Quick Start Guide

This guide walks you through the complete process from signing up to having a live chatbot on your website. Most users complete this in under 60 minutes.

Before You Begin

You'll need the following ready:

• An active NexChatBot account (sign up at NexChatBot.com)
• Access to your website's admin panel or CMS to paste a small code snippet
• Your business content — website URL, FAQs, or documents (PDFs, Word files)

Step-by-Step Setup

<script
  src="https://cdn.nexchatbot.com/widget.js"
  data-bot-id="YOUR_BOT_ID"
  async
></script>

Dashboard Overview

When you log into NexChatBot, the first thing you see is the main dashboard. This is your command centre — it lists all your AI agents and gives you access to account-level settings.

What You'll See on the Dashboard

The main dashboard has three core areas:

• AI Agents list — every chatbot you've created is shown as a card. Click any card to open that chatbot's individual dashboard where you can manage it fully.
• Create New Agent button — starts the process of building a new chatbot from scratch.
• Account & Settings menu — accessible from the top-right or side navigation, this is where you manage billing, your profile, and agency-level settings.

Navigating the Dashboard

Inside each individual chatbot, you'll see a set of tabs across the top or in a side menu. Here is what each tab does:

Key Actions Available on the Dashboard

• Share Chatbot — generates a public URL where anyone can test your chatbot in a standalone page, without it being embedded on a site. Useful for sharing a demo with a client or colleague.
• Clone Chatbot — creates an exact copy of your chatbot including all its knowledge base content, settings, and prompt. Useful when building a similar bot for a different client without starting from scratch.
• Set as Default — marks one chatbot as the default for your account.
• Chatbot UUID — a unique identifier for your bot used in API calls and embed codes.
• Delete Chatbot — permanently removes the chatbot and all its data. This cannot be undone.

Creating a Chatbot

Creating a new AI agent takes less than a minute. Here's exactly what happens during the setup process and what each field means.

Starting a New Chatbot

From your main dashboard, click "Create New AI Agent". You'll be asked to provide:

• Chatbot Name — this is your internal reference label. It does not appear to end users. Name it something clear like "Acme Corp Support Bot" or "Homepage Lead Bot".
• AI Model — choose which underlying language model to use. Each model has different capabilities and costs:

You can change the model at any time from the Settings tab of any chatbot. You are not locked in after creation.

What Happens After Creation

Once created, your new chatbot will appear in your dashboard list. At this point it has no knowledge of your business yet — it's a blank slate. Your next steps are:

1. Go to the Knowledge Base tab and add your business content
2. Go to the Prompt tab and write your instructions
3. Go to the Appearance tab and brand the widget
4. Test it in the live preview panel on the Dashboard tab
5. Go to the Integrations tab and get your embed code

AI Agent Dashboard

When you click into a specific chatbot from your main dashboard, you enter the AI Agent Dashboard. This is where you manage every aspect of that individual chatbot.

Dashboard Layout

The AI Agent Dashboard is split into two main areas:

• Left side — Live chat preview window. This is a real-time testing environment. You can type messages and see exactly how your bot will respond to visitors. If the bot gives a wrong or incomplete answer, you can update the knowledge base immediately from here and test again. Think of this as your quality-check panel.
• Right side — Management panel. This shows tabs for all the settings, content, and configuration options for the bot: Knowledge Base, Prompt, Appearance, Leads, Conversations, Scheduling, Integrations, and Summary/Analytics.

Sharing Your Chatbot

The Share Chatbot option generates a unique public URL for your bot — something like nexchatbot.com/chat/abc123. Anyone with this link can open the chatbot in a full-page interface without it needing to be embedded anywhere. This is perfect for:

• Sharing a working demo with a prospective client before you build their website
• Testing the bot from different devices
• Letting team members review the bot's responses before launch

Cloning a Chatbot

Cloning creates a complete copy of the selected chatbot — including all knowledge base content, the system prompt, appearance settings, and integrations configuration. The clone is created as a new, independent chatbot that you can modify without affecting the original.

Common reasons to clone:

• You're building a similar chatbot for a new client and want a starting point
• You want to test changes to a live bot without risking the production version
• A client's chatbot serves as a template for variations (e.g. different languages or product lines)

Chatbot UUID

Every chatbot has a UUID (Universally Unique Identifier) — a unique string like bot_a1b2c3d4e5f6. This is the technical identifier used in embed codes, API calls, and webhook configurations. You can find it by clicking "Chatbot UUID" in the dashboard actions menu.

