Issues & occurrences

Issue tracking for support, for the problems that keep coming back

A label tells you what a conversation is about; an issue tells you which problem came back, how many times, when and at what cost to your team. In Hodhod every recurring problem gets an issue, with occurrences you can measure separately and in total.

Why aren't labels enough?

A "payment" label tells you 40 conversations were about payments, but not whether they all came from one gateway outage or were forty separate problems. In Hodhod, an issue is separate from a label: one specific problem, with a title, that ties together different conversations and tickets.

Occurrences: when the same problem happens again

If the same problem happens again later, you don't need to create a new issue. An issue has occurrences: separate time windows for the same problem. When you link a conversation to an issue, you say whether it belongs to the current occurrence or a new one has started. Each occurrence:

  • Has a start and end time and notes (cause, fix).
  • Can close automatically if no new conversation is linked for a while; if a conversation is linked later, a new occurrence starts.
  • Can be closed by hand, or you can move conversations to another occurrence to split some of them off.

Measure each occurrence and the total

For every issue you see both per-occurrence and overall numbers, and you can compare occurrences side by side:

  • Satisfaction survey: satisfied (CSAT) and dissatisfied (DSAT) counts and the average score.
  • SLA compliance.
  • First response time and resolution time.
  • Mean time to resolve (MTTR) and recurrence rate: see how long a problem usually takes to fix and how often it comes back. MTTR is shown as mean, median and 90th percentile (p90), alongside the mean time between recurrences (MTBR) and the recurrence rate per 30 days.

Issue reports

The issue report shows new and ongoing issues, repeats, linked conversations, issues with the most conversations, and customer satisfaction and SLA broken down by issue. The "Recurring issues" section tells you which problems came back in the selected period, with the most frequent first, so you know which root causes to close for good. The "Chronic issues" list shows problems that occurred at least three times in the last 90 days, and a recurrence timeline lays the occurrences out on one axis.

Automatic linking

A new conversation that matches an ongoing issue is linked to it automatically, so agents don't have to search by hand every time. An admin writes a rule per issue, based on keywords (with Persian text normalised), labels, inbox or ticket category, and sees a live preview over recent conversations showing which ones would have been linked before turning it on. The conversation panel also suggests up to three issues. If the issue's last occurrence is closed, the rule by default only suggests; the admin can choose whether a new occurrence starts or the previous one continues.

An alert before your team is caught off guard

When the number of conversations about one issue suddenly climbs, Hodhod alerts the people in charge so they can recognise the incident before the queue grows. For each issue you set a threshold, a time window and a cooldown between alerts. The alert appears inside the panel as a notification, banner and toast for the owner and admins; there is no email, SMS or Telegram alert. If you want it to reach another tool, you can set an https webhook target. Someone acknowledges the alert with a button, and when the spike subsides the incident ends automatically. See also incident alerts in customer support.

Tell customers before they ask

For an ongoing issue you can show a notice (for example "The payment gateway is being looked into") so customers see the status instead of messaging you. This is optional and off by default. The notice appears as a banner in the widget of each inbox you choose, and after the issue is resolved you can also send a one-time message to the open or pending conversations linked to it. When you close an issue manually, you confirm the recipients yourself.

Send the problem to engineering in Jira

From an issue, create a Jira issue with a table of the issue's metrics, or link an existing Jira key. Hodhod works with Jira Cloud (email and API token) and Jira Server or Data Center (personal access token). The link goes both ways: through a Jira webhook, the occurrence closes when the Jira issue is Done; and if the problem comes back, a new comment is added to the Jira issue and it can reopen one that was Done. Credentials are stored encrypted and only an admin sets up the connection. Full guide: connecting support to Jira.

Linked to conversations and tickets

From the sidebar of any conversation and any ticket, you can link it to an issue or create a new one. That way support, engineering and the product manager talk about one "problem", not ten scattered conversations. Review the key support metrics in customer support KPIs.

Frequently asked questions

How is an issue different from a label?

A label only categorizes and counts. An issue is one specific problem with its own occurrences, customer satisfaction, SLA and resolution time.

Are issues created automatically?

Your team creates the issue, but you can define a rule so similar conversations are linked to the right issue automatically, and an agent can change it whenever they like.

What if the problem comes back months later?

You start a new occurrence on the same issue. The history of earlier occurrences stays, and you can compare against it.

Do spike alerts arrive by email or Telegram?

No. The alert appears inside the Hodhod panel for the owner and admins. If you want it elsewhere, set an https webhook target and Hodhod will send the alert there.

Which Jira versions does the connection support?

Jira Cloud (email and API token) and Jira Server or Data Center (personal access token). Only an admin sets up the connection, and credentials are stored encrypted.