How to Automatically Reduce Your Atlassian License Costs in 2026

Question

If you manage Atlassian products for an organization of any size, there's a good chance you're paying for users who haven't logged in for months. It's one of those invisible costs that compounds quietly until someone in finance asks why the Atlassian bill keeps climbing.

This post walks through why inactive users are a real cost problem, what MQB billing means for you, and practical steps to reduce spend.

THE INACTIVE USER PROBLEM

When someone joins a company, they get provisioned on Jira and Confluence. When they leave or move roles, the account usually stays active. Research shows 10-30% of licensed users are inactive at any given time.

For 100 users at Standard pricing ($8.15/user for Jira, $5.75 for Confluence):

- 10% inactive = ~$1,600/year wasted

- 20% inactive = ~$2,200-$3,200/year wasted

- At 500 users with 20% inactive = $11,000-$16,000/year wasted

MQB BILLING MAKES IT WORSE

Since July 2025, Atlassian bills based on your HIGHEST seat count at any point during the month, not the end-of-month count. Even adding users temporarily for a few days means paying for the full month. Quarterly cleanups are no longer enough.

MANUAL CLEANUP DOESN'T SCALE

For 200+ users, manual audits take 4-8 hours per cycle. Different admins apply different standards. Between audits, inactive accounts pile up again. And Jira and Confluence have separate user lists, so you need to check both.

AUTOMATION IS THE ANSWER

What to look for in a Marketplace automation tool:

- Covers Jira AND Confluence (and ideally all Atlassian products) in one install

- Forge-native (data stays within Atlassian)

- SCIM compatible

- Audit logging

- Free tier to prove ROI

Full disclosure: I built User Management Automation (https://marketplace.atlassian.com/apps/1230326/user-management-automation-for-jira-confluence) which covers all of the above, including managed and unmanaged accounts. But regardless of tool choice, the key is moving from quarterly audits to daily automation.

PRACTICAL STEPS

1. Export your user list from Atlassian Admin, filter by last login (60-90+ days)

2. Segment by product -- a user active in Jira but inactive in Confluence is still partial waste

3. Define an inactivity policy and document it

4. Send warning emails before deactivating (reduces false positives)

5. Automate for ongoing enforcement

6. Review your tier -- removing users may drop you to a cheaper pricing band

Has anyone here done a full user audit recently? What percentage of inactive users did you find? Curious to hear real numbers from different org sizes.

Jonas Nilsson _ Unitlane · Answer

The key difference I would add after the MQB change is that cleanup value is less about end-of-month heroics and more about preventing avoidable users from entering the next billing cycle in the first place.

The hard part in practice is not only detecting inactive users. It is explaining why they are still billable and separating real humans from service accounts, contractors, and temporary exceptions before you act. That is usually where manual cleanup turns into a governance problem instead of a simple cost problem.

James - Seats Management · Answer

Good point, Arkadiusz. You're right that the slot-based billing changes the math -- removing a user mid-month no longer gives you an immediate credit.

But I'd argue that makes proactive automation more important, not less. The key is preventing inactive users from entering a new billing cycle in the first place, rather than cleaning them up reactively after the fact.

For companies under 1,000 users without major staff fluctuations, the savings per individual user are smaller -- agreed. But in practice, even stable organizations accumulate 10-15% inactive accounts over time through natural attrition (role changes, quiet departures, forgotten contractor accounts). At 200 users, that's 20-30 seats. At Standard pricing, that's still $2,000-4,000/year.

The question is whether that amount justifies the 15 minutes of setup time for an automated tool. For most teams, the answer is yes -- especially since the security/compliance benefit of removing dormant accounts is independent of the billing model.

Where I agree with you: if an organization has strong offboarding processes and genuinely low turnover, the ROI is lower. But those organizations are rare in my experience.

Arkadiusz Wroblewski · Answer

@James - Seats Management

In my view, constant optimization made more sense before. Since the billing changes a few months ago, from license-based billing to “license slot billing,” where you effectively still pay for the slot until the end of the next billing month, the benefit is not as strong anymore for companies with fewer than 1,000 users and no major staff fluctuations.