Summary & Analytics

The Summary view inside the agent dashboard shows high-level performance data for that specific bot:

• Total conversations started
• Messages sent and received
• Leads captured
• Date range filters to compare performance over time

Knowledge Base

The knowledge base is the brain of your chatbot. It contains all the business information the bot uses to answer questions. The quality and completeness of your knowledge base directly determines how well your bot performs.

How the Knowledge Base Works

When you add content to the knowledge base, it is processed and stored in a way that allows the AI to search through it semantically — meaning it finds relevant information based on meaning, not just exact word matching. When a visitor asks a question, the system searches the knowledge base for the most relevant chunks of information, then uses those chunks to generate an accurate, grounded answer.

This is why the chatbot stays on-topic and doesn't make things up — every response it gives is based on what you've put in the knowledge base. If something isn't in there, the bot will tell the visitor it doesn't have that information.

Ways to Add Content

1. Scrape a Website URL

This is the fastest way to get started. You paste your website URL (or any specific page URL) and the system automatically visits the page, reads all the text content, and adds it to the knowledge base. You can:

• Scrape a single page — e.g. just your /pricing or /about page
• Scrape your entire site by starting from your homepage and enabling the sitemap crawler
• Enable 24-hour auto-scraping on any URL so the knowledge base stays up to date automatically when your website content changes
• Scrape a Google Doc URL (the doc must be set to "anyone with the link can view")

2. Upload Documents

Upload files directly from your computer. Supported formats include:

• PDF files (product manuals, brochures, policies, reports)
• Word documents (.docx)
• Plain text files (.txt)

The system extracts all the text from your uploaded file and adds it to the knowledge base. This is ideal for internal documentation, detailed product specs, or any content that isn't on a public-facing webpage.

3. Copy & Paste Text

Type or paste any text directly into the knowledge base editor. Use this for:

• Custom FAQs you've written specifically for the chatbot
• Policies or terms that aren't on your website
• Pricing details or offer information
• Anything that doesn't exist as a file or web page

4. Question & Answer Pairs

Q&A pairs are the most precise way to control responses to specific questions. You write the exact question and the exact answer you want the bot to give. When a visitor asks something that closely matches your question, the bot will use your pre-written answer rather than generating one from general content.

Examples of good Q&A pairs:

Q: What are your opening hours?
A: We're open Monday to Friday, 9am–6pm GMT. We're closed on weekends and bank holidays.

Q: Do you offer a free trial?
A: Yes! All plans include a 14-day free trial with no credit card required. You can start today at NexChatBot.com.

Q: How do I cancel my subscription?
A: You can cancel anytime from your account settings under Billing. There are no cancellation fees and your access continues until the end of your current billing period.

Data Storage (Characters)

Storage in the knowledge base is measured in characters. One character = one letter, number, space, or punctuation mark. As a rough guide:

• A typical web page = approximately 2,000–5,000 characters
• A 10-page PDF = approximately 20,000–50,000 characters
• A detailed FAQ document = approximately 5,000–15,000 characters

Your plan includes a character limit for the total knowledge base across all your chatbots. You can see your current usage in the account settings under Billing Usage. Need more room? Additional storage can be added to your account in 1-million-character blocks, billed through your account wallet.

Tips for a High-Quality Knowledge Base

• ✓
  Be specific and factual. Write "Our standard delivery takes 3–5 business days and costs £4.99" rather than "We offer fast delivery at competitive prices."
• ✓
  Phrase FAQs the way customers actually ask them. "How long does delivery take?" performs better than "Delivery Timeframe" as a knowledge base entry heading.
• ✓
  Keep content current. Outdated pricing or discontinued products will cause the bot to give wrong answers. Set URL auto-sync or review monthly.
• ✓
  Avoid contradictions. If one part of your knowledge base says "we ship to 10 countries" and another says "we ship worldwide", the bot will get confused. Audit for inconsistencies.
• ✓
  Use Q&A pairs for edge cases. For questions you know visitors always ask where you want a precise, controlled answer, use Q&A pairs rather than relying on the bot to infer from general text.
• ✓
  Review conversation logs weekly. Real visitor questions reveal gaps in your knowledge base you never anticipated. Add entries to address them.

System Prompt (Instructions)

The system prompt is the set of instructions that defines who your chatbot is, how it behaves, and what rules it follows. Think of it as the chatbot's job description and operating manual combined.

What Is the System Prompt?

The knowledge base tells the bot what to know. The system prompt tells the bot how to act. Both are necessary for a well-functioning chatbot.

When you first create an AI agent, it has no personality, no idea what your business is, and no rules to follow. The system prompt is where you provide that context. Think of it like briefing a new employee — they might be very capable, but they still need to know what company they work for, who their customers are, and how they should speak to people.

