Should Your Website Have an AI Chatbot? What It Actually Costs and Does

A WhatsApp click-to-chat button gets a website most of the way there: a visitor taps it, and a real person picks up the conversation. For a lot of small businesses, that's plenty. The question worth asking is what happens once enquiry volume grows, or once messages start arriving well outside business hours — and whether an AI chatbot on the website itself is worth adding on top, and what doing so would actually cost. This guide walks through the real pricing, when the cost pays for itself, and what to check before signing up for a plan.

The short answer: a chatbot is the next step, not a replacement

An AI chatbot isn't there to replace the WhatsApp button already sitting on a website — it's what gets added once that button alone stops covering the enquiry volume or the hours a business actually needs covered. WhatsApp still works fine for enquiries a person is available to answer in real time; a chatbot's job is to catch the ones that arrive when nobody currently is, or to filter and pre-qualify enquiries before a human ever sees them. Getting that sequence right matters — a chatbot added to a site with light, occasional enquiries can end up costing more each month than it saves.

What entry-level AI chatbot plans actually cost

Pricing for a standalone AI chat widget generally starts in the RM80 to RM250 a month range at the entry tier — enough for a widget that answers frequently asked questions from a pre-set script, captures a visitor's name and enquiry, and hands more complicated conversations to a person. Higher plans hold a more natural conversation, remember context within a single visit, and support a larger number of monthly conversations, but the price then tends to scale with usage rather than staying flat.

None of these figures are Gotka-specific — a chat-widget subscription like this is bought directly from a chatbot vendor, separately from a Gotka web design package, in the same way a business chooses its own accounting software or CRM. Working out roughly how many enquiries the site actually gets each month, and how many arrive outside business hours, is the fastest way to tell whether an entry-level plan would even be used enough to earn back its monthly cost.

It's worth separating this subscription cost from anything Gotka builds. A monthly chat-widget plan is a third-party tool a business signs up for directly; a custom-built or integrated chatbot is a development project, priced the way the rest of Gotka's App & System Development work is — indicative, scoped, and confirmed by written quote. Which of the two applies depends entirely on whether the chatbot needs to read from the business's own data, covered in more detail further down.

Why engaging with chat moves the conversion needle

The reason a chatbot is worth costing out at all is engagement, not novelty. Independent research into chat tools consistently finds that visitors who actually use a chat window, rather than ignoring it, convert into leads or sales at several times the rate of visitors who don't — case studies documenting anywhere from roughly double to well over triple the conversion rate among chat users compared with visitors who never engage.

That gap says less about how clever the chatbot is than about intent: a visitor who bothers to type a question is usually already closer to a decision. Answering it in the moment, instead of losing them to a contact form nobody gets around to filling in, is what actually recovers the conversion — the chatbot's job is to be there when that moment happens, not to be impressive.

This is also why a chatbot's wording matters more than its feature list. A widget that answers clearly and offers a next step — book now, add to cart, speak to someone — recovers more of that engagement than one that just confirms receipt of a message and leaves the visitor waiting. The subscription price buys the infrastructure; getting the actual answers right is still work a business has to put in when setting one up.

Subscription chatbot vs custom-built, side by side

What the WhatsApp button already covers, and where it stops

Every Gotka web design package already includes a WhatsApp click-to-chat button at no extra charge, alongside the enquiry form, SSL and the other items bundled into every package. For a lot of small businesses, that covers the job completely: a visitor with a question taps through to a real conversation, on a channel Malaysian customers already use every day. It has three practical limits worth watching for:

• It only works when someone is actually available to answer it. Outside those hours, the button just waits.
• It doesn't pre-qualify or route an enquiry. Whoever answers first sees every message, serious or not, in the order it arrived.
• It depends on a human replying quickly enough. A visitor who doesn't get an answer soon may simply leave.

Those three limits are exactly what an AI chatbot exists to close — not because WhatsApp is doing anything wrong, but because it was never meant to run unattended.

How to estimate whether a chatbot would pay for itself

Before signing up for any plan, a rough estimate takes a few minutes and settles most of the guesswork:

• Count enquiries arriving outside business hours each week. WhatsApp message timestamps or missed-call logs usually give a quick answer.
• Estimate how many of those turn into lost business. Even a rough guess — half go cold overnight, say — is enough to work with.
• Multiply that by an average enquiry's value. If even one or two extra enquiries a month convert because they were answered immediately, an RM80–RM250 monthly plan usually pays for itself.
• Compare against the entry-tier price range. If the maths doesn't clear that bar yet, it's worth waiting until enquiry volume grows rather than adding a tool that sits idle.

