ChatGenius Pricing About Blog Login

How to Migrate from ManyChat to ChatGenius

An honest, step-by-step guide written by ChatGenius's CTO. What carries over, what does not, and exactly where to rebuild each ManyChat feature inside ChatGenius. No marketing fluff.

Read this first

Most "migration guides" overpromise and bury the catches. This one tells you the catches up front.

The honest reality

There is no one-click ManyChat-to-ChatGenius importer, and there cannot be one. Three things block it: Meta's policy on subscriber consent, Meta's API rules on historical messages, and the schema gap between ManyChat's flow data and ours. Anyone who promises a CSV-import button is either lying or about to ship something broken.

What you can do is recreate your ManyChat setup in ChatGenius in two to four hours, manually, using the walkthroughs below. For most small-business setups that is faster than fighting an importer would have been anyway.

What actually moves over

ManyChat artifact Migratable? How it works in ChatGenius
Subscribers / contacts No Meta policy requires fresh consent. Each subscriber must DM your page once before ChatGenius can reply.
Conversation history No Meta API does not allow historical writes. New conversations sync going forward via webhooks.
Custom fields No ChatGenius does not have arbitrary per-subscriber custom fields. Use lead capture (auto-extracted email and phone) and tags instead.
Sequences (drip campaigns) No Not implemented. AI follow-ups handle the most common drip pattern (re-engaging silent leads). Roadmap item.
Audiences and segments No Tag-based filtering exists in the conversations inbox, but ManyChat-style rule-built audiences do not.
Visual flows (full) No Different schema, different node types. Rebuild in Flows section. Walkthrough below.
Keyword automations (text reply) Yes Direct mapping to Quick Replies. Walkthrough below.
Comment-to-DM triggers (concept) Manual rebuild Same concept, different setup. Walkthrough below.
AI Step prompts (text) Yes Paste your AI Step instructions into AI Configuration. ChatGenius is AI-first, so the entire conversation runs on this prompt without flow nodes around it.
Knowledge / FAQs Yes Upload your FAQ docs to ChatGenius's knowledge base. RAG-powered retrieval, native to all paid tiers.

Why subscribers and history cannot transfer

This is the part most guides skip. It is the most important part.

Meta's 24-hour messaging window

Meta only lets a business app message a user within 24 hours of that user's last inbound message. Consent is bound to the app that received the message. ManyChat held your subscribers' consent. ChatGenius will not, until each subscriber sends your page a fresh DM. There is no way around this without violating Meta's platform terms.

Conversation history lives on Meta's servers

Old conversations are owned by Meta, not ManyChat. ChatGenius syncs new conversations via webhooks, but Meta's API does not expose a "write historical messages" endpoint. If you need conversation logs, export them from ManyChat to your CRM before cancelling. After cancellation, that data is no longer accessible to either tool.

Flow JSON schemas don't match

ManyChat's flow data uses node types we do not have (Card Galleries, Buy Buttons, Smart Delays, External Requests, Telegram-specific nodes). Even a translator script would lose the majority of nodes in any moderately complex flow. The honest answer is to rebuild the strategic intent of each flow in ChatGenius's Flows section, which is faster than you'd guess once you start.

Step-by-step migration walkthroughs

For each common ManyChat use case, here is exactly where it lives in ChatGenius and exactly how to recreate it.

3.1 First-DM auto-reply (the AI's opener)

In ManyChat: Default Reply or Welcome Message inside Settings, usually wired into a flow.

Portal → ChatGenius → AI Configuration → Identity + Behavior
  1. Log in to your ChatGenius portal at sumgenius.ai/portal/login.
  2. In the left sidebar under "AI & CONFIGURATION", click AI Configuration.
  3. Open the Identity subsection. Set your AI's name and tone (casual, professional, formal). This shapes how every reply opens.
  4. Open the Behavior subsection. This is where your AI Instructions live. Paste in your ManyChat welcome copy plus any additional context about your business, voice, and goals. The AI uses this on the first DM and every DM after.
  5. Save. Run a test conversation from the inbox to verify the AI's first-DM reply matches your brand.
What changes vs ManyChat: ManyChat treats the welcome message as one node in a flow. ChatGenius doesn't have a separate welcome-message field at all. Instead, the AI's first reply is generated from your Identity (name, tone) and Behavior (instructions) settings, which then continue driving the rest of the conversation. You don't have to build out a multi-node flow for normal back-and-forth.

