Shopify
Stripe
Slack
Notion
Google Workspace
Confluence
Intercom AI Implementation (2026): Set Up Fin Yourself, Hire a Partner, or Have It Done for You?
Intercom AI implementation runs one of three ways: set up Fin, Intercom's own AI agent, with your team; pay Fin Deployment Services or an Intercom Solution Partner to set it up; or have an AI vendor's team build and run a different agent inside your Intercom inbox. Fin costs $0.99 per outcome on top of seats from $29 a month, and Fin says it can be set up in under an hour. Published partner builds run $3,750 to $17,500 over 2 to 6 weeks. A vendor-led setup such as Macha's reaches safe mode within a week, with tuning included in the monthly price.
Key takeaways
- Intercom AI implementation runs on one of three routes: Fin set up by your own team, Fin Deployment Services or an Intercom Solution Partner, or a vendor-led agent inside the Intercom inbox.
- Fin charges $0.99 per outcome, at most once per conversation, on top of Intercom seats at $29, $85 or $132 a month billed annually, with a 14-day free trial.
- Fin Deployment Services promises an enterprise rollout in under 90 days with no published price, while Smart Instinct publishes Intercom builds at $3,750 to $17,500 over 2 to 6 weeks.
- A B2B SaaS team with 6 Advanced seats and 3,000 conversations a month pays $1,188 to $1,782 in Fin outcomes at 40% to 60% resolution, on top of $510 in seats.
- Macha's team promises live in safe mode within 7 days or the first month free, at about $0.40 per ticket with setup and a dedicated success manager included.
Disclosure: Macha publishes this blog, sells the vendor-led route, and competes with Fin. Fin is a strong product, and if you already pay for it, the first question is whether Fin just needs better setup. Every outside price below links to the page it came from, read on October 14, 2026. This page is the Intercom chapter of our AI customer support implementation guide; the Zendesk and Gorgias versions cover the same decision on those desks.
What does Fin's own setup involve?
Fin's setup follows four areas in the product, which Fin's Deployment Services page also uses as its method: Analyze, Train, Test and Deploy. Per Fin's own overview, checked October 14, 2026:
| Area | What you set up | What it decides |
|---|---|---|
| Train: Content | Help Center articles, internal articles, PDFs, webpages | What Fin knows |
| Train: Guidance | Plain-language rules on tone, policy, clarifying questions and escalation | How Fin behaves |
| Train: Procedures and data connectors | Document-style steps that call your systems (Shopify, Stripe, Salesforce, your own API) | What Fin does |
| Test | Simulations of full conversations, and Preview | Whether it's ready |
| Deploy | Simple deploy per channel, or a "Let Fin handle" step in Workflows; audience by users, leads, visitors, region and channel | Who sees Fin, and where |
| Analyze | CX Score, performance dashboards, Topics Explorer | What to fix next |
Procedures replaced the older Tasks: Smart Instinct's Intercom page dates the end of new Tasks to March 12, 2026, and Fin's overview now lists Tasks as managed availability only. This is the decision page, so we stop at the summary. Our Fin explainer covers how Fin reasons, Fin's handoff setup covers escalation rules and workflow branches, and automating refunds in Intercom shows a Procedure with data connectors.
The "under an hour" claim is true for a first answer from your Help Center. The weeks go into what follows: filling gaps in the content, writing Guidance that doesn't contradict itself, and building Procedures that touch billing or account data. When Fin answers from an article instead of the Procedure, or loops on the same answer, the fix is content and Guidance work, and somebody has to own it.
What Fin costs while you implement it
On October 14, 2026, fin.ai/pricing listed $0.99 per outcome, charged at most once per conversation: a resolution, a Procedure handoff, a disqualification or a self-serve routing. A sales qualification costs $9.99. Escalation to a person without an outcome is free. On Intercom's own help desk there's no outcome minimum, because you pay seats; Fin on another help desk carries a 50-outcome monthly minimum. Intercom's pricing page shows seats at $29 (Essential), $85 (Advanced) and $132 (Expert) a month billed annually, or $39, $99 and $139 monthly. The Workflows builder starts on Advanced. There's no setup fee and the trial runs 14 days with no credit card. Our Fin pricing breakdown models the bill at several volumes.
