Rovo usage quotas, indexing behavior, and future cost model

[Chris Price]

Hi Atlassian Community

I’m trying to get a clearer understanding of how usage-based costs for Rovo will work, particularly in relation to indexed objects, connected apps, and the upcoming consumption-based billing model. 

Right now, Rovo is included for free under our Premium plan, but we know that metered pricing is coming and we want to make informed decisions before rolling it out more widely across our organization.

Here are a few specific areas I’m hoping someone from Atlassian or the community can help clarify: 

Indexed Object Quota

• Can you confirm if Atlassian-native content (Jira issues, Confluence pages) does not count toward the indexed object quota?
• If we connect something like a Google Shared Drive with 2,250 documents, does that immediately consume our quota of 50 object credits?
• Does adding a new file later increase the count cumulatively (e.g., now 2,251)? Or is there a monthly reset?

Slack Connector Behavior

• If we connect Slack, does Rovo index every individual message and thread reply as its own object?
• Is it possible to limit indexing to specific channels, or is it currently all-or-nothing?

Gmail Connector

• Can someone confirm that Gmail is queried live and not indexed?
• Does this mean Gmail messages do not consume object quota?

We’re really keen to use Rovo, especially for internal Atlassian content, but we want to be cautious before enabling high-volume connectors like Drive or Slack only to end up with unexpected costs and users becoming dependent on features we might need to pull back later. Any clarity or shared experience would be appreciated!

Thanks in advance!

[Chris Price]

.

[Phil Shakhovskoy]

Hi Chris,

Most of your questions should be answered by the Rovo quota usage documentation.

• Atlassian objects don't count towards your quota. 
• Indexed objects are per active subscription, so you get 50 objects per user. In your example, if you had 45 licensed Jira or JSM or Confluence users, that would cover the 2250 documents. 
• The indexed objects quota is cumulative, and does not reset per month. If you were to remove a Google Drive document from the Drive, that number would go down. It's a running total. This differs from the AI credits quota, which does reset.
• We can't quite (yet) answer the specifics of how the Slack message objects will calculate. Once we have that detail, we'll be sure to add it to the page above.
• Currently, Slack is all or nothing - we're looking to make that more flexible, but it's not available yet. We currently support Google Drive and SharePoint blocklists. 

• Correct about Gmail - we integrate directly through Gmail's search and the content is not indexed by Atlassian, and therefore does not count towards your quota. 

If you think there could be anything changed on the support page where we've covered most of the above, let us know.

Thanks!

[Chris Price]

Thanks again for the follow-up. I’ve now reviewed the documentation on blocklists and wanted to raise a few further points regarding the Google Drive connector and quota control.

 As I understand it, blocklists currently support:

• Shared drive URLs (Google Drive)
• Up to 50 entries per connector
• No support for individual folders or files
• No way to block access to “My Drive” content

 This leads to two key concerns for us:

Quota risk from personal My Drives
Many of our users have 1,000–2,000+ items in their personal Google Drives. If individual users connect their own My Drives (as I am sure there is valuable content in there that would be beneficial to index), we could quickly and significantly exceed our pooled indexed object quota, even if we only intended to surface content from Shared Drives. Is there currently any way to restrict or disable indexing of My Drive content at the org level?

Lack of fine-grained control
The inability to block individual folders or files, and the 50-entry blocklist limit, makes it difficult to manage Drive usage at scale. Are there any plans to expand the blocklist functionality or introduce allowlists in the future (only allow indexing of specific Shared Drives)?

Appreciate your help and happy to clarify anything if needed. We’re trying to avoid putting ourselves in a position where feature usage outpaces quota visibility and control.

[Chris Price]

.