Share product ideas and upvotes with our product team
Hey team! This is something quite straightforward, but I thought I’d share it anyway: I’m building multilingual workflows based on the user's browser language and currently have to make duplicates for other language versions since I can only start applying conditions after the initial message (Path A) I would ideally prefer to have different language versions within one workflow as it allows better language/version control.
Assign X random conversations per teammate for review (including phone calls and Fin not involved)I need some kind of tool that assigns an specific amount of conversations from my teammates for my review on a weekly basis. Some of these conversations are phone calls, some of them do not have 2 messages from Fin nor 2 responses from the customer.Monitors is a great tool, but it only works for conversations with at least 2 messages from Fin and 2 messages from the customer, and I need a sample of all conversations, not only those with the previous criteria.
Today, Workflows can’t be started just because a user/ticket attribute changes. It would be really useful to have a trigger that fires a Workflow when a specific attribute is updated (for example, “plan”, “status”, “region”, or a custom attribute). This would let us automate follow-ups, routing, or internal actions based on attribute changes, without needing a customer message or other event.
When discovering issues with Product Tours, it tells you which step they failed on, e.g. ‘Step 7’ but the steps in teh builder aren’t labelled as such. It would make it so much easier if they were when making edits, additions, and finding errors, etc.
When Fin is deployed over E-mail there is a DMARC security (setting) before Fin will answer to the e-mail. In my case; like 20-30% of the incomming e-mails aren’t getting answered because of this problem: "Fin did not try to answer because the sender could not be authenticated"Fin escalates the conversation immediately, even when Fin could answer the e-mail so easily. Right now, there aren’t any options to troubleshoot this problem or any triggers to solve it automatically. So my question is: Can you guys find a way to automate the process what fixes the error and so can Fin the e-mail that was sent in the first place?
Edit: seems like despite it saying it’s making an update, it does ask for confirmation once it’s finished the proposed update. The thought process as it loads give the impression it’s going to make the update without confirmation,Just like with Claude, the ask feature is too eager to make changes to things. I want guardrails in place that it always asks before updating procedures, guidance, etc rather than just updating itself and having to click “stop” before its too late.
Sometimes, we decide to convert an email conversation into a ticket in the middle of an ongoing conversation. It feels odd to the customer when they see an email notification that a ticket is created under a submitted status, as they have already been in touch with us. I wish we could select the ticket status while converting it
Is anyone else noticing that the Intercom Phone, when customers call, when it's connecting it sounds like the UK Dial Tone (https://www.youtube.com/watch?v=0aJYmQp4Q7U) and not a US dial tone (https://www.youtube.com/watch?v=j2Hck4Bmifw)I understand Intercom is compatible in other countries but how can we establish a clear US based ring?
When showing users (e.g. in the conversation list and in the "User" column), when there are multiple users, it should prefer to show the customer's user, not users from our own domain (which appear e.g. when replying or forwarding via email), so that one can see who the customer is. Right now we have the problem that when one of our team forwards an email to Intercom or replies via email, they show up as colleague@example.com, and in the conversation list, often that then shows up as the primary users. This means we cannot easily see in the conversation list who the actual customer is; instead we see primarily our own colleague. It would be great that if Intercom could solve it like this: In the conversation’s participants list, sort users that are not from our own email domain _before_ users that are from our email domain.For example, if our email domain is @example.com, and the participants list is [me@example.com, colleague@example.com, customer@company.com], it should sort it as [customer@company.com, colleague@example.com, me@example.com], so that customer@company.com is the first one that determines what’s shown in the conversations list (“customer@company.com and 2 others” is a lot more useful than “me@example.com”).
At the moment we’re unable to add Albania as a supported language within the help centre.Adding it would be much appreciated. Thanks!
Users should be able to automatically merge open conversations. Turning on ‘Prevent visitors from having multiple inbound conversations; within the messenger configuration is not enough as customers may open a conversation via email, or phone. This should be a setting to opt in or out.