The system prompt defines:

• The chatbot's name, role, and company it represents
• Its personality and tone of voice (friendly, formal, concise, detailed, etc.)
• Guidelines for how it responds (e.g. always answer in bullet points, always keep answers under 3 sentences)
• Guardrails — topics or questions it should decline to answer
• How to handle edge cases (e.g. what to say when it doesn't know the answer)
• Instructions for lead collection — when to ask for name/email and how
• Instructions for scheduling — when to offer to book a call
• Language instructions (respond in the visitor's language, or always respond in English, etc.)

System Prompt Template

Here is a solid starting template you can copy and customise:

## Role & Identity
You are [Bot Name], a helpful AI assistant for [Company Name].
Your job is to answer customer questions accurately and helpfully 
using the information in your knowledge base.

## Tone & Style
- Be friendly, clear, and professional
- Keep responses concise — avoid unnecessary padding
- Use plain language; avoid jargon unless the customer uses it
- Respond in the same language the customer is using

## What You Can Help With
- Answering questions about [Company Name]'s products and services
- Explaining pricing and plans
- Helping customers find the right solution for their needs
- Booking appointments or calls (if applicable)

## What To Do When You Don't Know
- If a question is outside your knowledge base, say:
  "I don't have that information right now, but our team would 
  be happy to help — can I take your email address and have 
  someone get back to you?"
- Never guess or make up information

## Lead Collection
- After answering the customer's first question, ask for their 
  email address so we can follow up if needed
- Only ask once — do not repeatedly prompt for contact details

## Rules
- Do not discuss competitor products
- Do not give medical, legal, or financial advice
- Always stay in character as [Bot Name]
- If asked who built you or what AI you use, say: "I'm [Bot Name], 
  [Company Name]'s AI assistant — I'm not able to share 
  technical details about how I work."

Prompt Writing Tips

• ✓
  Build iteratively. Make one change at a time, then test it in the live preview before making the next change. This way you know exactly what caused any improvement or regression.
• ✓
  Use section headers (##). Breaking your prompt into clearly labelled sections (Role, Tone, Rules, etc.) helps the AI follow the instructions more reliably.
• ✓
  Write in English first. Even if your chatbot serves non-English speakers, write the prompt in English and instruct the bot to translate. This generally produces better results.
• ✓
  Comments are ignored. Lines starting with // are treated as comments and will not be read by the AI. Use these to leave notes for yourself or your team about why certain instructions were added.
• ✓
  Be specific about edge cases. "If someone swears, calmly ask them to rephrase" is more useful than "handle difficult users professionally".
• ✓
  Don't write a novel. Very long prompts can cause the AI to prioritise some instructions over others. Keep the prompt focused on what matters most.

Chat Widget Appearance

The chat widget is the interface your website visitors see and interact with. You have full control over its visual appearance through the Appearance tab inside any chatbot, with a live preview updating in real time as you make changes.

Visual Settings

Chatbot Name & Subheading

Set the name that appears at the top of the chat window — this is what visitors see. You might use your brand name, or something like "Support Bot" or "Ask [Brand]". The subheading appears directly below the name and typically says something like "Online", "Available 24/7", or "Usually replies instantly".

Chatbot Icon (Header Image)

Upload a logo or avatar image that appears in the chat window header. Supported formats: PNG, JPG, JPEG. Recommended size: 32×32 pixels. This could be your company logo or a custom illustrated character.

Chat Avatar Icon

This smaller image appears next to each message the bot sends inside the chat window — like a profile picture next to messages in WhatsApp. It gives the chat a more personal, human feel. You can upload a separate image for this or use the same image as the header icon.

Chat Bubble Icon (Launcher Button)

This is the button that sits on your website before a visitor clicks to open the chat. By default it's a speech bubble icon. You can upload a custom image to replace it — for example, your brand logo, a photo of a team member, or a custom illustration.

Font Size & Font Family

Set the font size (in pixels) for the chat text. You can also choose the font family to match your website's typography.

Chat Input Placeholder

This is the grey hint text that appears inside the message input field before a visitor starts typing. Default is something like "Type a message…". You can customise this to something more inviting: "How can we help you today?" or "Ask me anything about our products…".

Send Message Button

You can upload a custom icon for the send button (the button a visitor clicks to submit their message). Supported formats: PNG, JPG, JPEG. Recommended size: 32×32 pixels. If not customised, a standard arrow icon is used.

