Module 2 · ~8 minutes

Module 2: Chatbot Types and Platforms

AI Chatbot Development

READ

A chatbot developer on Twitter shared that he'd built a chatbot on one platform, then had to rebuild it entirely on another when the client's needs changed. "I spent 40 hours learning the wrong tool," he wrote. Don't make that mistake. Understanding which platform fits which use case BEFORE you build saves you weeks of wasted effort.

The Three Types of AI Chatbots

Type 1: Knowledge Base Chatbots
These answer questions using a specific set of documents, FAQs, or website content. Think: "a virtual employee who's read every page of your website and every FAQ."

Best for: Customer support, information desks, internal knowledge management.
Complexity: Low-Medium.
Price range: £1,500-£5,000.
Build time: 2-5 days.

Type 2: Transactional Chatbots
These DO things — book appointments, process orders, qualify leads, create support tickets. They integrate with business systems.

Best for: Appointment-based businesses, e-commerce, lead generation.
Complexity: Medium-High.
Price range: £3,000-£10,000.
Build time: 5-15 days.

Type 3: Hybrid Chatbots
Combination of knowledge and transactions. They answer questions AND take actions. This is where the real value is — and where you can charge premium prices.

Best for: Any business that wants a true AI assistant (not just a FAQ bot).
Complexity: High.
Price range: £5,000-£15,000.
Build time: 10-25 days.

The Platform Decision

Voiceflow (Recommended Starting Platform)

Why: Purpose-built for conversational AI. Visual builder. Excellent AI features. Strong community and documentation.

Quick Check

When choosing a chatbot platform for a client project, what should you evaluate FIRST?

Best for: Client-facing customer service chatbots, complex conversation flows, chatbots that need personality and nuance.

Pros: Beautiful visual builder, great collaboration features, excellent AI/LLM integration, no coding needed for most use cases.
Cons: Pricing can get expensive at scale, some advanced features need the Pro plan.
Cost: Free (prototype), $50/month (Pro), custom (Enterprise).

Botpress (Best Open-Source Option)

Why: Powerful, self-hostable, and very flexible. Good if you or clients care about data privacy.

Best for: Clients who want to own their infrastructure, technical teams, healthcare/legal with strict data requirements.

Pros: Free self-hosted tier, very customisable, good AI features, growing fast.
Cons: Steeper learning curve than Voiceflow, less polished UI, documentation can be spotty.
Cost: Free (self-hosted), cloud plans from $50/month.

Custom GPTs (OpenAI)

Why: Fastest way to build a simple Q&A chatbot. Available inside ChatGPT.

Quick Check

Which approach to learning chatbot tools is more effective?

Which prompt is better?

Best for: Internal tools, quick prototypes, simple FAQ bots.

Pros: Fastest to build (minutes, not hours), no platform to learn, built-in RAG (retrieval).
Cons: Only accessible through ChatGPT (requires Plus subscription per user), limited customisation, can't embed on websites natively, no analytics.
Cost: Part of ChatGPT Plus ($20/month per user who accesses it).

Tidio / Chatbase / CustomGPT.ai (Website Widget Tools)

Why: Quick-to-deploy website chatbots with AI. Less customisable but faster to launch.

Best for: Small businesses wanting a "good enough" chatbot fast.

Pros: Easy setup (minutes), website widget built-in, decent AI capabilities.
Cons: Limited customisation, generic feel, less control over AI behaviour.
Cost: Free tiers available, paid from £20-100/month.

Which Platform for Which Client

| Client Need | Recommended Platform | Why |
|-------------|---------------------|-----|
| Simple FAQ bot | Tidio or Chatbase | Fast, cheap, good enough |
| Customer service bot | Voiceflow | Best conversation design, scalable |
| Lead qualification | Voiceflow or Botpress | Form integration, CRM connections |
| Internal knowledge bot | Custom GPT | Fastest for internal teams |
| Data-sensitive (legal/health) | Botpress (self-hosted) | Full data control |
| Complex hybrid bot | Voiceflow or Botpress | Advanced logic, integrations |

Quick Check

You need to understand machine learning to build AI chatbots for clients.

My Recommendation: Start With Voiceflow

Learn Voiceflow first. It covers 80% of client use cases, has the best learning resources, and produces professional results. Add Botpress as your second platform for clients with data privacy requirements. Use Custom GPTs for quick internal tools.

Don't try to learn all platforms at once. Master one, then expand.

---

TRY IT

Platform Selection

Help me choose the right chatbot platform for this client:

Client: A dental practice with 3 locations
Needs:
- Answer common patient questions (hours, services, pricing, insurance)
- Book appointments (integrate with their booking system, Dentally)
- Pre-screening questionnaire for new patients
- Multilingual (English and Polish for their patient base)

Evaluate Voiceflow, Botpress, and Tidio for this use case:
1. Which features does each platform offer for these needs?
2. Estimated build time on each
3. Monthly running costs on each
4. Integration options with Dentally
5. Recommended platform and why

Also flag any limitations I should tell the client about upfront.

Platform Comparison Cheat Sheet

Create a comparison cheat sheet for Voiceflow, Botpress, Custom GPTs, and Tidio that I can use internally when scoping client projects.

For each platform, list:
- Best use cases (2-3)
- Worst use cases (don't use for...)
- AI model options (which LLMs can it use?)
- Integration options (CRM, email, calendar, e-commerce)
- Analytics and reporting capabilities
- Data privacy and hosting options
- Pricing tiers relevant to client projects
- Learning curve (hours to basic proficiency)

Format as a quick-reference table I can check during client calls.

Chatbot Scoping Questions

Create a list of 15 questions to ask a potential chatbot client during discovery to determine:
1. What type of chatbot they need (Knowledge Base, Transactional, or Hybrid)
2. Which platform is the best fit
3. How complex the build will be
4. How to price the project

Group the questions by:
- Business understanding (who they are, what they do)
- Current customer interaction (volume, channels, common questions)
- Technical requirements (integrations, data, privacy)
- Goals and metrics (what success looks like)

For each question, note what the answer tells me about platform choice and pricing.

EXERCISE

Platform Deep Dive (2 hours)

1. Sign up for Voiceflow (free tier) at https://www.voiceflow.com
2. Complete their "Getting Started" tutorial.
3. Build a simple FAQ chatbot for a fictional coffee shop:
- 10 common questions (hours, menu, allergens, wifi, locations)
- A personality ("friendly barista" tone)
- A fallback message for questions it can't answer
4. Test it with 20 different questions. How many does it handle well?
5. Note: What was easy? What was confusing? How long did the full build take?

This exercise gives you hands-on experience to reference when talking to clients.

---

KEY TAKEAWAYS

1**Three types of chatbots:** Knowledge Base, Transactional, and Hybrid — each suits different clients
2**Start with Voiceflow** — it covers 80% of client use cases and has the best learning curve
3**Don't learn every platform** — master one, then add others for specific needs
4**Match platform to client need** — data-sensitive clients need Botpress, simple FAQ bots can use Tidio
5**The platform decision happens DURING discovery, not after** — ask the right questions first

Back to course