What a Failed Sales Job in 2001 Taught Me About Building Business Software
I build TurboData, a reporting and analytics layer for businesses running Tally. Before that I spent twenty-five years in sales, data capture and data warehousing.
This is the story of the year that pointed me here. It did not go well at the time.
Before it went wrong
I started at Philips Semiconductors as a Sales Engineer. People arrived at nine and left at five. Payments were largely in advance. Commercial terms like FOB and FCA were researched properly, by people who cared about the difference. Processes were stable.
It was, to be honest, boring.
My second assignment there was estimating the lighting and semiconductor markets in South India. I built an Access database of product bills of material. It took weeks. But afterwards a product manager in Mumbai could see what was likely to be needed, in what quantity, in which market — instead of waiting for fragmented field reports.
The business head was not thrilled that a junior was spending so much time in the field. The objection, essentially, was that I was gathering too much data.
Then I left. For a real sales territory. Because the database job was boring and I was young and I wanted to sell things.
I did not understand what I was giving up until about a year later.
September 2001
I joined Philips Lighting in Mumbai as a Sales Engineer, with the confidence of a superstar and almost no training.
Within days, the World Trade Center towers fell. The world appeared to be changing, and I reported to the office to begin selling light fittings.
Mumbai was divided into sales territories: Fort to Dadar, Dadar to Andheri, Andheri and beyond, Thane. I was given the second one.
I was assigned two old stockists located next door to each other. Experienced, decent people. Their sales strategy was to sit inside the shop and wait for customers to arrive. When I suggested we go into the market and meet customers, they looked at me as though I had proposed moving the shop to Antarctica.
Enterprise archaeology
Soon after joining, my manager handed me a large accounts-receivable statement. Some invoices were more than four years old. The instruction was straightforward: make the receivables zero.
At that stage I sincerely believed that if management had given me the list, the money must be waiting somewhere.
One line was around ₹73,000. The sale had happened in 1993. Payment was expected by 1997. The customer had closed operations in 1999. I joined in 2001.
My first major assignment was to collect money from a company that had ceased to exist two years before I arrived.
I was not selling lighting products. I was excavating ancient invoices.
The employees who handled the transaction had left. The customer contacts had disappeared. Supporting documents were hard to trace. The instruction did not change. When I failed to produce a financial resurrection, the shouting began.
Old receivables are a data problem long before they are a collection problem. Nobody in that building could separate what was collectible from what was disputed, doubtful or effectively dead — so the entire ledger was handed to the newest employee as a single number.
Who actually owned the customer?
A lighting project could involve an architect, a lighting consultant, an electrical contractor, an end user, a dealer, a design engineer, several salespeople — and at least three people claiming ownership once the order arrived.
Each of them could sit in a different Mumbai territory.
I once started visiting architects and consultants near Dadar. They were close to my assigned stockist, so I assumed this was called business development. Within a short time I received an angry call from head office: each architect had already been allocated to a lighting design engineer. I could not approach an architect a hundred metres from my own stockist without causing a territorial crisis.
I had finally started meeting potential customers, and the organisation reacted as though I had crossed an international border without a passport.
Some companies do not have territories. They have arguments drawn on maps. A territory tells you where a salesperson may travel. It does not tell you who owns the relationship, which is the only question that matters when four people decide one project.
Pricing as a live performance
The central product team circulated standard prices and volume-discount slabs. I communicated them to my stockists.
One month later my Area Manager arrived with the Product Manager and announced an entirely different discount structure.
The stockist looked at me. I looked at the manager. The manager looked completely confident.
Different managers could quote different prices for the same product, customer and quantity. A salesperson could communicate an officially approved price and still appear unreliable a month later.
Customers wanted consistency. Dealers wanted margins. Salespeople needed authority. Operations needed forecasts.
We had volume. Of voice.
Zero out of one hundred
I managed to reach organisations like Indian Oil, Indian Oil Tanking and Bharat Petroleum. They issued complicated tenders with technical specifications running to more than a hundred lighting products.
