
We’ve all been there. A support agent opens a new ticket in Jira Service Management. On the surface, it looks simple enough, a password reset, maybe a question about a feature. But reading between the lines, you can tell the customer is at their wit's end. They’ve already tried the self-service portal, waited on hold, and this ticket is their last shot.
How can your agent spot the difference between a routine question and a customer who’s about to churn?
This is exactly where sentiment analysis comes in. It’s a handy tool for modern support teams, helping you get a read on customer emotions, prioritize work, and ultimately provide better service. Atlassian has its own version of this built right into Jira Service Management, called Atlassian Intelligence Customer Sentiment.
In this guide, we’ll walk you through the whole feature. We’ll talk about what it does well, where it falls a bit short in the real world, and how to build a smarter, more proactive AI strategy for your entire support team.
What is Atlassian Intelligence Customer Sentiment?
Atlassian Intelligence Customer Sentiment is a feature inside Jira Service Management that automatically tries to figure out how a customer is feeling. It's not reading minds, just reading words. It uses Natural Language Processing (NLP) to scan the text in a ticket’s description and any follow-up comments.
Based on what it reads, it slaps one of three simple labels on the ticket: Positive, Neutral, or Negative.
The neat part is that this label isn't set in stone. It updates in real time as the conversation unfolds. If a customer who started out neutral adds a frustrated comment, the sentiment will flip to Negative, giving your agents a live look at the customer's mood.
It's a useful feature, but there is one string attached: it's only available for teams on Jira Service Management’s Premium or Enterprise plans, and an admin has to switch on Atlassian Intelligence first.
What Atlassian Intelligence Customer Sentiment does well
Having a sentiment score right inside your help desk is a definite plus. It gives your agents immediate context where they’re already working, helping them make better decisions without switching screens.
A smarter way to sort tickets
One of the most practical perks is using the sentiment field to organize your support queues. Since "Sentiment" is a standard field in Jira, you can use Jira Query Language (JQL) to create custom filters.
For example, a support manager could quickly whip up a dedicated queue for all tickets marked "Sentiment = "Negative"". This makes sure your most unhappy customers get helped first, going beyond the usual P1 or P2 labels that don't really capture how someone feels. It’s a simple way to get ahead of escalations and show your at-risk customers you’re paying attention.
How it helps your agents sound more human
When an agent sees that a customer is already upset, they can change their tone right from the first reply. Instead of a generic "Hello," they can start with something more empathetic that acknowledges the customer's frustration.
For a "Negative" ticket, a simple, "I'm so sorry to hear you're running into this" can make a huge difference. On the flip side, if a ticket is marked "Positive," an agent can match that energy to build a better connection. It’s a small tweak that goes a long way in making support feel less robotic.
A decent first step into proactive support
For teams just dipping their toes into data-driven support, this feature is a solid start. It gives you a built-in way to flag at-risk customers without needing a complicated setup or a third-party tool. It gets your team thinking about customer emotion as a real metric, which is a great first step toward a more proactive support culture.
Where Atlassian Intelligence Customer Sentiment falls short
While having a native tool is convenient, it's not always the whole package. Once you start using it, you might notice a few gaps that can keep your team from building a truly efficient and proactive support system.
You can’t see the big picture
A major drawback people run into pretty quickly is the lack of reporting for the sentiment feature. You can see the sentiment on a single ticket, but you can't easily track trends over time.
This leaves managers unable to answer important questions like:
-
Are we getting more negative tickets this month than last?
-
Which of our products is causing the most headaches for customers?
-
Does sentiment change depending on the customer's plan or location?
Without answers to these questions, you're just putting out individual fires instead of finding out what's causing them. You can see one customer is unhappy, but you can't spot the patterns that would help you keep others from feeling the same way.
It only knows what's in the ticket
Atlassian Intelligence is pretty smart, but it can only analyze what it can see. In this case, its view is limited to the text inside a single Jira ticket. The problem is, most of the information needed to actually solve the customer's issue doesn't live there.
It’s spread out across Confluence pages, Google Docs, internal wikis, and thousands of old support tickets. The sentiment analysis can tell you a customer is annoyed about a billing problem, but it can't hand your agent the step-by-step guide from your internal docs to fix it. It points out the symptom (frustration) but doesn't help with the cure (the actual problem).
It's reactive, not preventative
By design, the tool only flags a negative sentiment after a customer has already gotten fed up and written an angry message. At that point, you're playing defense, trying to salvage the relationship.
A truly helpful AI strategy shouldn't just be about managing frustration; it should be about stopping it from happening in the first place. That means giving customers self-service tools, like an intelligent chatbot, that can pull answers from all your company knowledge to solve their problem instantly, so they never have to create a ticket at all.
How to build an AI support system beyond Atlassian Intelligence Customer Sentiment
The good news is, you don't have to settle for the limitations of a built-in tool. By adding a dedicated AI platform, you can fill in these gaps and build a support system that’s not just reacting to problems, but actively preventing them.