One date matters for anyone signing a long contract. Salesforce completed its acquisition of Fin on September 10, 2026, for about $3.6 billion. The company had renamed itself from Intercom to Fin in May 2026, and Intercom is still the help desk's name; our Fin explainer has the timeline. Prices hadn't changed by October 14, but Fin's roadmap now sits with Salesforce.
How do the three routes compare?
The routes differ most on who keeps the agent right after launch. Fin's pricing page counts a resolution when "no further help is requested after Fin's last answer", so a wrong answer someone gives up on can still bill, and somebody has to read those conversations every week.
| Route | Who does the work | Weeks to live | Cost shown (source, checked October 14, 2026) | Who tunes it after launch |
|---|---|---|---|---|
| DIY with Fin | Your Intercom admin or support ops lead | "Set up in under an hour" for a first deployment (Fin pricing); Procedures and data connectors take longer | No setup fee; $0.99 per outcome plus seats from $29 | Your team, using Analyze and Simulations |
| Fin Deployment Services, or an Intercom Solution Partner | Fin's own deployment team, or agency consultants configuring Fin | Sandbox day 1, pilot in weeks, enterprise under 90 days (Fin Deployment Services); 2 to 4, 3 to 5 or 4 to 6 weeks (Smart Instinct); 2 to 4, 4 to 8 or 8+ weeks (Cloudstride) | Fin Deployment Services "scoped to your goals", no price; Smart Instinct $3,750, $10,500 or $17,500; Cloudstride EUR 2,500 to 4,000, 4,000 to 7,000 or from 7,000, discovery from EUR 1,200, credited if you go ahead; plus Fin's outcomes | Fin's Scale engagement; Smart Instinct retainers from $1,000 a month; Faye's jumpstart covers up to 12 hours over 5 business days, then a managed service with no published fee |
| Vendor-led done-for-you, a different agent inside Intercom | The AI vendor's own team, working in your Intercom inbox | Safe mode within 7 days, public replies on the first category from week 2 (Macha); "live in 14 days" (Fini) | Macha $299/month for 750 tickets, no setup fee; Fini $0.89 per resolution on Growth, implementation included; the visible cards show no minimum, but the pricing page's structured data lists Growth at $3,600 a month for 2,000 resolutions (our Fini pricing breakdown); Intercom seats still apply | The vendor's team, inside the monthly price |
Here's the meter for a B2B SaaS team with 6 Advanced seats and 3,000 conversations a month. Seats cost 6 × $85 = $510 a month billed annually. If Fin resolves 40%, that's 1,200 outcomes at $0.99, or $1,188. At 60% it's 1,800 outcomes, or $1,782. Macha's plan for up to 3,000 tickets is $1,199 a month and bills every conversation the agent works, resolved or not, so Fin is cheaper below about 1,211 outcomes and Macha is cheaper above, with the same $510 of seats on both routes. The incentives follow the meter. Fin earns more as its resolution count rises, so Fin's tools push automation up. A partner earns on the build, so post-launch hours are capped or sold as a retainer. A vendor billing per ticket keeps the account only while the agent keeps answering well, so tuning is part of what it sells.
For partners and agencies by name, see our ranked list of AI implementation partners and who can set up AI customer support. On October 14 we found no public Intercom partner directory: intercom.com/partners returned a 404, and the Solution Partner Program page says only that "Agencies, BPOs, and SIs are welcome to apply." Agencies state their own status, so ask for proof.
How we researched this
On October 14, 2026 we checked fin.ai/pricing, Intercom's pricing page (including the monthly prices in its page data), Fin Deployment Services, Fin's overview article, Intercom's articles on deploying Fin over chat and email, and the Solution Partner Program page. We read the Intercom pages of Smart Instinct and Cloudstride, Fini's pricing page and Faye's jumpstart page. Handoff and data connector details come from our own Intercom guides, which cite Intercom's docs. The Macha timeline is the plan our team publishes on the done-for-you page. We didn't hire a partner or run an Intercom workspace for this page, and the 3,000-conversation example uses illustrative resolution rates.
What are the phases of an Intercom AI implementation, with days?
