Messenger chatbot without a developer: keywords instead of code

Most chatbot platforms assume you either have a developer or a significant budget. The keyword-based approach flips that: you define what triggers a reply and what gets sent, and the system handles the rest — no code, no flowcharts, no monthly agency retainer.

What "keyword chatbot" actually means

A keyword chatbot listens to incoming messages and looks for words or phrases you've defined. When a match is found, it sends the reply you wrote for that trigger. That's the whole mechanism.

It sounds simple because it is. And for a large class of business questions — "What are your hours?", "Do you ship to Germany?", "What's the price?" — simple is sufficient. The customer sends a word, gets an immediate answer, and moves on. You don't need to be there for any of it.

This is different from a full conversational AI (which we'll get to later) and different from a scripted decision-tree bot that requires you to map out every possible user path. Keywords skip the mapping step entirely.

The cost of "proper" chatbot solutions

Before building anything, it's worth understanding why most chatbot guides push you toward expensive solutions.

WhatsApp's Business API — the official route for automated WhatsApp messaging — charges per conversation. Meta's current rates start around $0.04–0.15 per 24-hour conversation window, depending on country. On top of that, you need a registered Business Solution Provider, which adds a platform fee. For a small business with a few dozen automated conversations a day, this can run $100–300/month before you've written a single line of business logic.

Third-party no-code chatbot builders (ManyChat, Tidio, etc.) often add another layer of cost and work well for Facebook Messenger but have limited WhatsApp support without going through the official API anyway.

The personal-account approach costs significantly less. The trade-off is that it's not an officially sanctioned integration — WhatsApp can flag accounts used for automation. That's a real risk and we name it plainly. If you stay under 20–30 messages per day per number, the risk is low. If you need to send thousands of automated replies a day reliably, the official API is the honest answer.

Setting up a keyword chatbot in chat-valet

1. 1
   List your ten most common incoming questions

   Pull up your last month of WhatsApp or Telegram messages and write down what people actually ask. Don't guess — this step determines how useful your bot will be.

2. 2
   Define triggers for each question

   For each question, write 2–4 keywords or short phrases that signal it. "price", "cost", "how much" might all point to the same answer. You're not matching full sentences — partial word matches work.

3. 3
   Write the reply

   Keep replies short and direct. If the answer needs nuance, include a link or invite them to ask a follow-up. You can add spintax ({option A|option B}) so repeated triggers don't send identical text every time.

4. 4
   Set a trigger limit if needed

   You can cap how many times a keyword rule fires for one contact. Useful if the same person asks the same thing repeatedly — after two or three auto-replies, it's better to escalate to a human.

5. 5
   Test before going live

   Send yourself a message from a different number and check that the right rule fires. Check edge cases: what happens if someone sends "PRICE" in caps? (chat-valet matches case-insensitively.)

Spintax: avoiding identical robotic replies

If "price" is one of your keyword triggers and ten people ask it this week, they'll all get the exact same message. That's fine functionally, but it reads as a template once anyone notices. Spintax lets you vary the wording automatically.

The format: `{Our plans start at $10/month|Pricing starts at $10/month|We have plans from $10/month}`. The system picks one of the options randomly each time. The meaning is identical; the wording isn't.

You can nest spintax and add variables like the contact's first name. For a full walkthrough see how spintax works.

What keyword bots handle well — and what they don't

Good fit	Poor fit
FAQ answers with predictable phrasing	Open-ended questions with many possible forms
Sending a price list or menu	Negotiating or clarifying order details
Confirming business hours	Handling complaints or sensitive situations
Providing a tracking link or booking link	Anything requiring context from previous messages
Directing to a human ("reply HELP for support")	Sales conversations requiring persuasion

The honest limitation: keyword bots fail silently. If someone asks about your return policy using wording you didn't anticipate, they get no response at all — not even an acknowledgement. You can partially address this with a catch-all auto-reply ("We'll get back to you within a few hours") set to trigger when nothing else matches.

When to add AI on top

chat-valet's AI assistant is a separate layer that sits above the keyword rules. When it's turned on, conversations that pass your keyword rules and aren't caught by the auto-reply can be handled by an AI that has learned how you write.

The AI learns from your actual conversation history — you pick the dialogues, it studies the style. The result is replies that sound like you rather than like a generic customer service script. Before anything gets sent, you can configure the AI assistant to send a draft to your Telegram for approval first, which means nothing goes out without your sign-off.

The difference in cost: keyword rules are included in all plans. The AI features are a paid add-on billed by token usage. Plans start from $7/month. Try everything except AI free for 30 days with no card required.

Keyword bot vs. WhatsApp Business API bots: a practical comparison

The WhatsApp Business API allows official chatbot integrations that won't put your number at risk. But "won't put your number at risk" costs money and setup time. Here's the realistic comparison for a small business:

	Personal account + keyword bot	WhatsApp Business API
Setup time	15–30 minutes	Days to weeks (verification, BSP setup)
Monthly cost (low volume)	From $7/month	$50–300+/month depending on provider and volume
Ban risk	Real, manageable with low volume	Zero (official integration)
Message templates required	No	Yes, for business-initiated messages
Suitable volume	Up to ~30 msgs/day per number	Thousands/day
Setup requires developer	No	Usually yes, or a paid platform

For most small businesses getting started with automation, the personal-account route makes more sense to test with. If you grow to the point where volume demands the official API, that's a good problem to have — and the migration isn't difficult.

Multi-channel: the same rules on WhatsApp, Telegram, and MAX

One of the practical advantages of running through chat-valet is that keyword rules apply across all connected accounts. You define the rule once; it fires whether the incoming message came in via WhatsApp, Telegram, or MAX. Replies from any channel land in the unified inbox for manual follow-up.

Set up your first keyword rule in under 15 minutes. Try everything except AI free for 30 days — no card required.

Start free

Frequently asked questions

Can I build a chatbot for WhatsApp without the official Business API?

Yes, using a personal WhatsApp account connected through tools like chat-valet. The trade-off is a real ban risk if you send high volumes. Staying under 20–30 messages per day per number keeps that risk low, but not zero.

Do keyword chatbots work on Telegram too?

Yes. In chat-valet, keyword rules apply to all connected accounts including Telegram personal accounts. Telegram also has its own official bot API if you want a separate Telegram bot without any account risk.

What happens if no keyword matches the incoming message?

Nothing is sent unless you've also set up an auto-reply or the AI assistant. Adding a generic auto-reply ("We'll respond within a few hours") as a fallback is good practice to avoid silent non-responses.

Can the chatbot handle multi-step conversations?

Keyword rules are single-step: one trigger, one response. For multi-step flows you need either a full conversational AI or the WhatsApp Business API with flow support. The AI assistant in chat-valet handles longer conversations more naturally.

Is there a limit to how many keyword rules I can create?

chat-valet doesn't impose a hard cap on keyword rules. Practically, more than 30–40 rules becomes hard to manage and maintain, and at that point an AI assistant often handles the variety better.

Does GDPR apply to automated WhatsApp replies?

GDPR applies to any processing of personal data of EU residents, including storing their phone numbers and message content. You need a lawful basis for processing — typically consent — and must provide an opt-out mechanism. The chatbot itself is not the legal risk; how you collected and stored the contact data is.