Customer Support and Voice AI
AI receptionist for small business: calls, bookings and handoffs
What an AI receptionist does on a real call, the rules that apply, a sample greeting and a pass/fail checklist to launch one safely.
Health & Risk
AI Customer Health Score for Intercom
Health scores built from Intercom conversations and tags, visible on the contact when that customer writes in.
Your support team talks to customers every day, yet they rarely know whether the person writing in is a happy account or one about to leave. This edition scores health from Intercom conversations, tone, contact frequency and open issues, joined with one more source such as billing. The score sits on the Intercom contact, so agents see it before they reply.
- Health score visible on the Intercom contact while agents reply
- Built from conversation tone, frequency and unresolved issues
- Joined with billing or usage data from one more system
$4,250
20–30 working days
Optional maintenance $249/mo
What it is
AI Customer Health Score for Intercom reads conversations, contacts and tags from Intercom, scores sentiment, contact frequency and unresolved issues together with one more system such as billing, and writes the result back to mapped contact attributes or tags. Support agents see account health inside the conversation. Agents see account health inside each conversation.
Why businesses buy it
Support conversations are often the earliest sign that an account is unhappy, but that insight rarely leaves the inbox. Agents treat a frustrated key account the same as anyone else because they cannot see the bigger picture. This edition turns conversation patterns into a health score and puts it where agents work, so they can prioritize and escalate. The written scope sets out the mapping, the second source and the alerts before you order.
Problems this solves
Agents answer without knowing the account is already at risk
The health score is written back to mapped Intercom contact attributes or tags. Agents see account health inside the conversation and can adjust their tone and priority.
Frustration and unresolved issues are not tracked per account
The score weighs sentiment, how often a customer gets in touch and unresolved issues from Intercom conversations. Repeat frustration builds into a visible signal rather than staying buried.
Billing trouble is invisible to the support team
One more system, such as billing, is added to the Intercom data. A customer with a failed payment and a complaint can be spotted as a combined risk.
A good fit if
Intercom is where customers talk to you, and the tone and frequency of those conversations is your earliest sign of trouble.
Support agents reply without knowing whether the customer is a healthy account or one that is close to leaving.
Conversations are already tagged by topic, so tags such as billing problem or bug report can count as signals.
An Intercom admin can issue an app access token and join a 30-minute setup call.
Not the right fit if
You want Intercom conversations answered and resolved, not scored. See AI Customer Support Agent for Intercom.
Your agents need replies drafted inside Intercom. See AI Support Copilot for Intercom.
Account ownership and renewals live in HubSpot, and that is where the score should sit. See AI Customer Health Score for HubSpot.
You want new users guided through onboarding inside Intercom. See AI Customer Onboarding Agent for Intercom.
How it works
- On a 30-minute call your Intercom admin agrees the field mapping: which conversation data and tags are read, and which contact attributes or tags receive the score.
- Intercom conversations become signals: how often a customer writes in, how long issues stay open, repeat topics and the tone of recent messages.
- One more system is joined, usually billing or product usage, so a quiet but non-paying account is not mistaken for a happy one.
- Up to 25 metrics are weighted and checked against three months of your Intercom history, including customers who later cancelled.
- Scores are written to the mapped contact attributes or tags, so agents see health in the Intercom inbox while they reply.
- Up to five alerts and one dashboard or recurring report reach your success team by email, Slack or the web.
What you provide within 72 hours
Your portal lists these the moment you order. Delivery starts the working day after everything is in and your access is checked.
- An Intercom app access token.
- An Intercom admin for a 30-minute setup call.
- Access to the one other system, typically your billing platform or product analytics.
- Your customer list with plan and renewal dates.
- The Intercom tags you use and what each one means for account health.
- At least three months of conversation history, your success playbooks and the report recipients.
- One decision-maker for approvals; tokens are never sent by email.
- An API key for your own AI model account (for example OpenAI or Google Gemini), created for this build with a spending limit.
The 72-hour rule. Send everything on your checklist, including access, within 72 hours of ordering so your build starts on time.
Timeline
Start
Intake verified, Intercom token tested, attribute and tag mapping agreed.
Weeks 1–2
Conversation signals and second-system connection built.
Week 3
Score write-back to Intercom contacts configured; alerts set.
Weeks 3–4
Back-test on three months of conversations; write-back tested on test contacts.
Week 4
Review with support and success leads; two rounds of changes.
