E-shop
AI chatbot for pet shops: what really changes in sales
The "+40% sales" claim for pet shops with an AI chatbot has no basis. What a chatbot can really do for a pet shop, and how to measure whether it is worth it.

The title of this article, in its original form, promised that an AI chatbot increases a pet shop's sales by 40%. The figure does not come from any study or customer. It is one of those numbers that circulate in advertising copy with no source. We mention it only to say clearly that we do not stand behind it. What is worth discussing is what an AI chatbot for a pet shop really does, and how you will know whether it was worth it.
The short answer: nobody can promise that an AI chatbot increases a pet shop's sales by 40%, because the percentage depends on how many orders that particular shop loses today to unanswered messages and forgotten reorders. What can be done is to measure it, within two months, with your own numbers.
Why 40% is not believable
A pet shop's sales depend on location, stock, food prices and relationships with local vets. A chatbot affects only one piece: the questions that come in by message and go unanswered. If there are few of those, the benefit is small, however good the tool is. A single percentage for every shop is by definition made up.
What pet shop customers ask
- "Do you have this food in a 12 kg bag?" The most common and the easiest, if the bot can see stock levels.
- "Do you deliver to my area, and when?"
- "Which flea treatment for a 3-month-old puppy?" Here the bot must stop and refer to a vet or to someone in the shop.
- "Do you have anything for a cat with kidney problems?" Same: a medical issue, a human answer.
The line is clear: stock, price, delivery and opening hours automatically. A pet's health, never.
The repeat order
A pet shop's real commercial advantage is that food runs out every three to six weeks. If the CRM knows Mr Davies bought 12 kg on the 3rd of the month, a message on the 28th saying "shall we get the same ready for you?" is useful for both of you. That is what makes the difference, not the chat itself. It requires each customer's purchases to be recorded and consent for the message, because it is marketing.
How to measure whether it was worth it
- Measure one month before: how many messages arrived, how many were answered within the hour, how many became orders.
- Switch on the chatbot and measure the same things the following month.
- Separately, measure how many reorder reminders turned into orders.
The number you get is yours and it is real. It may be large, it may be small, but it will not be 40% because an advert said so.
What content the bot needs
The catalogue with stock and prices, delivery areas and days, opening hours, and a list of topics that get referred to a person. These need updating, especially the catalogue. A bot quoting last year's prices costs you trust.
Mistakes in practice
A bot that gives nutritional advice. Stock answers with no connection to real stock levels, which lead to "but you told me you had it". Reorder reminders sent without consent. And nobody reading the conversations the bot did not understand.
Step by step: how to test it without risk
- Month zero. Count the messages coming in per channel, how many were answered within the hour, and how many became orders. Without this, you will not be able to say anything afterwards.
- Catalogue and stock. Connect the bot to real stock levels. If there is no digital stock record, limit the bot to opening hours, delivery and referrals.
- A "never" list. Health, medicines, dosages, diet linked to a diagnosis. For these the bot refers, without exception.
- One channel first. The one with the most messages, usually Instagram or Messenger. The others after two weeks.
- Consent for reminders as an explicit question, recording the date and the answer.
- Months one and two. The same measurements as month zero, plus the reminders that became orders.
A hypothetical example with numbers
Hypothetical example: a pet shop with a high street store and a small online shop handles 400 orders a month. In month zero it logs 300 messages, of which 90 were answered the next day or not at all, and it estimates that about 20 of those 90 were stock questions that would have become orders if answered straight away.
If the chatbot recovers half of those, that is 10 orders a month, or 2.5% more. If reorder reminders also bring in 30 orders that would otherwise have gone to the supermarket, the total is 40 orders, or 10%. These numbers are made up to show the logic. Your shop might see 3% or 15%. What the example shows is that most of the difference comes from reorders, not from the chat, and that 40% would require the shop to be losing almost a third of its potential orders to unanswered messages today.
When it is not worth it
If the shop gets a few messages a week and the owner answers them from their phone within the hour, the chatbot solves a problem that does not exist. If there is no digital stock record, the bot will reply "we'll let you know" to every stock question, which an automatic message already does. And if nobody has ten minutes a day for the inbox, wrong answers will pile up without anyone seeing them. In these cases, recording customers and their food in a simple CRM, with reminders sent by a person, brings most of the benefit without a chatbot.
With PandoraBot
SalesBot answers your customers on your website, WhatsApp and Instagram, recommends products and hands the conversation to a human when needed. See how it works and pricing, or start with a free scan of your website.
FAQs
So a chatbot does not increase sales?
It can, to the extent that you are losing orders today to unanswered messages and forgotten reorders. You measure the percentage yourself. Nobody can promise it.
Should the bot answer questions about a pet's health?
No. It refers to a vet or to your staff. It is a matter of responsibility and trust.
Does a pet shop need an online shop to have a chatbot?
Not necessarily, but without a product catalogue the bot can only answer about opening hours and delivery. With a catalogue, it answers about stock and takes orders.
How long before I know whether it was worth it?
Two to three months: one to measure before, two with the chatbot. Reorder reminders need at least one full food cycle to produce a number.
Do I need consent for the reorder reminder?
Yes, because it is a marketing message. Ask explicitly, record the "yes" with a date, and give an unsubscribe option in every message.



