From 3 projects to 15: what breaks first

Rohan and Priya started their studio in Gurgaon with a laptop, a carpenter they trusted and three projects from friends of friends. Priya designed. Rohan handled clients, sites and money. Their system was a WhatsApp group per project and one Excel file that Rohan called “the master sheet”.

It worked. Clients were happy, the photos looked great on Instagram, and the enquiries started coming. Within two years they were running fifteen projects with a team of twenty. Revenue had grown several times over. So had the number of things going wrong.

Looking back, Rohan can name exactly what broke and when. It was not random. Each stage of growth found the weakest link and pulled on it. If you are somewhere on this road, this is a map.

Stage one, 3 to 5 projects: follow-ups slip

The first thing to break was not on site. It was sales. With three projects running, Rohan was on site most mornings. An Instagram enquiry would come in at 11, he would mean to reply after lunch, and by the evening it was buried under forty site messages.

He only noticed when a former client mentioned that her sister had “tried to reach you” and then gone with another studio. There was no list of leads. There was no next step written against any of them. Leads were not lost because the firm was bad. They were lost because nobody was looking.

What to fix: one list of every enquiry, wherever it came from, with a next step and a date against each one. Check it every morning. When a lead is lost, write down why. Within a few months you will see a pattern, and it is usually fixable.

Growth does not break everything at once. It breaks the weakest link first, and then the next one.

Stage two, 5 to 8 projects: drawing versions

With more projects and a second designer, drawings started multiplying. A master bedroom wardrobe might go through three revisions as the client changed her mind about shutters, handles and the loft.

Then it happened. The carpenter built the wardrobe from revision one, which was still sitting in the site WhatsApp group. The client had asked for sliding shutters in revision two. Hinged shutters were already fitted. Redoing them cost the firm the material, the labour and nearly a week.

Nobody was careless. The new drawing had been sent. It had just been sent into a chat where three older versions also lived, and the carpenter picked the one he had saved.

What to fix: every revision gets a short note on what changed, and that note goes to site. Old versions stay on record but are clearly marked as replaced. And nothing goes into production until the final, good-for-construction set is approved by both the design lead and the client.

Stage three, 8 to 11 projects: site visibility

At eight sites, Rohan could no longer visit each one every week. He started relying on what the site engineers told him. And what they told him, most days, was “done, sir”.

“Done” turned out to mean many things. Sometimes it meant done. Sometimes it meant mostly done. Once it meant the false ceiling was up but the electrical behind it had not been tested, which was discovered after the ceiling was painted.

The site groups were loud but not useful. A real blocker, a missing tile shipment, sank under a hundred messages of photos and good mornings.

Rohan tried asking for more photos. He got them, hundreds a day, with no way to tell which room, which task or which day they belonged to. More information did not mean more visibility. It meant more scrolling.

What to fix: a simple daily rhythm from every site. In the morning, what is starting today. In the evening, what got done. Then a quality check with photos before any stage is closed, and especially before anything is covered up. Concealed work, like wiring and plumbing, must pass before the ceiling or tiles that hide it.

Stage four, 10 to 13 projects: vendor payments

With more sites came more vendors, and more vendors meant more money going out on trust. The carpentry contractor asked for a third advance “for material”. The painter wanted his final bill paid on the area he remembered, not the area anyone had measured.

Rohan was signing payments from his phone between meetings, often without seeing the work. Twice he paid in full for work that later needed rework. Once, two people raised the same payment, and it went out twice.

What to fix: a purchase order for every vendor job, agreed before work starts. Measurements taken on site and countersigned by the vendor, so the final bill is built from quantities both sides agreed. A rule that the person who raises a payment cannot approve it. And a payment becomes due only when the work has passed its quality check.

Stage five, 13 to 15 projects: cash

The last thing to break was the one that frightened Rohan most. On paper, the firm was profitable. Every project had a healthy margin in the master sheet. And yet, more Fridays than not, he was not sure he could cover the week.

The problem was timing, and disputes. Client instalments were tied to vague milestones like “carpentry 70% complete”, so every payment became a negotiation. Meanwhile vendor money went out on schedule, or early.

What to fix: tie every client instalment to an event both sides can see, like drawings approved, a stage passing its quality check, or handover. Record every receipt against the right stage. And look eight weeks ahead, not just at today's balance: money expected in, next to money committed out.

The order matters

Rohan's biggest lesson was not any single fix. It was that the fixes depend on each other. He had tried to sort out cash first, with a better spreadsheet. It did not work, because the cash problem came from disputed instalments, which came from unclear completion, which came from not knowing what was really done on site.

If he did it again, this is the order he would follow.

• First, leads. It is the cheapest fix and it protects future work.
• Second, drawings. Rework from wrong versions eats margin on every project.
• Third, site truth. Daily updates and quality checks tell you what is actually done.
• Fourth, vendor payments. Once you know what is done, pay for checked work, on measured quantities.
• Fifth, cash. With completion on record, client instalments stop drifting and the forecast becomes reliable.

You can start anywhere the pain is sharpest. But know that the money fixes lean on the site fixes, and the site fixes lean on the drawing fixes. If your cash is a mess, look upstream before you buy a better spreadsheet.

Fix it before the next stage arrives

The best time to fix a stage is just before you reach it. If you are at five projects and growing, the drawing problem is coming. If you are at ten, vendor payments are next. Each fix is far easier to put in place while the firm is calm than in the middle of a crisis with an angry client on the phone.

How Pakka handles it

A firm can start with any one desk and grow into the rest. The Sales Desk gives every lead a next step, with an 8 AM call list on WhatsApp and a reason recorded on every lost lead. The Design Desk requires a note on every revision, marks old versions superseded, and keeps production shut until the GFC set is approved.

The Ops Desk runs site from WhatsApp with a morning and evening tap and QC checklists with photo evidence. The Money Desk makes vendor payments payable when QC passes, approved by a named person who did not raise them, releases client instalments on events, and shows money in and out for the next eight weeks.