Share product ideas and upvotes with our product team
Currently, when setting up custom Fin identities for multiple brands, it's not clear in the user interface that these identities only work in Messenger and not in email channels. This leads to confusion when users expect their brand-specific Fin avatars and names to appear in email interactions, but they don't.Suggested improvement:Add clear indicators in the Fin Identity setup interface that specify:"Custom identities apply to Messenger only" "Email support planned for future release" Visual badges or tooltips explaining channel limitationsThis would prevent users from having to discover these limitations through trial and error, improving the overall user experience and reducing support inquirie
Currently, users cannot modify Fin's identity settings in [TEST] workspaces, but this isn't indicated anywhere in the interface. This creates confusion when users attempt to change these settings.Proposed solution:Add a clear notification in Fin identity settings indicating this feature isn't available for test workspaces Show a tooltip or banner explaining why this limitation exists Direct users to the production environment for configuring these settingsBenefits:Prevents user confusion Saves time through clear communication Improves user experience
To be able to move a guidance to another guidance category without deleting – something like dragging it to another guidance category.
Would it be possible to add some settings when creating a new snippet via “Inbox > Improve this answer” more like what folder it should be and what tags it can have?
At the top of the new conversation a summary generated by AI that says “this client has recently messaged us about XYZ, and also reported bugs ABC in the last 2 weeks” Giving more context on customer’s past interactions so that we reduce being repetitive
We would like to:Allow customization of time interval and frequency/amount of follow-ups Fin can send Allow customization of time to automatic close if client does not respond Why are we requesting this? (real scenario below)Fin followed up after 1 minute without a response Fin then followed up again, 4 minutes after the first follow-up The way Fin followed up (phrasing) was quite intense 10 minutes after the second follow-up, Fin closed the conversation without letting the user answer
Allow automatic ticket/conversation merging if two or more tickets/conversations from one user (the same email address) are currently open.
Please support Hebrew in Fin Voice!So many of us would love to give it a go!
We are using the Fin Multilanguage Beta and I would like to provide some important feedback on the support of German language. The problem is German has two ways to address people, a formal version and an informal version. The formal version is used in B2B and more traditional interactions for example.Thats why in most language selectors for SaaS tools you can select between German (formal) and German (Informal) to control how to address a customer. Fin Multilingual does not do this differentiation at the moment and is using the formal version only. This though is nearly inappropriate or at least "super weird" form of address for our customer group and makes Fin nearly not useable. So I would like to provide this as Beta Feedback to the Product team to fix it. What would be needed is a setting to decide if Fin should answer in formal or informal German. Another work around would be the possibility to provide a custom base prompt as part of the Fin profile as other AI chatbots allow to do. This would allow me to type in something like: Always remember your customers are students and you should always use informal German to address them. You can use a relaxed communication style and also do some jokes. This is what really works in ChatGPT or other bots based on chatgpt to set a tone of the AI profile and also to deliver important communication info as the use of informal German. As chatGPT is a language model it can easily handle the communication correctly after this input.
Problem / Current limitationWhen using AI Inbox writing features — such as Expand, Rephrase, My tone of voice, More friendly, More formal, More concise, and Fix grammar & spelling — the output is consistently rewritten in US English.This creates friction for teams operating in Australia, as the AI overrides our original spelling, phrasing, and tone even when the input is already written in Australian English.Examples include: “organisation” → “organization” “licence” → “license” “favour” → “favor” US-centric phrasing that feels slightly off for Australian customers This means agents must manually re-edit AI-generated responses, reducing the efficiency gains these tools are meant to provide.Desired behaviourAllow AI Inbox writing tools to respect and generate Australian English, including: Australian spelling conventions Australian phrasing and tone Consistency with how teams naturally write to customers in Australia Ideally, this could be enabled via: A workspace-level language setting, or A per-user preference, or Automatic detection based on workspace locale / help centre language Why this matters Customer trust & brand consistency: Customers notice spelling and tone mismatches, especially in written support. Operational efficiency: Agents currently lose time correcting AI output back to AU English. Adoption of AI features: Teams are less likely to rely on AI writing tools if they introduce extra editing work. Global scalability: Supporting AU English sets a strong precedent for other regional variants (UK, NZ, etc.). Who this impacts Support and operations teams based in Australia Businesses with Australian customers and brand guidelines Any team aiming for consistent, localised customer communication
We would like an enhancement to Fin to be able to understand when a customer responds to the CSAT rating with the corresponding Emoji. Often when customers are responding to a survey request for CSAT they give us the Emoji instead of the numerical value and Fin does not record the CSAT response and the team member misses out on the CSAT.
