Why I Turned Down ₹6L: The Junagadh Margin Math 2026

Quick Answer: I turned down a ₹6,00,000 project in 2026 because the real effective rate was ₹410 per hour after a 90-day payment delay, three revision rounds, and 40 hours of unpaid presales. The gross number looked like my best month; the margin math showed it was worse than a ₹2,40,000 project with a 14-day payment term.

Author: Deepak Bagada — founder of SaaS Next in Junagadh, Gujarat. I have accepted a bad-fit project and paid for it in sleep, and I have declined a bigger one and paid the bills with better work; the pricing mistakes from year one are the origin of this checklist. This is the checklist I now use before signing anything above ₹3,00,000.

Key Takeaways

• Gross revenue is not income. A ₹6L project with 1,460 hours of real work is a ₹410 hourly rate, not a premium engagement.
• Payment terms are part of the price. Ninety-day terms on a ₹6L invoice cost about ₹24,000 in working capital at a 16% credit line, before a single late fee.
• Scope ambiguity is a hidden discount. "We will define the details in sprint one" means the client buys an option to add work at your expense.
• A no is a portfolio decision. Declining a low-margin project leaves capacity for two better ones, and the referral cost of a resentful delivery is higher than the lost invoice.

The Offer on Paper

The enquiry arrived in June 2026 from a Mumbai founder building a B2B marketplace. The headline was attractive: ₹6,00,000 for a six-month build, paid in three milestones. I had capacity, and the stack was familiar: Laravel 13, Postgres, Redis, and a Next.js storefront on a 4GB VPS.

Three terms made the project dangerous: unlimited revisions, an undefined scope, and net-90 payment. Each one is a margin leak. Together they turn a fixed fee into an open-ended commitment.

The Margin Math

I estimated the work at 980 hours before the first call and 1,460 hours after mapping the actual marketplace flows. The difference came from seller onboarding, payout reconciliation, and an admin approval workflow that was not in the brief.

The effective rate is ₹305 per hour, not the ₹410 I quoted in the quick answer, because the revision overrun is a probability-weighted estimate. Even the optimistic version, with no revision overrun, lands at ₹346 per hour. My target for custom work is ₹1,000 per hour. The project was a 65% discount disguised as a large contract.

The comparison that made the decision obvious was a smaller project already in the pipeline: ₹2,40,000 for 210 hours, net 14, with a written scope. I track these numbers in a JSONL ledger, and the p95 API latency target for that build was 300ms on a 4GB VPS. That is ₹1,143 per hour before costs. Two of those projects fit in the same calendar window and produce more cash with less risk.

Why Net 90 Is a Price Cut

A ₹6,00,000 invoice paid in 90 days is not the same as ₹6,00,000 paid in 14 days. I run a solo studio, so the gap is funded by my own reserve or a credit line. At 16% annual interest, 76 extra days on ₹2,00,000 costs about ₹6,700 per milestone. Across three milestones, the working-capital cost is roughly ₹24,000.

The number is small against ₹6L, which is why founders ignore it. The risk is not small: if the client disputes the final milestone, the last ₹2,40,000 can sit for six months. A solo studio cannot absorb that while paying subcontractors and infrastructure.

I now price payment terms explicitly. Net 14 is the default. Net 30 adds 2%. Net 60 adds 5%. Net 90 is a no unless the client is a listed company with a procurement process I can verify.

The Scope That Was Not a Scope

"Full marketplace, details in sprint one" is not a scope; it is an invitation to negotiate every screen twice. I have learned to test this before quoting. I ask for three things: a list of user roles, a list of money flows, and a list of integrations. If the client cannot provide them, the project needs paid discovery first.

For this project, the three lists would have revealed the real size: four roles, three payout paths, and integrations with two logistics APIs and a GST service. That is not a six-month build with one developer; it is a product team.

The scope creep clause I use in Gujarat would have helped after signing, but it would not have fixed the initial estimate. A change-order process manages additions; it does not rescue a scope that was never defined.

Production War Story: The ₹4.2L Project I Should Have Declined

In 2025 I accepted a ₹4,20,000 project with terms similar to this one. The client was friendly, the demo looked simple, and I needed the cash after a slow quarter. The build ran five months, the scope grew by 60%, and the final payment arrived 118 days after launch.

