Don't build new live chat on Salesforce Live Agent. Live Agent was rebranded to Chat (and surfaced as Embedded Chat in Lightning), and Salesforce has since announced the retirement of Chat — the underlying Live Agent technology. The supported path for live chat today is Messaging for In-App and Web (MIAW) on Salesforce's messaging platform, paired with Einstein Bots or Agentforce for AI-assisted conversations.
If you run an older org that still has Live Agent enabled, it may keep working for now, but it is a legacy, end-of-life product. Plan your migration to MIAW rather than investing in new Live Agent build-out.
Key takeaways
- Live Agent is legacy. It was rebranded to Chat; Salesforce has announced Chat's retirement and directs customers to Messaging for In-App and Web (MIAW).
- MIAW is the current standard for web and in-app live chat, with persistent conversations rather than session-only chat.
- Route with Omni-Channel and automate with Einstein Bots / Agentforce.
- Confirm the exact end-of-life date in Salesforce's current retirement schedule for your edition — don't rely on a date quoted in a blog post.
- New projects should start on MIAW; legacy orgs should plan a phased migration.
How has Salesforce live chat evolved?
| Era | Product name | Status |
|---|---|---|
| Original | Live Agent | Retired technology — do not build new |
| Rebrand | Chat / Embedded Chat (Lightning) | Same Live Agent engine; retirement announced |
| Current standard | Messaging for In-App and Web (MIAW) | Recommended for all new live chat |
| AI layer | Einstein Bots, then Agentforce | Add automated, AI-assisted conversations |
Salesforce has publicly announced that Chat (the Live Agent technology) is being retired and that MIAW is the go-forward channel. Because retirement dates can shift by release and edition, confirm the current end-of-life date in Salesforce's official retirement documentation rather than assuming a fixed cutoff.
What is Messaging for In-App and Web (MIAW)?
Messaging for In-App and Web is Salesforce's current framework for embedding real-time conversations in your website and mobile apps. Unlike classic Live Agent chat — which was largely session-based and disappeared when the customer closed the tab — MIAW conversations can be persistent, so a customer and your agents can pick up the same thread later.
MIAW runs on the same messaging platform that powers channels such as SMS, WhatsApp, and Facebook Messenger, so a single Service Cloud setup can handle web chat and external messaging channels in a consistent way.
How is MIAW different from legacy Live Agent / Chat?
| Capability | Live Agent / Chat (legacy) | MIAW (current) |
|---|---|---|
| Status | Retirement announced | Actively developed |
| Conversation persistence | Session-based | Persistent / asynchronous supported |
| Deployment | Embedded Service chat | Embedded Service deployment for Messaging |
| Routing | Omni-Channel | Omni-Channel |
| AI and bots | Einstein Bots | Einstein Bots and Agentforce |
| Other channels | Separate setup | Shared platform (SMS, WhatsApp, social) |
Why migrate from Live Agent to MIAW?
- It is the supported path. Investing in a retiring product means rework later; MIAW is where Salesforce is adding features.
- Persistent conversations improve customer experience and let agents resume threads instead of starting over.
- One platform for many channels — web chat, in-app, SMS, and social — reduces fragmented, duplicated setups.
- A stronger AI story with Einstein Bots and Agentforce for deflection, greetings, and assisted responses.
How do you set up live chat with MIAW?
At a high level, standing up Messaging for In-App and Web in Service Cloud looks like this:
- Enable Messaging Settings in Setup, then turn on Messaging for In-App and Web.
- Create a Messaging Channel for your web or in-app deployment.
- Configure an Embedded Service deployment to generate the code snippet you place on your site.
- Set up routing with Omni-Channel — presence statuses, routing configurations, and queues — so conversations reach the right agents.
- Add automation with an Einstein Bot or an Agentforce agent for greetings, FAQs, and deflection before handing off to a human.
- Deploy the snippet to your website and test the full flow end to end.
The deployment produces an embedded messaging bootstrap script similar to the sample below, which you add to your site's pages.