Feature request: Load Messenger before cookie consent with contextual cookie promptCurrent behavior: The Intercom Messenger is currently only loaded after a visitor has given cookie consent via a cookie banner (e.g., OneTrust, Cookiebot, UserCentrics).If a visitor declines or dismisses the cookie banner, the Messenger does not load at all.The problem: This creates a poor user experience:Visitors who decline cookies have no way to contact support or ask questions There is no visible support channel on the page Users may not realize that accepting cookies would unlock the support optionFrom a business perspective, this means:Lost support inquiries Reduced engagement No alternative communication path for privacy-conscious usersProposed solution: Allow the Messenger to load in a "cookie-pending" mode before consent is given.In this mode:The Messenger widget is visible and can be opened When a user tries to interact (e.g., start a chat), the Messenger displays a contextual message explaining that cookie consent is required A button is shown that re-opens the cookie consent dialogThis approach:Maintains visibility of the support channel Educates users on why consent is needed Provides a clear path to grant consent in context Complies with GDPR (no cookies set before consent)Technical note: The Messenger could load in a "consent-required" state and trigger the cookie banner via standard IDs or API calls, for example:OneTrust: <button id="ot-sdk-btn">Manage cookie settings</button> UserCentrics: trigger via their JavaScript API or custom elementThis would work seamlessly with common consent management platforms.How this would support our workflow:Visitors always see a support option, regardless of their initial cookie decision We can offer a consent-in-context flow that's clearer and more user-friendly Support remains accessible for privacy-conscious users who may reconsider after understanding the value Reduces friction in the customer journeyReferences: This pattern is already used by some modern support tools and is considered a best practice for GDPR-compliant UX.
When website content is synced, it updates existing articles and adds new articles but does not automatically detect articles that no longer exist and remove them. This should happen during the website sync to avoid old/invalid content from staying active within the Interocm’s content side of things. During the sync, it should also remove articles that no longer exist.
It would sure be helpful to have the “drag to resize” functionality for images in Articles. While you can manually change the width and height, or enter in your desired values, there is no “lock” between the values to help it scale with a proper ratio, leading to some wonky images. Here is the current option: Only have the option to manually adjust the width and height with values...not dragging the image. Best use case: I add an image to an article, and then can drag/pull the corners to scale the image up or down to fit the content. This prevents accidentally breaking the image or making it wonky. We’re not all designers who know the ins and outs of image scaling and sizing rules, and the usability here would be great to allow documentation teams to self serve without bugging the design team for visual issues. Thanks for considering!
The end of the static Knowledge Base is here Currently, Fin is the best in the business at retrieving answers. But even the best librarian is limited by the books on the shelf. To win the AI race, Fin needs to stop just reading the library and start building it. 🏗️ The Vision: The "Self-Healing" Support Ecosystem 🧠 Right now, the Knowledge Base (KB) is a bottleneck. It’s manually written, often out of date, and disconnected from the "grit" of real-time customer struggles. The Future: Fin shouldn't just be an "Answer Engine"—it should be a "Knowledge Architect." 🏛️ How it Works (The Unlocked Power): Real-Time Knowledge Mining 💎: Fin should analyze every live conversation. When it identifies a successful resolution to a problem not in the KB, it should flag it instantly. Autonomous Article Drafting ✍️: Instead of a Support Manager spending hours writing docs, Fin should "draft" a new KB article based on the transcript of a perfectly handled chat or update existing. Semantic Refresh 🔄: If customers start using new slang or technical terms for an old problem, Fin should update the tone and keywords of existing articles in real-time to match the actual language of the user. Gap Identification 🔍: Fin should proactively tell us: "I’ve seen 15 people ask about [X] in the last hour; we have zero documentation on this. Should I create a draft?" Why This is the Intercom Moat 🏰 Velocity Beats All ⚡: Competitors are still asking humans to manually update "Help Centers." If Intercom can offer a Self-Optimizing Knowledge Base, it will reduce the "Human-in-the-loop" requirement to a simple "Approve" button. Accuracy at Scale 📈: By using Conversations as the Source of Truth, Intercom should ensure that Fin is never hallucinating based on 6-month-old data. It learns from what worked five minutes ago. The Growth Play 🚀: This transforms Intercom from a "support tool" into a Product Intelligence Engine. Intercom not just resolving tickets it’s mapping the customer’s pain in real-time. The "Mic Drop" Moment 🎤💥 By turning Fin into a Knowledge Architect, Intercom creates a platform that literally gets smarter with every single message sent. Let’s stop answering questions and start building the future of collective intelligence. ✅BTW: I’m aware Intercom has Content Gap Recommendations, but currently, that feature is a patch for the bot. It identifies where Fin failed so you can feed it a snippet to stop the escalation. My vision for a Knowledge Architect is different: it’s not about 'fixing Fin'; it’s about using the 'grit' of live conversations to automatically evolve the Help Center articles themselves. Intercom just trains an AI Agent; It should be using AI ensure that public knowledge is a real-time reflection of the product, written in the customer’s actual language. Stop Hiding Knowledge Gaps in the "Reports" Basement. 🏢📉Currently, most AI "insights" or "content gaps" are buried in an Analytics dashboard. A true Knowledge Architect lives when and where the work happens.Cheers,Roy 🇬🇧
I would like to have the possibility to include additional Attributes into a Macro Text.We would like to use a Macro to sum up all ticket information in order to hand it over to our DEVs in Jira. But therefore I would appreciate, if I could fill in additional attributes - other that People, App, Event, and Company - into a Macro.Also Fin and Conversation attributes would be needed.I hope, that this is something that could be done.
