One wrong cell

Arjun is the project manager on a villa in Gachibowli, and it is the week the carpentry vendor's purchase order goes out. He is reconciling the vendor's quantities against the client quote, line by line, with a cup of tea going cold. The master bedroom matches. The living room matches. Then he reaches the kids' room and stops.

The kids' room is in the quote. Eight lines: a study unit, a bunk bed, two wardrobes, a loft, a book wall, a window seat and the panelling. The lines are there, neatly typed, with rates and amounts. But when Arjun adds them up by hand and compares them to the carpentry total on the summary page, the numbers do not meet. The summary is short by ₹3.4 lakh.

It takes him ten minutes to find out why. The carpentry subtotal was a SUM over rows 12 to 48. The kids' room was added after the first meeting, in rows 49 to 56. Nobody extended the range. The client signed a quote that looked complete, line by line, and simply did not charge for one room.

The client is not going to pay ₹3.4 lakh more for a room they were already shown on paper. Arjun knows it. The owner knows it. The firm will build that room for free.

Spreadsheets fail even experts

If this sounds careless, it is worth knowing how ordinary it is. Spreadsheet errors are one of the better studied problems in business software, and the findings are humbling.

Raymond Panko, a researcher at the University of Hawaii, spent years collecting audits of real-world spreadsheets and running experiments on how people build them. The pattern in his work is consistent: when spreadsheets are examined carefully, most of the larger ones turn out to contain errors, and the people who built them are usually confident they are correct.

Two famous cases show that expertise is no protection. In 2013, a graduate student re-checking an influential economics paper by Carmen Reinhart and Kenneth Rogoff found that an Excel formula had left several countries out of an average. The paper had been widely quoted in debates about government debt. The authors acknowledged the error, while arguing about how much it changed their conclusions.

In 2012, JPMorgan lost billions of dollars on trades that became known as the 'London Whale'. The bank's own review afterwards pointed to a risk model that ran through spreadsheets, with figures copied by hand from one sheet to another and a formula error among the problems it found.

A spreadsheet never tells you it is wrong. It shows you a clean number, formatted in rupees, and lets you believe it.

Economists and investment bankers are not careless people. The lesson is not that someone in your firm is bad at Excel. It is that spreadsheets fail silently, and the tool gives you no warning when they do.

The four errors that cost interior firms the most

Interior quotes have their own favourite failure modes. If you have built quotes in Excel or Google Sheets for a few years, you have probably met all four.

1. The SUM that misses new rows. Arjun's error. Rows added below the range, or above it, are not counted. The line looks right, the total looks right, and they do not agree. This one is common because quotes grow after the first draft.
2. The hard-coded GST. Someone types 1.18 straight into a formula instead of pointing to one GST cell. Then a line that should be taxed differently is not, or the total is taxed twice because one subtotal already included GST. The number looks plausible, so nobody checks.
3. The rate copied from last year. The quote was started from an old one. The plywood rate, the hardware rate and the labour rate came along with it. Prices have moved since, and the quote has not.
4. The dragged formula that shifts. A formula is dragged down a column and its references move with it. A rate cell that should have stayed fixed now points one row lower. Half the column is multiplied by the wrong number.

There are smaller cousins too: a row hidden during a review and never unhidden, a quantity in running feet multiplied by a rate per square foot, a value pasted over a formula so it no longer updates. Each one is small. Each one can still be expensive.

What one wrong cell really costs

The painful part of Arjun's error is not the ₹3.4 lakh on the page. It is where that money comes from. The firm still has to build the kids' room, so it pays the full cost of that work with no revenue against it. That cost comes straight out of the project's margin.

One cell took away nearly half of the project's profit. And because it was found at the purchase order stage, months after signing, there was no graceful way to fix it. The alternative, going back to the client with a corrected quote, risks something worth more than ₹2.55 lakh: their trust in every other number you have given them.

Errors in the other direction cost money too, just differently. An over-quote because of a doubled GST line may lose you the job outright, and you will never know why.

Checks that catch errors before the client sees them

You do not need new software to catch most of these. You need a few habits that make errors visible. Good firms build them into every quote.

• Run a control total. Add the quote up two ways, by room and by trade. If the two totals do not match to the rupee, something is missing or counted twice.
• Keep GST in one cell. Every formula that needs the rate points to that cell. Nobody types a tax rate into a formula, ever.
• Date your rate card. Every quote pulls rates from one sheet with a 'last updated' date at the top. If the date is old, the rates are suspect.
• Lock what should not move. Use absolute references for rate cells, and protect formula cells so a stray paste cannot overwrite them.
• Compare against a sanity ratio. Divide the total by the carpet area and compare it to similar past projects. A number far outside your usual range deserves a second look.
• Get a second pair of eyes. The person who built the quote is the worst person to check it. A colleague with a fresh view finds errors in minutes.
• Read the final version as the client will. Print it or open the PDF, and read every line and total once, slowly, before it goes.

These checks will not make spreadsheets safe. They will make the common errors much less likely to reach a signed quote, which is the point at which they become expensive.

How Pakka handles it

Pakka's Sales Desk builds quotes from structured lines rather than free cells. Rates come from your own rate library of finishes, recipes and bundles, so nobody types a rate you already have, and nobody carries an old one forward from last year's file.

The payment plan, design fee, PMC and GST are built into the proposal rather than typed into formulas. Versions can be compared side by side, so you can see exactly what changed between v1 and v2. Low-margin quotes are flagged for approval before they go out, and salespeople never see cost or margin.

A person still checks every quote, and should. The difference is that the check is about the specification and the client, not about whether row 56 made it into a SUM.