Colours

The colour customisation section lets you change virtually every colour in the widget:

• Header background colour — the colour bar at the top of the chat window
• Bot message bubble colour — background of the bot's message bubbles
• User message bubble colour — background of the visitor's messages
• Text colours — both bot and user message text
• Widget background — the main body area of the chat window
• Input field background — the colour of the message input area
• Chat bubble (launcher) colour — the colour of the floating button on your website

Layout & Size

• Widget size — choose Small, Medium, or Large for the overall dimensions of the chat window
• Widget border radius — controls how rounded the corners of the widget are, in pixels. A value of 0 = sharp corners; 12+ = rounded corners
• Widget background preference — choose between a solid colour background or a subtle pattern/gradient
• Widget position — choose where the chat button sits on the page: bottom-right (default) or bottom-left

Pop-Up & Auto-Open Settings

You can configure the widget to automatically open or show a pop-up message after a certain amount of time. This is managed under the Pop-Up sub-section of Appearance:

• Enable pop-up — toggle whether the chat auto-opens after the visitor has been on the page for a set number of seconds
• Delay — set how many seconds the visitor has been on the page before the chat pops up automatically
• Initial messages — pre-set 1–3 messages that appear in the chat bubble before the visitor opens the window, acting like a conversation starter. For example: "👋 Hi there! Have a question about our pricing?" These are not sent to the AI — they're just display messages to encourage engagement.

AI Agent Status Messages

Enable this option to show status text to visitors while the AI is processing a response — for example, "Thinking…" or "Looking that up for you…". This prevents visitors from thinking the chat has frozen when there's a brief processing delay.

Tagline

A short line of text that can appear at the very bottom of the chat widget, below the input field. This is optional and often left blank, but you can use it for things like "Powered by AI" or a brief trust message.

General Settings

The General Settings tab gives you control over the technical behaviour of your chatbot — including which AI model it uses, message rate limits, and other advanced configuration options.

AI Model Selection

You can change the AI model powering your chatbot at any time from the Settings tab. The model determines the quality, speed, and cost of each response. Options typically available include:

Changing the model does not affect your knowledge base or prompt — only the underlying language engine changes. The exact list of models available on your account may vary by plan — check the model dropdown inside the Settings tab for what's currently enabled.

Message Rate Limits

Rate limits control how many messages a single visitor can send within a given time window. This protects your message quota from being consumed by bot traffic or one visitor sending thousands of messages. You can configure:

• Messages per window — the maximum number of messages allowed per user in a time window (e.g. 20 messages)
• Time window — the period over which the limit applies (e.g. per hour, per day)
• Rate limit message — the text the bot shows the visitor when they've hit the limit (e.g. "You've reached the message limit for this session. Please try again later or email us at support@company.com")

Other Settings

• Conversation memory — controls how many previous messages in a conversation the AI remembers when formulating its next response. Higher memory = more context-aware responses, but also higher token usage.
• Auto-delete conversations — optionally auto-delete stored conversation data after a certain number of days for privacy compliance.
• Language settings — configure specific language behaviour if needed. By default, the bot responds in whatever language the visitor writes in.

Lead Capture (Leads 2.0)

The Lead Capture feature lets your chatbot automatically collect contact information from visitors during a conversation — and send that data instantly to your email or CRM via webhook. This is one of the most powerful features for turning chat into business results.

What Is Leads 2.0?

Leads 2.0 is the current lead capture system. It extracts specific pieces of data from a conversation (like name, email, phone number, or any custom field) and sends it to you automatically. Unlike older approaches that required complex custom prompts, Leads 2.0 generates the extraction prompt for you based on the fields you define.

You can capture any data point from a conversation, including:

• Email address
• Full name
• Phone number
• Appointment date/time preference
• Specific interests or needs expressed during the conversation
• Budget range
• Any custom field relevant to your business

When Is Data Extracted?

You choose from three trigger options for when the system extracts and sends the lead data:

1. Immediate — As Soon as All Data Is Collected

The moment the visitor has answered all the questions you've configured, the system instantly extracts the data and sends it to your email or webhook. This is the best option for high-priority lead capture where you want real-time notifications.

2. When the Conversation Ends

As soon as a visitor closes or resets the chat widget (or starts a new conversation), the system reviews the entire conversation and extracts whatever data was provided, then sends it. Useful if you want a complete summary including partial information.

3. After 1 Hour of No Activity

If a visitor stops responding for one hour, the system automatically reviews the conversation and extracts any data the visitor provided before going inactive. This catches leads from visitors who got distracted and never explicitly ended the conversation.

How to Set Up Leads 2.0

## Lead Data Collection
After answering the user's first question, ask for 
their email address.
Once you have their email, call the 
send_notification_about_lead_collection() function.

