Close

Cloud migration models: which works for you?

Before the pandemic of 2020 hit and hit hard, 61% of companies were already planning a cloud migration. Post-pandemic, the speed of those migrations increased by a factor of 24, according to McKinsey. So, if you’re planning a migration, you are in good company. 

And once you know you want to migrate, the question becomes: Which cloud migration model is right for your business?

The best strategy for your company will vary depending on a few factors, including the size of your organization, which data you want to migrate, and the types of teams that will first use the cloud applications. Here are some of the most popular strategies, how they work, and who they’re right for.

Optimize and shift (recommended)

Step 1

Magnifying glass illustration

Audit data

Step 2

Bullseye on target illustration

Prioritization the data to migrate

Step 3

Rocket launching illustration

Migrate your optimized data set

Optimize and shift (also known as replatforming) is a strategy where you audit and prioritize which data to migrate to the cloud and then migrate it all at once. Whatever data you don’t choose to migrate can be stored on your servers in a read-only state for future reference. 

The benefit of this strategy (and the reason we recommend it for most Atlassian customers) is that you are only migrating what you actually need. This saves time, reduces complexity, and—in many cases—saves money. Think of it as a chance to “spring clean” your systems, taking what you need and leaving behind what you don’t. 

If you choose this strategy, your first step will be auditing the data you currently have. Next, you’ll prioritize which data you want to migrate, leaving behind data you no longer need. And then you’ll proceed to migrate with that smaller, higher-priority data set, keeping your migration lean.

Benefits

  • Everything is migrated at once, which means no lengthy transition process
  • Only migrate what you need
  • Shorter migration timeline and often reduced migration downtime
  • Makes cloud simpler for teams to navigate
  • May improve cloud performance
  • May decrease migration costs due to shorter timeline

Challenges

  • User onboarding happens all at once (and help desk resources may be particularly busy during those first weeks)
  • If you’re migrating a lot of data, an all-at-once approach may increase migration downtime
  • Requires up-front planning and strategic work

You should optimize and shift if…

  • You want to move all your users to cloud at once
  • Your migration is low-complexity (fewer than 5,000 users and 10 apps)
  • Your critical apps are available in cloud and have migration pathways
  • You are migrating Jira Service Desk or Advanced Roadmaps

Lift and shift (rehost)

Cloud migration models

Lift and shift is a cloud migration model where you migrate all your data at once and get your teams working in cloud together. This model is similar to optimize and shift, but without the up-front data audit and optimization. This means all product, user, and app data is shifted from on-premise to cloud. If you know you need to migrate all your data or if you need to make the move to cloud asap and plan on optimizing later, this is the strategy for you.

Benefits

  • Everything is migrated at once, which means no lengthy transition process
  • Shorter migration timeline
  • May decrease migration costs due to shorter timeline

Challenges

  • User onboarding happens all at once (and help desk resources may be particularly busy during those first weeks)
  • If you’re migrating a lot of data, an all-at-once approach may increase migration downtime
  • Moving unnecessary data to the cloud can increase costs

You should lift and shift if…

  • You need to move to cloud asap
  • Your migration is low-complexity (fewer than 5,000 users and 10 apps)
  • Your critical apps are available in cloud and have migration pathways
  • You do not need (or want) to clean up your data before migrating
  • You are migrating Jira Service Desk or Advanced Roadmaps

Phased migration

Cloud migration models

Phased migration is a strategy that migrates users and data in stages. By necessity, this means your migration would take longer. But it also means you can train and onboard users in small groups instead of all at once (potentially keeping your help desk from getting overwhelmed with the transition) and that you can work out any process or migration issues along the way while only impacting a small group of users.

Benefits

  • Phased user onboarding means a less frantic schedule for your help desk
  • Any issues that crop up during migration only impact a few users at once
  • Gives users time to adapt to the new way of working

Challenges

  • Not well supported if you need to migrate Jira Service Desk or Advanced Roadmaps
  • The longer migration timeline may mean more work for your tech teams and higher costs
  • This approach is more complex, as your teams manage multiple deployments during the lengthy transition
  • Requires careful planning and dependency mapping
  • Fixing broken links post-migration is more challenging when only some projects are migrated

You should phase your migration if…

  • You have a longer timeline for migration
  • Your migration is high-complexity (5,000+ users and 10+ apps)
  • Some teams’ critical apps are not yet available in cloud

Start fresh (repurchase or drop-and-shop)

Cloud migration models

Start fresh (also known as “archive and go”) is a cloud migration model that does exactly what it sounds like: starts your team in cloud without any legacy data. If you’re confident you can archive most (or all) of your existing server data to read-only mode and you want to get your teams into their new cloud tools asap, this strategy—which isn’t really a migration, but more of a process shift—is for you.

Benefits

  • No (or limited) migration downtime
  • If you have a server license, you can keep data archived on-prem

Challenges

  • Users won’t be able to edit old projects or other data
  • Archived data will be stored separately

You should start fresh if…

  • You want to expedite your move to the cloud
  • You know you won’t need to work with your existing server data
  • You’re a Jira Service Management customer and don’t need quick access to old tickets
  • You’re creating a new team or restructuring the way you work
  • You want to streamline your cloud footprint

Merge and consolidate

Cloud migration models

A merge and consolidate cloud migration model is designed for companies who want to consolidate multiple server and cloud sites into a single cloud. The goal is typically to simplify systems by bringing everything together.

Benefits

  • Less maintenance work for your IT teams
  • Easier access for users
  • Tools work together without complex integrations

Challenges

  • The more complex your current setup is, the more complex this migration will be
  • You’ll need to strategically audit all systems before migrating to make sure you don’t lose critical apps or functions in the shift

You should merge and consolidate if…

  • You want to simplify your setup
  • You need to free up tech teams from complicated maintenance requirements
  • Your critical apps are available in cloud and have migration pathways

Staged migration/permanent hybrid

Stage migration private cloud illustration

Private cloud

Stage migration public cloud illustration

Public cloud

If you already know a hybrid cloud strategy is right for your business, then a staged/permanent hybrid migration may be the right choice for you. In this migration model, some applications and data stay on a private cloud, managed by your team, and some are moved to the public cloud, managed by a vendor and easier to access.

Benefits

  • You choose which data to host in cloud and which to keep on-prem

Challenges

  • Tech teams will be responsible for more integrations and maintenance
  • Users may have more complex processes when jumping between cloud tools and on-prem tools, which may impact productivity
  • Hosting data on-prem is typically more expensive

You should stage your migration if…

  • You are subject to stringent legal guidelines that require you to host some data on-prem
  • You need an unusually high amount of customization or control over some data

Choosing your cloud migration model

You can prepare for a cloud migration by assessing the amount of data you have in active use, how often it’s accessed, where it’s currently located (in one place or many?), and how high-priority it is for the business. From there, it should start to become clear which model is right for your business.

And if you’re still unsure? Visit our migration center for more information about cloud migration.

Up next...

Cloud Platform

Visit the Atlassian Trust Center

Get the latest information on the security, reliability, privacy, and compliance of Atlassian's products and services.

Cloud Migration

Learn more about cloud migration

Find all the resources, tools, and support you need to start evaluating if cloud is right for your organization.