A person visits a counselling practice's website late at night and types: "I don't know if I can keep going." That message is not a booking enquiry. It needs an immediate, careful response that points to real help. The same website will also get messages like "How much is a session?" and "Do you offer evening appointments?" An assistant must handle both correctly.
Therapy practice websites deal with people at vulnerable moments. Here are the myths about what an assistant should do.
Myth 1: an assistant can offer some support while people wait
Reality: an assistant is not a therapist and must not act like one. Offering advice, coping techniques or emotional reassurance as if it were therapy is risky. It should help people reach a real person.
Myth 2: crisis messages are rare, so they can be handled later
Reality: they may be rare, but they matter most. Any message suggesting danger, self-harm or crisis must be met immediately with direction to emergency services or a crisis line. This must be tested before launch.
Visitor: I don't know if I can keep going. Assistant: I'm really sorry you're feeling this way. If you're in danger or thinking about harming yourself, please contact emergency services now or call a crisis line. The practice's page lists crisis contacts you can reach at any time. If you'd like, you can also leave your name and a way to reach you, and a therapist will get in touch.
Myth 3: practical questions are not worth answering carefully
Reality: practical questions are often how people take the first step. Fees, availability and what happens in a first session are the questions that decide whether someone books. First contact is covered in therapy enquiries from people who find it hard to ask.
Myth 4: asking for details upfront helps therapists prepare
Reality: a long intake form in the chat window can feel like an exam. The therapist will gather what they need in a proper conversation. Online, the lighter the request, the more likely someone is to send it.
Myth 5: tone matters less than accuracy
Reality: both matter. A cold or clinical reply can push someone away. Warm, plain language helps. Good practice for any assistant is covered in chatbot best practices every website owner should know.
Myth 6: an assistant can decide who needs urgent help
Reality: it cannot assess anyone. It should treat any mention of danger as a reason to direct the person to urgent help. What clinical assistants must refuse is covered in what a clinic assistant must refuse to discuss, and what no assistant should promise is covered in what an AI receptionist must never promise.
Privacy
Visitors often wonder whether a receptionist, a therapist or nobody at all will read what they type. Saying plainly who sees messages, and how long they are kept, lets people decide what they are comfortable sharing.
Review and testing
Before going live, write a handful of messages that hint at risk in different ways, some direct and some vague, and check each one. Then keep reading a sample of real conversations every week, because people rarely phrase things the way a test script does.
How SpideyChat handles it
SpideyChat for therapy and counselling practices replies using only what your practice has published, hands conversations to your team when a person is needed and points anyone at risk towards urgent help. It never offers therapy or a view on someone's mental health. The demo shows a handover in action, and plans are listed on the pricing page.
Getting started
Publish clear fees, availability, therapist profiles and crisis contacts, and test the assistant with difficult messages before going live.