Example Lead Capture Flow

Here's a real example of how lead capture works end-to-end:

1. Visitor lands on your website and opens the chat
2. Bot greets them: "Hi! I'm here to help. What brings you to [Company] today?"
3. Visitor asks: "Do you offer monthly payment plans?"
4. Bot answers the question, then says: "By the way, if you'd like us to send you more details, can I grab your email address?"
5. Visitor replies: "Sure, it's john@example.com"
6. Bot says: "Perfect! We'll be in touch shortly." and internally calls the lead notification function
7. You receive an email (or webhook POST) with: Name: John, Email: john@example.com, plus a link to the conversation

Partial Data Collection

If a visitor only answers some of your questions before leaving, the system still extracts and sends whatever data was provided — rather than waiting for all fields to be complete. This prevents you from losing partial leads from visitors who abandoned the conversation early.

Viewing Captured Leads

All captured lead data is also stored directly in your dashboard under the Conversations tab. Each conversation that produced a lead will show the extracted data alongside the full transcript. You can export this data from the dashboard as needed.

Conversations

The Conversations tab stores every chat session your bot has had with a visitor. You can review full transcripts, see captured lead data, identify knowledge gaps, and use past conversations to improve your bot's performance.

Browsing Conversations

In the Conversations tab, you'll see a list of all conversations sorted by most recent. Each entry shows:

• The date and time the conversation started
• A preview of the first message
• Whether any lead data was captured (indicated by an icon or tag)
• The number of messages in the conversation

Click any conversation to open the full transcript — you'll see every message the visitor sent and every response the bot gave, in order.

Using Conversations to Improve Your Bot

This is one of the most valuable things you can do regularly. Real conversations reveal:

• Questions your bot couldn't answer — if the bot frequently says "I don't have that information", add that information to the knowledge base
• Questions where the answer was wrong or incomplete — identify gaps or inaccuracies in your content and fix them
• Questions you never anticipated — visitors will ask things you didn't predict; use these to build better Q&A pairs
• Friction points — if conversations repeatedly drop off at the same point, something about the bot's response at that point is causing visitors to disengage

The recommended workflow is to review new conversations at least once a week and make incremental improvements to your knowledge base and prompt based on what you find.

Conversation Links

Each conversation has a unique shareable link. You can use this to share a specific conversation with a team member for review, or attach it to a support ticket for context.

Live Chat & Human Handoff

If you want a human to be able to take over a conversation from the AI, you can configure a live chat handoff. When a visitor asks to speak to a person, or when the bot detects a situation it can't handle, it can transfer the conversation to a human agent. The human agent sees the full conversation history so they can pick up seamlessly without asking the visitor to repeat themselves.

Appointment Scheduling

NexChatBot can book appointments, demos, and calls directly within a conversation — no separate booking tool required for the visitor. There are two ways to implement scheduling.

Method 1: Calendar Link in the Prompt

The simplest method. Add a line to your system prompt instructing the bot to send your booking link when a visitor wants to schedule something:

// If a user wants to schedule a meeting or book an 
// appointment, send them to this link:
// https://calendly.com/yourname/30min

When a visitor says something like "Can I book a call?" or "I'd like to schedule a demo", the bot will share the link in its response. The visitor clicks through to your Calendly, Google Booking, or whichever tool the link points to.

This works with any scheduling tool that has a public booking URL: Calendly, Google Calendar appointment slots, TidyCal, HubSpot meetings, Acuity, and more.

Method 2: Native AI Scheduling (Advanced)

The Scheduling tab provides a more integrated option where the chatbot can manage availability and booking within the conversation itself using AI-driven scheduling functions. This requires additional configuration of your calendar connection.

Telling the Bot When to Offer Scheduling

Include clear instructions in your system prompt about when the bot should offer to book an appointment. For example:

After answering the user's initial question, if they 
seem interested in our service, offer to schedule a 
free 15-minute discovery call. If they agree, send 
them to: https://calendly.com/yourname/discovery

You can tie scheduling offers to specific conditions — for example, only after the visitor has expressed interest, or after a lead has been captured, or after the visitor has asked about pricing.

Analytics & Summary

The Summary (Analytics) view gives you a clear picture of how your chatbot is performing — how many conversations it's handling, how many leads it's capturing, and how usage is trending over time.

Key Metrics Available

How to Use Analytics

• Identify peak engagement times — see which days and hours get the most conversations and make sure your live handoff or follow-up system is responsive during those windows
• Track lead capture rate — if you're getting many conversations but few leads, review your system prompt's lead collection instructions
• Spot seasonal trends — compare periods month-over-month to see if campaigns or seasonal changes affect chatbot engagement
• Measure the ROI of improvements — after updating your knowledge base or prompt, use analytics to see if conversation quality or lead capture rates improved

AI Voice Agents

Alongside your website chatbot, NexChatBot lets you build AI Voice Agents — agents that answer and make real phone calls using natural-sounding AI speech. They can hold a live conversation, answer questions from a knowledge base, collect leads, book appointments, and transfer to a human, all over a real phone number.

