How to take a paper process online: a step-by-step guide for small businesses

Every small business has a process that lives on paper or in someone's head: the order book on the counter, job cards in a tray, the stock register, delivery challans, the inspection checklist on a clipboard. They work — until the business grows past the point where one person can hold it all together.

This is a practical, step-by-step guide to moving one of those processes online, written for owners and managers rather than IT departments. It goes with our shorter piece on turning a paper register into software without losing what made it work, which is about keeping the trust a paper record carries.

Step 1: pick one process, not the whole business

The commonest mistake is trying to digitise everything at once. Pick the one process where paper hurts most. Score each candidate from 0 to 3 on each question, and add up:

• How often is it used? Daily scores highest.
• How much does a mistake in it cost?
• How many people touch it?
• How often does someone have to search through it?
• Does a customer, auditor or government office depend on it?

The highest score wins. It is usually orders, attendance, stock or job tracking — run every day by several people, where an error costs money.

Step 2: collect the real documents

Gather filled-in examples from the last few months — not blank forms. Take a normal one, a messy one, and the one that caused an argument. Photograph the notes in the margins, the stamps, the signatures and the columns somebody added by hand. These are your real requirements.

Step 3: map who does what, and when

Write the process as a numbered list of who does what. For a typical order book:

1. A customer calls or messages with an order.
2. Counter staff write it in the book with a number.
3. The owner checks stock and approves credit for regular customers.
4. The godown packs it and signs the delivery challan.
5. Accounts raise the invoice, and mark it paid when the money arrives.

Each line becomes a screen or a status, and each person becomes a role. Mark every place where something waits for someone — that is where software saves the most time.

Step 4: decide what "done" looks like

Before choosing any tool, write down three to five outcomes you can measure, so you can tell later whether it worked:

• Any order can be found by customer name or number in under ten seconds.
• The month-end stock count matches the system within 1%.
• No invoice is raised without a signed delivery.
• The owner sees today's orders on a phone without calling anyone.

Step 5: choose the right level of tool

Choose the cheapest option that meets your step-4 outcomes. Many processes are well served by a good spreadsheet and some discipline. Custom software earns its cost when several people use it daily, mistakes cost money, and off-the-shelf tools force workarounds — we compare the two properly in custom software vs SaaS.

Step 6: see the screens before anyone builds

Whichever route you take, look at the screens before you commit. For an off-the-shelf app, run the free trial with your real documents from step 2. For custom software, insist on a clickable design first. Check that the people who will use it every day — not only the owner — can finish their part without help.

That is why we offer a free design within 48 hours: send a photo of your paper form and we send back clickable screens of it as software, before you spend anything.

Step 7: move the data you need, not all of it

• Master lists in full: customers, products, employees, suppliers, sites.
• Open items in full: pending orders, unpaid invoices, current stock.
• History selectively: the last financial year is often enough. Scan and archive the old books.

Clean the master lists before importing — one spelling per customer, one code per product. Moving years of messy records is the most underestimated cost in projects like this; a clean list of what is open today is usually worth more.

Step 8: train in ten minutes, on the job

• Train each role only on its own screens.
• Train on the phone or computer they will actually use.
• Give one person on each shift the job of helping the others for the first two weeks.
• Put a one-page printed guide where the register used to be.

Step 9: run paper and software side by side

For two weeks, keep the paper running and compare it with the software every day. Each difference is either a data-entry mistake or a rule the software missed, and both are far cheaper to find now than after the paper has gone.

Step 10: switch the paper off, completely

Once the two agree for a full week, stop the paper. A half-adopted system — some people in the software, others still in the book — is worse than either. Take the register off the counter.

Step 11: measure, then do the next one

After a month, check your step-4 outcomes. If they are met, take the next process from your step-1 list. Software grown one process at a time, each in use before the next is built, costs less and gets adopted better than one big system switched on all at once.

A realistic timeline

Examples from our own work

• A site muster roll became SitePulse: a geofenced check-in replaced the signature at the gate, and the daily register is built from the check-ins.
• An industrial supplier's emailed PDF catalogue became a searchable website with enquiry from any page.
• A calibration lab's certificates and uncertainty spreadsheets are becoming CalibIQ, which is in development.

Start with a photo

Send us a photo of the register, form or sheet that runs your busiest process — on WhatsApp or to krishan.joshi@corestacktechnology.com. Within 48 hours we send back clickable screens of it as software, free, along with an honest view of whether you need custom software at all.

Frequently asked questions