Fin Product Area

Hello!This quarter we’re working on moving to a more proactive support approach, an element of this is building out our helpcentre/FAQs and utilising the AI bot.However I recognise the importance of studying our support queries to really understand what people are asking, learn what’s missing from the help centre etc.But how do people tend to do this?Does intercom have any functionality which reports on questions people are asking which the bot can’t answer? Or do people download conversation data and manually read through conversations?I’m uncertain atm the best approach within intercom.

My org would like to explore getting Fin in front of calls/emails. What resources should I consume before surfacing a request for consultation on our specific use case? Is it trustworthy? What are the hallucination implications? 

Is it possible to create an article from a document uploaded to our Knowledge Base? If so, could you please explain how to perform this action?"

When trying to create a new linked customer conversation through a back-office ticket, the new conversation will open up and we can change the subject and content of the message as usual, but when we hit Send, the message disappears, the subject line changes back to the default subject, and nothing actually gets sent to our customer. I have tried this across multiple tickets and customers. This issue only began this morning.

Hi guys,I know that conditional attributes for conversations are available, but what about tickets? can i set conditional attributes for them?  Thank you

Hi I’m getting unsafe-eval CSP error from script-src policy.Can this be fixed somehow without adding unsafe-eval to the policy?Do you have any other script version without eval inside your JS code? Attaching screenshots.  Regards

Hi!I want to design a system where any employee can ask Fin questions and get answers based on our internal resources and customers can ask questions and get answers based on what’s in our help center. One idea was creating two help centers (one for employees and one for users) but looks like Fin can only be trained on one help center. Anyone have success with creating a system that uses Fin trained on two different knowledge bases to answer questions? Or other general thoughts about how to go about this? We use confluence for our internal resources and have a help center that we directly edit for customer questions. Also don’t want to add every single employee as an intercom user 

From time to time, agents cannot log-in and see different errors:Developer login page instead of regular one Or error “You’re running old version of the Intercom web app...” (hard clear of cache and cookies works from 3-4 attempt)Does anyone else encounters such behavior, or are we the only lucky one? :) What helped you with that?For the context, it happens with several people at once, different devices, different IP, location. The browser is updated, no internet connection issues. We’re using EU Data Hosting, btw

Hi Intercom community, I am wondering if there are keyboard shortcuts for the Help Center formatting features like the note callout, embed videos, dividers, bulleted lists, and so on. The keyboard shortcuts for standard formatting features (Ctrl + B on Windows for bolding text) seem to work. However, I am not seeing any evidence that these shortcut keys exist for the Intercom-specific features.If it has not been implemented, hotkeys would be a really nifty tool that would improve quality of life and productivity!

Hi, I am wondering how users that have multiple products manage their Help Center when using the Related Articles feature.Because of the way the AI works, the product name has to be referenced in the title of the article to have that article pull up in the Related Articles table.For example, the breadcrumbs will look like this:[Product Category] > [Product Name] > User Guide > [Product Name Feature X]How are others titling their articles to avoid redundancy? I would like to avoid referring the product name in the article title, since the way the collections are configured in our environment, it’s already implied the article is related to the product. Thank you!

We want to configure the layout of Apps in the Details panel for team inboxes. At the moment each team member can update their personal view, but we want to set up a standard layout for our team. I haven’t found a solution yet in the documentation. Is this possible at the moment or will this need to be a wishlist item?

Please someone help me out about this community

loginIdentifiedUser / loginUnidentifiedUser works properly on app launch. The issue is when you carry out these actions in sequence -  loginIdentifiedUser, logOut and loginUnidentifiedUser, I get this UI error with console error errorCode: 0, errorMessage: The operation couldn’t be completed. (Network request succeeded but the response was ignored error 0.)  

Hi everyone, I’m a WFM employee assigned to a new client who’s using Intercom, and it’s my first time to use it. I am looking into building a payroll report, currently they pay the employees just based on how long they are logged in which I can easily find the report in intercom (Adherence Report), however, they wanted to change it and reference it with their schedules. The only way to do it is to use the Login/Logout report of each employee which shows what exact time an employee logged in/out. I can’t seem to find the report in intercom. Is there anyone here who happen to find that report and if you can share where in intercom can I get it?

