Back to Insights

You sold the build. Now nothing comes next month

TL;DR

  • A one-off build pays once. The income stops the day the project ends.

  • The founder goes back to hunting, and the hunt starts from zero every time.

  • Indian automation agencies now publish the shape openly: a build fee plus a monthly retainer.

  • One Bengaluru engineering team found partnerships earn two to three times a one-off project, with better margins.

  • The fix is not a bigger project fee. It is a second line on the quote: care, updates, and improvements, every month.

The last week of a project feels good. The work is done, the client is happy, the payment lands. You breathe out. Then Monday comes, and there is nothing in the diary that pays. No invoice to send. No retainer ticking over. Just the hunt again, from zero.

This is the part of project work nobody puts in the quote. The fee looks big. The gap after it is invisible until you are standing in it.

Why does the month after a big project feel empty?

Because the project was the whole business. You sold one thing, delivered it, and the relationship ended at handover. There is nothing left to bill for, and no reason for the client to call you until something breaks.

So you do what every founder in this spot does. You start selling again. New calls, new proposals, new scoping, new waiting. The sales work that paid for the last project now has to pay for the next one, and the clock is running. That is why you cannot take a day off without the pipeline going quiet behind you.

The cycle has a name in agency circles: feast and famine. A busy month, then a quiet one. It feels like a marketing problem. Most of the time it is a shape problem. When every rupee you earn needs a fresh sale, every month starts at zero, and a busy delivery month is the same month your marketing stops. The famine is scheduled the day the feast begins.

What does a one-off build really cost you?

More than the hours you spent building it. Count the full cycle. A boutique engineering team in Bengaluru wrote up its own numbers after it stopped taking one-off projects in 2024, and the shape is familiar to anyone who sells builds (source).

Their typical one-off project ran 2 to 4 weeks of sales and scoping, 3 to 6 months of development, and then a handoff and silence. When the client came back 8 months later with a new requirement, the original engineers had moved on, the context was gone, and the team spent 3 to 4 weeks rebuilding knowledge it already had. The client paid for that ramp time. Nobody won.

The maths was just as blunt. A one-off project of $40,000 to $80,000 cost roughly 70 to 75 percent of the revenue to deliver, once sales, scoping, onboarding, development, QA, and deployment were all counted. Margin: 25 to 30 percent. That is before you count the quiet month after handover, when you are selling the next project for free.

There is a second cost, and it is the one founders feel most. You get good at starting. You never get good at keeping. The clients who already trust you are the easiest revenue you will ever earn, and a one-off shape gives them no way to keep paying you. It is the same trap as quoting three prices to the same client: the price was never the real problem. The shape was.

What are Indian automation agencies charging now?

The market has started publishing its numbers, and the shape is the same everywhere. A build fee, then a monthly retainer. Here are the bands published by Indian automation agencies this year.

| What | Published band | Source |

| Small automation job | ₹10,000 to ₹30,000 | Touch The Skies |

| Full system build | ₹50,000 to ₹2,00,000 and above | Touch The Skies |

| Monthly retainer | ₹15,000 to ₹60,000 a month | Touch The Skies |

| Build packages | From ₹20,000 | RisonAI, via Digital Patron |

| Monthly retainers | From ₹15,000 a month | RisonAI, via Digital Patron |

The sources are Touch The Skies, Branofy, and Digital Patron.

Two details matter more than the numbers. First, Branofy lists the price as three separate lines: a one-time onboarding fee, a monthly service retainer, and the platform cost billed on its own. The build is a line. The care is a line. The tools are a line. Nothing is hidden inside anything else.

Second, the advice attached to those bands is consistent: price on the value of the time saved or the revenue recovered, not on the hours you spent building. A system that saves a team ten hours a week is not worth your hourly rate. It is worth a share of those ten hours, every week, for as long as it keeps saving them.

That is the part founders keep missing. The build is a one-time event. The saving repeats. So should the fee.

Why does a retainer change the maths?

