A Civil Engineer's Estimation Freelancing: Where the Work Came From

The invoice that changed my mind about freelancing was for $45.

A builder in Texas needed a quantity takeoff for a garage addition. I spent 6.5 hours on it, most of that fighting with software I had learned three weeks earlier from YouTube tutorials. After Upwork's cut and currency conversion, the amount that landed was about Rs 3,300. For one drawing set.

The same week, a local contractor paid me Rs 2,700 for a full house estimate that took 13 hours. Both clients were happy. Only one of those rates made long-term sense, and it took me embarrassingly long to see which.

Some background. I have spent 7 years as a civil engineer on residential and small commercial sites. Billing and estimation were always part of my office work. The discovery, and it genuinely felt like a discovery, was that people outside my office pay separately for the thing I did every day between site visits.

---

The first client called at 9:40 pm

A contractor I knew from an old site called one night. His bank wanted an item-wise estimate for a loan file, due in two days. He offered Rs 2,500. I countered with Rs 2,700 for no strategic reason, mostly because round numbers felt strange to say out loud. He agreed instantly, which told me I had quoted too low. That lesson arrived about a year later, though.

The work itself surprised me. Extracting quantities from the architect's drawings was the easy half. Building a bank-acceptable format, doing the rate analysis, and figuring out what the loan officer actually wanted took longer than the engineering. Two late nights. Thirteen hours.

But that project produced two assets that outlasted the fee: a reusable Excel format I still use on every local job, and a contractor who started telling other contractors about me.

---

Word of mouth works, until it doesn't

Within about 9 weeks, referrals from that one contractor brought me 5 more estimates, all bank-loan cases, all between Rs 2,500 and Rs 4,000.

And then a month came with zero inquiries. Nothing. I remember refreshing WhatsApp like the messages were hiding somewhere.

That silent month taught me the real lesson of this whole story. Small contractors only need formal estimates when a bank, a tender, or a government file demands one. The demand is lumpy. One client type, one work type, one city: that is not an income stream, that is a lottery ticket that occasionally pays.

---

Architects turned out to be the steady source

The fix came from a direction I had never considered. An architect who worked with one of my contractor clients asked if I could prepare BOQs for his projects on a regular basis, because he delivered drawings to clients but hated doing quantities.

Rs 3,200 per BOQ, 2-3 projects a month, every month. Suddenly there was a floor under the income: Rs 7,000-9,500 monthly from that single arrangement.

To be fair, I did nothing clever to find him. He found me. But the pattern was now visible, so I approached 6 more architects cold. Four never replied. One said no. One said yes and is still a client. A 1-in-6 conversion rate on cold outreach sounds bad until you realize each yes is worth years of repeat work.

---

The online half: one platform failed, one didn't

My Fiverr gig for BOQ preparation sat at zero orders for two months. The only inquiry wanted a full commercial estimate for Rs 800, which I declined without negotiating.

Upwork was a different market entirely. Builders in the US and Australia routinely outsource quantity takeoffs, and the account approval grind was worth it. The catch: clients there expect Bluebeam Revu or PlanSwift, and I had used neither. Three weeks of trial versions and tutorials got me functional. Not fast, functional.

That first $45 garage takeoff led to a five-star review, and the review led to more invitations. Now 2-3 international projects arrive per month at $40-120 each. The trajectory looked a lot like the mechanical engineer CAD freelancing story on this site: two dead months, then slow compounding through repeat clients.

Actually, "compounding" overstates it. It is more like a tap that drips faster each month if you don't break anything.

---

The current numbers, and the mistake that cost three weekends

Over the last 6 months the freelance side has averaged around Rs 16,800 a month on top of my salary. Best month: Rs 21,460, when two commercial estimates and three takeoffs collided. Worst: Rs 6,300. The spread is the honest part of that sentence.

Time cost is 8-10 hours a week, mostly weekends plus two or three late nights.

My worst mistake deserves its own paragraph. I once quoted Rs 4,500 for a "small commercial building" without seeing the drawings. The drawings arrived: four floors, a basement, and structural details that were half guesses. Three full weekends disappeared into that estimate. I now refuse to quote before seeing drawings, no matter how urgent the client claims the bank deadline is.

And one thing nobody mentions about estimation work: a client's "small change in the drawing" is never small. A revised drawing means re-measured quantities. Every quote I send now includes one free revision, and a per-revision charge after that. The number of "small changes" dropped sharply once that line appeared.

Next on the list is a paid Bluebeam license. The price stings in rupees, but every workaround on the trial version costs me time I could bill. One of these months I'll stop flinching at that invoice and just