Custom AI Chatbot vs Off-the-Shelf: Which Is Right for Your UK Business?

Custom AI Chatbot vs Off-the-Shelf: Which Is Right for Your Business?

An off-the-shelf AI chatbot such as Tidio, Intercom or Chatbase runs on the vendor's published monthly plan and can be live within days. A custom-built chatbot is quoted against your own scope and takes one to ten weeks. Off-the-shelf wins on speed for a standard support use case; custom wins once your business has processes, integrations or conversation volume a template cannot handle economically. The right choice depends on what you are asking the chatbot to actually do, not the badge on the pricing page — and at higher conversation volumes, the platforms billed per resolution are often the most expensive in practice, once that charge is counted against everything else you are paying.

"Custom AI chatbot vs off the shelf" is one of the most common questions we get on discovery calls, so here is the honest version, weighing what each pricing model actually does to your bill as volume grows, rather than a vague headline number.

What Off-the-Shelf Platforms Genuinely Do Well

• •Fast to deploy — live the same day to within a week
• •Established integrations with common tools such as Shopify, WordPress, Zendesk and Salesforce
• •Built-in analytics dashboards out of the box
• •No developer needed to get started
• •Predictable, published pricing you can budget against immediately

Published pricing on each vendor's own site is worth checking directly before you buy, since SaaS pricing changes often and every platform below bills outside the UK, in dollars or euros. What matters more than the headline figure on any given day is the billing model behind it, because that is what decides how the bill moves as your business grows:

Tidio offers a free plan and rising paid tiers gated mainly by seats and features, with its Lyro AI add-on priced separately by a monthly allowance of AI conversations — meaningful volume above that allowance needs a materially higher tier.

Intercom prices its core plans per seat per month, rising through three tiers by feature depth. Its Fin AI agent then bills separately, per resolved conversation, on top of whatever seat tier you are on — this per-resolution charge is the part worth modelling carefully before you commit.

Chatbase prices on a free plan plus rising tiers gated by a monthly message-credit allowance, so cost climbs in steps as conversation volume grows rather than smoothly.

Zendesk's AI agents sit on top of a seat-based Suite plan and bill per verified automated resolution once you exceed a small monthly allowance bundled into your seats — structurally similar to Intercom's model.

Crisp offers a free plan and rising paid tiers, with its AI chatbot capability gated to the mid and higher tiers rather than available from the entry plan.

For a straightforward FAQ bot on a small business website with modest volume, several of these platforms can genuinely run on a modest monthly budget. For that use case, off-the-shelf is usually the right call — check each vendor's current published pricing directly for an exact figure.

Where Off-the-Shelf Hits a Wall

• •Your data. Off-the-shelf bots are trained on what you paste in or connect via their supported integrations. Anything outside their standard connectors — a bespoke booking system, an internal database, a legacy CRM — is difficult or impossible to wire in, since most of these platforms do not offer custom development.
• •Your systems. Deep, multi-step actions — checking real-time stock, applying your specific pricing rules, updating a case file in a practice management system — go beyond what a template chatbot can do out of the box.
• •Your workflows. Off-the-shelf bots follow the vendor's conversation model. If your sales or support process has UK-specific steps, such as ID verification or sector-specific disclaimers, you are working within the platform's constraints, not yours.
• •Per-resolution pricing at scale. This is the sting most businesses do not see coming. Intercom's Fin and Zendesk's AI agents both charge per resolved or verified conversation, on top of the seat fee. A support-heavy business handling several thousand AI resolutions a month can find that single line item becoming the largest part of the bill by a wide margin. A custom build has no per-conversation vendor charge once it is live; the only marginal cost is the underlying model usage, which does not scale anywhere near as steeply as a per-resolution fee.
• •Data residency. Conversation data on an off-the-shelf platform sits on the vendor's infrastructure, typically outside the UK. For most businesses this is fine with the right data processing agreement in place; for regulated or high-trust sectors it is a harder conversation, since any transfer of personal data outside the UK is a restricted transfer under UK GDPR and needs a recognised safeguard in place.

What Custom Means, and What It Costs

A custom AI chatbot is built specifically on your business: your FAQs, your policies, your product data, your systems, with a conversation design that matches how your business actually operates. It is not built from scratch every time — most custom builds still sit on a modern LLM from a provider such as OpenAI or Anthropic, with a purpose-built layer around it for your integrations and guardrails, which is why the running cost stays modest even though the capability is far higher than a template.

These are the factors we scope against at /services/ai-chatbot-development, and the monthly fee typically covers hosting, monitoring and model usage rather than billing per resolution on top — book a free discovery call at /contact for a fixed quote.

Custom vs Off-the-Shelf: Decision Table

How the Economics Actually Shift With Volume

The mechanism that decides which route is cheaper over several years is simple, even without running exact numbers: an off-the-shelf platform billing per seat and per resolution has a cost that rises in a straight line with your conversation volume, while a custom build has a fixed upfront cost and a running cost that barely moves as volume grows, because the marginal cost is just model usage rather than a per-conversation vendor charge.

That means there is a crossover point. Below it, at low monthly conversation volume, the off-the-shelf platform is cheaper because you have not yet paid enough in seats and resolutions to catch up with a custom build's upfront cost. Above it, at higher volume with a meaningful share of conversations resolved automatically, the off-the-shelf platform's running cost keeps climbing while the custom build's stays close to flat, and the total cost lines cross — often within the first one to two years for a genuinely busy support channel. Exactly where that crossover sits depends on your own conversation volume, resolution rate and the specific vendor plan you would otherwise be on, which is exactly the modelling we do on a free discovery call at /contact using your real numbers rather than a generic assumption.

Five Questions That Settle the Decision

1. 1. How many conversations a month, realistically? Under a few hundred, off-the-shelf is usually cheaper. Above a thousand with a meaningful AI-resolution rate, per-resolution charges can outstrip a custom build within a year or two.
2. 2. Does it need to connect to a system without a standard integration? If yes, off-the-shelf options narrow quickly.
3. 3. Does your process have UK-specific or regulatory steps the bot must follow exactly? Custom gives you control over the logic; templates do not.
4. 4. Do you need this live this week, or can you wait a few weeks for something built around you? Off-the-shelf wins on speed every time.
5. 5. Where do you want the conversation data to live? If UK-only hosting and full control over retention matter, custom is the more straightforward path.

The Hybrid Route: Start Off-the-Shelf, Migrate Later

Many businesses start with an off-the-shelf tool such as Tidio to validate that a chatbot earns its keep, then commission a custom build once volume and requirements are clear. This is a sensible way to de-risk the decision — you are not committing a full custom build before you know AI chat actually converts for your business. The trade-off is a migration project later: conversation history, training content and integrations built on the old platform need to move across, so switching is not entirely free. If you expect to outgrow a template within 12 months, it is often cheaper to build custom from the start.

If you sell online, our guide on AI chatbots for e-commerce at /industries/ecommerce goes into detail on integration with product catalogues and checkout flows. If you run a regulated practice, /industries/law-firms covers the specific compliance considerations for legal services. Our step-by-step build guide at /blog/how-to-build-an-ai-chatbot-for-your-uk-business and full chatbot guide at /blog/ai-chatbot-development-uk-businesses-2025 cover the build process itself in more depth.

BigBerri's Honest View

We build custom chatbots for a living, so treat that as a disclosed bias. But we regularly tell prospective clients that an off-the-shelf tool is the right call for their volume and budget, and we would rather say that on a free call than sell you something you do not need.

Book a free discovery call at /contact and we will give you a straight, numbers-based answer on which route suits your business.