Spreadsheet, CRM, or Custom? A Simple Rule

Every few months a business owner tells us they need a CRM — customer relationship management software, meaning a system that keeps a record of every enquiry, who owns it, and what happened next. Usually the trigger is that somebody has dropped a lead, or the sales person who left took their contacts with them.

That is a real problem. But the solution is almost never the one being asked for. A CRM is not a cure for an undefined process. It is a container for a process you already have. If nobody can describe, in one sentence, what happens between an enquiry arriving and a quote going out, no software will invent that sentence for you.

So here is the decision rule this article argues, and it has nothing to do with how big you are or how much you turn over:

The question is not how many leads you handle. It is how unusual your process is. Common process, off-the-shelf tool. Genuinely unusual process, consider a build. Unclear process, spreadsheet.

And most businesses — including profitable ones with twenty staff — should start with the spreadsheet.

Why the spreadsheet is the honest first answer

A shared Google Sheet with ten columns will run a sales pipeline for a surprisingly long time. Date, name, phone, source, what they asked for, owner, stage, next action, next action date, notes. That is it. Add conditional formatting so anything with a next-action date in the past turns red.

What you get for free, and would pay for elsewhere: everyone can see it, nobody needs training, it works on a phone, it costs nothing, and you can change a column at 11pm because you thought of something better. That last one matters more than anyone admits. In the first year of any process, you are wrong about the stages. A spreadsheet lets you be wrong cheaply.

What the spreadsheet genuinely cannot do:

• Stop two people calling the same lead. There is no record locking in practice. People overwrite each other.
• Remind anyone of anything. Red cells only work if someone opens the sheet.
• Keep an audit trail you can trust. Version history exists, but nobody reads it.
• Restrict who sees what. If your sales team should not see each other's margins, a sheet is the wrong place.
• Capture the enquiry automatically. Somebody types it in, or it never arrives.

Read that list honestly. If none of those five is currently costing you money, you do not have a CRM problem. You have a discipline problem, and buying software to fix a discipline problem gives you the same problem plus a monthly bill and a login nobody uses.

The spreadsheet stops working at a fairly identifiable moment: when more than about three or four people are writing to it at once, or when the cost of one forgotten follow-up is larger than a few thousand rupees. A property broker losing one site visit has crossed that line. A tuition centre with fifteen enquiries a month has not.

What off-the-shelf actually buys you

The ready-made CRMs — Zoho, HubSpot's free tier, Freshsales, Pipedrive, and the dozens of smaller Indian tools — are all solving the same five gaps listed above. They give you one record per person, ownership, reminders, permissions, and a form-to-record connection so web enquiries land without retyping.

They are, genuinely, good value. Zoho in particular is priced for the Indian market in a way the American tools are not, and it bills in rupees. For a business doing ordinary business — an enquiry comes in, someone qualifies it, a quote goes out, it is won or lost — an off-the-shelf CRM will fit about ninety per cent of what you do on day one.

The trap is the other ten per cent. Every off-the-shelf CRM assumes a shape: a linear pipeline of stages, one contact, one deal, one value, one close date. If your work does not have that shape, you will spend the next two years bending the tool, and the bending is where the money goes.

Signs off-the-shelf will fit you well

• One enquiry becomes one quote becomes one order.
• Your stages have names a stranger would understand — new, contacted, quoted, won.
• Pricing is a list, or a list plus a discount.
• Handover to delivery is a WhatsApp message, not a system.

Signs it will fight you

• One enquiry produces several quotes at different specifications, and the customer picks one.
• Your price depends on a calculation — square feet, weight, distance, GST slab, material rate that moves weekly.
• Deals take eighteen months and pass through four departments, each with their own checklist.
• You sell through dealers, so the person enquiring is not the person buying and is not the person paying.
• The same customer has a service relationship, a recurring order and a project running at once, and they must not be conflated.

Note that none of those is about size. A four-person fabrication shop quoting on drawings has an unusual process. A hundred-person distributor selling a fixed catalogue does not.

