Time is more valuable than ever. More than half of product managers agree that “insufficient time” is one of the biggest challenges in their day-to-day operations.
“Product managers believe they should spend 53% of their time on strategic activities, but only 8% can focus at this level.”
Pragmatic Institute
But who will do all the manual, time-consuming work that still needs to get done?
You might’ve guessed the answer: AI.
At Canny, we asked ourselves:
“How can we free up product managers’ time to help them focus on that creative and strategic work?”
This is how we came up with Autopilot. It’s an AI-powered suite of tools that simplifies and automates feedback management. One of Autopilot’s key features is Feedback Discovery. It finds, extracts, and deduplicates feedback from customer conversations.
Now, it can find feedback on ten review sites:
- G2
- Capterra
- App Store
- Google Play
- Trustpilot
- Trustradius
- Product Hunt
- Salesforce AppExchange
- WordPress
- Shopify
I’m excited to walk you through this new feature.
Why automate feedback management?
When you dreamed of being a product manager…we bet you didn’t dream of this:
- Manually copy-pasting feature requests from G2 to a spreadsheet
- Sifting through that spreadsheet with hundreds of lines of customer feedback
- Desperately trying to find patterns and commonalities
- Attempting to find duplicated ideas and merge them
- Building features no one ends up using
There are more issues with managing feedback manually, including:
- Inefficiency. Sorting through large volumes of feedback can be unproductive and overwhelming.
- Human error. It’s easy to miss important feedback and make mistakes.
- Inconsistency. Manual processes can lead to inconsistencies in feedback categorization and management. This creates even more manual work down the line.
- Lack of scalability. As the amount of feedback grows, it becomes increasingly difficult to manage manually.
- Delayed responses. Manually sorting feedback can delay the response time to customer issues and feature requests.
- Lack of comprehensive view. It’s challenging to get a holistic view of all feedback across different platforms.
- Difficulty with tracking trends. It’s harder to identify patterns and trends in feedback.
Let’s see how automated feedback management can solve all these issues.
What is Canny Autopilot?
Canny Autopilot is an AI-powered extension of our core product. Our suite of AI tools includes:
- Feature Discovery – to find, extract, and deduplicate feedback in customer support, sales communications, and now in public reviews
- Smart Replies – to respond to your users and ask for more context
- Comment Summaries – to summarize busy comment threads and give you the main outcomes
Our product has always been feedback-first – we believe it’s the foundation of all good product management. We built Autopilot’s Feedback Discovery to help product managers find, extract, and deduplicate that feedback faster and easier.
We spent months testing, iterating, and perfecting it to ensure it extracts feedback with very high accuracy. To achieve this, we used a multi-layered AI approach. This means we don’t just ask AI: “Is there feedback here?” We keep asking questions and iterating on results until we get the most accurate output possible.
“We don’t just have a single-stage process. We’re not going straight to the API and asking: ‘What is the feedback here?’ or ‘Is there a bug report in this?’ Instead, we have a multi-stage process. We ask one small question at a time and try to get the most accurate response possible. This is how we get higher fidelity and accuracy rates.”
Niall Dickin, engineer at Canny
We’ve received lots of positive feedback so far.
“I LOVE how it auto-scans our support tickets and magically finds customer feedback without me lifting a finger. We’ve been able to 10x the feedback coming into Canny & remove many duplicate posts, with only a few minutes of work a week.”
Keenan Jones, VP of Product, Credit Repair Cloud
Finding feedback in customer reviews
But we didn’t stop there. Today, Autopilot finds feedback not only in your customer communications. Now, it can scan your public reviews and find feedback there.
After Autopilot detects feedback, it extracts and imports it into your Canny portal. Then, it finds duplicate requests and merges them for you.
Note: if you don’t want Autopilot to automatically merge duplicated feature requests you can turn off automated mode. This way, you’ll see what Autopilot recommends to merge, but you’ll be able to make those decisions yourself.
“I thought, surely I can’t just turn it on, and it’ll do its magic. But that’s exactly what it’s doing. We’re seeing hundreds of support tickets turned into actionable insights…with very high accuracy.”
Matt Cromwell, senior director of customer experience at StellarWP
Why public review sites?
Reviews often contain feature requests that slip by unnoticed. That’s why we connected Autopilot to ten public review sites (and we’ll keep adding more). Now, you can capture all the feedback right where it lives. You no longer need to:
- Manually scroll through and read every review
- Look for feature requests and feedback in those reviews
- Copy-paste each idea to your feedback portal
- Compare these ideas to your existing feature requests
- Identify and merge duplicates one by one
Autopilot now does all that automatically.
The truth is, your feedback is out there. It lives in your App Store, Product Hunt, and Trustpilot reviews. And right now, you’re relying on humans to find and pass it on.
It’s easy to let that feedback slip through the cracks. When it does, you don’t have the full picture of your customer sentiment. And you can’t make objective data-based decisions for your product.
Autopilot automatically captures your feedback from anywhere it lives. There’s no limit to how many sources you can connect to Autopilot either.
“Canny’s Autopilot ensures feature requests never fall through the cracks. We’ve seen an 80% increase in requests logged since introducing Autopilot.”
Owen Doherty, COO at OrcaScan
Getting started with AI
We get it – trusting AI might be daunting. You might not want to use customer chats, sales call recordings, and other private data just to test an AI tool.
Instead of testing Autopilot with your customer communications, try it with public reviews first. You probably don’t need to get your team’s approval for testing Autopilot with public reviews.
Use your free AI credits – every account has them, even our free plan! Once you see the quality of feedback detection and extraction, you can decide whether or not you want to connect more sources.
We did this specifically for those who are nervous about fully trusting AI right away.
How does it work?
Autopilot’s Feedback Discovery can find feedback in minutes. All you need to do is connect a review site and watch it happen. Here’s a quick step-by-step tutorial.
Once you connect a review site, Autopilot will scrape it to identify possible feedback. Then you can either:
- Use Autopilot in automatic mode to let it merge duplicated posts (you’ll still see an audit log and will be able to adjust what Autopilot did)
- Review Autopilot’s suggestions and decide what to do with each item
Conclusion: save time and focus on strategic tasks with Autopilot
We know that manually collecting user feedback might not be your idea of strategic work. We get how much time feedback management takes. And we want to make this simpler for you.
Autopilot’s Feedback Discovery can free up your time for strategic tasks. Try it today with public reviews and see how much feedback is out there.
