Asks BANT-style questions (budget, authority, need, timeline), scores each answer, and routes hot leads to a calendar booking while parking the rest in a nurture list. Qualification happens in conversation, so sales only ever sees leads that survived it.
Greets every visitor, answers company and product questions from a knowledge base, and hands serious enquiries to a human with full context. Replaces the contact form for the 80% of questions that never needed one.
Catches shoppers abandoning a cart, asks what stopped them, answers the objection - shipping cost, sizing doubt, payment worry - and applies an incentive only when the objection warrants one. It recovers carts a discount email cannot, because it learns the reason first.
Interviews the shopper - use case, budget, constraints - then narrows the catalogue to two or three picks with reasons. The in-store assistant pattern: choosing from three beats scrolling through three hundred.
Roughly 40% of website traffic arrives outside business hours. This bot answers the FAQs, files a ticket for anything it cannot resolve, and commits to a morning follow-up - the overnight queue arrives triaged instead of raw.
Detects the visitor's language, answers in it, and escalates to an agent with a translated transcript - multilingual coverage without multilingual headcount, the single most requested support capability we see.
Shows the live schedule, checks capacity, books the slot and sends the reminder. Swap 'class' for 'table', 'court' or 'consultation' and this is the pattern behind every appointment bot on the internet.
Qualifies the job - what, where, how urgent - offers real time slots, confirms the technician visit and chases the reminder. For trades this one bot is routinely the difference between answered and missed demand.
Answers treatment, pricing and insurance questions and books appointments - the front-desk load - while the conversation stays on your own widget rather than a third-party messenger. The same skeleton serves clinics, dentists and physio practices.
Collects requirements - beds, budget, area, move-in date - matches them against listings and schedules viewings. Agents get a qualified viewing calendar instead of a voicemail box.
Explains course options, checks prerequisites, answers fee and schedule questions and walks the student to a completed enrolment - education teams use it to stop losing applicants between interest and paperwork.
Answers the benefits questions every HR inbox answers weekly - coverage, enrolment windows, claims - from one maintained source. The internal-bot pattern: same builder, private audience.
Strip the industry paint off any chatbot on this page - or any chatbot anywhere - and one of five conversational patterns is underneath. Knowing which one you need shortens every build.
| Pattern | Job | Success metric | Where to start |
|---|---|---|---|
| Qualify | Turn anonymous traffic into scored leads | Qualified leads per 100 visitors | Lead generation guide |
| Deflect | Answer repeat questions without a human | Resolution rate (aim 40-60% at launch) | Support chatbot guide |
| Book | Convert interest into a calendar slot | Bookings per conversation | Calendar booking |
| Recommend | Narrow a catalogue to a confident pick | Assisted conversion rate | Ecommerce templates |
| Escalate | Hand hard cases to a human with context | Handover rate and CSAT after handover | Live chat handoff |
1. Open the closest example and click Use Template. The whole flow copies into a drag-and-drop visual builder - every question, branch and handover in the preview is a node you can edit.
2. Give it your knowledge. Point the AI knowledge base at your site or documents so the bot answers your questions, not the template's placeholders.
3. Put it where your customers are. The website chatbot is one script tag; WhatsApp, Telegram, Messenger and Slack connect from the dashboard. The free plan covers 600 conversations a month while you tune it - see the 2026 benchmarks for what resolution rate to expect in month one.
For context on what the patterns above look like at scale - widely documented deployments, described here, not affiliated with Conferbot.
A chatbot example is a working demonstration of a conversational flow built for a specific job - qualifying a lead, recovering an abandoned cart, booking an appointment, answering support questions. Every example on this page is rendered from its template's real flow data and opens as a live template you can interact with and copy, rather than a screenshot of someone else's bot.
Five patterns cover most deployments: lead generation and qualification bots, customer support and FAQ bots, appointment booking bots, ecommerce assistants (recommendations, order status, cart recovery), and internal bots for HR or IT questions. The gallery above has a live example of each.
Frequently cited production examples include Bank of America's Erica for in-app banking assistance, Sephora for product matching and booking, H&M for order and returns support, Klarna for AI-first support at scale, and Duolingo for conversational practice as a product feature.
Yes. Each example links to a live template with a preview - open it, click through the conversation, then use it as the starting point for your own bot on the free plan (600 conversations a month, no card required).
Three things: a single clear job (book, qualify, deflect - not all three at once), an escape hatch to a human at every step, and a first message that states what the bot can do. Every template in this gallery follows all three; most bad chatbots in the wild fail the first one.
Industry benchmarks put newly deployed AI support bots at a 40-60% resolution rate, climbing past 60-70% after 6-12 months of tuning the knowledge base. Treat any example or vendor claiming 90% on day one with suspicion - the honest number grows with iteration.
Yes - every template here deploys to the website widget and to WhatsApp, Telegram, Messenger, Instagram, Slack, Discord, LINE and Teams from the same flow. Build once, connect the channels your customers already use.
Open any example, click Use Template, and edit the flow in a drag-and-drop builder - no code. Connecting it to a website is one script tag; Telegram, WhatsApp, Slack and other channels connect from the dashboard.