Atlassian has set the dates: no new Data Center sales after 30 March 2026, last Data Center renewal on 30 March 2028, retirement in 2029. Those facts sit in the full Atlassian Data Center end-of-life guide. This article is the schedule for acting on them, built so that every negotiation moment happens while you still hold a credible choice rather than after the clock has taken it away.
Why does timing decide the leverage?
Because optionality is the asset, and the clock spends it. Today you can credibly renew Data Center, stay put, and decide your Cloud move on your schedule, which means Atlassian has to earn the migration with incentives and protection. By late 2028 that choice is mostly gone, the deadline does the talking, and Atlassian no longer needs to offer much because you have nowhere else to be. The buyer-side rule is simple: negotiate while you still have a No in your pocket.
What does the migration timeline look like, phase by phase?
Work backward from your intended cut-over, and anchor the whole plan ahead of the 30 March 2028 renewal wall. The phases below assume a 12 to 18 month runway; compress them at your own cost in leverage.
| Phase | Window before cut-over | What you do |
|---|---|---|
| 1 · Scope and inventory | 12 to 18 months | Map users, projects and the full Marketplace app estate; classify each app keep, replace or retire |
| 2 · Model the destination | 9 to 12 months | Size the Cloud user tier against band boundaries; price the plan and app re-licensing into a costed business case |
| 3 · Negotiate the terms | 6 to 9 months | Trade migration timing for a multi-year price hold, capped uplift, step-up credits and migration assistance |
| 4 · Execute the migration | 2 to 6 months | Run the technical move; minimize any dual-licensing overlap where you pay both platforms |
| 5 · Optimize post-move | After cut-over | Reconcile actual Cloud usage, retire unused apps, confirm the tier still fits |
Phase 1 · Scope and inventory
You cannot negotiate or cost a migration you have not measured. Inventory the user base, the projects, and every Marketplace app, then classify each app as keep, replace or retire. App re-licensing is the most commonly underestimated line in an Atlassian Cloud move, so do this first and do it thoroughly.
Phase 2 · Model the destination by tier
Cloud prices step at user-count thresholds, so the destination tier is the cost cliff. Size your user count against the band boundaries deliberately and choose the plan on evidence, not on the vendor's recommendation. Build the costed comparison that finance will sign, loading in app re-licensing and any AI add-ons, so the business case reflects total cost and not just the subscription line.
Phase 3 · Negotiate from the timing you control
This is the phase the whole schedule exists to protect. Because Atlassian wants estates on Cloud before 2029, your decision to commit early is worth something, so trade it: a multi-year price hold, a capped annual uplift, step-up credits that carry your Data Center value forward, and migration assistance. The companion discipline of timing the Data Center side is in how to time an Atlassian Data Center renewal against the 2028 deadline.
Our Jira Service Management Negotiation Guide includes the Data Center-to-Cloud cost model and the phased migration checklist behind this timeline.
Book a renewal review.
We phase your Atlassian migration so every negotiation happens while you still hold leverage, then trade your timing for price protection. Fixed fee or gainshare, no savings no fee.
Book a renewal review →Phase 4 · Execute and avoid double payment
During cut-over you may briefly run both Data Center and Cloud, and paying for both is a cost to negotiate down rather than accept. Keep the overlap short and ask Atlassian to cover or credit the dual-licensing window as part of the migration deal.
Phase 5 · Optimize after the move
A migration is not finished at cut-over. Reconcile actual Cloud usage against what you committed, retire apps you carried over but do not use, and confirm your user count still sits in the tier you sized for. This is where the last of the saving is captured.
What are the timing mistakes that cost the most?
Three recur. The first is drift: doing nothing, renewing Data Center at rising list prices, and arriving at 2028 with no plan, which hands the deadline to Atlassian. Drift is the default outcome and the most expensive one, because it spends every month of leverage for nothing in return.
The second is compressing the runway. A migration scoped, modeled and negotiated in three months instead of twelve forces every decision under time pressure, which is exactly the condition that produces an over-sized tier, an unreconciled app estate and a weak negotiating position. The technical move can be fast, but the scoping and negotiation that protect cost cannot be rushed without paying for it.
The third is negotiating the Cloud deal and the final Data Center renewal separately. They are one decision. The terms you accept on your last Data Center renewal and the terms you secure on Cloud should be negotiated together, because your willingness to commit to Cloud early is the concession that earns protection on both. Splitting them lets the vendor win each in isolation. Treating the wind-down as a single, sequenced negotiation is what keeps the leverage compounding rather than leaking.
What to do next
Set your intended cut-over date, then work the phases backward from it, anchored ahead of 30 March 2028. The single most valuable move is to start Phase 1 now, because every month of runway is a month of leverage. We have run this schedule against deadline-driven migrations and $420M-plus of ITSM contract value since 2019, and the engagement runs through our renewal advisory service on fixed fee or gainshare. The commercial hub for the platform is the Jira Service Management platform page.
Frequently asked questions
- When should you start an Atlassian Cloud migration?
- 12 to 18 months before cut-over and well before the 30 March 2028 last renewal date, so you can scope, model and negotiate while you still hold a credible option to stay on Data Center.
- How does the end-of-life clock affect leverage?
- It cuts both ways. Plan early and your migration timing is a concession you trade for protection; plan late and the deadline negotiates for Atlassian.
- Should you migrate before the 2028 deadline?
- Decide deliberately on a costed model. An earlier planned migration can be traded for incentives and a price hold; a late one risks rising Data Center list prices and a move under pressure.