Skip to content
    Get in touch

    AI lead qualification applies the same criteria to every enquiry, scores it and routes it to a next step. This post covers the criteria, the routing, the handoff to staff, and the rules on human oversight and personal data.

    Response speed

    The Harvard Business Review analysis covered 1.25 million sales leads received by 29 B2C and 13 B2B companies in the US. It defined qualifying a lead as having a meaningful conversation with a key decision maker.

    Companies that tried to contact a lead within an hour were nearly seven times as likely to qualify it as companies that tried an hour later. They were more than 60 times as likely as companies that waited 24 hours or longer. The study dates from 2011.

    Qualification criteria

    A lead score combines signals that show how well an enquiry matches the business. The owner sets the criteria for each service.

    • Stated intent: a booking request or a price enquiry only.
    • Service match: whether the business offers the service asked for.
    • Location fit against the service area.
    • Timeline.
    • Budget alignment.
    • Completeness of the details given.

    A new system can start with explicit rules defined by the owner. The rules can be refined later using the business's own enquiry history.

    Channel flows

    The qualification flow follows the channel. A chat channel such as WhatsApp or web chat collects requirements through a few short questions in sequence. A form already holds fixed fields, so the flow asks only for missing details. Enquiries from every channel feed one inbox with one prioritised list.

    Routing thresholds

    Routing by confidence.
    ConfidenceNext step
    HighRouted to booking
    MediumA follow-up question, or a call from staff
    LowAn informational reply, flagged for monitoring

    Human handoff

    The system hands the conversation to staff on defined triggers.

    • Job scope above a set value.
    • Custom or complex requirements.
    • An upset or frustrated tone.
    • A request to speak to a person.
    • A score that further questions cannot resolve.

    The handoff carries the full conversation, the extracted lead data, the score, the trigger reason and a suggested next action. Staff continue without re-reading the thread.

    The WhatsApp Business Messaging Policy requires automation to have prompt, clear and direct escalation paths. It lists in-chat human agent transfer, a phone number, an email address, web support, in-store visits and a support form.

    Human oversight

    The National Guidelines on AI Governance & Ethics, published by the Ministry of Science, Technology and Innovation in 2024, define three oversight arrangements.

    • Human-in-the-loop: the capability for human intervention in every decision cycle of the AI system.
    • Human-on-the-loop: the capability for human intervention during the design cycle and for monitoring the system's operation.
    • Human-in-command: the capability to oversee the overall activity of the AI system and to decide when and how to use it.

    Personal data

    Scoring an enquiry processes personal data. Section 6(1)(a) of the Personal Data Protection Act 2010 requires the data subject's consent unless an exception in section 6(2) applies. One exception covers processing necessary for steps taken at the data subject's request with a view to entering into a contract.

    Section 7(1) requires a data user to inform the data subject by written notice that personal data is being processed and for what purposes. The reprint cited is as at 1 July 2023.

    What OtomoAI builds

    OtomoAI's qualification layer routes enquiries across the channels a client connects. High-priority leads reach the team. Lower-priority enquiries are handled automatically until a booking decision or a natural close.

    Get in touch and we quote for your case.

    Sources

    Ritchie Boon Win Yew ·