We Built a Self Service Portal and People Kept Opening Tickets Anyway

Description and use

The self‑service portal was redesigned based on ticket analysis, improving transparency and reducing reliance on manual requests.

Problem

Despite a self‑service portal, most requests continued to arrive as tickets or direct messages, indicating low adoption and inefficiency.

Outcome

Redesigning the portal around actual ticket data increased transparency, reduced reliance on tickets, and created a clear backlog for future automation, improving overall efficiency.

Workflow

• 1. Analyze six months of ticket data to identify high‑volume requests.
• 2. Map each high‑volume request to a catalogue item or a human‑handled route.
• 3. Redesign portal UI to display progress and expected completion time for each request.
• 4. Implement a fallback path for requests outside the catalogue that routes to a human without stigma.
• 5. Monitor adoption and backlog to prioritize future automation.
• 6. Iterate based on user feedback and ticket volume changes.

Evidence

The portal was good work. A catalogue of the things people asked us for constantly, each one wired to a pipeline that provisioned the thing without a human in the middle.

Source

Six months after launch it was handling about twelve percent of requests.

Source

The portal replied with submitted successfully and nothing else.

Source