When something built for you is the cheaper answer

A custom system is justified in one situation: your process is a real competitive advantage, it does not match the shape any standard tool assumes, and you have run it long enough to describe it precisely.

All three conditions. Miss the third and you will pay to have your confusion encoded in software, which is much harder to change than a spreadsheet column.

The honest cost comparison people get wrong: off-the-shelf looks cheap because you see the subscription. A custom build looks expensive because you see the whole price at once. But the subscription is per user per month, forever, rising, and the customisation work to force it into your shape is billable too. Over five years with fifteen users, the numbers are often closer than they appear — and with a build you own the thing.

The honest cost comparison people also get wrong in the other direction: a custom system needs somebody to maintain it. Servers, backups, browser changes, a staff member who wants a new field. If you are not prepared to budget for ongoing technical support, do not commission a build. An unmaintained custom system is worse than any off-the-shelf tool, because when it breaks there is no help documentation and no forum.

The middle path almost nobody offers, and the one we suggest most often: keep a standard CRM for contacts and pipeline, and build only the one unusual piece. The quotation calculator. The dealer portal. The approval chain. Custom where you are unusual, bought where you are ordinary. That is usually the cheapest total answer, and it is what most of our CRM work ends up looking like.

What we would not buy

Things sold in this category that we would tell you to decline, including things an agency could happily bill you for:

A CRM bought before the process exists. The single most common waste. If you cannot draw your pipeline on a napkin, you are not ready. Run the spreadsheet for three months first, then buy. The spreadsheet is the requirements document.

Data migration of your entire history. You will be quoted to move eight years of contacts. Most of them are dead. Move the last twelve months of active records, archive the rest as a CSV file, and get on with it. Migration budgets consumed by junk data are the reason CRM projects overrun.

A fully custom CRM for a business with a normal pipeline. If you sell a catalogue to walk-in and web customers, Zoho will do it. Anyone quoting you for a bespoke replacement is selling you a slower, more fragile version of something that already exists.

Per-user licences for people who will never log in. Count the people who genuinely type into the system. Your accountant probably needs a report, not a seat. Delivery staff probably need a WhatsApp message.

Automation bought at the same time as the CRM. Automating a process you have not yet run inside the new tool means automating guesses. Get the records clean, watch what humans actually repeat, then look at automating the repetition. That order round matters.

A CRM as a fix for lost web enquiries. If your contact form is silently failing or landing in spam, a CRM will faithfully record zero leads. That is a website problem, and it is cheaper to fix.

A test you can run this week

Before you speak to anyone selling software, do this. Take your last twenty closed enquiries — won and lost. For each one, write down the steps that actually happened, in order, in your own words.

Then compare the twenty lists.

• If they are broadly the same four or five steps: your process is common. An off-the-shelf CRM will fit. Choose on price, Indian support hours, and whether it connects to your website and phone system.
• If they diverge wildly and you cannot see a pattern: you do not have a process yet. Spreadsheet. Three months. Come back.
• If there is a clear pattern but it does not look like a pipeline — branches, loops, several quotes per enquiry, a calculation at the centre: you are the genuine candidate for something built, or for a standard tool plus one custom piece.

That exercise takes an afternoon and it is the most useful thing you can do. It also means that when you do talk to a vendor, you are describing your business rather than being told what your business should look like.

What to do next

If you are on paper, WhatsApp and memory: build the ten-column sheet this week. Do not buy anything yet.

If your sheet is straining — people overwriting each other, follow-ups missed, no way to restrict visibility — shortlist two off-the-shelf tools and trial both with real leads for a fortnight, not with demo data.

If you have done the twenty-enquiry exercise and your process genuinely does not fit a pipeline, that is worth a conversation. Bring the twenty lists. Tell us what you found and we will tell you honestly whether you need a build, one custom piece bolted onto a standard tool, or another six months on the spreadsheet. Sometimes the answer is the last one, and we would rather say so than sell you something you will abandon.

Questions we get asked