Go back to the Bengaluru team's numbers. Their partnership engagement ran $8,000 to $15,000 a month, or $96,000 to $180,000 a year, against a one-off project at $40,000 to $80,000. Total cost to deliver fell to 55 to 65 percent of revenue, because there was no repeated sales and onboarding cycle. Margin rose to 35 to 45 percent. The partnership generated two to three times the revenue of a project, with better margins. And year two of a partnership had near-zero sales cost and near-zero onboarding cost, so the margin kept improving the longer the relationship ran.

That is the real difference. A project resets your costs every time. A retainer lets them fall.

There is a second number worth knowing. Matt Inglot, who rebuilt his agency after burning out on the feast and famine cycle, found that 20 percent of his clients produced 80 percent of his profits (source). His answer was a small, loyal set of clients, taken on rarely and carefully. A retainer is how you build that set on purpose instead of hoping it happens.

The retainer is not extra money for the same work. It is different work, and the client can feel it: the system stays updated, the small fixes happen before they become big ones, the improvements land every month, and the client calls you first instead of shopping around. A system with no care attached goes stale within a year. The same team put it plainly: a $60,000 custom system with no updates, no patches, and no iteration for 12 months is a liability, not an asset.

How do you move a client from a build to a retainer?

You do not sell the retainer at handover. You sell it on day one, in the same quote, as its own line. Three steps.

First, split the quote into three lines before the client ever sees it: the build, the monthly care, and any platform or tool cost. When the care is a line from the start, it is part of the deal, not an upsell at the end.

Second, name what the care covers, in plain words. Updates, fixes, small improvements, and a reply when something breaks. Then name what happens without it: no patches, no fixes, no help when the team changes. Clients do not buy a retainer. They buy the absence of that list. The same logic applies when your best person leaves: a system with no one caring for it rots quietly, and the rot is discovered at the worst moment.

Third, offer a short first term. A quarter is enough for the client to feel the difference: the fix that arrived before they asked, the small improvement that saved an hour a day. A quarter of care sells the next year better than any proposal will.

One warning. Do not fold the care into the build fee to make the quote look simpler. A folded-in fee is a one-off fee with a prettier name, and it ends the same way: the money stops, the work continues for free, and you are back to hunting. If the advance is not the objection, the retainer is not either. The objection is a quote that asks for trust and offers nothing to hold on to after the handover.

A question to sit with

Look at your last twelve months of income. Not the total. The shape. How many months started at zero? How many months had money arriving from work you had already done?

If the honest answer is "almost none," you do not have an income problem. You have a shape problem. The work is good, the clients are happy, and the money still stops every time a project ends. That is not a sign you need to charge more for the next build. It is a sign the next build needs a second line.

And that is also when winning a client stops turning into chasing them, because the relationship has a reason to continue after the handover, and the income has a reason to repeat.

Frequently Asked Questions

Why does project income feel so unstable?

Because every rupee needs a fresh sale. A one-off build pays once, then the income stops at handover. While you are busy delivering, marketing stops, so the next sale starts late. That is the feast and famine cycle, and it is a shape problem, not a marketing problem. Income repeats only when the offer repeats.

What is a fair retainer for a small automation build in India?

Published Indian bands put monthly retainers at ₹15,000 to ₹60,000, with build packages from ₹20,000 and small automation jobs at ₹10,000 to ₹30,000. The fair number depends on what the system saves: price on the time saved or revenue recovered each month, not on the hours you spent building it.

Will clients pay a monthly fee after paying for the build?

Yes, when the fee is its own line from day one and covers something real: updates, fixes, small improvements, and help when things break. Clients resist a retainer sold at handover as an upsell. They accept one that was part of the deal from the first quote.

How do I price the build and the retainer separately?

Use three lines: a one-time build or onboarding fee, a monthly care retainer, and any platform or tool cost billed on its own. The build covers delivery. The retainer covers keeping the system alive and improving. Keeping them separate is what makes the retainer survive the first year.

What if a client only wants a one-off build?

Take it if you need the work, but price it honestly: a one-off build with no care becomes a liability within a year, with no updates and no support. Tell the client that plainly. Many will still say no to the retainer, and that is fine. The point is that you offered the shape that lasts.