Used For

• +**Your agents answer the same 20 questions 200 times per day** → AI handles routine stuff like password resets and billing questions automatically
• +**Customers wait 4 hours for basic subscription changes** → Connects to your systems to actually update accounts, not just provide links
• +**Support costs are eating your margins alive** → Deflects 60-80% of tickets so humans only handle complex issues
• +Simulates results on your historical data before you buy - shows exactly which tickets it would've caught
• +No-code builder means support managers build flows without developers

Best For

• >Your support team is drowning in 500+ chats per day and humans can't keep up
• >You're bleeding money on support staff but customers still wait hours for responses
• >Enterprise with existing Zendesk or Salesforce that needs serious automation, not band-aids

Not For

• -Small teams handling under 100 tickets per day — you're paying enterprise prices for a fire hose when you need a garden sprinkler
• -Companies wanting plug-and-play simplicity — this requires dedicated setup time and ongoing AI coaching
• -Anyone on a tight budget — voice and social channels cost extra, plus integration work adds up fast

Pairs With

• *Zendesk (where tickets actually live and get routed after Ada handles the easy stuff)
• *Salesforce (to update customer records when AI resolves billing or account issues)
• *Slack (where your team gets alerts when AI can't handle something and needs human backup)
• *Stripe (for subscription changes and billing updates that customers request through chat)
• *Intercom (if you're switching from their basic chatbot to something that actually resolves issues)
• *Tableau (to build executive dashboards showing ticket deflection rates and cost savings)

The Catch

• !Voice and social media support are billed separately from the base plan - the sticker price is just chat and email
• !Custom integrations and handoff rules require their Solutions team, not self-service setup
• !You'll spend ongoing time reviewing failed chats to coach the AI - it's not set-and-forget automation

Bottom Line

The heavy-duty chatbot for enterprises with massive support volumes - overkill for most, but actually works at scale.