Hi Intercom Team 👋🇬🇧, ℹ️ Feature Request: ✨ AI-Powered Semantic Prioritization & Smart Sorting 🤯 The Problem: Chronological BottlenecksCurrently, conversation sorting is limited to chronological or rigid SLA-based metrics (e.g., "Last activity" or "Next SLA"). In a high-volume environment with 30-40 open chats, this "dumb" sorting treats a simple "Thank you" with the same urgency as a "System Down" emergency. Agents waste significant mental energy manually triaging lists to find critical issues, leading to burnout and delayed responses for high-value customers. ✍️ The Solution: Semantic AI SortingWe propose adding a ✨ AI Priority sorting option to the conversation management dropdown. This feature would utilize a "Middleware Triage" to analyze the intent and sentiment of every incoming message in real-time. 😇 Dynamic Triage: The AI automatically identifies and bubbles up frustrated users or high-intent leads (e.g., "Payment failed" vs. "Just checking in").✅ Seamless Integration: A new "AI Recommended" sort option should be placed at the top of the existing sorting menu. 🍳 Low-Level Effort: This can be implemented using existing LLM APIs (like Gemini) to return simple priority tags (e.g., priority: high) without restructuring the entire database. 👔 Why This Matters for the Product Efficiency: Agents move from "First-In-First-Out" to "Value-First" processing, ensuring the most impactful tasks get immediate attention. Customer Retention: High-urgency tickets are resolved faster, directly reducing churn risk for frustrated users. 🏦 Scalability: As chat volume spikes, AI sorting acts as an automated supervisor, preventing critical messages from being buried in the "Documentation Black Hole. Cheers,Roy 🇬🇧
When a conversation shows “seen by” right now it does not detail which participant in the conversation has opened the email. Ideally, there would be the ability to show the list of all participants to display whether each person has opened the conversation. Sometimes we are attaching internal company personal emails to conversations to ensure that our team knows we have completed a reach out. But if the internal employee opens the email, we’re unable to see if the customer actually opened the email.
It would be good to show the 'X' (top-right) on the post preview/builder before setting it live. Had to test one in prod with me being the only audience member to ensure the user has a way to exit the post without the CTA link button being the only way out.
IT would be nice to be able to set a message as unread after reading, if you don’t have time to answer right away.
Request: give the option to change the font colour for the dismiss X button on banners etcUse case:One of our customers has been in touch to flag up that the dismiss X button on our banners disappears when she switches on accessibility modes in Windows 11. The customer is using the ‘Desert’ visual theme to improve contrast on her system, as she is visually impaired.As you can see from the screenshot, the X button on our banner is no longer visible. Giving us the option to change the button to a black font or similar may fix this.
Many conversations are a shared responsibility between team members as we try to resolve an issue. In such cases, we use “assign to” after writing a message, so that team mate sees the message in their inbox. It would save us a lot of keystrokes if there was a quick action to “assign conversation back to previous assignee”, or if that was at least possible via a macro.
When inserting images into Intercom articles, there is currently no option to apply a drop shadow effect. A shadow helps images stand out from the white article background, giving them a more polished, professional look, especially useful for screenshots, product UI captures, and diagrams that would otherwise blend into the page.We'd love to see a simple shadow toggle or style option in the toolbar (similar to how you can align or resize an image).Expected benefit: This would improve the visual quality of help articles and make it easier for customers to distinguish UI screenshots from the surrounding content, reducing confusion and improving readability.
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.