Go-live
Health visible in Intercom, recorded handoff, support starts.
| Option | Delivery time | Price |
|---|---|---|
| Standard | 20–30 working days | $4,250 |
| Express | 15–25 working days | $5,000 |
| Rush | 10–20 working days | $5,750 |
What you receive at handover
- AI Customer Health Score configured and writing to your Intercom workspace
- A documented field mapping for the two-way Intercom connection, plus up to one more system
- Access through user invites or keys you control, with no keys sent by email
- One dashboard or recurring report with up to five alerts
- A test report on test records, with scores checked against three months of history
- An admin guide, the recorded 60-minute handoff session and 30 days of post-launch support
Expected benefits and how to measure them
- Helps agents treat at-risk customers with extra care in the moment.
- Makes it easier for success and support to share one view of risk.
- Ratings from at-risk customers
- Filter Intercom conversation ratings by the health attribute and compare the first months with earlier ratings.
- Repeat contacts on the same issue
- Count customers with several open conversations about one topic, monthly, before and after.
- Churn among low-score accounts
- Track cancellations in billing for accounts tagged low health over the following quarter.
Before you start: Export recent Intercom ratings and repeat-contact counts. Note any current tags your team uses to mark unhappy customers.
These are things to track, not promised results. Outcomes depend on your data, your team and your market.
Running costs
The build price is a one-time payment. Expect roughly $20 to $150 a month in your own accounts for AI model usage, which grows with conversation volume, plus hosting or scheduling. Your Intercom plan and seats are unchanged, and any Intercom upgrade sits outside this build.
Usage during the build and testing is billed to you too.
Ask about running costsCustom requirements
Custom Intercom objects, data outside the agreed mapping, a third system or a score per company workspace need a Change Quote. Add a note at checkout for a written quote. Larger or bespoke projects go through the Custom AI request form instead.
Request a custom versionKeep it running
Monthly maintenance covers monitoring, fixes, updates, a health report and change hours, such as new conversation tags to weigh or a revised alert threshold. The hours scale with your build.
$249/mo (the minimum fee, includes 2 change hours)
Optional. Starts after the post-launch support period. Cancel any month.
How maintenance worksHow this compares
| Do it yourself with a SaaS tool | Hire a custom agency | Upmark Labs | |
|---|---|---|---|
| Price | A CS platform charges another seat license beside Intercom | An Intercom specialist quotes per project | Ours is one fixed price, shown before you buy |
| Conversation signals | Most CS tools count tickets but ignore tone | Depends on the brief | Ours weighs tone, frequency and open issues together |
| Seen by support | Health lives in a tool agents never open | An agency build depends on what was agreed | Ours writes health onto the Intercom contact |
| Delivery | Syncing a new platform takes weeks | An agency usually takes several weeks, set per project | Ours has a fixed delivery time shown before you buy, with faster options |
Questions about this product
What Intercom data does the health score read and write?
It reads conversations, contacts, tags and help center articles through a two-way connection, and writes the score to the contact attributes or tags named in your documented field mapping. Data outside that mapping is not touched, and your Intercom workspace settings stay as they are.
Can support agents see a customer's health while replying in Intercom?
Yes. Because the score is written to the contact as an attribute or tag, it appears alongside the conversation in the Intercom inbox. An agent can see at a glance that a customer is at risk and handle the reply, or escalate it, with that in mind.
How is conversation tone turned into a score?
Recent messages from each customer are assessed for tone, alongside how often they write in, how long their issues stay open and whether the same topic keeps returning. You decide how much each signal weighs, and the three-month back-test shows whether tone actually predicted cancellations for you.
What does AI Customer Health Score for Intercom cost?
Each delivery speed has one fixed price. The price panel shows standard, express and rush side by side before you pay, and the cart confirms your delivery date. Running costs of roughly $20 to $150 a month are separate and paid to your own providers.
Which second system pairs best with Intercom?
Usually billing, such as Stripe, because Intercom shows how customers feel but not whether they are paying, downgrading or failing payments. Product usage data is the other common choice. One extra system is included; a third would need a Change Quote or an integration add-on.
What Intercom access does our admin need to give?
An Intercom app access token and 30 minutes of an admin's time to agree the field mapping. The token is used only for the conversations, contacts and tags in that mapping, and it is never sent by email. Your Intercom plan and seats remain your own.
Guides on this topic
All guides Let support see account health before they reply.
Order AI Customer Health Score for Intercom and conversation tone, frequency and open issues become a health score on every Intercom contact, visible before your agents reply.
Works well with
Health & Risk
AI Churn Prevention
Predicts churn risk and recommends the retention play for each at-risk account.
Health & Risk
AI Customer Risk Agent
Flags accounts showing churn signals and recommends save actions.
Front Line
AI Customer Support Agent
Resolves customer questions across chat and email using your knowledge base and hands off complex cases.