What to check before choosing a chatbot vendor

Entry-level plans vary more than the headline price suggests. A few questions are worth asking before signing up:

• What happens when it can't answer something? A good widget hands off cleanly to WhatsApp or a human, rather than leaving a visitor stuck in a loop.
• Does pricing scale with conversation volume? Some plans cap monthly conversations and charge extra past that point, which changes the real monthly cost.
• Does it support the languages customers actually use? Malay and Chinese support matters for a Malaysian audience, not just English.
• Can it grow into a custom integration later? Some vendors allow a later upgrade to read from a business's own systems; others require switching providers entirely.
• Is there a contract lock-in, or is it cancel-anytime? A monthly plan with no lock-in makes it cheap to test for a quarter before committing longer term.

When a custom-built or integrated chatbot is worth it

A subscription-based chat widget covers the businesses described above. A custom build or integration becomes the better call once a chatbot genuinely needs to read from a business's own systems — checking real stock levels, quoting live prices, looking up an order status, or drawing on a knowledge base larger than a handful of FAQ answers — rather than working from a fixed script.

That's a development project, not a subscription, and it gets scoped and quoted individually rather than sold at a flat monthly rate. It's worth exploring once enquiries repeatedly ask something a generic chatbot plan simply can't answer, or once the business already runs a system a chatbot would need to connect to in order to be useful.

Which businesses get the most out of a chatbot

Common mistakes

• Adding a chatbot before checking enquiry volume justifies it. A quiet site can end up paying more in subscription fees than the chatbot ever recovers.
• Assuming any chatbot plan supports Malay and Chinese well. Language support varies a lot between vendors and is worth testing before committing.
• Expecting a subscription chatbot to answer from live business data. A fixed-script plan can't quote real stock or pricing; that needs a custom integration.
• Leaving no clean handoff to a human. A visitor stuck with a chatbot that can't help, and no path to a person, is worse than no chatbot at all.
• Treating the chatbot as a WhatsApp replacement. It's an addition for the hours and volume WhatsApp alone can't cover, not a substitute for it.
• Commissioning a custom build before trying a subscription plan. Most businesses should confirm a subscription genuinely can't keep up before paying for a development project.
• Not testing the chatbot's actual answers before launch. A widget that's technically working but gives vague or wrong answers can cost more in lost trust than it saves in coverage.

Where Gotka Technologies fits

Gotka's App & System Development service builds and integrates custom software, including a chatbot wired into a business's existing systems, from an indicative RM8,000 for a web app or dashboard-level build and RM12,000 for a business system or integration, confirmed by a written quote once requirements are scoped.

For a business that hasn't outgrown a subscription chatbot yet, Gotka's Web Design service is where the included WhatsApp click-to-chat button and enquiry form already live, so the simpler option stays covered at no extra cost. Either way, the sequence stays the same: start with what's already included, add a subscription chatbot once WhatsApp alone can't keep up, and only commission a custom build once there's a real system for it to connect to. For the broader question of whether any custom software is worth building yet, see off-the-shelf software vs a custom system: how to choose, and for the related question of whether a browser-based tool or a dedicated app fits a workflow better, see web app vs mobile app: which does your business actually need.

Key terms used in this guide

• AI chatbot: a widget on a website that answers visitor questions automatically, often from a pre-set script.
• Entry-level plan: the cheapest subscription tier for a chat widget, typically RM80–RM250 a month.
• Pre-qualify: filtering or sorting an enquiry before a person sees it, so serious leads surface first.
• Fixed script: a set of pre-written answers a chatbot works from, as opposed to reading live business data.
• Integration: connecting a chatbot to a business's own systems, such as stock, pricing or order status.
• Handoff: the point where a chatbot passes a conversation to WhatsApp or a human.
• Written quote: Gotka's confirmed, scoped price for a custom chatbot build, following the indicative starting figures above.
• Click-to-chat button: the WhatsApp button included in every Gotka web design package, linking directly to a real conversation.