Custom Applications vs Readymade Software

Most businesses do not choose between custom and readymade software. They drift into readymade because it is the thing you can start using this afternoon, and then discover three years later that they are paying every month to work in a way that does not quite fit, on data they cannot easily take with them.

That is not an argument against subscriptions. It is an argument for making the decision deliberately, once, with the five-year number in front of you rather than the monthly one.

The monthly fee is the whole argument

A subscription is priced to feel small. That is the design. The number that matters is not what it costs this month, it is what it costs across the life of the business, and that number has three properties owners consistently underestimate.

1. It never stops. A custom build is a one-time cost with maintenance after it. A subscription is rent, and rent does not end while you are still using the room.
2. It rises. Renewal increases, tier changes, and features quietly moving into a higher plan are all normal in this market. You have no leverage in that conversation, because leaving means migrating your data and retraining your staff.
3. It scales with your success. Per-user and per-transaction pricing means the better your year, the larger the bill. You are, in a small way, sharing your growth with a company that did nothing to cause it.

Run the arithmetic before anything else. Take the monthly figure, multiply by twelve, multiply by five, then add a modest annual increase and the extra seats you expect to need. Put that total next to a one-time build cost. For a lot of businesses the two numbers cross somewhere in the second or third year, and everything after that is money that could have been an asset.

You end up shaping the business around the software

This is the cost nobody puts on a spreadsheet, and in our experience it is the larger one.

Readymade software encodes somebody else's idea of how your work should be done. Where your process differs, and it always differs, one of two things happens. Either you change how you work to suit the software, or a member of staff invents a workaround — a parallel register, a WhatsApp group, a spreadsheet that the real numbers live in. The second is far more common, and it means you are now paying for a system that does not hold the truth.

• A field you need that the form does not have, so it goes in the notes box and can never be reported on.
• An approval step your business genuinely requires, which the tool has no concept of, so it happens over phone calls.
• A discount structure, a credit arrangement or a seasonal rate that the pricing screen cannot express.
• A report the owner wants every Monday that has to be assembled by hand from three exports.

Each of these is a few minutes a day. A few minutes a day, across a few staff, is a salary. Custom software starts from what you actually do, so those minutes never get spent.

The local fit that global software gets wrong

A great deal of readymade software is built for a market that is not this one, and the gaps show up in exactly the places that matter to a business in Ranchi.

• GST structures, invoice formats and the state-wise rules that go with them.
• UPI as a first-class payment method rather than an afterthought bolted on beside cards.
• WhatsApp as the channel customers actually reply on — not email, which many will never open.
• Bilingual and Hindi-first screens for staff who are far faster in one language than the other.
• Screens that tolerate a dropped connection, because the person using them is on mobile data in a market or on a site, not on office fibre.
• Serviceability by pincode, local courier partners, and cash on delivery with a confirmation step.

Some readymade products handle a few of these well. Very few handle all of them, and the ones that do are usually the expensive tiers.

Your data, and how hard it is to leave

Ask any prospective vendor one question: if we leave in three years, what exactly do we get back, in what format, and how long does it take?

The answers are revealing. A full structured export, on demand, with the relationships between records intact, is a sign of a confident product. A CSV of the main table only, or an export that requires raising a support ticket, is a lock-in mechanism whether or not anyone would describe it that way. Your customer history, your pricing, your order records and your documents are the business — not a feature of the software you happen to be renting.

With a custom application the data sits in a database you control, on hosting in your name, and you can take a backup of the whole thing tonight without asking permission.

Things a vendor can do to you that you cannot prevent

• Discontinue the product, or the plan you are on.
• Change the terms, the pricing model, or the storage limits.
• Remove a feature you depend on, or move it into a higher tier.
• Be acquired, after which any of the above becomes more likely.
• Suffer an outage on the day you most need the system, with no recourse beyond a status page.

None of these are hypothetical and none of them are unusual. They are simply the terms of renting rather than owning. Custom software has its own risks — mainly that you need a developer who will still be there — but the failure modes are ones you can act on.

What "custom" actually buys you

• Software shaped to your process, so nobody keeps a parallel register.
• Only the features you use, which means less training and fewer places to make a mistake.
• Integration with what you already run, rather than another system to reconcile by hand.
• One-time cost, then maintenance — a capital item, not an indefinite operating expense.
• The code and the data in your name, with ownership transferring on final payment.
• The ability to add the thing you need next quarter, instead of waiting for a roadmap you do not control.

That last one is worth more than it sounds. The competitive advantage in most businesses is a handful of small things done differently. Readymade software, by definition, gives you the same capabilities as every competitor who bought the same product. Custom software is the only kind that can encode the part you do better.

When readymade is genuinely the right answer

Often. We would rather say so now than have you find out after paying for a build.

• The need is a commodity. Email, accounting, payroll and document storage are solved problems, solved well, and building your own would be a mistake.
• Your process is genuinely standard, and you can say so honestly rather than hopefully.
• You are early and still changing your mind weekly. Rent while the shape of the business is still moving; build when it settles.
• The budget is not there yet. A subscription now beats a half-finished custom build, every time.
• You need it working this week. A build takes weeks; a subscription takes an afternoon.

There is also a middle path that suits a lot of businesses: keep readymade for the commodity parts, and build custom only for the one workflow that is actually yours. A clinic can run its accounts on off-the-shelf software and still need a booking and records system built around how it really operates.

How to decide, in five questions

1. What is the five-year total of the subscription, including expected growth in users and price?
2. How many minutes a day do staff spend working around the software today? Multiply that out.
3. If you had to leave in three years, what would you get back, and in what format?
4. Is the thing you do differently from competitors something this software can express, or something you do despite it?
5. Is the process settled, or still changing month to month?

If the five-year number is large, the workarounds are real, the export answer is vague and the process is settled — build. If the need is a commodity, or the business is still finding its shape — subscribe, and revisit in two years with better information.

The wrong answer is the one nobody chose: drifting into a subscription because it was easy, and still paying for it in year six while a spreadsheet quietly holds the real numbers.