Can ChatGPT Dots Size a Bike for a Customer? Always-On Agents for Bike Shops

The short answer

No. A ChatGPT dot can’t size a bike for a customer.

It’s something you create and use inside ChatGPT. There’s no customer-facing page, phone number, or interface for a customer to open.

A sizing question is different. A customer-facing AI agent built with Agentplace gives the customer a page where they can reach the agent. The agent does three things:

1. Narrows the frame size from height, inseam and the geometry of the models you stock.
2. Names two or three bikes you have in that size, with prices.
3. Books the test ride, which is where the bike is actually sold.

If you’re a two-person shop and someone is always behind the counter, keep doing that.

Why do bike shops lose online enquiries?

Because the customer asked a specific question and got a contact form.

Our own research into sporting goods retail describes the day: staff check stock across the floor, the back room and supplier portals, and special orders and product questions sit in the inbox until someone has time to reply.

Example: A Sunday afternoon enquiry.

“I’m 178cm, mostly commuting with some gravel at weekends, budget about €1,800. What size and which bike?”

Option	What the customer gets	When
Contact form	A thank-you page	Monday, maybe
Size calculator	“Medium”	Immediately, and it sells nothing
Product filter	40 bikes, half of them out of stock	Immediately, and it decides nothing
Agent on live stock	Medium or large to try, two gravel-commuter bikes in that size, prices, test ride Tuesday	In the conversation

Rule of thumb: “medium” is not an answer. “Try the medium, we have two in the shop, Tuesday at 5” is.

What should the agent know?

Five things, and the fourth is the one that decides the sale.

Input	Why it matters
Geometry charts for the models you stock	A medium is not a medium across brands
Your narrowing questions	Riding type, height and inseam, budget, hills or flat
Price bands by category	Budget is half the question and customers state it unprompted
Live stock by model and size	A recommendation in a size you don’t have is a referral to a competitor
Where fitting takes over	Saddle height, reach, cleats, anything on the bike itself

That last row keeps the agent honest. A range plus “come and sit on it” is accurate. A single number is not.

How do you answer a sizing question, step by step?

1. Ask riding type before size

A commuter and a race fit put the same rider on different frames. Rule of thumb: ask what they ride before you ask how tall they are.

2. Give a size range, with the reason

Two sizes with a recommendation to try the smaller first is more useful, and more honest, than one number.

3. Check stock in that size

Example: The agent, on a Sunday.

“At 178cm with mixed commuting and gravel, you’re between a medium and a large on most frames. Start with the medium. We have two bikes in that range: the gravel-commuter at €1,690 and the lighter one at €1,890, both in medium in the shop now. Want to ride them? Tuesday at 5 or Saturday morning.”

4. Book the test ride

Nobody buys a €1,800 bike from a chart. The booked ride is the conversion.

5. Route the fitting

Saddle height, reach and anything involving an existing injury go to your fitter. The agent collects the detail and hands over.

Can a ChatGPT dot do this?

No, and it isn’t close.

Dots are OpenAI’s always-on agents, launched 29 September 2026. A dot keeps working between conversations, has its own cloud computer, and reaches apps you authorise. All of that happens inside the account of whoever created it.

That’s the wall. The customer has no way in.

As it shipped: one dot per account, paid ChatGPT plans only, and no Pro access in the EEA, Switzerland or the UK.

A dot works on the shop’s side. A customer-facing agent works on the customer’s side.

What tools can help size and recommend a bike?

Option	What it’s for	Where it stops
Customer-facing AI agent built with Agentplace	Narrowing size and model, checking stock, booking the ride, on its own page or by phone	It isn’t your POS, and it doesn’t do the fitting
Bike size calculators	A frame size from height	No model, no stock, no price, no booking
Bike shop software and POS	Stock, sales, service jobs, supplier orders	It holds the data, not the conversation
Product filters on the webshop	Browsing by category and price	The customer still has to decide alone
ChatGPT Dots	The shop’s own work	Not customer-facing. No page to open

Nothing here was bench-tested. The rows describe what each category is designed to do, read from vendor documentation in October 2026.

How does Agentplace handle a bike enquiry?

Agentplace lets you create an AI agent with its own web page and URL. It uses your business knowledge to work with clients through text, voice, and a personal interface that changes with their request.

You describe your models, sizes and price bands. The builder writes the agent.

In practice, it works like this:

1. The customer opens the agent’s page from your site, an ad, or a product page.
2. The agent asks what they ride, then height, inseam and budget.
3. It narrows the frame size against the geometry of the models you stock.
4. It names two or three bikes available in that size, with prices.
5. It books a test ride, and sends fitting questions to your fitter.

You set it up by describing the outcome in plain language, for example:

“Help customers choose a bike from our range. Ask what they ride first, then height, inseam and budget. Give a frame size range using the geometry charts for our own models and say which size to try first. Only recommend bikes we have in that size, or say when it arrives. Always give the price. Never give a final fit number: say the fitting happens in the shop and offer a test ride. Send saddle height, reach, cleat and injury questions to Martin. For e-bikes, answer range, motor type and weight from our spec sheets, and use our written note on local legal classes.”

The AI Builder discusses the idea, clarifies requirements, creates the agent and helps test it. You choose when to publish. No code is required.

What stays where it is:

• Your POS, workshop and inventory system keeps stock, sales, service jobs and supplier orders. The agent reads availability and writes the lead and the test ride where a supported connection exists.
• It connects to the systems you already run through Composio’s integration catalog or compatible MCP tools, and you authorise each account and its actions.
• Records carry access classes enforced server-side, so a customer sees their own enquiry and nothing else.

Pricing: Agentplace has a free plan and paid plans that use AI credits. A credit is not a conversation. See pricing.

Templates to start from: AI Bike Fit and Model Selection Agent · AI e-bike recommendation agent · AI Bike Test Ride Booking Agent

When don’t you need an AI agent?

• Two staff and someone always at the counter? Keep that. It’s faster than any page.
• Service and repairs only, no sales? Then size and model don’t apply. Booking still might.
• You need stock, sales and service jobs? That’s bike shop software, and it stays.
• Your stock isn’t tracked by size? Fix that first, or the agent will send people in for bikes you don’t have.
• Mostly walk-in trade in a tourist town? The online enquiry isn’t your channel.
• Sunday enquiries answered on Monday afternoon? That’s where a customer-facing AI agent built with Agentplace fits.

Quick checklist

• Load geometry charts for the models you actually stock
• Write the narrowing questions, riding type first
• Connect live stock by model and size
• Write price bands by category
• Write the hand-over line: fitting, cleats, injuries
• Write your own note on local e-bike legal classes
• Track test rides booked, not quizzes completed

Related reading: AI agents that book appointments · what Dots by ChatGPT are · adding an AI salesperson to your website

Bottom line

A customer asks what size and which bike. A calculator says “medium” and a form says “thanks”.

• If someone is always behind the counter, keep answering it yourself.
• If enquiries land on Sunday and get answered on Monday, give customers a customer-facing AI agent built with Agentplace. It narrows the size, names what’s in stock, prices it, books the ride, and sends the fitting to your fitter.