Improved recommendations and rich support insights
After a short [solo] retreat in Frankfurt, we're stepping on the gas and shipping a bunch of updates to Ferndesk (German beer is amazing btw).
Here are the highlights:
🍱 Improved recommendations - Fern now creates fewer, smarter recommendations
📊 Rich support insights - Get a birds-eye view of what's happening in your support inbox
🌿 Task improvements - Fewer hallucinations, TONS of bug fixes
Customer spotlight: Cap.so
Cap is the open-source alternative to Loom. It's a fast, lightweight, privacy-first screen recorder for teams and creators with one-click recordings, secure sharing, the ability to store your data in your own S3 bucket, and a Pro tier that layers on AI niceties.
In this week's case-study, Richie McIlroy shares how he went from zero docs to 20 articles in minutes with Ferndesk.
Improved recommendations
In the past, Fern would generate recommendations for almost every customer request in your inbox.
For companies receiving a few dozen tickets per month, this format worked really well.
But for companies getting thousands, and even tens of thousands of customer requests per month, Fern would generate hundreds of recommendations for you to review.
This was unworkable. There needed to be a way to surface the changes with the greatest ROI.
So we've updated Fern to prioritize recommendations based on:
volume: how many similar cases
time to resolution: how long did it take to resolve the customer's issue
ease of implementation: how much effort will it take Fern to draft an update
and many other factors
The quality of recommendations is also much higher.
Fern now understands which tickets are relevant to your company VS which tickets are spam, marketing content etc. You will also notice a lot fewer duplicate recommendations.
Recommendation summaries are also terser, and easier to scan.
Try new recommendations now by creating an audit.
Rich Support Insights
In audits, Fern now does much more than tell you common customer requests. Now Fern can show you:
Common requests by category (bug reports, questions, feedback etc)
Common requests by feature/product area (billing, account questions, etc)
Average time to resolution
Escalation rate. etc
Also, by clicking on any of the cards in the audit you can browse customer requests matching the card's criteria with a summary for each request.
For example, you can get a birds-eye view of all bug reports related to CSV imports, and ask Fern to generate a troubleshooting guide for them.
Try it now by creating a new audit! You can also learn more about using analytics to improve your content.
Task improvements
Fern now hallucinates much less during tasks, and finds relevant content more reliably
Summaries in tasks are now less verbose, and formatting has been improved dramatically
Codebase research is now slightly faster, with greater accuracy
Have feedback for tasks? I'd love to hear your thoughts.