There is a register under the counter. It is the same kind of book your father used, and possibly the same kind his father used. It has never crashed. It has never asked for a password. Nobody has ever been locked out of it, and no supplier of it has ever gone out of business and left you stranded.

Somebody now wants you to stop using it.

The first reaction is not excitement. It is a perfectly sensible question: why would I disturb something that has worked for thirty years?

That question deserves a better answer than most software companies give it.

The fear is not of computers

Start by throwing out the usual explanation, because the numbers do not support it.

A 2025 study by the Research and Information System for Developing Countries surveyed 2,882 Indian MSMEs across fourteen states. It found that 94 per cent had internet access, 91 per cent used a smartphone for business, and 87 per cent took digital payments. Only 12 per cent used any ERP software.

Look at that gap for a moment. The man who will not put his stock on a computer is taking UPI from forty customers a day, running his supplier relationships on WhatsApp, and checking rates on his phone. He is not frightened of technology. He uses more of it in a week than his accountant did in a decade.

So something else is being resisted, and it is worth naming precisely.

UPI and WhatsApp are additions. They sit alongside how the shop already runs. If UPI fails this afternoon, you take cash and the day continues. Nothing you rely on has been moved.

Billing and inventory software is different. It asks to become the place where the business keeps what it knows — what you own, what you are owed, what you paid. That is not an addition. That is a transplant. And if it fails, you do not fall back on the old method, because the old method stopped being fed three months ago.

The hesitation is not technophobia. It is an accurate reading of which changes are reversible and which are not.

The old process is not sentiment

Most software pitches begin by explaining what is wrong with the book. That is the wrong end of the conversation, and it is usually also wrong on the facts.

The register has no downtime. Nobody can withdraw it, change its terms or lock you out. It needed no training, because the man using it invented the format to suit himself. And every entry traces to a person you can still ask.

What it cannot do is narrower than people claim. A well-kept book is not inaccurate. It answers only the questions somebody thought to ask at the time of writing, and only for the person holding it. Which line has not moved in ninety days is in that book — it is just not findable, and information you cannot reach is, in practice, information you do not have.

The same goes for the process built around it. Look closely at any long-running shop and you will find steps that make no sense until somebody explains them. The purchase book is checked twice on Tuesdays. Nothing moves between two of your own godowns without a slip. One supplier's goods are counted at the gate and never at the rack. Every one of those oddities is a scar: something went wrong once, and the process grew a step to stop it happening again. Software that calls all of that inefficiency is throwing away thirty years of debugging.

Which is why the first question should never be "what does the software do" but "how does your shop actually work" — and the honest version of that takes an hour and a notebook, not a demo. We have written separately about how to tell a software team what your business actually does. The job is to keep what the process is protecting and automate only the part that is repetitive, error-prone or invisible.

The fear of a failed changeover is statistically well founded

Here is where owners are usually told they are being irrational. They are not.

Bent Flyvbjerg and Alexander Budzier of Oxford examined 1,471 IT projects and found an average cost overrun of 27 per cent — and, more to the point, that one project in six ran about 200 per cent over cost and almost 70 per cent over schedule. That was published in the Harvard Business Review in 2011 and nothing since has overturned it.

Those were large corporate projects, not a kirana shop buying billing software, so the scale does not transfer. The instinct does: an owner who suspects a changeover can go badly wrong is quoting the base rate correctly.

A second finding is worth sitting with. Gartner surveyed 1,120 technology buyers in 2022 and found that 56 per cent reported high regret about their largest technology purchase of the previous two years — and that regret peaked before implementation had even begun. Being unsure at the point of buying is not a character defect. It is the most common state a buyer is in.

The failures that get remembered also share a pattern. Hershey went live in July 1999, weeks before its Halloween season; order processing broke, deliveries stretched out, and fourth-quarter sales fell about 11 per cent. The lesson to take from that is not don't. It is not now — not in your season.

If you sell crackers, do not change systems in October. A sweet shop, not before Diwali. A uniform and stationery dealer, not in May. Switch in your quietest month, when a bad week costs you a bad week instead of a year. That single decision removes more risk than any feature comparison.

Go live in stages, and keep writing the book for a month

The second thing that removes risk is refusing to do it all at once.