I would like to suggest adding Haitian Creole (Kreyòl Ayisyen) as an official language that Fin AI can understand and respond in.Haiti has millions of native speakers, and a large part of customer support interactions in the Caribbean happen in Haitian Creole. For businesses like mine, being able to serve our customers in their own language is extremely important, especially when dealing with support, instructions, refunds, account information, or any customer-care situation.Right now, Fin works perfectly in English and French, but Haitian Creole is not yet supported. Adding this language would allow thousands of companies to offer fully accessible, inclusive, and natural support experiences to our audiences.Haitian Creole is simple, structured, and one of the fastest-growing languages online. Supporting it would open Intercom’s ecosystem to a whole new market.If Fin AI could understand and speak Haitian Creole natively, it would be a game-changing upgrade for customer support in Haiti and the diaspora.Thank you for considering this important addition
Informal German is not consistently being applied. Sometimes the text is formal, even though “German” is selected and not “Formal German” Presumably, it is because a polite phrase was used in the original text.
idea to add human approved check mark visible for customer in answers which were checked by person.in this case customer will know that answer is checked and can be taken spuriously
Hey everyone,I started working with FIN a few weeks ago and I noticed FIN doens’t reply to as many mails as I wished.With the help of your amazing Customer Support (he knows) I managed to improve the response rate, but I got an idea while replying to mails/conversations/tickets that FIN didn’t pick up.Since the reply I gave to customers was the right answer (ofcourse right?) I missed a button where I could instantly make a Snippet of the conversation.It would be awesome if there was button, or possibility, to instantly add the question and answer as a Snippet, in stead of me making copy pasting it to a Snippet.Glad to hear how I’ve thought of a major breakthrough in efficiency 🤝🏼Kind regardsMichael
Hi, we would like to suggest that the Privacy Policy Notice remains visible and static in every conversation (Or be able to choose that it’s always there), including when returning to the messenger in the same browser or tab. Currently, it appears that once a user dismisses the notice, it does not reappear when they return later. As it stands today, you can close the Privacy Policy notice, but it does not reappear the next time you access the messenger in the same browser. At least, that is the impression we have. It only seems to return after clearing browser data.
Hi, we would like to request a preview or test window for Copilot, similar to what is available for Fin AI. This functionality would make it feel safer and more convenient to experiment and test Copilot without affecting live conversations. Currently, many users hesitate to open an existing customer interaction just to test Copilot, which limits adoption and confidence in the tool. A dedicated preview environment would encourage learning and exploration in a controlled setting.
Hi, another suggestion we would like to add: Access to Copilot Conversation History Linked to Customer's conversation (For backoffice tickets)We would like to request a feature that allows us to view the conversation history between the customer advisor and Copilot, directly linked to the customer's conversation. In our workflow, we actively handle back-office tickets related to errors detected in sources and the way Fin AI responds. We rely heavily on conversation history attached to tickets to correct mistakes and improve content, as this helps us avoid misunderstandings. Currently, we have a back-office ticket type called Report on Fin AI, which is an effective way for us to improve our content and processes. We also created a ticket type called Report on Copilot, but so far, this does not function in the same way. Having Copilot conversation history linked would make this process much more efficient and help us maintain accuracy. I recall that last year there was a “Report” button in Copilot to flag mistakes to the back office, if my memory serves me correctly. It would be very helpful to have a similar feature again. Our internal routines are extremely strict and must remain accurate at all times.
It would be great to be able to create an escalation rule that will work only for one workflow. Or maybe to create an escalation rule based on the teammates acitivity - if active then escalte to team, if not active - let fin answer? What I want to do is to collect Fin attributes in every single conversation but I don't want Fin to handle all chats.
I would love if we were able to help “train” Fin beyond content/guidance by giving a rating to past conversations that it’s been involved in. In other words, give Fin a way to learn off its previous answers, reinforcing answers we really like and highlighting/giving pointers on responses that weren’t exactly what we were hoping for. I see this as an iteration/update to the existing “Improve answer” button that appears on Fin conversations, prompting you to create new content. Instead, it could be a simple thumbs up or thumbs down, weighing the answer accordingly for future similar inquiries.
Issues:Once you add a content to a missing one and create either a snippet or article from it- you can never go back or see an audit trail of what you have actually done. There is no list anywhere where you can see which content gaps you have actually optimized. We would like to store them somewhere which them might be helpful for configuring another area. Once you add the content and hit save, it opens the next suggestion instead of taking you back to the Optimize page, this is quite confusing and unrelated from what you were doing. No export buttons anywhere There is a table where they give percentages of conversations, but you can not click open any of these to see where they are stemming from- again, also no export option located.
What the title says. Saw this post https://www.chatbase.co/changelog/agent-page Would love to have something similar from intercom without bolting on an additional piece application onto our site.
Already have an account? Login
No account yet? Create an account
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.