When Great Features Are Hard to Reach: The Rovo Agent Onboarding Challenge

Rovo
,
1

Hi there, @AdamMoore, @dmorrow and the Rovo team

I am working on a feature and considering using a Rovo agent to give teams guidance on a specific topic - but it seems Rovo is not ready for Prime time yet.

I noticed the following answer - leaving the Rovo Agent unaccessible for teams not on the right plan: What happens if a user doesn't use rovo tries to install an marketplace app which use rovo agent?

Furthermore, how can I effectively onboard and market an App - feature (Reporting and chat agent) that:

  1. Is only available for sites with Premium and Enterprise plans
  2. Is only available for sites with Rovo enabled
  3. Getting started with a Rovo Agent is as follows - users need to:
    a. Understand they are on a plan that includes Atlassian Intelligence
    b. Send an email to the Jira admin - ask them to Enable Rovo for the site
    c. Jira Admin will need to get approval from the Security team
    d. Once granted - Admin needs to inform the user they now got Rovo enabled
    e. User needs to search for and find the name of the App’s Rovo Agent
    f. Open the Rovo chat to find the correct Rovo Agent using the name
    g. Start using the amazing new Rovo Agent

I am afraid majority of users will get lost and only small fraction end up completing all steps up to 3g. I am afraid this means Rovo is not ready for it’s prime time.

What if?

  1. Atlassian Intelligence was available on all plans - limited usage based on plan → allow customer to upgrade to enable more usage
  2. Atlassian Intelligence was by default Enabled rather than Disabled - When and why defaults influence decisions: a meta-analysis of default effects | Behavioural Public Policy | Cambridge Core
  3. User could click a button within the UI - to open the Rovo chat - with the correct Rovo Agent

Are there plans to improve on this? Or do you have any insights showing app developers are not struggling getting users onboarded?

2

@bjornbrynjar,

Thanks for the constructive “what if” statements. Let me start there.

The Atlassian Intelligence brand and the associated features are effectively being folded into Rovo. Going forward the effective “toggle” for Rovo will allow customers to turn on or off LLM-based features, which is where the greatest concerns remain for customers. That includes Agents, including Forge-built ones, but also Chat and a few more features that were labeled Atlassian Intelligence.

To my knowledge, we tried Atlassian Intelligence “on by default”. I think that was already true last year, well before the Rovo announcement at Team. I believe that even Rovo is now “on by default”, for the Premium & Enterprise tiers. New trials and sign-ups on those tiers should have Search, Chat, and Agents.

I think your “getting started” rings true, even if some of those steps are a function of the current state of roll-out, rather than the end-state targeted to finish before July. Conversely, I think your statement about “ready for prime time” is overly broad. The core of your argument seems to be that a Rovo Agent alone is not a viable Marketplace offering.

That’s not entirely unique to Rovo’s Forge modules. There are many modules that don’t make sense to use alone to build an app. That doesn’t mean those modules aren’t “ready for prime time”. Indeed, I’m working with customers where Forge-based Agents are going into production and solve significant problems. That’s what “prime time” means to customers. In the short-term, our assumption about Rovo Agents is they would be part of existing Marketplace Apps, rather than something to build as a distinct business offering.

There are plans to improve Rovo’s extensibility via Forge. One direction is the ability to allow customers to use Forge-defined Actions in their own user-defined Agents. Another is to have Forge-based Search Connectors. For your specific needs of user onboarding, I like your suggestion to open Chat with an Agent active. And generally, we would welcome all suggestions that solve mutual onboarding: both driving Rovo usage and your specific agents.

1 Like
3

Thanks for the thoughtful and detailed reply @ibuchanan — and I agree, the title and phrasing of “not ready for prime time” was too broad. I’ll update the post title accordingly to better reflect the core issue I’m trying to work through.

“The core of your argument seems to be that a Rovo Agent alone is not a viable Marketplace offering.”

That’s not quite it. The Rovo Agent I’m building is designed to enhance an existing app by helping teams interpret current status and take action based on them. It’s not meant to stand alone — it complements a broader user workflow. The functionality itself is working well and is genuinely useful.

However, my core challenge — and reason for the post — is about how I can effectively sell and onboard users in a way that minimizes Time to Value and gets them to the Aha Moment quickly , despite the reality that:

  1. The app offers a great new feature — but it might not be available to you, depending on your plan or Rovo status.
  2. The process of experiencing the Aha moment is multi-step, requires update plans for many, and may take days.
  3. As a result, many users are likely to drop off before they ever experience the value the feature offers.

That’s the tension I’m trying to solve: the feature works and provides great value — but how to:

  1. market a feature that might not be available to large part of Atlassian Customer base (assuming most customers are on other plans than Premium and Enterprise)
  2. shorten the user path to the aha-moment

.. is the problem.

Solutions
Thinking about solutions:

  1. :white_check_mark: Rovo enabled for Enterprice and Premium
  2. Rovo enabled for all customers by default
  3. Allow Users to click a button within the UI to open a specific Rovo Agent
1 Like