A sensible order for most shops:

1. Billing first. It is the most repetitive, the most visible, and the easiest to judge. You will know within a week whether it is faster.
2. Then stock, once billing has been feeding it for a few weeks and the opening counts have settled.
3. Then purchases, which is where the money actually leaks.
4. Then receivables and follow-ups, which is where it sits stuck.
5. Then reporting, mobile access and everything else, if and when you want it.

At each stage the owner should see the benefit before the next one is asked of him. If he cannot, stop and find out why rather than pressing on.

Then the unglamorous advice that nobody selling software wants to give: keep writing the register for the first few weeks. Yes, it is double work. It is also the cheapest insurance there is, and it ends by itself — one morning somebody reaches for the book, realises the screen already told him, and stops. That day is the real go-live date, and it should be allowed to arrive on its own.

An honest note here: there is no good study proving that phased rollouts fail less often than all-at-once ones — people assert it constantly, but the research is not there. The reason to phase is different and simpler. It does not make failure less likely. It makes failure small.

"Will my employees be able to see my margins?"

This is asked quietly, usually near the end, and it is often the real question behind all the others. It deserves a straight answer, and the straight answer is that on this point paper is the weaker option.

A register sits on a counter. Whoever is standing at that counter can read it, including the purchase rates written on the facing page. The man who has worked for you for eleven years has known your landed cost on every line for about ten and a half of them.

A system does the opposite of what the fear expects, because access follows the job. In Stock2Track the person billing sees selling rates and stock; he does not see purchase cost, margin, or what the business made yesterday, unless you decide he should. That is not a restriction placed on you. It is a restriction becoming available to you for the first time.

A second half gets mentioned even less. In a book, an altered figure looks like an altered figure and that is all you will ever learn about it. In a system, who changed what and when is recoverable — which matters most for the thing owners worry about and rarely say: not outsiders, but somebody inside. We have gone through that properly in the controls every growing shop sets, and how each one gets tested, including a customer who traced a ₹40 lakh loss in transfers between his own locations.

And when an employee leaves — which he will — a book system loses whatever he was carrying in his head. A recorded one does not.

"Will I get my money back?"

Small business owners do not ask about digital transformation. They ask a better question: what do I get for this? The honest answer is that the return almost never shows up as extra sales. It shows up as losses that did not happen:

• A purchase not placed on a line that already has four months of cover sitting in the godown.
• A payment chased in week three instead of remembered in month four.
• A billing error not made, because nobody typed the rate.
• A short delivery caught at the gate instead of a fortnight later.
• A slow line spotted while it is still slow rather than while it is dead.

None of those are dramatic. A handful a month, in a shop doing a few lakhs a week, recovers the cost and then keeps going. So the real question is not "how much extra did it sell" but "how much did it stop me losing, and how much of my attention did it give back".

Now the counterweight, because this article is not a sales pitch. If your shop is small enough that you genuinely do hold all of it in your head, and it is working, the return may be thin. One counter, sixty or seventy lines, one person who sees every transaction — that business may be better served by a good manual discipline than by software it will half-use. We would rather tell you that than sell you a licence you resent in year two.

The calculation changes the moment a second person, a second location, or a real jump in line count arrives. That is when the head stops being enough, and it rarely gives notice.

"I am not comfortable with computers"

This may be the largest barrier of all, and it is almost never said out loud, because saying it costs something.

A man has run a business for thirty years. He negotiates better than anyone in the room and knows hundreds of rates from memory. Then somebody puts a laptop in front of him, in front of his own staff, and he feels like a beginner again.

That feeling is the resistance. Not the software. Two things follow from taking it seriously.

The software has to speak business, not computing. Sale. Purchase. Stock. Payment. Customer. Supplier. Outstanding. Not "master data", not "entity", not "sync". A man drives a car without knowing what the engine control unit does, and he is not a worse driver for it.

And the training should happen on his terms. We have taught Stock2Track to people who had never used a laptop. For daily operations, a focused session of two to four hours with our team is usually enough to get going, with the rest arriving through ordinary use. If the owner would rather do that after the shutter is down, with nobody watching, that is a completely reasonable request and one that we are asked more often than you would think.

Nobody needs to become a technical person. The aim was never to teach the shopkeeper technology. It was to get the technology to learn the shop.

"My accounts have always been in my language"

This one is usually treated as a minor convenience. It is not minor.