Bring all your knowledge together for better answers
First things first: break down those information silos. A platform like eesel AI connects to all the tools your team already relies on. It works with Jira Service Management, but it also learns from Confluence, Google Docs, Slack, your help center, and your past ticket history.
By learning from everything, the AI gets a complete picture of your business. This unified knowledge can then power an AI Agent to handle frontline questions or an AI Copilot to give your human agents instant, correct answers. This helps them solve problems faster, which is the best way to turn a negative feeling into a positive one.
Test and report with confidence
Instead of crossing your fingers and hoping your AI works, you should be able to test it. The simulation mode in eesel AI lets you run your AI setup on thousands of your past tickets in a safe environment. You can see exactly how it would have replied, get solid predictions on how many tickets it can solve, and find gaps in your knowledge base before it ever talks to a customer.
Once you go live, the reporting dashboard gives you the insights that JSM's native tool is missing. You can track automation rates, see what your customers are asking about, and get a clear picture of how to improve your knowledge base and support.
Go live in minutes, not months
Getting started with AI shouldn't feel like a massive project. People in community forums often seem confused about how to even turn on Atlassian’s sentiment feature. In contrast, eesel AI is designed to be completely self-serve and straightforward.
You can connect your help desk with a click, train the AI on your knowledge, and launch a working agent in just a few minutes. No mandatory demos or long sales calls needed. This lets your team start small, show some quick wins, and then scale up your AI strategy when you're ready.
Pricing for Atlassian Intelligence Customer Sentiment: JSM vs. eesel AI
As mentioned, Atlassian Intelligence Customer Sentiment isn't part of every Jira Service Management plan. To get it, you need to be on the Premium or Enterprise tier, which can be a big price jump for some teams.
Here’s a quick look at what it takes to get Atlassian's feature:
Atlassian JSM Plan | Price (per agent/month, annual) | Sentiment Analysis Included? |
---|---|---|
Free | $0 (up to 3 agents) | No |
Standard | $22.05 | No |
Premium | $49.17 | Yes |
Enterprise | Contact Sales | Yes |
eesel AI offers a more predictable and transparent way to budget. All of our main products, including the AI Agent, AI Copilot, and AI Triage, are part of every plan. The big difference is that we have no per-resolution fees. Our pricing is based on the capacity you need, so you won't get a surprise bill after a busy month. Plans start at $239 per month for your whole team, and you can choose a flexible month-to-month plan you can cancel anytime.
Move from sensing emotion to solving problems
Atlassian Intelligence Customer Sentiment is a nice feature to have. It gives agents in JSM a basic, real-time clue about a customer's mood, which can help with sorting tickets and being more empathetic.
But on its own, it’s a reactive tool that doesn't solve the bigger problem. It doesn't have the reporting you need to make smart decisions, it only works with the information in the ticket, and it only tells you someone is frustrated after the fact. A better customer experience doesn't just come from spotting frustration; it comes from fixing the root cause quickly and efficiently.
This requires a dedicated platform that can unify all your company knowledge, give you useful reports, and is simple enough to set up and manage without needing a team of developers.
Ready to go beyond just Atlassian Intelligence Customer Sentiment?
If you're looking to build a support system that actually prevents frustration, resolves issues faster, and gives you clear insights, it might be time to look beyond built-in features. eesel AI plugs into tools you already use, like Jira Service Management, in minutes to create a unified, powerful, and easy-to-manage AI support setup.
Try eesel AI for free or book a demo to see how you can transform your customer support.
Frequently asked questions
Atlassian Intelligence Customer Sentiment is a feature within Jira Service Management that uses Natural Language Processing (NLP) to scan the text in a ticket’s description and comments. Based on this analysis, it assigns a sentiment label: Positive, Neutral, or Negative.
Support teams can leverage Atlassian Intelligence Customer Sentiment by using its sentiment field with Jira Query Language (JQL). This allows you to create custom filters, such as a dedicated queue for all tickets marked "Negative," ensuring that unhappy customers are addressed promptly.
No, Atlassian Intelligence Customer Sentiment currently lacks built-in reporting features. While it displays sentiment on individual tickets, it does not provide aggregated data or trends that could help managers track changes in customer emotions over time or across products.
To utilize Atlassian Intelligence Customer Sentiment, your team must be on Jira Service Management’s Premium or Enterprise plan. Additionally, an administrator needs to activate Atlassian Intelligence within your Jira instance.
Atlassian Intelligence Customer Sentiment is primarily a reactive tool. It identifies negative sentiment after a customer has already expressed frustration in a ticket, rather than proactively preventing issues from escalating or providing solutions before a ticket is even created.
No, Atlassian Intelligence Customer Sentiment is limited to analyzing the text directly within the Jira ticket. It does not integrate with or analyze content from external knowledge bases or documents, which means it can't provide context from your broader company knowledge.