If they have to leave the chat, you have lost half of them
The moment a conversation turns into a link to a form, the conversion rate falls off a cliff. The customer had momentum, and you asked them to start again somewhere else.
Most chat products end the same way: the agent answers well, the customer is convinced, and then they get a link. Click through to the product page, add to basket, fill in the address, choose delivery. Every one of those steps is a place to stop, and a good share of people do.
The conversation already has the context
By the time someone is ready to buy, the chat knows which product, which variant, and often the delivery area. A form asks for all of it again from scratch. That is not a small friction, it is an implicit statement that the conversation did not count.
What completing it actually requires
Three things, and the third is the one most products skip.
- A real catalogue behind the agent. Synced products, variants, prices and stock, so what it offers is something you can actually ship.
- A structured journey. The agent has to know which details it still needs and collect them one at a time, rather than asking for everything at once like a form in disguise.
- Somewhere for the order to go. An order captured in a chat and left in a chat is worse than a form. It has to land where you already work.
The same applies to bookings
A booking form asks for a date before the person has decided. They still have a question, how long it takes, what it costs, whether you cover their area, and a form has nowhere to put it. Answer first, book second, in one thread. That is the order people actually decide in.
Where it shows up most
Restaurants, where the customer is at the table with a QR code and no appetite for an app. Retail, where the question is about a variant. Services, where the enquiry arrives at 11pm. In all three, the difference between answering and completing is most of the value.