I had never been trained in lighting design. I was uncomfortable with the design software. I did not know how to configure a hundred products in a technical bid. I was competing against companies with dedicated design teams and several sales officers covering the same market.
One tender contained more than a hundred items. We won zero.
Winning nothing out of a hundred requires remarkably consistent performance. Our pricing was not competitive, availability was uncertain, technical support was weak and coordination was poor.
One part of the process, however, worked perfectly. The blame reached me by evening.
Nobody asked which items we lost on price and which on specification. Nobody asked what the winning rate had been. Nobody asked which products we simply did not have. A tender where you win zero of a hundred should produce a win-loss analysis. Ours produced a meeting.
Responsibility travels faster than material
Since my stockists preferred waiting for customers, I opened the Mumbai Yellow Pages and looked for alternatives. There was no LinkedIn, no prospecting software, no CRM intelligence. There was a large yellow directory and optimism.
I found a former high-performing stockist of a competing brand and persuaded them to sell our products. Together we pursued a project near Thane. Unknown to me, my own manager was chasing the same opportunity through another channel partner. We won anyway.
For one brief moment I felt like the superstar I had imagined myself to be. Then delivery began.
The material was not available locally and had to come from Kolkata. The shipment was delayed. The customer complained directly to the CEO.
I had found the dealer, developed the opportunity and helped win the order. Inventory, dispatch and logistics were outside my control. The blame came to me.
Responsibility travels downward considerably faster than material travels from Kolkata.
Not fit for sales
By early 2002 I had reached my limit. Around twenty-four people had reportedly left before me. Turnover was high, though management did not seem curious about why.
Management refused to settle my expenses unless I reduced the old receivables to zero. My reimbursement depended on collecting invoices from customers who no longer existed.
Eventually, in a meeting attended by the HR head, the Regional Manager told me I was not fit for sales.
At the time, I believed him. I went home to Gurgaon convinced my career had failed before it properly began.
What I did instead
I documented all of it — the absent training, the incoherent territories, the conflicting channel ownership, the pricing by announcement, the uncollectable legacy receivables, the weak design support, the tender failures, the delivery problems, the turnover, and the gap between responsibility and authority.
I circulated it internally. Then I took it to Bajaj, Crompton and Thorn — not as an interview, but as a diagnosis of the market they were competing in.
I received two offers.
The man declared not fit for sales had sold his analysis of a broken sales system to three of its competitors. That was, in retrospect, my first consulting assignment.
Why any of this matters now
Every problem in that year was a visibility problem.
They had receivable reports but could not tell collectible amounts from dead accounts. They had territories but no account ownership. They had prices but no record of what was approved, by whom, from when. They had orders but no view of what was actually going to ship. They had data. They did not have information anybody could act on.
Twenty-five years later, that is still the most common condition I find inside a well-run business with a well-maintained Tally installation. The data is all there. The answers are not.
That is what TurboData is for. It makes the state of the business provable — which receivables are genuinely open and traceable to the bills and payments behind them, and, for contractors and developers, what a slipped milestone actually cost, read from vouchers rather than estimated in a meeting.
Instead of telling a salesperson to make receivables zero, management should be able to see which amounts are collectible, disputed, doubtful or dead. That is not a sophisticated ambition. It is the report I needed in 2001 and did not have.
The final irony
They told me I was not fit for sales, and they may have been partly right.
I was not fit for a sales system in which customers had no clear ownership, prices changed without control, impossible receivables were handed to new employees, training was absent, delivery failures became salesperson failures, and data existed but insight did not.
It was an expensive management course. No certificate, no refund and, for some time, not even my expenses.
But it taught me exactly what a poorly instrumented business needs. Clear data. Clear ownership. Clear pricing. Clear accountability. And considerably less shouting.
If any of this sounds like your own ledger, the fastest way to find out is to look at your actual numbers. We run a 30-minute session on your own Tally data and show you what is currently invisible in it — no slides.
Receivables and collections → · Construction and project costing → · Talk to us →