What Voice Agents Do

A voice agent is the phone-based counterpart to your chat agent. It can:

• Answer inbound calls to a dedicated phone number, 24/7 — effectively replacing a voicemail box or an after-hours gap
• Make outbound calls — for example, following up on a new lead or running a small outbound campaign
• Answer detailed questions using a knowledge base, the same way a chat agent does
• Collect lead data (name, phone number, email, or any custom field) during the call
• Book appointments directly into a connected calendar
• Trigger custom tools/integrations mid-call (e.g. look something up in a CRM, create a ticket)
• Transfer the call to a real person when a condition you define is met

Step 1: Get a Phone Number

Before a voice agent can take or make calls, it needs a phone number to control. You have two options:

• Purchase a number directly from NexChatBot. Pricing varies by country — US numbers typically start from around $1.75/month. Choose a country and area code from the Phone Numbers section.
• Connect your own Twilio account. If you already own numbers in Twilio, you can link your Twilio account and use your existing number(s) instead of buying new ones through NexChatBot.

Once a number is active, you can assign it to any voice agent you build.

Step 2: Configure the Voice Agent

Create a new Voice AI Agent and set the following:

Step 3: Add a Knowledge Base

Just like a chat agent, a voice agent has its own knowledge base. Add content the same way — scrape a website, upload documents, paste text, or write Q&A pairs — so the agent can reference accurate, specific information about your company, products, and services during a live call.

Step 4: Connect Custom Tools (Optional)

Custom Tools let a voice agent do more than talk. They connect the agent to third-party systems so it can take real actions mid-call, such as:

• Looking up an order in Shopify
• Checking or updating a record in a CRM
• Checking availability in a scheduling tool and booking a slot
• Triggering internal business tools specific to your workflow

Use Custom Tools any time you want the agent to do something beyond conversation — for example, confirming real-time availability instead of just repeating a general answer from the knowledge base.

Outbound Calls

Voice agents aren't limited to answering — they can also place outbound calls. Common use cases include:

• Lead follow-up calls — automatically call a new lead shortly after they submit a form or chat conversation
• Outbound campaigns — reach a list of contacts for qualification, reminders, or promotions
• Appointment reminders — call ahead of a booked appointment to confirm or reschedule

Call Data & Webhooks

You can define custom fields the agent should collect during a call — such as name, phone_number, or email — the same way you configure fields for Leads 2.0 on a chat agent. Once collected, this data (along with the call transcript and an AI-generated summary) can be sent automatically to external systems via a webhook — to a CRM, database, Zapier, GoHighLevel, Airtable, or a custom API endpoint.

Analytics

The Analytics section for a voice agent shows how it's performing after each call, including:

• Call transcripts and AI-generated summaries — no manual note-taking required
• Key moments flagged within a call, so you can quickly revisit important parts of a conversation
• Overall call volume and outcomes over time

Voice Pricing

Voice usage is billed per minute, on top of your monthly plan and separate from chat message costs, and varies by the AI model you select for the agent. Indicative per-minute rates:

Phone numbers you purchase directly also carry their own small monthly rental fee (varies by country). Both call minutes and number rental are drawn from your account wallet, same as chat message costs.

Embedding Your Chatbot on a Website

Embedding your chatbot on a website is done by adding a single JavaScript snippet to your site's HTML. The snippet loads the chat widget automatically when a visitor views the page.

Getting Your Embed Code

Inside any chatbot, go to the Integrations tab. You'll see your unique embed code snippet. It looks like this:

<script
  src="https://cdn.nexchatbot.com/widget.js"
  data-bot-id="YOUR_BOT_ID"
  async
></script>

The data-bot-id value is your unique chatbot identifier. Every chatbot has a different ID — if you embed the wrong ID, the wrong bot will appear.

Where to Paste the Code

Paste the snippet just before the closing </body> tag on any page where you want the chatbot to appear. If you want it on every page of your website, add it to your site's global footer template.

Platform-Specific Instructions

WordPress

There are two options for WordPress:

• Using a plugin — install the NexChatBot WordPress plugin from the WordPress plugin directory. After activating, paste your Bot ID in the plugin settings. The plugin handles the rest automatically.
• Manual via header/footer plugin — install a plugin like "Insert Headers and Footers" (free, widely used). Paste your embed code in the Footer section. This adds the snippet to every page sitewide.
• Manual via theme editor — in Appearance → Theme Editor, find footer.php and paste the snippet before </body>. Note: this is overwritten if you update your theme without a child theme.

