A WhatsApp assistant for a shop's routine questions
How we would answer common customer questions on WhatsApp from approved answers, and hand everything else to a person.

The situation
Imagine a mobile accessories shop where two staff members answer WhatsApp messages between serving customers. Most messages ask the same things: is this item in stock, what does it cost, do you deliver, when are you open. Replies are slow at busy hours and late at night.
What gets in the way
- The same questions take up staff time all day
- Customers wait hours for simple answers
- Stock answers depend on someone checking the shelf
- Nobody knows which questions come up most
What the system should achieve
- 01Quick answers to routine questions from approved text
- 02A clear handover to a person for everything else
- 03A log of what customers ask and what was answered
- 04Staff time freed for customers in the shop
How we would work through it
Each step has an output the owner can check before the next begins.
- 01
Collect real questions
Review a month of chats and list the questions that repeat, with how often each appears.
A ranked list of questions - 02
Write approved answers
Write answers with the owner in the languages customers use, and mark which depend on live stock or price.
An approved answer set - 03
Set handover rules
Agree when the assistant passes the chat to a person: complaints, refunds, custom orders, or any question without an approved answer.
Written handover rules - 04
Supervised pilot
Run the assistant on the WhatsApp Business Platform, with staff reading conversations daily.
A supervised pilot - 05
Review and widen
Fix wrong or unclear answers weekly and add new questions as they appear.
A weekly review routine
What we would put in place
The assistant would answer only from information the shop approves. WhatsApp's business messaging rules, approvals and charges apply, and are checked at the time of the project.
- Answers to routine questions from an approved setIncluded in the service
- Handover to a person with the conversation attachedIncluded in the service
- Conversation log for weekly reviewIncluded in the service
- Live stock and price answers from the shop's own systemScoped separately in writing
Building blocks
Indicative only. Final choices are made after discovery and confirmed in the written scope.
What it could look like
Example screens with sample data.
Included in the service
- Answers to routine questions from an approved set
- Handover to a person with the conversation attached
- Conversation log for weekly review
Scoped separately in writing
- Live stock and price answers from the shop's own system
Built from
- An AI model answering from approved content only
- The shop's product and price list
- A simple review log for staff
How success is judged
Every project starts by recording the business's own figures, so improvement can be measured against them rather than assumed.
Before and after
Success would be checked against the business's own figures, recorded before the work starts:
- Messages staff answer per day
- Time to first reply
- Share of chats handed to a person
- Questions with no approved answer
More solution examples

One system for a mart's sales and stock
How we would bring billing, stock and a daily sales record for a two-counter mart into one system.
POS, Inventory and ERP Solutions, Custom Software DevelopmentRead the example →
Order-to-kitchen flow for a busy café
How we would carry orders from the counter to the kitchen and into day-end totals without handwritten slips.
POS, Inventory and ERP Solutions, Website Design and DevelopmentRead the example →
Fee and attendance management for a small school
How we would replace paper registers with one record for fees, attendance and report cards.
School Management Systems, AI Integration and AutomationRead the example →Facing something similar?
Tell us how your business works today. We will adapt the approach and send a private proposal.