<!-- Embedded Messaging for Web bootstrap (sample - replace placeholders) -->
<script type="text/javascript">
function initEmbeddedMessaging() {
try {
embeddedservice_bootstrap.settings.language = 'en_US';
embeddedservice_bootstrap.init(
'YOUR_ORG_ID',
'YOUR_DEPLOYMENT_DEVELOPER_NAME',
'https://your-domain.my.site.com/ESWyourdeployment',
{ scrt2URL: 'https://your-domain.my.salesforce-scrt.com' }
);
} catch (err) {
console.error('Error loading Embedded Messaging: ', err);
}
}
</script>
<script
type="text/javascript"
src="https://your-domain.my.site.com/ESWyourdeployment/assets/js/bootstrap.min.js"
onload="initEmbeddedMessaging()">
</script>Where do Einstein Bots and Agentforce fit?
For automated, AI-assisted chat, Einstein Bots handle structured dialogs — greetings, qualification, FAQ deflection — and hand off to a human agent over MIAW when needed. Agentforce is Salesforce's newer AI agent layer, enabling more autonomous, generative responses grounded in your Salesforce data and knowledge. Both sit on top of MIAW, so you start with the supported messaging channel and add automation incrementally.
How should you plan a Live Agent to MIAW migration?
- Inventory what you have — deployments, chat buttons, routing rules, quick text, and any custom Live Agent code.
- Map routing to Omni-Channel in the MIAW model, including presence and queues.
- Rebuild deployments as Embedded Service for Messaging, then re-add automation with Einstein Bots or Agentforce.
- Run both in parallel during cutover where possible, then retire the legacy Live Agent deployment.
If you are documenting the legacy setup for reference, our Basic Live Agent Implementation walkthrough shows the older Live Agent configuration — useful for understanding what you're migrating away from. For where live chat fits in Salesforce's customer-service product, see Sales Cloud vs Service Cloud. And if you also capture written inquiries through forms, Salesforce Web-to-Case pairs well with messaging to route requests into cases.
Need help moving off Live Agent without disrupting your support team? Our Salesforce consulting services cover Service Cloud, MIAW, and Agentforce migrations.
Frequently Asked Questions
Is Salesforce Live Agent being retired?
Yes. Live Agent was rebranded to Chat, and Salesforce has announced the retirement of Chat — the Live Agent technology. Salesforce directs customers to Messaging for In-App and Web (MIAW) as the replacement. Because the exact end-of-life date can vary by release and edition, confirm it in Salesforce's current official retirement schedule rather than relying on a date quoted elsewhere.
What replaced Salesforce Live Agent?
Messaging for In-App and Web (MIAW) is the current standard for live chat on your website and in mobile apps. It runs on Salesforce's broader messaging platform alongside channels like SMS and WhatsApp, and pairs with Einstein Bots or Agentforce for AI-assisted conversations.
What is the difference between Chat and MIAW?
Chat is the rebranded name for the legacy Live Agent engine and is largely session-based. MIAW is the modern messaging framework that supports persistent, asynchronous conversations, shares a platform with other messaging channels, and is where Salesforce continues to add features. New live chat should be built on MIAW.
Can I still use Live Agent in my existing org?
If your org already has Live Agent (Chat) enabled, it may continue to work for now, but it is a legacy, end-of-life product. You should avoid new investment in it and plan a phased migration to MIAW. Always check Salesforce's current retirement timeline for the cutoff that applies to your edition.
Do I need Einstein Bots or Agentforce to use MIAW?
No. You can deploy MIAW with human agents only and add automation later. Einstein Bots and Agentforce are optional layers on top of MIAW that handle greetings, FAQ deflection, and AI-assisted responses, then hand off to a human agent when the conversation needs one.
What drives the effort of a Live Agent to MIAW migration?
The main factors are the number of deployments and chat buttons, the complexity of your Omni-Channel routing, the volume of custom Live Agent code or quick text to recreate, and whether you are also adding Einstein Bots or Agentforce automation. A simple single-deployment migration is far quicker than a multi-channel rebuild with custom automation.