Custom Software vs Spreadsheets: When to Build

There is a moment in most small businesses when the spreadsheet stops being the solution and starts being the problem. The tabs multiply, one person becomes the only one who understands the formulas, and every new hire learns the shortcuts by osmosis instead of reading a manual.

The question that follows — buy more SaaS, hire more people, or pay to build something custom — is one of the most expensive calls a founder makes. The posts that rank for "custom software vs off the shelf" are almost all written by agencies, which means they almost all end with the same answer: build. This one is written by an agency too, so it is only fair to tell you upfront where we disagree with our peers.

I'm Tarun, founder of Gigzman. We build custom software for Indian SMBs starting from ₹2,00,000 with a typical timeline of 8 to 16 weeks. We also talk at least a third of the people who ask for a quote out of commissioning one, because for their business the honest answer is a SaaS subscription. This post is how we get to that answer.

Start with one question, not a feature list

Before the pros-and-cons table, run the test Martin Fowler laid out in his classic note on utility versus strategic software: it is all about whether the underlying business function is a differentiator or not. If the answer is no — if a competitor doing the exact same thing would not change who wins the deal — the function is a utility, and the rule is to buy the package and adjust your business process to match the software.

Fowler's own caveat is worth posting on the wall: people tend to overrate how much of their work is strategic. Payroll, accounting, basic CRM, e-commerce checkout, email, scheduling, HR leave — these are utilities for almost every business. They feel personal because you live inside them, but no customer picks you because your accounting is unique. Buy the package, bend your process.

The strategic part — the thing a customer would actually notice — is usually a lot smaller than it feels. It is the part worth building.

When custom software is the right call

These are the signals we take seriously, lifted from the business owners we have actually built for. If three or more sound like you, custom is on the table.

• A core process runs on spreadsheets and memory. The written process and the real process differ in every business we have worked in, and it is the real one that has to be built.
• You are paying for several tools that do not talk to each other. New software that cannot see the existing data just adds a place to type things twice.
• Headcount is growing to cope with manual work. If the next hire's job description is mostly re-typing things from one system into another, you are about to put payroll behind a workflow.
• The workflow is yours, not the industry's. A travel marketplace that routes thousands of operators by a trust score you invented is not going to find that in an off-the-shelf tool.
• Per-seat costs or export limits are deciding your org chart. When you start leaving people off the SaaS seat list to save money, the tool is making your decisions for you.
• Permissions and audit are actually required. Who can see what, who changed what, and when — retrofitting this into a live system is painful and expensive. If a regulator is going to ask, build it in from the start.

None of these alone forces a build. Three of them, plus a process that is actually yours, usually does.

When it is not the right call

This is the section the top five ranking posts leave out. If any of these is true, keep your money.

• An off-the-shelf product covers most of what you need. Buy the product. Custom software is worth it for the part of your business that is genuinely yours, not for the parts everyone has.
• You are still figuring out the workflow. Locking a half-understood process into custom software is how you pay twice: once to build it, once to rewrite it. Use SaaS to learn how your business really operates first.
• One person in the team wants to replace the tool because they do not like it. Dislike of Tally or Zoho is not a build case.
• The pain is a configuration problem, not a product problem. Half the "we need custom software" conversations we have end with a two-hour session configuring the existing tool properly.
• You cannot name three people who will use it daily for two years. If adoption is uncertain, pilot with a SaaS tool and a few spreadsheets, not with a build.
• The budget assumes a one-time cost. Custom software needs a maintenance line in next year's P&L; if that line scares you, stay on SaaS.

If none of these is true and three of the "build" signals are, you are probably at the point where custom software pays off.

Score your own business

Fill this in before you ask anyone for a quote. Every "yes" tilts you further toward building.

• Spreadsheets as system of record? A core process depends on a workbook only one or two people understand.
• Three or more tools that do not integrate? You are already paying integration tax in people's time.
• Unique process that affects revenue? A customer would notice if you did it the way everyone else does.
• More than ten seats across your SaaS stack? Per-seat pricing is starting to drive decisions.
• Audit, compliance or permissions requirement? Someone will ask who changed what, and when.
• Workflow stable for a year? You have stopped reshuffling how it works on a weekly basis.
• A named team who will use it daily? Not a hypothetical team — an actual roster.

