Almost every growing business we meet runs on the same setup. CRM in one tool. Invoicing in Tally. Inventory in a Google Sheet. Customer support in WhatsApp. Bookings or jobs in a third app. Reports cobbled together once a month from CSV exports.

It’s nobody’s fault. Each tool got added because it solved a real problem at the time. Then the next problem came, and the next tool was the fastest fix. After three or four years you wake up paying for five or six different SaaS subscriptions, and your team is the integration layer.

For a long time this works fine. The question we keep getting is: when does it stop working?

The point where the maths flips

The honest answer is: it stops working before anyone notices it has. Most teams keep going for another year or two on willpower.

We’ve seen it play out the same way three or four times now. There’s no single moment. You just start hearing the same phrases in standups every week:

  • “I’ll have to check Tally and the sheet to confirm.”
  • “That’s already there in WhatsApp, did nobody update it?”
  • “The dashboard doesn’t match what I’m seeing in the system.”

When those phrases hit a critical mass, you’ve already lost the productivity gain that buying the tools was supposed to give you. The team has built a parallel manual workflow around the software. The software is now a tax, not a tool.

What the patchwork actually costs

People usually price their stack by the line items on the monthly invoice. Five tools at four to ten thousand rupees each works out to roughly thirty to fifty thousand a month. That feels manageable.

The cost that isn’t on the invoice:

  • Double data entry. Same customer, same job, entered into three places. We’ve timed this in client offices. For a 10-person team it’s typically 6–9 hours per person per week. That’s a full extra hire’s worth of time, gone.
  • Reconciliation work. Someone’s spending two or three days a month making the numbers in tool A agree with tool B. If that person is your accountant, it’s quiet. If it’s your operations head, it’s loud.
  • The decisions you can’t make. This is the expensive one. You can’t see real margins by branch, can’t see which lead source actually converts, can’t see which technician is twice as productive as the others. The data exists, but it’s in five places, so you make the call on instinct.

That last cost is invisible in any spreadsheet, which is why most owners catch it last. By the time you can put a number on it, the patchwork has already cost you a year of growth.

When custom actually wins

Custom software is a real commitment. Money, time, and the discipline to stick with one platform when shiny new SaaS tools keep appearing. It’s not for everyone, and we tell people that on the discovery call.

A quick rule of thumb we use in client conversations:

  1. You have three or more user roles doing genuinely different jobs in the same system (not just “manager” and “user”).
  2. You’re paying for four or more separate SaaS tools that don’t sync, or sync badly.
  3. Your business has specific workflow rules that no off-the-shelf tool models — GST compliance, multi-branch inventory transfers, industry-specific approvals, anything similar.
  4. You’d want the same system to scale with you for at least the next three years.

If three of those four are true, custom is worth a conversation. If only one or two are true, fix the SaaS integration first. There are usually 80% solutions hiding in Zapier or n8n that buy you another year cheap.

What custom actually looks like

The phrase “custom CRM” gets thrown around a lot. What we usually end up building is closer to an operations platform — a single system that maps to how your business actually runs, role by role.

For a multi-branch service business it tends to look like this: one job ID per customer interaction, traceable from intake through delivery; one customer record with full history visible to anyone who picks up the phone; inventory, invoicing, and reporting all reading from the same database. Eight to fourteen modules, depending on the business, all sitting on a single Laravel or Node.js codebase that you own.

The cost works out to somewhere between five and twenty lakhs depending on scope. The trade-off is straightforward: you pay it once instead of every month, and you stop being the integration layer.

Whether that’s the right move for your business depends on the four questions above, not on a sales pitch.

If you’ve been staring at your SaaS invoice wondering whether it’s time to consolidate, book a free 30-minute call. We’ll talk through your specific setup and tell you honestly whether custom makes sense, or whether you should just fix the integrations for another year.