Kantar and the IAMAI reported in 2024 that of India's roughly 886 million internet users, 98 per cent access the internet in Indian languages — and that even in urban India, 57 per cent prefer content in an Indian language over English. The EF English Proficiency Index for 2025 places India 74th out of 123 countries, in the "low proficiency" band, and EF is careful to note that its own sample over-represents exactly the connected, educated people most likely to score well. The true picture is worse than the ranking, not better.

So English-only software is not a neutral default. It is a filter on who in your shop is allowed to use the system — and it usually filters out the people who have been there longest.

Stock2Track works in Indian regional languages for this reason. A person's ability to run a business has never depended on his English, and his software should not be the first thing to suggest otherwise.

Automation is not handing the business to a computer

Software reminds, calculates, compares and records. It does not decide. You decide, exactly as you did before, with better information in front of you — and the pattern is the same every time:

• You already know a line sells in season. The system tells you how much, and how that compares with last year.
• You already know one supplier runs late. The system shows you the pattern, in days, which is a different conversation to have with him.
• You already suspect stock is walking. The system shows you where and when, which is the only version of that suspicion you can act on.
• You already know some customers delay. The system makes sure nobody forgets to call them.

Experience and records are not competitors. The owner who has both is harder to catch out.

And the risk is slowly moving to the other side of the argument. GST taxpayers have gone from about 66.5 lakh in 2017 to about 1.65 crore by mid-2026, and the e-invoicing threshold has come down in steps from ₹500 crore in October 2020 to ₹5 crore since August 2023. That argument is routinely oversold, though: below the ₹40 lakh registration threshold for goods none of it is forcing your hand, and under ₹5 crore, e-invoicing is not your problem this year.

The pressure that actually bites is quieter. The number of things happening in a day keeps going up — more lines, more payment methods, orders on WhatsApp, a second location, staff who change. Memory scales beautifully until it doesn't, and it gives no warning on the way.

Which is worth sitting with, given how this article started. The register your father kept was a system of record he could hand to you. The version that now lives only in your head is not.

Start by removing one headache

None of this requires automating a business in one go. It requires picking one thing.

Pick whichever of these you would most like to stop doing:

• Writing bills by hand at the busiest hour.
• Counting stock to find out what you have.
• Working out what a customer still owes you.
• Deciding what to buy without knowing what did not move.
• Knowing what happened in the shop on a day you were not there.

Fix that one. Let the owner see the work go down before anything else is asked of him. Once somebody has watched automation remove work rather than add it, the fear stops being about software and becomes an ordinary question of what to do next.

The best first question is not "how do I change my business to use software".

It is "which part of what I already do could be easier by Friday".

The short version

• The resistance is not fear of technology — these are people already running on UPI and WhatsApp. It is a correct instinct about replacing the system of record.
• The old process is accumulated repair, not sentiment. Keep what it protects.
• Never change systems in your season. More small-business risk lives in that one decision than in any feature comparison.
• Go live in stages and keep writing the register for a few weeks. Phasing does not make failure less likely; it makes it small.
• On privacy, software is the stronger option. A book on a counter has no permissions at all.
• The return shows up as losses prevented, not sales gained — and if your shop is small enough to run from memory, that is a fair reason to wait.
• Language and vocabulary matter more than the feature list, and so does training the owner when nobody is watching.

If you recognise your own shop in any of this, the useful next conversation is not a demo. Tell us how your counter actually runs, what you would most like to stop doing by hand, and which month is your quietest. The answer usually falls out of those three.

---

Where the figures come from: the MSME technology-adoption figures are from MSME Digitalisation in India: Current Status and Challenges (Research and Information System for Developing Countries, 2025, n=2,882); the IT project overrun figures from Flyvbjerg and Budzier, Harvard Business Review, September 2011 (n=1,471); the purchase-regret finding from Gartner's 2022 technology buyer survey (n=1,120); the Indian-language internet figures from the Kantar–IAMAI Internet in India 2024 report; the English proficiency ranking from the EF English Proficiency Index 2025; and the GST taxpayer and e-invoicing figures from Government of India press releases of June 2026. The Hershey and Sainsbury's cases are drawn from contemporaneous press reporting in 2000 and 2004. Thresholds and compliance rules change — check your own position with your tax adviser rather than with us.