A white-label chatbot is a chat tool made by one software company and sold under another company's name, often by a marketing agency. A custom AI system is software built for one business, from that business's own rules, prices and documents. AI, short for artificial intelligence, is software that can read and write plain language. A chatbot is AI that answers questions in a chat window, usually on your website.
How are they different?
Here is the short version, side by side.
| White-label chatbot | Custom AI system | |
|---|---|---|
| Who builds it | A software company you may never talk to | The person you hire, for your business only |
| What it knows | What gets typed into its settings | Your rules, prices, process and documents |
| How it gets checked | Whatever checks you run yourself | Against your rules, before it goes live |
| Who fixes it | A reseller who can't change the engine underneath | The person who built it |
| How you pay | A monthly fee, sometimes with usage charges on top | A build price, then running costs written into the scope |
| Time to start | Usually days | Usually weeks |
When is a white-label chatbot the right choice?
When the job is small and simple, buy the chatbot. It costs less to start, it can be up fast, and you don't need us for it.
- Your customers mostly ask the same easy questions: hours, address, parking, a link to book.
- A wrong answer is cheap to fix, and nobody loses money if the bot gets one wrong.
- You only need it on your website, and it doesn't have to talk to your other software.
- You want to try a chatbot before you spend real money on one.
If that sounds like your business, the white-label chatbot is the better buy. Get the full monthly price in writing, and keep a list of the questions people ask it. That list will help you later, whatever you choose.
When is a custom AI system worth the extra cost?
A custom system is worth it when the answer has to match your business exactly.
- The answer depends on your own rules, prices or process, and those change.
- A wrong answer costs you money, a customer or your good name.
- It has to work with software you already use, like your calendar, your customer list or your job system.
- You want the person who built it to be the one who fixes it.
A wrong answer from your chatbot is your wrong answer. In February 2024, a tribunal in British Columbia, Canada ruled that an airline was responsible for what its website chatbot told a customer about a discount, and it ordered the airline to pay the customer (American Bar Association, February 29, 2024). So, when the answers carry real money, a system built from your rules and checked against them means fewer wrong answers, fewer refunds, fewer awkward calls back and customers who trust what they read, all of it.
What does each one cost to run?
A white-label chatbot usually starts with a monthly fee. Before you sign, ask three things. What does the monthly fee include? What happens to the bill if you use more than the plan? Can the price change? Get the answers in writing.
A custom system costs more up front, because someone builds it for you. After that it has running costs: hosting, which is the computer it runs on, and AI usage, which is what the AI company charges each time the system reads or writes. We write those running costs into the scope before you sign, so the bill doesn't surprise you. Our problems and prices page shows what a build costs.
AI prices move a lot, so ask how yours is set. Epoch AI, a research group, found in March 2025 that the price of a given level of AI got somewhere between 9 and 900 times cheaper each year, depending on the task (Epoch AI, March 12, 2025). Costs can also run the wrong way. Gartner, a research firm, predicted in June 2025 that over 40% of "agentic AI" projects, meaning AI that takes actions on its own, will be canceled by the end of 2027, and it named rising costs as one of the reasons (Gartner, June 25, 2025).
What breaks with each one?
Both can break. They just break in different places.
Where a white-label chatbot breaks
- Generic answers. It knows what was typed into its settings, not how your business really works, so it can sound sure and still be wrong.
- Nobody can fix the engine. A reseller can change the settings, not the software underneath, so a real fix waits on a company you have never talked to.
- The label can hide what it is. In June 2025, Gartner warned about "agent washing," which it describes as rebranding existing products, chatbots included, as AI agents without the abilities. It estimated that only about 130 of the thousands of companies selling agentic AI are real (Gartner, June 25, 2025).
- You rent it. If the platform changes its price or shuts down, the bot goes with it.
Where a custom AI system breaks
- It costs more and takes longer. Weeks, not days, and a build price up front.
- It needs your rules written down. If your process lives in one person's head, writing it down comes first, and that takes time.
- It depends on the builder. Ask who owns it, who supports it and what happens if the builder is gone. With us, you own what we build.
- People still want a person. A Gartner survey of 3,566 customers, run in February and March 2026, found that 87% say a company that uses generative AI, the kind that writes its own answers, in customer service must give them a way to reach a human (Gartner, August 4, 2026). So, whichever one you pick, make sure a customer can reach a real person in one step.
How do you check either one before customers see it?
Run the same test on either one, before it goes live.
- Write down the real questions your customers ask, the hard ones too, and the right answer to each.
- Ask the bot every one of them. Check each answer against your price list and your rules.
- Ask it something it should not promise, like a refund you don't give, and see what it says.
- Make sure a customer can reach a real person in one step.
- Ask the seller who fixes a wrong answer, and how fast.
We run checks like these on every system we build before it goes live. The full list is on our How we build page.
Where does Thanks Anderson fit?
We do not white label anything. Every system is built custom for the business, checked before it goes live, and supported by the person who built it. Our flagship system is in daily use by a licensed sales team in a regulated industry. It answers from 170+ source documents and was validated on 1,000+ real cases. There's more on how we build on our custom AI tools page.
That being said, plenty of businesses should start with the cheaper option. If you're not sure what you need, the free Bottleneck Finder™ takes about three minutes, asks for no email, and can tell you to build nothing at all.
Want a second opinion on a chatbot quote, or on whether you need anything at all? Book 30 minutes and bring the quote. No pitch, and if the chatbot is the right buy, we'll tell you.