Every route runs the same six phases. Partners publish only totals in weeks, so the day ranges below are the plan our team publishes; the third column is what you supply on any route.
| Phase | Days (Macha's published plan) | What you provide | Intercom-specific work |
|---|---|---|---|
| 1. Kickoff and conversation audit | Day 1 call, then days 1 to 3 | Intercom access; a 30-minute call on how support runs | Rank a month of conversations by topic and channel (Messenger, email, WhatsApp); note which need account or billing data |
| 2. Knowledge | Days 3 to 5 | Help Center access, internal articles, saved replies, product docs | Fix articles for the first category; move policy out of saved replies and Slack threads into a source the agent reads |
| 3. Actions and integrations | Days 3 to 5 | Stripe, HubSpot, Salesforce or Jira access; refund and plan-change limits | Decide which writes run without a person; map the contact and company attributes the agent should read |
| 4. Test on past conversations | Days 3 to 5 | Agreement on tone and handover rules | Replay real conversations; check Workflows, assignment rules and Fin's audience don't fight the agent |
| 5. Safe-mode launch | Days 5 to 7 | Someone to review drafts, about 5 minutes a day | Drafts as internal notes on live conversations; no customer sees them |
| 6. Tuning | Week 2 onward, for as long as it runs | Feedback, product releases, the next category | Public replies on approved categories, then the next one |
Before launch, test four kinds of conversation from your own Intercom history:
- Normal case. "How do I add a teammate?" from a logged-in user on a plan that allows it. The answer should match the docs and your tone.
- Escalation. A security report, a data deletion request, an enterprise admin threatening to cancel. The agent should hand over with a note.
- Edge case. A lead (not a user) asking about their invoice, a user whose company is on a legacy plan, a feature that shipped yesterday.
- Tone. A frustrated developer pasting an error, a one-line message in German, a procurement question. The facts can be right and the voice still wrong for that customer.
On Fin's route, Simulations run full test conversations, and Intercom suggests setting Fin live to a small internal audience first. Macha runs this step as simulations on past conversations, described in testing agents on real data before launch. Rollouts that skip it are the ones that get switched off, as we covered in why AI support pilots get rolled back.
After launch, someone reads the agent's conversations each week, sorts the misses (wrong answer, missed handover, right answer in the wrong tone), fixes the article, Guidance or instruction, and picks the next category. Watch three numbers weekly: handovers by reason, conversations reopened after an AI answer, and CSAT on AI-handled conversations against human ones. Your support lead signs off each new category before it goes to public replies, on every route. On B2B SaaS the work never ends, because every release changes the right answer, so put the support lead on the release notes list.
Who does each job on each route?
Bring this table to every sales call and ask the provider to fill in its column.
| Job | DIY with Fin | Fin Deployment Services or a partner | Vendor-led (Macha) |
|---|---|---|---|
| Classify past conversations and pick the first category | You, with Topics Explorer | Partner, in discovery or the success scorecard | Macha's team |
| Fix Help Center articles | You | Partner, if scoped | Macha's team, with the Knowledge Builder |
| Write Guidance or agent instructions | You | Partner | Macha's team, from how your team already answers |
| Connect Stripe, CRM and product data | You, with data connectors | Partner, within the tier's integration count (1, 5 or 8 at Smart Instinct) | Macha's team: Stripe, HubSpot and Jira built in; Salesforce and your own API as custom API tools |
| Set refund, credit and plan-change limits | You | You | You |
| Adjust Workflows, assignment and audiences | You | Partner | Macha's team, with your admin's sign-off |
| Test on past conversations | You, in Simulations | Partner, with your testers | Macha's team |
| Approve tone and go-live | You | You | You |
| Read live conversations and fix misses | You | You, a retainer, or the Scale engagement | Macha's team and your dedicated success manager |
| Update answers after each product release | You | A new project or the retainer | Macha's team, included |
Two rows stay with you on every route: what the agent may change in a customer's account, and the decision to let it talk to customers.
What Intercom-specific problems slow an implementation down?
These are the issues that add weeks on Intercom in particular.
