Share product ideas and upvotes with our product team
I am requesting a date only attribute that can be used in workflows. We onboard new customers and have them fill out a date that they plan on transitioning into our product. The time zone is irrelevant for what we need. Currently, we have to use the text attribute to capture the date and then a human transfers that date into a date/time attribute. I know that there are outbound messages we can use with the date/time attribute but that is not ideal from a customer experience perspective. This will also be beneficial to use with Procedures.
Guidance sections support version history and save-time notes, but Escalation Guidance doesn't. Would be great to have parity — these features are useful for tracking changes and context.
It would be extremely helpful if AI Compose could be configured to follow company-specific branding, terminology, and capitalization guidelines.Many teams rely on AI Compose to rephrase and improve responses, but the generated content can sometimes modify approved brand names, product terminology, capitalization, or formatting standards. This requires additional manual review and correction before responses can be sent.A feature that allows teams to provide custom branding rules, terminology dictionaries, and capitalization guidelines would help ensure that AI-generated content remains aligned with internal standards while still benefiting from AI-assisted writing.Benefits:Improved consistency across customer communications Reduced manual editing and review time Better adherence to company branding standards Increased confidence when using AI Compose at scaleIt would be great if workspace admins could define approved terms, capitalization rules, and brand-specific language that AI Compose automatically preserves when generating or rephrasing content.
Fin/Intercom encourages its clients to make these elaborate workflows, but I see a huge missed opportunity whenever I’m branching off and allowing Fin to handle.Currently, each “Let Fin handle” instance in a workflow can be customized with a select few traits. You can enable Sales agent vs. Service agent, for instance, or you can customize how long a conversation stays open before Fin closes.However, one thing users can’t customize is the content that Fin serves up during a specific path.The idea: Allow users to customize each “Let Fin handle” instance with specific Collections or Help Centers. If a user chooses to go down the subscription management path, limit content to just the subscription management Collection. If a user goes down the “Help with Product A” path, limit content to just Product A.Why it matters: The functionality would dramatically improve our ability to personalize the Fin experience based on the user’s intent. Right now, Fin throws all content at the user, regardless of which path they take in a workflow. It makes open-ended workflows seem pointless -- each “Let Fin handle” results in exactly the same experience. I’m positive that this change would help us (and other clients) improve resolution rates and derive more value from Fin.
I’d love to be able to understand our customers sentiment based on keywords in their body of their message to us. For example if we update our product being able to do a keyword search to understand how many customers are asking us about this and their sentiment would be a great tool for sharing insight across the wider business. Appreciate we could tag but this is very manual and relies on keeping tagging data clean and maintained
Allow selecting languages you speak in https://app.intercom.com/a/apps/X/settings/personal/multilingual so they do not get auto translated
The IssueFin automatically sends the "Did that help?" prompt after every response. When customers say "Yes," the conversation closes immediately. This breaks workflows where Fin provides an explanation but still needs to collect data or perform a handover.Example: Customer: "I want to change my email." Fin: Explains the requirements and asks: "Please provide your new email. Did that help?" Customer: Clicks "Yes" (out of politeness). Result: Fin closes the ticket before the customer can provide the email. This leads to negative CSAT and duplicate tickets. Requested Improvements Toggle Auto-Reply: Ability to disable the "Did that help?" prompt for specific Procedures or Intents. Logic-Aware Closing: Fin should recognize if he just asked the user for information; in these cases, the success prompt should be suppressed. Custom Follow-ups: Allow the "Yes" button to trigger a specific next step in some cases (not for all conversations) instead of an immediate close. Business ImpactReducing "premature closures" will significantly improve CSAT and prevent customers from feeling "shut out" by the bot, while maintaining the efficiency of Fin’s pre-qualifying explanations. Previous Workaround (Inefficient)To prevent this, we previously marked these intents as direct handovers. However, this meant agents had to handle the basic explanations themselves. We want to keep Fin’s efficiency in explaining the process (which works in 90% of cases) without him prematurely "killing" the conversation via the success prompt.
It would be very helpful to be able to manage the snippets via the intercom API for a variety of maintenance tasks and for changes in operations to review and purge outdated snippets as well or to transform that content into guidance or procedures etc where needed. List snippets using the Developer API
Addition of domain field for companies with automatic email domain identification.When someone writes in from a domain that is clearly associated with a company, it would be helpful if the system automatically detected this. This would not only be helpful for reporting purposes but also for workflows where certain content is specific to a segmented audience. As an example, if Big Health’s domains are sleepio.com, trydaylight.com and bighealth.com, allow the team to specify those associated domains in company details and when someone writes in from teammate.name@sleepio.com, automatically assign the company “Big Health”. This would allow Fin to point to guidance specific to the “Big Health” audience in knowledge base and in workflows
Currently, it appears that the Recommendations tool within Fin > Analyze only considers Snippets and public articles (i.e. public synced webpages), but not internal articles. We hold many of our Fin resources as synced Guru pages, which are considered internal articles, so the recommended snippets are not able to consider what already exists there. The Recommendations tool would be so much more useful if it considered all of what Fin can actually use when constructing answers.Here’s what Fin shared with me when I asked why internal articles weren’t showing - thanks for considering!Internal articles that are accessible to Fin are not used when generating the recommendations in Fin > Analyze. Recommendations are generated by comparing failed Fin responses to successful human replies, and they’re only generated from your public articles and snippets (not internal articles). That’s why you’ll only see Snippets as “related content” in those recommendations, even if you’ve enabled internal articles for Fin to use in answers.
