What the August 2025 Outage Actually Taught Us
Ten days.
That's how long Drift was offline following the August 2025 data breach. Every visitor who hit the chat bubble during that window received silence. No automatic rerouting to a backup form. No fallback message. No proactive communication from Salesloft explaining what happened or when service would return.
For teams that had built their inbound lead capture around that chat bubble, ten days was not an abstract inconvenience. It was ten days of visitors arriving on high-intent pages and leaving without a single conversation. No meeting booked. No contact created. No signal sent to the CRM.
G2 reviewers who lived through it were direct: customer service response was inadequate. There was no plan in place for a critical infrastructure failure on a system they treated as mission-critical.
The lesson isn't that Drift had an outage. Every platform has outages. The lesson is what the response revealed: Drift's infrastructure no longer holds the same priority inside Salesloft's roadmap that it held when your team chose it. That's useful information for a risk assessment.
What Lives in Drift That You Haven't Backed Up
Before you can protect your pipeline, you need to know what your pipeline is actually built on.
Most teams significantly underestimate how much lives inside Drift in a format that doesn't transfer cleanly. A Drift shutdown doesn't just turn off a chat widget. It takes several layers of data and logic with it.
Playbook and bot flow logic. Drift's playbooks live in a proprietary format. The conditional logic, the routing rules, the qualification branches your team built over months or years: none of it exports to a portable format that other platforms can import. If Drift shuts down before you document those flows, you rebuild from memory.
Conversation history and enrichment data. Every lead conversation, every enriched contact record tied to Drift's identity layer, the context behind how each contact entered your funnel. Some of this overlaps with your CRM. Some of it doesn't.
Meeting routing rules and calendar integrations. The rules that send a booked meeting to the right rep, the calendar connections, the round-robin logic: these are configurations that live in Drift and need to be mapped before you can recreate them elsewhere.
CQL scoring and definitions. If your team built Conversation Qualified Lead definitions inside Drift, those scoring models need to be documented. They represent strategic thinking about what constitutes a qualified conversation.
Contact and company data tied to Drift's enrichment layer. Drift enriches visitor data through external sources. Some of that enrichment feeds your CRM directly. Some of it lives only in Drift's records. Know which contacts in your CRM came through Drift and how much of their enrichment data is replicated elsewhere.
What Breaks First When Drift Goes Away
If Drift shuts down without a planned migration, the failures don't all happen at once. They happen in a sequence.
Inbound chat goes dark immediately. The most visible failure. Every visitor who arrives at a high-intent page and looks for the chat bubble gets nothing. No conversation. No qualification. No meeting booked.
Booked meetings decline. Reps lose the calendar booking layer. Visitors who would have booked through a Drift playbook now need to find a form or navigate to a scheduling page directly. Conversion drops.
Lead routing to HubSpot or Salesforce breaks. If Drift served as the intermediary between visitor intent and your CRM record creation, that pipeline breaks. Leads that would have been created and routed through Drift simply don't appear.
Attribution gaps widen. Contacts that entered your funnel through Drift conversations become harder to track. The channel attribution that showed chat as a lead source disappears or becomes unreliable.
Any Drift-dependent nurture flows stop. If your marketing automation includes triggers tied to Drift conversation events, those triggers stop firing. Contacts that should enter nurture sequences don't.
The compounding effect is what makes unplanned Drift shutdown genuinely damaging. Each failure individually is manageable. Together, they create a gap that takes weeks to diagnose and longer to recover from.
The 48-Hour Pipeline Protection Checklist
If you're reading this with urgency, start here. These actions protect your pipeline before any migration happens.
Export conversation history now. Go to your Drift account and pull your full conversation history. Do this before anything else.
Document every playbook flow. Screenshot each playbook in your account. Capture the conditional logic, the routing rules, the messages at each step. Drift's format won't import elsewhere, but the logic can be rebuilt if it's documented.
Audit your CRM for Drift-sourced contacts. Identify which contacts in HubSpot or Salesforce entered through Drift conversations versus forms. Confirm their enrichment data exists in your CRM independently of Drift.
Map your calendar and routing rules to a spreadsheet. List every routing rule, every rep assignment, every round-robin configuration. You'll need this to recreate routing in your next platform.
Identify your top three integrations and confirm each has a direct alternative. Most Drift integrations connect to platforms you already own. Verify those connections can be rebuilt without Drift as the intermediary.
Put a temporary backup form on your highest-traffic pages today. Before you have a new chatbot live, a simple form on your top landing pages ensures no lead goes completely uncaptured during the transition window.
What You Actually Need Running While You Migrate
The backup form is a bridge. Your actual replacement needs to cover more ground.
Visitors arriving during your migration window need a real conversation experience, not a static form. The capabilities that matter most during the transition period are the ones Drift was supposed to provide: real-time capture, routing to the right person, and CRM record creation without manual intervention.
Your replacement needs to go live within hours, not weeks. It needs to connect to HubSpot or Salesforce with field mapping that matches your existing structure. Routing rules need to follow your territory model, not require a full reconfiguration. And it needs to handle the variation in how visitors phrase questions, not break when someone asks the same question twelve different ways.
The platform that goes live fastest without engineering involvement wins the transition.
How Expertise AI Covers the Gap Immediately
Expertise AI was built for exactly this situation: teams that need to be live now, not after a 6-week implementation project.
The setup speed is documented by the teams that ran it. Imagine AI Live added Expertise AI to two marketing websites, configured custom Q&A, and was capturing leads within 30 minutes. No developer. No implementation contract. No waiting.
The AI handles the kind of question variation that breaks scripted playbooks. Swimmingly found that visitors asked the same question in as many as fifteen different ways. Their previous platform failed on the variation. After switching to Expertise AI, automated resolution climbed from 70-75% to 90-95% within 30 days. The AI processes what a visitor actually means, not whether their phrasing matches a preset path.
The HubSpot integration maps directly to your existing fields. First name, last name, and custom fields like "Student type" write automatically to HubSpot without manual cleanup. No data wrangling. No field-mapping project.
And for teams that need to know who is visiting, not just who converts: using IP-level enrichment, Expertise AI surfaces company-level buyer intent even when visitors don't fill out a form. One customer identified visiting companies from 369 separate accounts that would have otherwise remained invisible.
GIC-GREEN came in with a fragmented setup. Reps missed conversations. Response times lagged. After switching, Expertise AI became their largest lead volume channel, with the highest lead value and strongest close rate of any source in their pipeline.
What Migration Actually Looks Like
The migration itself is not the hard part. The documentation before it is.
Show up to your Expertise AI setup call with your Drift flows documented: the logic maps, routing rules, and qualification criteria. The Expertise AI team will work with you to document and transfer everything into the new setup, so you don’t have to rebuild from scratch. You hand off what you already built, and they carry it forward.
Most teams are live within the hour. HubSpot or Salesforce connects in the same session. Routing rules go in directly. The AI trains against your existing website content, so it answers questions about your product from day one.
You can run Expertise AI alongside Drift during the transition window. Both chat layers stay live until you're confident the new one is performing. Then you remove Drift. No gap in coverage.
The work you did to build your Drift setup was real work. The routing logic, the playbook thinking, the qualification criteria. That work doesn't disappear in the migration. It moves forward.
We'll get you live in under an hour. Book a setup call and we'll migrate your core Drift flows for free.
Your pipeline doesn't need to wait for a formal shutdown notice to get protected.