3.2 Keyword automation (text reply)

In ManyChat: Automation → Keywords → Set up keyword.

Portal → ChatGenius → Quick Replies
  1. From the sidebar, click Quick Replies.
  2. Click + New Quick Reply.
  3. Enter the trigger keyword exactly as you had it in ManyChat (e.g. pricing, hours, book).
  4. Enter the response text. You can include line breaks, links, and emoji.
  5. Save. Repeat for every ManyChat keyword automation that is text-only.
If your ManyChat keyword launched a multi-step flow: create the Quick Reply as a stub for the first message, then build the rest in Flows (covered in 3.4). For most use cases (FAQ replies, hours, pricing, contact info), a single Quick Reply is enough.

3.3 Comment-to-DM triggers

In ManyChat: Growth Tools → Comment Trigger.

Portal → ChatGenius → Comment Triggers
  1. From the sidebar, click Comment Triggers.
  2. Click + New Trigger.
  3. Choose the post you want to trigger from (specific post or any post).
  4. Enter the keyword that should fire the trigger when found in a comment.
  5. Enter the DM message that goes out to the commenter when the keyword fires.
  6. Enable the trigger and save.
What's the same as ManyChat: trigger fires when a comment matches your keyword, then sends a private DM to the commenter. Same concept, different config screen. Use the Comment Manager section to monitor and respond to inbound comments separately if you want to handle public replies manually.

3.4 Multi-step flows (the visual flow builder)

In ManyChat: Automation → Flows.

Portal → ChatGenius → Flows
  1. From the sidebar, click Flows.
  2. Click + New Flow and give it a name matching your ManyChat flow.
  3. Drag a Trigger node onto the canvas. Set its keyword to match the original ManyChat trigger.
  4. Add Send Message nodes for each step the user should see. Connect them in order.
  5. For branching (Yes / No options, choices), use the buttons configuration on a Send Message node — each button output port becomes a separate path you can wire to different next steps.
  6. For more complex flows, you have additional node types available: Smart Delay (timed pause), Collect Input (capture a user reply into a variable), Condition (branch on a value), AI Step (let the AI handle a turn inside the flow), HTTP Request (call an external API), Book Appointment (Calendar booking inline), Randomizer (A/B split), AI Handoff (hand control to the AI brain), and Go-To Flow (jump to another flow).
  7. Save and toggle the flow on. Test it from your own Instagram or Facebook account.
What you DO have that maps to ManyChat: Smart Delay (= ManyChat's Smart Delay), HTTP Request (= ManyChat's External Request), AI Step (= their AI Step), Condition (= their Condition), Collect Input (= their User Input). These are 1:1 conceptually.

What you DON'T have: Card Galleries (visual carousels of items), Buy Buttons (in-chat checkout), and Sequences (drip campaigns). For card-style product showcases, send a list of items as text or send media one at a time. For checkout, link out to your existing payment page using a Send Message node.

3.5 AI conversations (replacing ManyChat's AI Step)

In ManyChat: AI Step inside a flow, plus the $29/mo AI add-on subscription.

Portal → ChatGenius → AI Configuration → Behavior
  1. Open AI Configuration from the sidebar, then the Behavior subsection.
  2. Copy your ManyChat AI Step prompt and paste it into the AI Instructions field. Add any business context that was scattered across other AI Step nodes — you only need one prompt for the whole conversation.
  3. Open the Identity subsection: set tone (casual, professional, formal) and AI Name if you give your assistant a persona.
  4. Open the Restrictions subsection to define off-limits topics (things the AI should refuse to discuss).
  5. Save. The AI is now active on every conversation, not just at specific flow nodes.
What you get for the same money: ChatGenius AI is included in every paid tier from $29/mo. There's no separate AI add-on. The model is GPT-5 with full context (your AI Instructions, your knowledge base, the conversation history).

3.6 Booking and appointments

In ManyChat: Acuity or Calendly templates wired into a flow.