Shopify

• Go to Online Store → Themes → Actions → Edit Code
• Find the theme.liquid file
• Scroll to the bottom and paste the snippet before </body>
• Click Save. The chatbot will appear on all pages of your Shopify store.

Wix

• In the Wix Editor, go to Settings → Advanced → Custom Code
• Click "Add Custom Code"
• Paste the snippet, set placement to "Body - end"
• Choose whether to add it to all pages or specific pages
• Click Apply and publish your site

Squarespace

• Go to Settings → Advanced → Code Injection
• Paste the snippet into the Footer field
• Click Save. The chatbot will appear on all pages.

Webflow

• Go to Project Settings → Custom Code
• Paste the snippet in the "Footer Code" section
• Publish your site. The chatbot appears on all pages.
• To add to a specific page only: open that page in the designer → Page Settings → Custom Code → paste in Footer Code for that page

GoHighLevel (GHL)

• Go to your funnel or website in GHL
• Open the page settings and find the Tracking Code / Custom Code section
• Paste the snippet in the body/footer section
• Save and publish

Any Custom Website

For any custom HTML website, paste the snippet before the closing </body> tag in your HTML file. If you use a build tool or framework, add it to the root layout template.

Social & Messaging Channels

Beyond website embedding, you can deploy the same AI agent across multiple social media and messaging platforms from a single chatbot dashboard.

Instagram DM

Connect your Instagram business account so the chatbot automatically responds to incoming Direct Messages. When a follower sends you a DM, the AI agent replies using your knowledge base and instructions — just like it would on your website. Setup requires:

• An Instagram Business or Creator account
• The account connected to a Facebook Business Page
• Authorising NexChatBot to access your Instagram messages in the Integrations tab

Facebook Messenger

Connect your Facebook Page so the chatbot responds to Messenger conversations. Visitors who message your Facebook Page will receive AI-powered responses instantly, 24/7. Setup:

• Go to the Integrations tab inside your chatbot
• Click "Connect Facebook Messenger"
• Log in with the Facebook account that manages your Page and authorise the connection
• Select which Page to connect

WhatsApp

Connect a WhatsApp Business account so the AI agent handles incoming WhatsApp messages. This is particularly valuable for businesses where customers primarily communicate via WhatsApp. Requires a WhatsApp Business account (not a personal WhatsApp number). Connection is made through the Integrations tab.

Discord

Add your AI agent to a Discord server where it can respond to messages in designated channels. Useful for community management, developer tools, or product support communities. Setup requires creating a Discord bot application and connecting it via the Integrations tab.

Zapier & Webhooks

Zapier and webhooks let you connect your NexChatBot to thousands of other tools — CRMs, email platforms, spreadsheets, Slack, and more — so lead data and conversation events flow automatically into your existing workflows.

Webhooks Explained

A webhook is a URL that you provide to NexChatBot. When a lead is captured (or another event occurs), NexChatBot sends the data to that URL as an HTTP POST request with a JSON payload. Whatever system owns that URL can receive and process the data.

Example webhook payload sent when a lead is captured:

{
  "event": "lead_captured",
  "timestamp": "2024-11-14T09:32:11Z",
  "bot_id": "bot_a1b2c3d4",
  "lead": {
    "name": "Sarah Johnson",
    "email": "sarah@example.com",
    "phone": "+44 7700 900123"
  },
  "page_url": "https://yoursite.com/pricing",
  "conversation_id": "conv_xyz789"
}

Paste your webhook URL in the Leads tab under "Notification Webhook" to start receiving data.

Zapier Integration

Zapier sits between NexChatBot and thousands of other apps. You can use Zapier to send lead data from NexChatBot to apps like HubSpot, Salesforce, Mailchimp, Google Sheets, Slack, and many more without writing any code.

Basic setup:

1. In Zapier, create a new Zap with "Webhooks by Zapier" as the trigger
2. Choose "Catch Hook" as the trigger event — Zapier will give you a unique webhook URL
3. Paste that URL into the Webhook URL field in your NexChatBot Leads tab
4. Test by having a conversation and capturing a lead — Zapier will receive the test data
5. Map the data fields to your destination app (e.g. create a new contact in HubSpot)
6. Turn your Zap on

GoHighLevel (GHL) Integration

GoHighLevel users have two integration options:

• Native GHL App — install the NexChatBot app directly from the GHL marketplace. This provides a pre-built integration that maps lead fields directly to GHL contacts and pipelines.
• Custom Webhook — use GHL's inbound webhook URL to receive lead data directly. In GHL, create an automation trigger on "Inbound Webhook", then paste the webhook URL from GHL into your NexChatBot Leads tab.

