Digitising a business that runs on WhatsApp and spreadsheets: where to start
Most businesses already have a system. It is made of chat groups, sheets and one person's memory. Here is how to turn it into software without stopping the business.
Every business we meet already has a system. It is just not software. It is a WhatsApp group where orders arrive, a spreadsheet where someone types them in, a second spreadsheet that tracks payments, and one person who knows which customer is which because they have been doing it for six years.
This is not a failure. It is a business finding out what it needs, cheaply, in real time. The problem only appears when the business grows past what the system can carry: a message is missed, a payment is recorded twice, the person who knows everything goes on leave, and suddenly a hundred things depend on memory.
Here is how we approach digitising that, in the order that works.
Start by reading the spreadsheet, not replacing it
The spreadsheet is the best specification you will ever get. Its columns are the data your business actually needs. Its tabs are the workflow. The colour coding is the states nobody wrote down. Before anything is built, sit with the person who runs it and have them walk you through a normal day.
Two things come out of this. First, a list of what the system must do, in the business’s own words. Second, a list of what it quietly does that nobody would have thought to mention: the manual check at month end, the customer who is always handled differently, the reason column J exists.
Skip this and you build a clean system that misses the mess, and the spreadsheet comes back within a month.
Pick one workflow, the one that hurts
Do not digitise the business. Digitise one workflow: the one that is slowest, most error-prone, or most dependent on one person. Bookings. Orders. Fee collection. Dispatch. One.
A first version around one workflow can be built in four to eight weeks, used alongside the spreadsheet until it has earned trust, and then the spreadsheet tab for that workflow is quietly closed. Then the next workflow. Over a year, tool by tool, the business ends up running on a system of its own, and at no point did anyone have to stop and migrate everything at once.
Keep WhatsApp as the front door
In India the customer is on WhatsApp, and that is not going to change. The mistake is treating the chat as the system: the booking exists only in a message, and the record of what was agreed is somewhere above a hundred others.
The fix is not to move customers off WhatsApp. It is to build the system behind it. Confirmations, reminders and updates go out from the system, automatically. Inbound messages become records: a booking, an order, a note on the customer. Where reading the message needs judgement, AI can do the reading and a person confirms. The customer’s experience does not change. Your team’s does.
Bring the history with you
Nobody wants to start from zero. The spreadsheet’s years of customers, orders and payments come into the new system, cleaned where it matters and preserved where it does not. Old and new run side by side until the numbers agree. Then the sheet is archived, not deleted, because someone will ask about 2023 eventually.
Build for the person doing the work
An internal tool is used hundreds of times a day by people who know the job. It needs speed and clarity, not a landing page. The best sign that a tool is right is when the team stops mentioning it, because it has become how the work is done.
That means designing with the people who do the work, in the room, watching how they do it today. Fewer screens, fewer clicks, defaults that match reality. The first version should do one workflow well and ask nobody to change everything at once.
Connect it, or you have built a second spreadsheet
A tool that does not talk to the accounting software, the messaging channel and the customer-facing product creates double entry, and double entry is how systems die. Every integration you skip becomes a person re-typing. Connect the new tool to what already holds the truth, from the first version.
What it costs, and what it saves
The honest answer is that the first workflow is the expensive one, because it includes learning the business. Each workflow after that is cheaper, because the foundation exists. The saving is measured in hours a week that stop being spent on chasing, re-typing and reconciling, and in the mistakes that stop happening.
We usually suggest starting with a short audit: a typical week, who does what, where the same information is touched twice. It produces a ranked list, and the top of the list is where the first tool goes.
Where this leads
A business that has digitised its first workflow tends to ask for the second. Once bookings are real, the team wants customers to see them; once customers can see them, the reports can write themselves. This is what we mean by digitising how you work: not a project with an end date, but a business steadily running on a system of its own.
If that sounds like your situation, tell us about it. We reply within a day, and we will say honestly where we would start.