Portal → ChatGenius → Calendar
  1. Click Calendar in the sidebar.
  2. Connect your Google Calendar via OAuth. ChatGenius reads your real availability and respects your existing events.
  3. Set your booking rules: business hours, buffer time, appointment duration, services menu.
  4. The AI now offers booking automatically when conversations indicate interest. Booked appointments sync back to your Google Calendar instantly.
Why this is cleaner: ManyChat hands the booking off to a third-party tool (Acuity, Calendly), then the user leaves the chat to complete it. ChatGenius does it in-conversation, in-DM, and writes directly to your existing Google Calendar with no third party.

3.7 Lead capture (name, email, phone)

In ManyChat: Custom Fields plus flow steps that ask the user for their info.

Portal → ChatGenius → Leads
  1. You don't have to build anything to start. ChatGenius's AI auto-extracts emails, phone numbers, and names from natural conversation flow.
  2. Click Leads in the sidebar to view the running list of captured leads with the source conversation linked.
  3. For more aggressive capture, edit your AI Instructions in AI Configuration → Behavior to tell the AI when and how to ask for contact info (e.g. "always ask for email before booking").
  4. For deterministic capture inside a Flow, use the Collect Input node to ask a specific question and store the answer.

3.8 Knowledge base (upload your FAQs, services, and docs)

In ManyChat: not natively supported. Most users hard-code FAQs into flow text.

Portal → ChatGenius → AI Configuration → Knowledge
  1. From AI Configuration, open the Knowledge subsection.
  2. Upload your business FAQs, your services or service menu, and any product docs (PDFs, Word, text). You can also paste URLs of your business website.
  3. ChatGenius indexes the content and the AI uses it as grounding when answering customer questions, so it pulls real answers from your real docs instead of guessing.
  4. Knowledge limit by tier: Free 15 items, Creator 40, Professional 100, Business 300, Enterprise unlimited.
This is where switching pays off the most: instead of anticipating every possible question inside flows, you upload your real business docs (FAQs, services list, pricing sheet, hours, policies) and the AI handles the long tail of "how does X work" naturally. Most ChatGenius clients shrink their AI Instructions dramatically once they get real knowledge docs uploaded.

3.9 Webhooks and external API calls

In ManyChat: External Request nodes inside flows. Most users hit the wall here because the External Request node is finicky and can race-condition with ManyChat's field-write latency.

Portal → ChatGenius → Webhook + Send API
  1. Subscribe to the Webhook + Send API add-on ($29/mo on top of any tier). Click Webhook + Send API in the sidebar.
  2. For inbound (we send events to your server): paste your webhook URL, pick which events you want (message received, lead captured, etc.), and copy your HMAC signing secret to verify requests on your side.
  3. For outbound (your server sends DMs to users): generate an API key, then POST to our send endpoint with the conversation ID and message body. Same idempotency, retry, and dead-letter handling as the rest of the platform.
  4. For SMS: same add-on covers SMS send and receive (refill notifications, order updates, etc.).
Why we built this: ManyChat's External Request is synchronous and waits for a response, which causes timing issues when messages come in fast. ChatGenius's webhook flow is async with idempotent retries. If you were duct-taping ManyChat + Make + Airtable + OpenAI together, our Webhook + Send API replaces that whole stack for one $29 add-on.

3.10 Broadcasts (Pro tier and above)

In ManyChat: Broadcasting tab, send to selected audiences.

Portal → ChatGenius → Broadcasts
  1. Available on Professional ($99/mo) and above.
  2. Click Broadcasts in the sidebar.
  3. Compose your message, pick the channel (Instagram, Facebook, WhatsApp, SMS).
  4. Pick your audience from the conversations you've already had on ChatGenius (subscribers must have engaged with you on our system to be eligible — Meta's window rules apply here too).
  5. Schedule or send now.
Important: your ManyChat audience cannot be imported. You'll be broadcasting to people who have already messaged you on ChatGenius after migration. For most businesses, the audience rebuilds within 30 to 60 days as your normal inbound DM flow continues.

3.11 WhatsApp (where the math gets interesting)

In ManyChat: WhatsApp connection through their BSP (Business Solution Provider) status. They mark up every WhatsApp message you send.

