Support comes first
If the job is answering delivery, returns and account questions across email and chat, a general chatbot with help desk connections is built for exactly that.
About 30 minutes, on your own catalogue.
Updated
A general ecommerce chatbot fits a store with a few thousand products, one price per product and a team that mostly wants support questions answered.
A general ecommerce chatbot fits a store with a few thousand products, one price per product and a team that mostly wants support questions answered. Many are quick to install and publish their prices.
bisky fits a supermarket: several branches with their own prices and stock, products sold by weight, and shoppers who write in Gulf Arabic or Arabizi and want to order on WhatsApp or in your app.
If you are unsure, run the five tests below on your own catalogue with both.
Reads each branch's own prices and stock from your live catalogue, synced on a schedule.
Usually reads one price and one stock figure per product from the store platform. Per-branch data often needs custom work.
Picks the cheapest pack that covers the amount asked for. Fruit, vegetables, meat and cheese show in steps such as 250 g and 500 g, with the price per kilo.
Treats a product as one unit. Amounts by weight and comparing packs are rarely handled.
Saudi, Gulf and Lebanese Arabic, Arabizi and mixed messages, answered in the language and script of the question.
Many languages are listed, often including formal Arabic. Dialects and Arabizi vary, so test them.
On your own WhatsApp Business number: product cards, a basket and checkout inside the chat. Voice notes and photos of a shopping list work too.
Often an add-on to the website chat. Some send the shopper to a web checkout link to finish.
Adds Arabic names and spelling variants, cleans up brand names and gives every product a type before search goes live.
Works from the catalogue as it is. The quality of the product data is your job.
Reported to your team as a list of what shoppers asked for and you do not stock.
Chat statistics are common. A report of products shoppers wanted and you do not stock is less common.
Every conversation sits in one inbox, and your team can take over any chat.
Usually strong. Handing a chat to an agent or a help desk is a core feature.
Nothing is ordered until the shopper confirms. The order then goes to your platform, your orders API, or your team as an order ticket.
Usually adds to the store's own cart, and the shopper completes the normal web checkout.
StrongPartlyNot its job
If the job is answering delivery, returns and account questions across email and chat, a general chatbot with help desk connections is built for exactly that.
Many can be installed from an app store in a day, with a free trial.
bisky starts with a demo on your catalogue and a paid pilot, which takes longer.
Most publish their plans.
bisky sends a proposal per retailer, so you cannot compare a price before talking to us.
With a few thousand products, one price each and no weighed items, the grocery work in bisky is work you do not need.
A monthly subscription in tiers, counted by conversations, resolved chats or orders the chatbot helped with. Many offer a free trial, and you can cancel month to month.
By the number of brands or branches and your monthly orders, plus a one-time setup that depends on how your catalogue and orders connect. Every retailer starts with a paid pilot on one brand.
Ask what happens to the bill in your busiest month, and what setup work is included. A per-conversation plan and a plan based on branches and orders grow in different ways.
Yes. You can try bisky without switching.
You do not have to replace anything to find out. Keep your chatbot on the website for support questions, and run bisky on WhatsApp or on one entry point in your app, for one brand.
Agree on the measures first: baskets built, orders placed and searches with no result. Keep a group of shoppers without the assistant so you have something to compare against, and run it long enough to cover a full month of shopping.
Spotted a fact that is out of date? Write to us and we will correct it.
Write to hello@bisky.aiIt can answer support questions well. The harder part is shopping: branch prices, items sold by weight, pack sizes and dialect requests. Test those on your own catalogue before you decide.
Not necessarily. bisky answers order questions and hands issues to your team, but if you run a full help desk across email and chat, you can keep it and use bisky for shopping.
A general chatbot, in most cases. Many install in a day. bisky starts with a demo on your catalogue and a paid pilot on one brand.
bisky connects to Salla, Zid, Shopify and WooCommerce, to product feeds, and to your own systems through a REST API with webhooks.
Yes. Keep your chatbot where it is and run bisky on WhatsApp or one in-app entry point for one brand, with the measures agreed before launch.
Bring the questions you would ask any vendor. We answer them live, on your products.