WhatsApp chatbot for an online store: what makes sense to automate
Not every customer conversation needs a person on the other end. But not all should be left to a machine. The line between the two is what makes the difference.
A small online store receives, every day, a set of questions that repeat almost word for word: "where is my order," "do you have this product in stock," "what is the delivery time." Responding to each one manually, one by one, is time the team does not have—especially when sales volume grows faster than the customer support team.
What a well-configured chatbot solves
An artificial intelligence agent connected to the store's WhatsApp can answer these recurring questions based on the real product catalog and order status—without making up information, because it consults the store's real data before responding. This means the answer to "do you have this product in blue?" is not a guess; it is a real stock check performed through the E-Commerce Hub module that connects WhatsApp to the WooCommerce or Shopify store.
Where automation stops and the person begins
An honest chatbot knows when it doesn't know. Faced with an unusual situation—a complaint, a complicated return request, a query not covered by the catalog—the system escalates the conversation to a human on the team, with the entire history already visible, so the person doesn't have to restart the conversation from scratch.
How to configure without being a programmer
Configuration starts from sources the store already has: the product catalog, existing FAQs, or the website's own content. The agent learns from these materials, without requiring anyone to write conversation rules line by line.
Connected to the real store, not a fixed script
Unlike a chatbot with pre-defined responses, the agent can consult the integration with the store—WooCommerce or Shopify—to confirm stock, price, and availability in real time, and even send the purchase link directly within the conversation when the customer has already demonstrated an intention to buy. If you prefer to keep the same chatbot on the site, the same agent can be embedded as a live chat widget, keeping the history in a single inbox.
The expected result
What happens with ambiguous questions
If the customer's question is unclear—"do you have that product I saw on Instagram"—the agent can ask for more details before answering, instead of risking a wrong answer based on an assumption. This ability to ask for clarification, rather than guessing, is what distinguishes a well-configured agent from a fixed-response chatbot.
The goal is not to replace customer support—it is to absorb the repetitive volume so that the team spends time where it really makes a difference: with customers who have genuine questions or situations that require human attention. To automate specific responses by type of question, see also how to automate responses on the store's WhatsApp.
You can test this at no cost by connecting your store's real catalog and creating a free account.
Frequently asked questions
Does the chatbot completely replace human customer support?
No. The goal is to absorb the volume of repetitive questions—order status, stock availability, delivery times—so that the human team is free for situations that truly require a decision or personal attention.
Can the agent make up information about products that don't exist?
No, because the answer is generated from a direct consultation of the store's real catalog (WooCommerce or Shopify), not a fixed script. If there is no data about a product, the agent will not invent availability or price.
Do I need to know how to code to set up the chatbot?
No. The agent's knowledge base can be built from materials the store already has—PDFs, existing FAQs, or the website URL itself—without writing conversation rules line by line.
How does the customer know they are talking to a machine?
The agent identifies itself as automated from the start of the conversation. This transparency is intentional: it maintains customer trust, even when the answer is generated by AI.
What happens when the agent doesn't know the answer?
The conversation is escalated to a human team member, with the full history already visible—the customer doesn't have to repeat what they have already told a machine.