criteria

Define your go/no-go thresholds before the day starts so the decision is not made under pressure. Typical thresholds: answer rate within 10% of baseline from the old system, Wrap-up time under 60 seconds, zero unanswered inbound calls to your queue, and no SIP errors in the first 50 outbound attempts. Write these down and assign a named person - not a committee - to make the call.

Monitoring the first live calls

Open the VICIdial real-time report in a separate browser tab and leave it visible throughout the shift. Watch the lines-per-agent ratio stay steady, confirm that dispositions are being recorded correctly, and have a supervisor listen to two or three live calls via the monitoring interface in the first 15 minutes. Problems usually surface immediately or not at all.

Also review building a VICIdial migration rollback plan before cutover day so the steps are already written if you need them.

If you are migrating to a managed VICIdial environment, VICIfast provisions a fully configured dialer in under 40 seconds so you can run your pilot campaign the same day you decide to move. See VICIfast managed plans for sizing options that fit your agent count.