The work itself was fine. The cost was the opportunity: I turned down two smaller projects during the same period, and I spent the last six weeks in a resentful sprint instead of selling the next engagement. The invoice cleared; the referral did not. The client never introduced me to anyone, because the project had become a negotiation rather than a win.

That experience is why I built the checklist below. It is not a moral position; it is a cash-flow and energy position.

The Six-Question Qualification Checklist

I run this before any quote above ₹3,00,000; the SME consulting playbook explains how the same questions apply to advisory work. A "no" on budget, decision-maker, or payment terms ends the conversation. A "no" on scope starts a paid discovery instead of a build quote.

The data question is the one that saves the most time. A project that depends on an API the client has not accessed yet is a project with an unknown integration risk. I have seen a "simple" Tally sync turn into three weeks because the client's Tally version did not expose the endpoint the brief assumed.

The Email I Send When I Decline

I decline with a specific reason and a smaller alternative. This keeps the relationship alive and tests whether the client is flexible or only shopping for a discount.

Subject: Re: Marketplace build — scope and timing

Hi [Name],

Thank you for the detailed brief. I am going to pass on the
six-month build as scoped.

The reason is the fit between the fixed fee and the undefined
scope. With unlimited revisions and a 24-week timeline, I cannot
commit to a fixed price without either raising the fee or limiting
the revision rounds, and I do not want to start a project with that
disagreement unresolved.

If it is useful, I can do one of two smaller things:

1. A paid two-week discovery at INR 45,000 that produces the role
   map, money flows, integration list, and a fixed build quote.
2. A single seller-onboarding module at INR 1,80,000 with two
   revision rounds and net-14 payment.

Either one gives you a real deliverable and gives us a chance to
work together without a six-month bet.

Deepak

The alternative matters. A flat "no" is easy to ignore. A smaller paid next step gives the client a path and gives me a low-risk way to test the relationship.

When NOT to Use This Checklist

Do not turn it down if you need the cash to make payroll this month. A ₹305 hourly rate is better than zero, provided the payment terms are enforceable and the scope is capped. Survival is a valid strategy; just do not call it a premium project.

Do not turn it down if the logo is a strategic reference you cannot buy otherwise. A first enterprise logo can unlock three similar clients. Price that referral value explicitly and cap the loss. I have taken one discounted project for a category leader and it paid back within two quarters.

Do not turn it down if the work teaches a capability you plan to sell. A complex payout system is a reusable asset. If the learning is worth more than the margin, take it, but write down what you will reuse.

Do not turn it down if the client accepts your terms after one conversation. Some large projects start badly on paper and improve once the boundary is set. The test is whether the client moves on scope and payment, not whether the first offer is perfect.

Bottom Line

I turned down ₹6,00,000 because the effective rate was under ₹350 per hour, the payment terms financed the client with my reserve, and the scope was a negotiation waiting to happen. The smaller ₹2,40,000 project with a written scope and net-14 payment was the better business. Gross revenue is a vanity number; margin, cash timing, and the energy left for the next sale are the real ledger. The Junagadh build playbook shows how I protect the calendar around that ledger.

Frequently Asked Questions

Why would you turn down a ₹6L project?

Because the real hourly rate was under ₹350 after 1,460 engineering hours, a ₹48,000 design subcontract, and a 90-day payment delay. A smaller ₹2,40,000 project with 210 hours and net-14 terms paid more per hour and carried less risk. Deepak Bagada uses a six-question qualification checklist before quoting above ₹3,00,000.

How do you calculate the real value of a project?

Divide the net fee after direct costs by the actual engineering hours, then add the cost of payment terms. A ₹6,00,000 fee with 1,460 hours and ₹24,000 in working-capital cost is not a premium project. Compare it with the next-best project you could deliver in the same calendar window.

What payment terms should a solo developer ask for?

Net 14 is the default for a solo studio, with a 40/40/20 milestone split. Net 30 adds roughly 2% to the price, and net 60 adds 5%. Net 90 should be treated as a price cut unless the client is a listed company with a verifiable procurement process.

When should you accept a low-margin project?

Accept it when you need the cash to survive, when the logo unlocks a category you cannot reach otherwise, or when the work teaches a reusable capability. Cap the scope, enforce the payment terms, and write down the strategic value. Do not accept it because the gross number looks impressive.