Messenger and email behave differently
Fin on chat runs once it's deployed, but Fin on email needs its own Simple deploy or a "Let Fin handle" step in an email workflow; per Intercom's email deployment article, email forwarding alone isn't enough. On email Fin sends one combined answer instead of several short messages, and it treats the sender as the customer, which goes wrong when a reseller or an assistant writes in for someone else. Messenger customers expect a reply in seconds and often leave mid-conversation. Plan the two channels as two launches, and ask any vendor which Intercom channels it replies on.
Workflows that fight the agent
Intercom's chat deployment article says Simple deploy "will always take precedence" over Messenger new-conversation Workflows. Teams with years of Workflows find that some never run again, or that a "Customer sends any message" workflow pulls Fin back in after a person took over. Before launch, list every workflow that fires on a new conversation or a customer reply and decide which still apply. A third-party agent needs the same audit, plus a clear slice of conversations Fin never touches.
B2B data lives in the CRM and billing system
The answer to "why was I charged twice" sits in Stripe, and "what does my plan include" sits in your CRM or product database. Fin reaches them through data connectors, and our refunds guide found that Intercom's Stripe connector templates only read data, so a Fin refund through Stripe needs a connector you build. Clean up the contact and company attributes first, because the agent can only use what's synced to Intercom or callable by API; our guide to custom attributes vs tags covers which to rely on. For CRM data, see connecting Salesforce to Intercom and connecting HubSpot to Intercom.
Multiple brands and workspaces
One Intercom workspace can hold several brands, each with its own Help Center and email address, and Fin picks the Help Center by the conversation's brand. That only works if each brand is set up in Brand settings and its email address is linked to it. Companies that grew by acquisition often run separate workspaces instead, and each one is its own setup, its own billing and its own connection for any outside agent. Our Help Center setup guide covers multiple Help Centers.
Handover to humans
Fin decides when to hand over through Escalation rules and Guidance, and Workflow branches decide which team gets the conversation. Guidance can't name a destination. Office hours are respected on a workflow handover, which matters if your SLAs pause out of hours (SLA and office hours setup). A third-party agent hands over by assigning the conversation and leaving an internal note, so the person who picks it up doesn't have to ask the customer to repeat themselves.
When do you need a partner instead?
An Intercom partner or Fin Deployment Services is the better hire when the project is bigger than the AI:
- You're moving to Intercom. Conversations, contacts, companies and articles are partner work, and so is redesigning the inbox. Our Intercom migration guide shows what moves and what doesn't.
- You're committed to Fin and want Intercom's own engineers on it. Fin Deployment Services knows the product better than anyone, and its Launch engagement includes post-launch optimization.
- You run several teams, markets or workspaces and need routing, SLAs and reporting designed across them. Cloudstride's complex tier exists for that.
- Your help center is empty. Smart Instinct lists it as a reason it isn't a fit for Fin, and it applies to every agent. Write the core articles first.
- You want the configuration documented and owned in-house. That's how partners hand over, and it suits a team with an admin who'll keep it up.
The vendor-led route isn't a fit if you want Fin's native features (Fin Voice, product tours, Fin's own reporting) or if your admin wants to edit every rule personally. It also isn't a fit if you want a resolution percentage in the contract. Smart Instinct lists that as a reason it isn't a fit either, and nobody can promise one before seeing your conversations.
The routes also combine. A partner can design your Intercom while Macha's agent works the conversations inside it, as long as Fin and the other agent are split by channel, audience or team so they never answer the same customer.
Which route fits your situation?