Improvment : export a single consolidated list including all your escalation guidance and all your escalation rules from the interface
At the moment, when using Operator to update article content, it would be great to have the ability to be able to edit additional sections of the article. Often when reading through the change proposals, I spot additional elements that require small updates, but I have to deal with that separately afterwards in the article editor. Having the ability to make these additional changes within the existing flow in Operator would be ideal.
Hi Intercom team, we’d love to see an auto-translate option for bulk (broadcast) messages, so one message can be automatically sent in each recipient’s language. This would be a huge time saver for multilingual outreach. We can manually set )or better with API) set the language to each customer. Thanks for considering it.
We would like more control over what is presented in the AI Summary for inbound phone calls.Today, the AI Summary for inbound calls is automatically generated and cannot be customized. It would be very valuable if admins could influence or enrich the generated summary based on captured attributes during the call.Example use cases:Identification number Store number Customer ID Order number Selected menu option / routing pathIf these attributes are captured during the conversation or workflow, it would be extremely helpful to:Include them automatically in the AI Summary Control how they are formatted/presented Prioritize important operational data at the top of the summaryExample:“Customer called regarding invoice issue. Store number: 12345678. Routed via Billing queue.”This would make inbound call summaries much more actionable for support teams and reduce the need to manually listen back to calls or search for key identifiers.
Gostaria de sugerir uma evolução para o FIN voltada ao acompanhamento da qualidade das tratativas humanas durante os atendimentos.A ideia é que o FIN consiga atuar de forma semelhante a um CX Analyst em tempo real, analisando a interação entre cliente e analista ao longo da conversa.Funcionalidades sugeridasMonitorar se todas as perguntas realizadas pelo cliente foram devidamente respondidas pelo analista; Notificar o analista em tempo real quando identificar alguma pergunta pendente ou informação não respondida; Avaliar a qualidade da tratativa com base no contexto da conversa; Identificar o sentimento/emocional do cliente durante o atendimento (positivo, neutro ou negativo); Sinalizar para o analista se o atendimento está caminhando para uma experiência ruim, boa ou excelente; Gerar insights sobre empatia, clareza na comunicação e nível de satisfação percebida do cliente.ObjetivoO principal objetivo seria apoiar os analistas durante o atendimento, reduzindo falhas de comunicação, aumentando a qualidade das respostas e ajudando a melhorar a experiência do cliente em tempo real — não apenas após o encerramento da conversa.Isso permitiria uma atuação mais proativa do suporte, além de contribuir diretamente para indicadores como CSAT, tempo de resolução e satisfação geral do cliente.
Currently, FIN Voice supports keypad input (DTMF) only for limited use cases such as call ratings (1–5) and simple reply options (press 1, 2, etc.).We would like support for free-form multi-digit keypad input during voice conversations, for example:Entering an identity number Entering a store number Entering customer/account IDsUse case:When callers are asked for numeric identifiers, many users prefer entering the numbers on their phone keypad instead of speaking them aloud. This would improve:Accuracy User experience Accessibility Handling of noisy environments Recognition reliability for long numeric sequences
Our team is taking over AI reviews and needs to be able to switch between Monitors and Inbox effortlessly. Right now, if someone is reviewing conversations in Monitors, they can see the Inbox on the left. But if they click into a conversation to reply to a customer from the Inbox, it takes them out of Monitors, and getting back to it requires 5 clicks: Fin AI > Analyze > Monitors > Select Monitor > Filter Reviewer. They will be going back and forth between these all day and that's just not a workable workflow.Using multiple tabs side by side isn't a great workaround either. We've had issues in the past with away and available statuses not syncing correctly across multiple tabs, so staying within one window is the goal.Ideal solution: each user keeps their existing Inbox, but also gets a Monitors Inbox that pulls in all conversations assigned to them for review.
This was a good option for making the chat more user-friendly. You didn't have to remove it. It was one of the best options for adjusting our replies to customers. I was looking for a similar tool for the other services I use, and it was just like yours. That was a bad move, and you don't need to do that. I like the new UI and Intercom overall, but that was a shame, to be honest. Bad move.
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.