Help & Support

After your account is created, go to Projects and create your first project. We then connect your lead sources to it for you — each source gets its own secure intake URL. Configure your system prompt and scoring rules, and you're live.

A Project is a top-level business context — think of it as a campaign or client. Each project has its own scoring rules, handoff threshold, system prompt, and AI settings. If you run an agency, you'd create one project per client.

A Tube is a lead intake channel within a project. Each tube has its own webhook URL, field mapping, and optional prompt overrides. A project can have multiple tubes — for example, one tube per ad campaign, or separate tubes for different lead sources.

Every time the AI responds to a lead, it assigns a score from 0-100 based on your configured scoring rules and the conversation context. Scores only go up (ratchet) — they never decrease. When a lead's score crosses your handoff threshold, the system triggers a handoff recommendation.

A handoff is when a lead is qualified enough to be connected with a human sales agent. When the AI scores a lead above your threshold, it creates a handoff recommendation. You can review the full conversation transcript and confirm the handoff, which notifies your sales agent via email.

Nurture is a holding status for leads who aren't ready to buy yet but haven't disqualified themselves. The AI will send periodic follow-up messages on a schedule you configure (e.g., 24h, 72h, 1 week) to re-engage these leads.

Lead Tube recognizes common opt-out keywords, including STOP, CANCEL, and QUIT, and records suppression so automated follow-up is withheld. Operators remain responsible for consent and the applicable messaging requirements.

The system prompt tells the AI who it is, what it's trying to accomplish, and how it should behave. Think of it as the AI's personality and instructions — its role, tone, industry knowledge, and goals. You set this at the project level and can override it per tube.

Flow Purpose is a separate field from the system prompt that defines the business objective of your messaging flow — for example, "qualify leads for AI workflow consulting" or "book discovery calls." It's rendered as a dedicated prompt block so the AI can orient around outcomes.

The Library is where you upload FAQs, product information, pricing details, objection handlers, and other reference material. The AI retrieves relevant library entries during conversations so its responses are accurate and on-brand.

In Auto mode, the AI sends responses immediately. In Manual Review mode, every AI-generated message goes into a pending queue where you can review, edit, and approve before it's sent. This is useful when you're first setting up or for sensitive campaigns.

Flow Templates let you design multi-step conversation sequences. Each step has conditions (like keyword matches or score ranges) and a prescribed response or action. The AI follows these templates while still adapting naturally to each lead.

Your operator configures the messaging provider and sending number for your project. Lead Tube supports different provider paths, so contact support for the setup that applies to your account.

A2P (Application-to-Person) 10DLC is a registration process for business messaging over US local phone numbers. Registration and approval are handled through the applicable provider. Contact support about the requirements for your sending setup.

The greeting delay adds a configurable pause (in seconds) between when a lead arrives and when the first AI message is sent. This prevents the message from arriving suspiciously fast after the lead submits a form. Set to 0 for immediate response.

Analytics shows lead volume over time, conversion rates, average scores, handoff counts, response times, and AI performance metrics. You can filter by date range and project.

QA Review lets you audit AI conversations. You can review individual AI responses, rate them with thumbs up/down, and provide feedback that helps improve the AI's performance over time through the AI Learning system.

AI Learning shows the feedback loop from QA reviews. When you rate AI responses, the system learns patterns about what works well and what doesn't, helping to continuously improve conversation quality.

Yes. Each tube has a generic webhook URL that accepts JSON payloads. You can POST leads from any source — CRMs, landing pages, other ad platforms — as long as the payload includes at least a phone number or an email address. Use the field mapping in the tube settings to map your payload fields.

If a webhook delivery fails (network issue, server restart, etc.), Webhook Replay lets you resend the original payload. This is useful for recovering leads that were lost during a brief outage.

Yes. We use encrypted connections (TLS), httpOnly session cookies, parameterized queries to prevent SQL injection, role-based access controls, and tenant isolation to keep your data secure. See our Privacy Policy for full details.