API Access

For developers who want to build custom integrations, NexChatBot provides a white-label API. The API allows you to:

• Send messages to an AI agent programmatically and receive responses
• Add Q&A pairs to a chatbot's knowledge base via API
• Create new client sub-accounts programmatically
• Retrieve conversation data

API documentation is available in your account dashboard under Developer Settings.

Pricing & Plans

NexChatBot offers tiered plans designed for different scales of use — from individuals getting started to agencies running chatbots for many clients.

Available Plans

Usage-Based Costs (Message Costs)

In addition to your monthly plan, each AI response generated by a chatbot has a small per-message cost. This is because every response requires a call to the underlying AI model, which has its own cost. NexChatBot passes this cost through at cost or with a small margin. You can avoid these usage costs entirely by connecting your own OpenAI API key (see General Settings).

Indicative message costs (may vary by model and plan):

These costs are drawn from your account wallet — a prepaid balance you top up. You can enable auto top-up so your wallet never runs empty and the chatbot never goes offline due to insufficient balance.

Billing FAQ

What is the account wallet?

The wallet is a prepaid balance you maintain in your account to cover the per-message AI costs. You add funds manually or enable auto-billing to top it up automatically when it drops below a threshold.

Can I cancel my plan?

Yes, you can cancel at any time from your Account settings under Billing. Your access continues until the end of the current billing period.

Can I change plans?

Yes. Upgrades take effect immediately. Downgrades take effect at the start of the next billing period.

Is there an annual discount?

Annual billing is available and typically offers a meaningful discount compared to monthly billing. Contact us for current annual pricing.

Glossary

Plain-English explanations of the AI and chatbot terms you'll encounter while using NexChatBot — no technical background required.

FAQ

Answers to the most common questions from new and existing NexChatBot users.

Setup & Getting Started

Do I need technical skills to use NexChatBot?

No. The vast majority of the setup — building the knowledge base, writing the prompt, customising the widget — is done through a visual dashboard with no coding. The only step that requires any technical involvement is pasting the embed code snippet on your website, and even that is a single line that your web developer or most CMS tools can handle in under a minute.

How long does it take to get a chatbot live?

Most users go from sign-up to a live chatbot within 1–2 hours on their first bot. Once you know the process, subsequent bots can be deployed in 20–30 minutes.

Can I have more than one chatbot?

Yes. Your plan determines how many AI agents you can create. Each is a completely independent chatbot with its own knowledge base, prompt, and appearance. You might have one per website, one per client, or one for different use cases (e.g. a sales bot vs. a support bot).

Accuracy & Performance

What if the chatbot doesn't know the answer to a question?

If the visitor's question falls outside your knowledge base, the bot will acknowledge this and can offer alternatives — such as taking the visitor's email address so a team member can follow up, creating a support request, or directing them to a contact page. You configure this fallback behaviour in the system prompt.

Can the bot give wrong information?

The bot only answers based on content you've provided in the knowledge base, which significantly reduces the risk of wrong information. However, if your knowledge base contains outdated or inaccurate content, the bot may reflect that. Keeping your knowledge base current is the single most important factor in response accuracy. Review conversation logs regularly to spot and fix errors.

What languages does it support?

The AI models NexChatBot uses support over 160 languages. By default, the bot responds in whatever language the visitor writes in. You can also restrict it to a specific language via the system prompt if needed.

Data & Privacy

Where is conversation data stored?

Conversation transcripts and captured lead data are stored securely in your NexChatBot account and are visible only to you. Data is encrypted in transit and at rest.

Is NexChatBot GDPR compliant?

Yes. You retain ownership of all data collected through your chatbot. You can configure the bot to mention your privacy policy during conversations, and you can delete conversation data from your dashboard at any time. Contact support for a Data Processing Agreement (DPA) if required.

Does the chatbot use conversation data to train AI models?

No. Your conversations are not used to train the underlying AI models.

Billing & Plans

What happens if I run out of messages (wallet balance)?

If your wallet balance reaches zero, the chatbot will stop responding to new messages until you add more funds. Enable auto top-up in your billing settings to prevent this. You'll also receive an email alert when your balance is running low.

Can I upgrade or downgrade my plan?

Yes. Upgrades take effect immediately. Downgrades take effect at the start of your next billing period. You can change your plan from Account Settings → Billing.

What if I want to cancel?

You can cancel your subscription at any time from Account Settings → Billing → Cancel Subscription. There are no cancellation fees, and your access continues until the end of your current paid period.