Portal → ChatGenius → WhatsApp Templates
  1. ChatGenius is a Meta Tech Provider, not a BSP. You connect your own WhatsApp Business Account (WABA) to your own Meta Business account.
  2. You pay Meta directly, at Meta's published rates. Zero markup from us. We only charge the flat $99/mo Pro tier (or $297/mo Business).
  3. WhatsApp connection guides walk you through the WABA setup and template approval. Templates submitted from WhatsApp Templates in the sidebar.
  4. WhatsApp DMs flow into the same inbox as Instagram and Facebook DMs. AI handles them with the same configuration.
The cost difference, concretely: if you send 5,000 WhatsApp service conversations per month, ManyChat's BSP markup adds roughly $50 to $150 in pure middleman cost on top of Meta's actual rate. ChatGenius cuts that to zero. The break-even on switching for any heavy WhatsApp user is usually under one billing cycle.

3.12 Content Studio (a feature ManyChat doesn't have)

In ManyChat: not available. Users typically pair ManyChat with Later, Buffer, or Hootsuite for scheduling.

Portal → ChatGenius → Content Studio
  1. Click Content Studio in the sidebar.
  2. Compose a post. Upload media (image, video, carousel) or generate AI images.
  3. AI captions and hashtags auto-suggest based on your business context.
  4. Schedule for any future date or post immediately. Posts publish to Instagram and Facebook through Meta's official Graph API.
  5. View post performance analytics inline. No separate scheduling tool needed.
If you were paying for Later or Buffer separately: that subscription becomes redundant. Content Studio is included in Creator ($29/mo) and above. Schedule limits scale with tier.

What you gain by switching

The honest list. Each one is a verifiable difference, not marketing language.

  • Flat pricing. $29 Creator, $99 Professional, $297 Business. No per-contact ladder, no per-channel surcharges. AI is included in every paid tier (no $29/mo AI add-on).
  • WhatsApp without the BSP markup. Connect your own WABA, pay Meta directly. We're a Tech Provider, not a Business Solution Provider.
  • Click-to-WhatsApp ad attribution. ChatGenius captures the full referral data (ad ID, headline, body) from CTWA leads — ManyChat cannot.
  • 13 languages with auto-detection. The AI detects each user's language per message and replies in kind. No separate flows per language.
  • GPT-5 AI in every conversation. Not a flow node. Not an add-on. The whole conversation runs on AI with full business context.
  • RAG-grounded knowledge base. Upload your real business docs; the AI uses them as factual grounding. Ends the "we just guessed at FAQs" problem.
  • Native Google Calendar booking. No third-party booking tool needed. Real availability, in-conversation, syncs to your existing calendar.
  • You can talk to a human. Onboarding is handled personally by Jared (CTO). ManyChat's support is a ticket queue.
  • Content Studio built in. Schedule and publish to Instagram and Facebook with AI captions. Replaces a separate Later or Buffer subscription.

What you lose by switching

The honest list of tradeoffs. If any of these are deal-breakers, ChatGenius is not the right tool for you.

  • No Sequences (drip campaigns). ManyChat's day 1, day 3, day 7 follow-up sequences do not have a direct equivalent yet. AI follow-ups handle the most common case (re-engaging silent leads), but scheduled multi-day sequences are roadmap, not live.
  • No arbitrary Custom Fields. ChatGenius captures lead email, lead phone, conversation tags, and notes — but you cannot define unlimited per-subscriber fields the way ManyChat does.
  • No rule-built Audiences. Filtering in the conversations inbox uses tags and search; there's no standalone Audience-builder UI with stacked rules.
  • No TikTok DMs. ManyChat supports TikTok messaging. ChatGenius does not (yet).
  • No mass DM tools. ChatGenius will never offer "DM 10,000 strangers per day" features. We work strictly within Meta's official Graph API and within Meta's policies. If unauthorized mass outreach is your use case, ChatGenius is not for you.
  • Smaller flow node library. ChatGenius's Flows section covers triggers, messages, buttons, and branching — solid for the 80% case. ManyChat's flow builder is more mature with Card Galleries, Buy Buttons, Smart Delays, and more.
  • No iOS or Android app yet. The portal is a responsive web app. Works fine on mobile browsers, but no native app store presence.

The migration playbook (six days)

A realistic timeline for the average small business. Power users with complex flows should plan for one focused day or two split sessions.

