Ladybug 🐞

What problem we're working on, and why does it matter

When public officers use government tools like Plumber, they sometimes get stuck. Not because of bugs, but because they are unsure how to do something. They submit support tickets, but these often lack enough detail for support teams to help quickly. Our research found that 25–50% of tickets need follow-ups just to understand the problem, and support staff often re-read tickets two to three times before they can respond.

This matters because we want users to get assistance quickly, and we want product teams to spend less time trying to interpret incomplete tickets. When support is slow and frustrating, it negatively affects how users perceive the product. Faster, clearer support should improve the overall user experience.

What we've built over the last month

Add Ladybug to your app with one line of code to collect screenshots, recordings, and rich context from users.

We built a working MVP: a lightweight widget that lets users screen-record their issue (with cursor tracking) and submit it directly to support. Recordings appear on an admin dashboard and trigger an email notification to the product team. The widget is functional but has not yet been deployed on any live product.

Our key pivot this month: we shifted from positioning this as a "bug catcher" to focusing on helping users explain product usage problems, which turned out to be the more common pain point.

A ladybug widget appears on the right side of the page which enables one-click screen shots and privacy-enabled screen recordings.

Users can submit support tickets with screenshots or screen recordings directly from the widget.

We were intentional about not creating yet another Zendesk. Instead, the widget integrates seamlessly with CRMs and workflows that admins already use.

Connect Ladybug to your existing tools like Zendesk or Gitbook so support tickets land where your team already works.

Additional features we added to further furnish support tickets:

Auto-generated draft response from Gitbook MCP

Auto-tagging and analytics

How real people are using it, and what is happening as a result

There is no production usage yet. However, we interviewed four product support teams (Kampung Spirit, Isomer, LetterSG, Bright) and all expressed interest in having this widget on their product. Support leads shared that screen recordings with voice narration would "significantly reduce the need for clarification calls." One team rated the potential impact at 6.5/10, noting it could shorten resolution times.

Next step: deploying on existing platform products to get feedback and iterate.