| You are | Best route | Why |
|---|---|---|
| A B2B SaaS team, 3,000 conversations a month, already on Fin | Keep Fin; spend on tuning first, from your own admin or a partner | Fin's price at 40% resolution sits close to a per-ticket plan; better content usually moves the number more than a switch |
| A B2B SaaS team on Intercom, no one with time to run the AI | Vendor-led done-for-you | Setup and weekly tuning sit in the subscription, and product releases get handled |
| A startup on Essential, under 500 conversations, good docs | DIY with Fin | Seats from $29 and $0.99 outcomes stay small at that volume; setup really is quick |
| An enterprise rolling out across regions and teams | Fin Deployment Services or a Solution Partner | Rollout plans, routing and change management across teams are the real work |
| A team whose answers depend on Stripe, HubSpot or Jira data | Test the data path first, on whichever route | The agent is only as good as the account data it can read |
| An ecommerce brand on Intercom with Shopify | Fin with Intercom's Shopify app and data connectors, or a vendor with a Shopify connector | Order lookups and refunds are the first categories either way |
| A team whose Fin bill keeps climbing with resolutions | Compare a per-ticket vendor at your volume | Per-outcome pricing scales with every resolution; per-ticket pricing is capped by tier |
Where Macha fits
Macha is one of the vendor-led options on Intercom, and it's ours. The Macha team classifies your past Intercom conversations by topic and channel, fills the Help Center gaps for the first category, builds the agent inside your existing Intercom, and runs it in safe mode, with drafts as Intercom internal notes, until your team approves them. Live in safe mode within 7 days, or your first month is free. Your part is a 30-minute call, about 15 minutes to connect Intercom and the tools answers come from, and about 5 minutes a day reviewing drafts.
The Intercom connector signs in with OAuth. It reads and searches conversations and contacts (including custom attributes and companies), searches Help Center articles, replies as the connected Intercom admin, adds internal notes, assigns, tags and closes, and runs on triggers such as First Customer Message. Stripe, HubSpot and Jira connect as built-in tools; Salesforce or your own product API connect through a custom API tool that Macha's team sets up during onboarding. See Macha on Intercom for the tool list, or Fin vs an AI agent layer for the wider comparison.
It costs about $0.40 per ticket, which includes setup and a dedicated success manager who handles new categories, product changes and new tools, on one plan priced by monthly ticket volume. The free trial includes the same done-for-you setup: $50 of free usage (about 125 tickets), no credit card, no time limit. Macha suits B2B and ecommerce teams on Intercom that want conversations resolved without anyone on staff maintaining the AI. It doesn't replace a partner for a migration or an Intercom build, and if Fin is already working for you, keep it. To see which categories an agent could take first, request the free support audit.
Common questions
Who can help me implement Fin?
Your own Intercom admin, Fin Deployment Services, or an agency in Intercom's Solution Partner Program, such as Smart Instinct or Cloudstride, which say so on their own pages, or Faye, which calls itself Fin's 2026 Service Partner of the Year. Intercom had no public partner directory on October 14, 2026. A vendor such as Macha or Fini sets up its own agent inside Intercom instead of configuring Fin.
How long does a typical Fin implementation take?
Fin says a first deployment takes under an hour. Fin Deployment Services quotes a pilot in weeks and an enterprise rollout in under 90 days, and partners publish 2 to 8 weeks or more by scope. Macha's team commits to safe mode within 7 days of connecting Intercom, or the first month is free.
Do I need Fin Deployment Services or a partner?
Not to start. Fin's Deployment Services page says "You can set Fin up yourself, and many teams do." Buy help when Procedures touch billing or account systems, when several teams or regions are involved, or when nobody internal has time to tune Fin every week.
How much does an Intercom AI implementation cost?
Fin has no setup fee: $0.99 per outcome plus seats from $29 a month. Published partner builds run $3,750 to $17,500 (Smart Instinct) or EUR 2,500 to 7,000 and up (Cloudstride), plus Fin's outcomes. Macha includes setup in $299/month for 750 tickets.
Can Fin and a third-party AI agent run in the same Intercom workspace?
Yes, if they never answer the same conversation. Narrow Fin's audience or channel in Deploy, or remove the "Let Fin handle" step from the workflows that route to the other agent, and give the other agent its own trigger scope.
Who maintains the AI after go-live?
On the DIY route, your Intercom admin, using Analyze and Simulations. With a partner, you after handover, or a retainer from $1,000 a month at Smart Instinct; Fin Deployment Services sells a Scale engagement. With Macha, a dedicated success manager handles new categories, product changes and new tools inside the per-ticket price.
What resolution rate should we expect from AI on Intercom?
Nobody can tell you before reading your conversations, and anyone quoting a number up front is guessing. Fin's Deployment Services page cites a 72% average resolution rate and Salesforce's release cites 76%; both are averages across Fin's customers, and yours depends on your content and your mix. The conversation audit at the start of every route gives a realistic share per category, and the first category you launch tells you more than any benchmark.