Six-day migration sequence

  1. Set up the bot's brain. Sign up for ChatGenius free (no credit card). Open AI Configuration and fill in Identity (AI name, tone) and Behavior (paste your ManyChat AI Step prompts here, plus any business voice/instructions). Set Restrictions for off-limits topics. Don't connect your Meta pages yet — you want the bot fully prepared first.
  2. Upload your knowledge. In AI Configuration → Knowledge, upload your FAQs, services or service menu, and any product docs. Add your website URL too. The AI will use these as factual grounding once it's live.
  3. Test the bot in the sandbox. AI Configuration has a built-in test sandbox where you can chat with your bot exactly as a real customer would, before any Meta page is connected. Throw realistic customer questions at it. Tweak Behavior + Knowledge until the answers sound right.
  4. Build your Flows. In the Flows section, recreate any deterministic ManyChat flows that need to be exact (booking funnels, qualification questionnaires, specific YES/NO branches). Most simple keyword replies belong in Quick Replies (next step) rather than Flows.
  5. Build Quick Replies and Comment Triggers. In Quick Replies, create one per text-only keyword automation from your ManyChat setup. In Comment Triggers, recreate your ManyChat comment-to-DM rules. Test each by triggering them in the sandbox or by commenting on a private test post.
  6. Connect your pages and disconnect ManyChat. Once the bot is fully configured and tested, connect your Instagram and Facebook pages to ChatGenius via OAuth. Disconnect ManyChat from those pages immediately afterward (only one tool can be connected at a time — this is the cutover moment). Watch the inbox for the first day of real DMs to confirm everything's wired correctly. Cancel your ManyChat subscription once you're confident.
Why this order matters: connecting your Meta pages is the LAST step, not the first. If you connect pages before the bot is configured, real customer DMs hit a half-built bot and may get poor replies. Set up Identity, Behavior, Knowledge, Flows, Quick Replies, and Comment Triggers first, test them in the sandbox, then flip the switch.

Want me to walk you through it?

I'm Jared, the CTO and one of the two co-founders. I personally handle onboarding for new clients. If you'd rather have someone do the migration with you on a screenshare than do it solo, book a free 30-minute session and we'll get you fully set up live.

Migration FAQ

The questions we get most often from people considering the move.

Can I import my ManyChat subscribers into ChatGenius? +

No. Meta's messaging policy locks subscriber consent to the app that captured it. Your existing ManyChat subscribers will need to send your Instagram or Facebook page a fresh message before ChatGenius can respond to them. This is a Meta-level restriction, not a ChatGenius limitation.

Will my ManyChat conversation history transfer? +

No. Meta's API does not allow writing historical messages into a conversation. New messages on ChatGenius will sync via webhooks, but past conversations stay in ManyChat. We recommend exporting any important threads to your CRM before cancelling.

Will my ManyChat flows convert automatically? +

No, flows require manual rebuild. ManyChat and ChatGenius use different node types and a different data model. Simple keyword automations (one keyword equals one text reply) carry over cleanly as Quick Replies. Multi-step flows must be rebuilt in our Flows section.

How long does the full migration take? +

For a typical small-business setup (one page, ten or fewer keyword automations, one comment trigger, AI instructions): two to four hours. Power users with dozens of flows, sequences, and custom fields should plan a full day. The CTO will walk you through it personally on a free call.

Should I keep ManyChat running during migration? +

Yes, run them in parallel for at least one week. Connect ChatGenius first, test with your team, then disconnect ManyChat from your Facebook page once ChatGenius is handling messages reliably. You cannot have both connected to the same page at the same time.

What about my ManyChat custom fields? +

ChatGenius does not have arbitrary per-subscriber custom fields. Email and phone are auto-captured into Leads. Conversation-level tags and notes cover most use cases. If you relied on custom fields for complex routing logic, you'll need to either (a) drop that complexity in favor of AI handling, or (b) push the data to your own backend via the Webhook add-on.

Will I save money by switching? +

For most users with growing audiences, yes. ManyChat charges per contact, which scales aggressively past 1,000 contacts. ChatGenius charges per active conversation per month at flat tiers. WhatsApp users save the most because we don't add BSP markup to message costs. Run the math on your specific volume and current ManyChat bill before committing.

What if I just want to try it before fully migrating? +

Sign up for the free tier (10 conversations per month, no credit card). You can connect a different Facebook page or a test page to evaluate ChatGenius without disconnecting ManyChat from your main page. When you're ready, switch the connection over.

Looking for a side-by-side feature comparison instead? Read the full ChatGenius vs ManyChat comparison →