Updates

It’s now possible to set granular permissions to specific users for almost every action that can be performed on Fernand. We’ve done this to give our users more flexibility regarding the kind of things they want to allow or not at different levels of collaboration, but also because it allowed us to push for something very exciting.

We believe that customer support should be a company-wide effort, so our new pricing model makes it easier for companies to adopt Fernand across the entire company.

As of now, only active users with permission to reply directly to customers are paying for a seat on Fernand! This means you can invite all your team members, including your entire engineering team, legal, security, marketing, sales, accounting, your office mascot dog or whoever else, so they can access the context of customer support tickets and collaborate internally on issues. They can access conversations, collaborate via notes, assign, tag, snooze, and much more — all of that, for free.

We’ve also fixed several issues and optimized flows.

• Optimized rendering on iPhone for Safari users
• Migrated our Super Replies to ChatGPT API for even better, more accurate generations
• Added a shortcut to compose new emails, as well as a new action in our command palette
• Rules, channels, and smart snippets are now sorted in alphabetical order
• Enhanced the edge case of having an email hard-bounce in a conversation
• 8 other minor bug fixes 🎉

Fernand web application is now responsive and scales for both tablets and smartphone screens. While most users access Fernand on their desktop, it’s common to respond quickly to a conversation on the go. The experience is much better now, but please note that this is an ongoing process, and we’ll improve the experience over time.

We’ve also fixed several issues and optimized flows.

• Added the ability to Compose new emails and start conversations from scratch
• Fixed a problem when an email bounced in a loop in multiple-recipients conversation
• Refreshed our inbox zero imagery for even better “Woah” effects
• Fixed a bug with our Snooze action in smart snippets
• Fixed a bug where invoices wouldn’t load properly in the billing section
• Improved the experience of pasting images in the editor
• Customized the error message when OpenAI’s API is over capacity
• 55 other minor bug fixes 🎉

Today, we’re introducing the most exciting feature so far for Fernand.

We’ve leveraged OpenAI‘s best-in-class API to bring our vision to life. We believe that AI can enhance processes and help automate mundane tasks while still keeping the user in charge of the final output. Hit SHIFT+R or click on the magic sparkles icon of Fernand’s editor. You’ll see an input field with a prompt to give further instructions to the AI, based on the content of your customer’s request. Some example prompts you could use:

• “Decline politely”
• “Interested”
• “Will investigate”
• “Escalated to engineers”

With just a small prompt like the ones above, the AI will be able to provide a fully-fledged, well-written email answer that retains the context of the initial message. This supports also multiple languages, including French, German, Spanish, Russian, Chinese, and many others.

We’ve also fixed several issues along with the release:

• Better dark-mode handling of smart snippets
• Dedicated “Open full view” item for the smart snippets in the command center
• An improved visualization of rules
• 6 minor bug fixes

We pushed a major revamp of the Smart Snippets on production today. The most significant improvement is the introduction of the ability to send data using webhooks. You can use it to send a message, update a customer’s attributes in your CRM, or take just about any action on a customer that you’d want to do.

You can choose between different call types, including GET, POST, PUT, PATCH, and DELETE. You can also now specify up to 4 custom headers with a name/value pair.

We’ve also fixed several issues and optimized flows.

• Ability to set a custom FROM name for automated emails and replies, including the use of variables like organization or agent name
• Fixed an infinite loop in our Websocket implementation
• Pasting an image in the editor now adapts automatically for high-density displays like Retina screens
• Clearly labeled the “Open” status in the conversation sidebar for related tickets
• Better UX of the DNS verification to avoid confusion due to propagation issues
• Show / Hide the events in a conversation using CMD+E
• Increased the max size of attachments
• 18 other minor bug fixes

We’ve been hard at work listening and implementing feedback from our first users. This update brings many experience improvements, new impactful features, and general bug fixes. We’ll list some of the highlights below.

• A completely revamped emoji picker on the editor
• Improved Markdown support on the editor
• Better bounce handling of messages that couldn’t be delivered
• Stripe integration is now fully functional, along with a few key changes to pull up Stripe customer data even when no subscription is active
• “All” now lists all conversations, even those that are closed
• Notification email to alert when a rule is executed successfully for the first time
• Notification email to alert you when your channel DNS settings have been properly verified
• Better UI for team member permissions
• A delightful confetti animation when you reach 100% on the progress bar
• Better UI for selecting multiple agents in the rules and smart snippets creator
• Cross-browser optimizations, especially for Safari on Mac OS Ventura
• Improved dark mode handling of some components
• 23 other minor bug fixes

92% of customer support reps experience anxiety daily. This has been the accepted norm for far too long. Today, we’re releasing Fernand, the calmest support experience ever made. Here’s everything we did to design an intentionally calm product.

We’ve done customer support ourselves for the past 10 years and experienced first-hand what a soul-crushing thing it can be. We set ourselves a rule: each flow should only show one thing at a time. No 5-column layouts, and no side panel with infinite scrolling. We do this to counter Hick’s law, which dictates that the more choices you have to make, the more time you’ll need to make a decision.

We’ve designed with psychological levers in mind to try to influence habit-forming, squash procrastination, and increase focus. For example, we communicate your progress throughout your session so that you feel a sense of completion toward a goal. Calm is predictability.

Visually, we made sure to use UI elements as sparingly as possible. We’ve used soft pastel background colors to bring a sense of depth while still feeling soft and invisible. Copy has also been crafted with calmness in mind: simple, honest, humane, and straight to the point.

Our pricing model is dead simple ($29/user/mo). One plan means no decision paralysis: it’s simple and predictable. Most importantly, it lets us include everything unlimited so you don’t have to worry about hitting a quota or how your bill will look at the end of the month.

We also make sure we’re monitoring our infrastructure and app uptime 24/7. No matter how well you’ve designed your product to be calming, everything can be destroyed in a split second. We know that nothing is more stressful than an outage, and we promise to make everything in our power to avoid incidents from occurring. We know that stuff inside out because we’re running an infrastructure-critical company as well.

We practice what we preach: our company itself is calm. We’re async first, remote first, and on a 4-day workweek schedule.

Watch this space for updates, or follow @AntoineMinoux, @cx42net or @getfernand on Twitter