Golden Horizons
CAPABILITY · SALES & LEAD-GEN
FAQ Chatbot
Site chatbot grounded in your own content. Cites sources. Pings hot leads instantly.
$5,500 build · $1,500–3,500/mo
Talk to us about a FAQ Chatbot build →What it does
Site chatbot grounded in your own content. Cites its sources. Captures intent and pings hot leads to Slack or SMS within seconds.
Every front desk has the same twenty questions. What are your hours? Do you accept my insurance? How much does the procedure cost? Can I bring someone with me? What should I avoid beforehand? The questions aren't hard. But they come in on the phone while someone's with a patient, over email at 9pm, through the website contact form on a Sunday. Someone has to answer them. Usually that someone is the most expensive possible option — a staff member pulling their attention away from a patient or a client who is physically in the room.
The FAQ chatbot is trained on your actual content: your service descriptions, pricing pages, policies, intake forms, FAQs, and any other documents you hand over. It doesn't guess. When someone asks a question the chatbot can answer from your material, it answers and cites the source — so the patient or client knows exactly where the information came from and can verify it. When a question is outside the scope of what you've documented, the chatbot says so and routes the person to a human instead of fabricating a response.
After the initial training session, Golden Horizons wires the chatbot into your website as a widget. SMS and email routing are available if your team handles questions through those channels. Hot-lead signals — a visitor asking about pricing, availability, or a specific procedure by name — trigger an immediate Slack or SMS ping to whoever on your team handles new business. The person who was just browsing your site at 11pm gets a response before they've opened a competitor's tab.
The owner side is a plain dashboard: every conversation logged, every question flagged as unanswered, and a simple interface to update answers when your policies change. No engineering ticket required to edit a response. If you raise your prices in March, you update the document, the chatbot reflects it. The build is two to three weeks from kickoff to live widget on your site. The retainer covers prompt tuning when your service mix shifts, integration upkeep, and answer reviews when the dashboard flags gaps.
Use cases
- A med-spa stops losing evening inquiries: the chatbot handles pricing questions for injectables and laser treatments after hours, flags contraindication questions as needing a licensed consult, and pings the front desk via SMS when a visitor asks about booking a specific treatment package.
- A family restaurant with common allergen questions trains the chatbot on its current menu and ingredient list. Diners get accurate answers about gluten, nuts, and dairy before they arrive, reducing mid-service interruptions for the kitchen and floor staff.
- A residential real estate team uses the chatbot to field showing-availability questions on listing pages. The bot checks the agent's published schedule and routes confirmed showing requests directly to the agent's calendar, capturing leads that used to slip through weekend contact forms.
- An HVAC company fields service-area and emergency-call questions through the chatbot on nights and weekends. Customers in the coverage zone get a call-back confirmation. Customers outside the area get a clear answer instead of a voicemail that never gets returned.
- A veterinary clinic answers vaccination schedule, flea-prevention protocol, and boarding policy questions through the chatbot so front desk staff can focus on check-ins and calls that require clinical judgment.
- An independent insurance broker trains the chatbot on the product lines it carries. Prospective clients asking about coverage types and general eligibility get factual answers drawn from carrier documentation, with a clear handoff to the broker for anything requiring a personalized quote.
What’s included
- Fixed scope with written acceptance criteria before any build starts
- Customization layer for your brand voice and business rules
- Clean handover with documented runbook and live training
- Monthly ROI report for three months post-delivery
- Source code delivered to your GitHub on handover
What’s NOT included
- Third-party API subscription costs (billed to your accounts)
- Data migration from legacy systems
- Ongoing infrastructure costs after handover
Retainer
Monthly retainer covers monitoring, prompt tuning, config refinement, and minor integration additions. Range: $1,500–3,500/mo.
How clients use this
Fixed-scope build with clean handover, then an optional monthly retainer covering maintenance, monitoring, and minor changes. Most clients move to retainer within 60 days of delivery.
Part of
Used in: Law Firms , real-estate-agents , Dental Practices , restaurants
Questions FAQ Chatbot clients ask
How do you train the chatbot on our specific policies and content?
You share the documents you already have — your existing FAQ page, service descriptions, pricing PDFs, intake forms, policy documents, whatever you've written down. We run those through a structured ingestion process that breaks them into retrievable chunks and indexes them so the chatbot can pull the right passage when a question comes in. If you don't have things written down yet, we run a one-hour intake session where we ask you the questions your customers ask most often and build the document together. The chatbot is never trained on generic internet content about your industry — only what you've explicitly provided. That's what keeps it accurate and on-brand instead of generic.
What stops the chatbot from making something up if it doesn't know the answer?
The architecture is retrieval-first: before the chatbot generates any response, it checks your indexed content for a matching passage. If it finds one, it answers from that passage and cites the source. If it doesn't find a strong match, it's configured to say so explicitly and route the person to a human — a contact form, a phone number, or a calendar link depending on what makes sense for your operation. It does not speculate, it does not pull from general internet knowledge about your industry, and it does not invent policies you haven't documented. Every answer is traceable back to a specific document you provided. The dashboard logs every conversation, so you can see exactly what the chatbot said and flag anything that needs a correction.
What happens when a visitor asks something that needs a human — how does the handoff work?
Two triggers. First, unanswerable questions: anything outside your indexed content gets a clear 'I don't have that information — here's how to reach us' response with your preferred contact method. Second, hot-lead signals: if a visitor's questions indicate purchase intent — asking about availability, specific service pricing, booking timelines — the chatbot completes the answer and simultaneously pings your team via Slack or SMS with a transcript of the conversation. Your team gets context before they pick up the phone. The handoff point is configurable. Some operators want every unanswered question routed to a form; others want a phone number. Some want lead pings for any pricing question; others only for certain services. We set it up based on how your team actually works.
How much staff time does this actually save, and how long until we see it?
That depends on your current volume of repetitive inbound questions and how consistently your team is answering them now. For most small service businesses fielding fifteen to thirty FAQ-type contacts per week across phone, email, and chat, the chatbot handles the majority of those without staff involvement — typically within the first two weeks of go-live, once the initial training content is dialed in. The clearer your existing documentation, the faster the chatbot gets accurate. The after-hours capture is usually the most immediate win: questions that were previously going unanswered until the next business day get a response within seconds. What we won't do is quote a specific percentage reduction or an exact number of hours saved before we've seen your actual volume — that kind of upfront precision is a sales pitch, not an honest projection. The $99 audit maps your current contact patterns first, and the build scope comes out of that picture.