We’re using react-native and the SDK version 8.5.0 on iOS 15+When our app boots, the user is logged out, so we run `Intercom.loginUnidentifiedUser()`The user then clicks Chat and we present intercom via `Intercom.present();`We then have a workflow/setting enabled for intercom to ask for the users email address (they’re not logged in)But when the input appears, you cannot click on it, if you do, it just toggles the timestamp on the input.  No matter how often we try click it, it doesn't work, we’ve tried on simulator, on device - multiple devices of diferent models and non of them work. Please advise.

I'm currently trying to set up an outbound template that includes the following dynamic data from every colleague of mine:(full) NameJob titlePicture Last year I talked to the Intercom support about this, but it was too much work back then. Now I want to try it again. The last tip about it I got from support was to use something like this:{{message.author.first_name}} from {{app.name}} and also use the variable for fetching the pictures: <img src="https://www.example.com/avatars/{{ message.author.first_name }}.png" height="40" width="40" class="avatar" alt="intercomavatar"> My first tries today weren't really successful, any other tips for me?

We can choose which attributes are shown (or hidden) but how can we re-order the list? For example, we might want the first 3 items to be:  phoneconnection countjob title but display a lot more than this below. At the moment we need to scroll to find these higher value fields in a longer list. We'd rather see them at the top.  

I'm trying to get to know about you intercom. Please be nice to me. Thanks 

We use Docusaurus for all our documentation and would like to centralize onboarding by embedding Intercom checklists and product tours directly into our docs.Our engineering team also built a custom search bot that scans our Docusaurus portal, but its button overlaps with the Intercom Messenger since both use similar placement and styling. To resolve this, we’re exploring two options: Embed Intercom content (like checklists and product tours) directly in Docusaurus pages, without relying on the Messenger or Help Center. Integrate our custom search bot into the Intercom Messenger via API, so we can unify both experiences in a single UI element. Are either of these approaches supported? Or is there another recommended way to host or display Intercom-created content outside of the default Help Center and Messenger?Thanks!

Hey team! Is there a way to somehow use multiple trigger words within branch for OR conditions? My goal is to build a custom action (data connector nowadays?) that would trigger once one of the keywords is used by the prospect, however that would be over 100 items. Is there any way to achieve this result without having to type in the word for every trigger word despite what article from help centre suggests?  

I would love to have a block in Workflows where I can define a set of topics like:Cancelation Product Idea Feedback Return  Other (always exists) …. Each one opens a new branch in the workflow. Which path the user is channeled through is decided by an AI content analysis with Fin. If Fin is not sure, it always uses “Other” as fallback branch. This would really help qualifying the inbound message, and start follow up flows, route to Fin, route to Team etc,

We use chat with Fin AI and we are glad of how AI answers. Also, we have an technical Documentation page (API docs). Users often asks support about code examples. Fin AI answers perfectly. But chat window is too small for this purpose. Would be perfect to let us open chat window in a full-screen mode to let user to ask AI and to view answers in a handy view.Code example:window.intercomSettings = {  app_id: 'abc123',  alignment: 'full-screen'};The view example: 

I’m trying to determine how to define success with outbound messages. Are there guidelines for the minimum open/click rate required for success or the percentage of the goal that should be reached? I’d also like help with goal setting - is there a recommended period in which the goal should be reached? For example, if the goal isn’t reached within 3 days, then is it best to try something else?

Hi, has anyone implemented Intercom in a FlutterFlow app (Web, Android, and iOS)? I'm facing CORS issues. And I'd like to know how to embed the chat instead of using the Intercom activator.

We are just getting started on FinAI. When nobody is marked as available, but a user selects “talk to a person”, we’d like to train it to tell the user that nobody is here right now and we’ll get back to them soon.  I have given FinAI guidance in the Communication Guidance and Handover & Escalation sections, but it doesn’t give the answer I trained it to give, and our users are left hanging with the expectation that a live agent will be with them soon.  How do I train it to give specific information to a user to let them know we aren’t available right now and to leave their info?  Or is there a way to automatically turn off “talk to a person” if nobody is marked available?