All seven yes — commission a build. Two or three — buy the SaaS, configure it properly and revisit next year. In between — start with a paid discovery, not a quote.

What a sensible build in India actually costs

The custom software posts that rank for this query quote either US dollars or very wide ranges. Here are the real starting floors we publish on our own pricing page, in rupees, alongside the common SaaS alternative so you can run the comparison.

What you are considering	Buy the SaaS (typical range)	Build it, starting at (Gigzman)
CRM / pipeline	₹500–₹5,000 / seat / month	₹1,25,000 for a CRM built around your sales motion
Internal operations tool (jobs, inventory, scheduling)	₹1,000–₹15,000 / month, per module	₹2,00,000 for software development, 8–16 weeks
First version of a product you want to sell	SaaS does not apply	₹1,50,000 for an MVP, 6–12 weeks
Customer-facing e-commerce	Shopify / Woo plans	₹85,000 for e-commerce development
Multi-tenant AI SaaS	SaaS does not apply	₹3,00,000+

These are floors, not averages — a project with more integrations, more complex permissions or a design from scratch lands above them. The final quote follows a documented scope, because a number given before anyone knows the scope is a number that changes later. The two drivers that move the build number the most are the integration list and whether anyone has written down what the thing actually does yet.

The break-even sanity check we run before quoting: add up your current and projected SaaS bills, the salary cost of the manual work the new system would absorb, and the cost of the next hire you would make instead of building. If the three-year total is well below the build number, buy the SaaS.

If the answer is "build", start with a discovery, not a quote

A quote for custom software that does not come after a discovery is a guess. The way we scope these is a paid discovery phase first, with its own fixed fee and its own deliverable — the process map and the phased plan. You can take that and build it anywhere. Only then do we quote the build, and by then it is a real number rather than a guess.

The discovery is also the test: if sitting with the people doing the job does not change the written process in at least a few places, you did not need custom software — you needed to write down what you already had. In our experience, the real process and the documented one differ in every business.

Phased delivery is not optional

The way most two-year internal software projects happen is big-bang scope. We build the highest-pain module first, in production, used by real people, before the second one starts. This is scoped in our process stage too: every screen, integration and dependency written down with an estimate beside it, so nothing lands as a surprise mid-build. Each cycle lives on a staging URL you can open at any time.

Two things come with that choice. First, you can stop after module one if your business changes and keep the value already in production. Second, your early users shape the second module, so it fits the real job instead of the brief you wrote six months ago.

What you walk away with

Whichever module we start with, the handover is the same. You get a process map of how the work is done today, a phased plan with the first module dated, the system in production with your data migrated, admin tooling so you are not dependent on us for routine changes, and documentation written for your team, not for developers.

Two deliberate design choices behind that list: the code is in your repository from the first commit and the documentation is written for a developer who has never met us. Maintenance from us afterwards is on a monthly retainer, and it is optional — plenty of clients run their own.

Before you brief any agency

Write these down first. Send them to every agency you ask for a quote.

• Which processes you want replaced, in priority order, with the current tool for each.
• The three to five SaaS tools the new system has to read from or write to.
• Who edits data, who approves it, and who audits it.
• The team who will use it daily, and the training time you can spare.
• Whether you want the first module live in a quarter or the whole thing live in a year.
• Whether you plan to maintain it in-house afterwards.

Agencies that come back with a flat number and no questions have guessed. Ones that come back with a discovery proposal and a timeline for the quote itself have read the brief.

See what Gigzman's engagement looks like

If after this you think custom is the right answer, our software development page lists what is in the engagement, and the pricing page carries the floors used in the table above. If you think it is not, keep the SaaS and the two hours this saved you.

Either way, if you want a second opinion on your own list, send us the brief and you will have a clear plan, a timeline and a price within two working days.