Chapter 2 — The Wild West of Code (1980)

The Evolution of Accounting — A Real Analytics 101 Perspective

The 1980s didn’t arrive quietly; they arrived with the clatter of keyboards, the glow of monochrome screens, and the promise that a desktop machine could change the economics of work.


The Noise Before the Revolution

The early 1980s turned garages and back rooms into laboratories. Anyone who could string together a few lines of BASIC suddenly had a title: programmer. Personal computers had landed—small enough to sit on a desk, yet large enough in imagination to launch a revolution. Software escaped its mainframe confines and entered homes, classrooms, and boardrooms. A single floppy could now replace an entire filing cabinet. The pioneers weren’t celebrities in hoodies; they were hobbyists, accountants, engineers, and dreamers—underpaid, overworked, and underestimated. Few realised that the people typing on green-glowing screens were wiring the backbone of the next century.

Boom Time, Bug Time

Markets were hungry for automation. Accounting packages, inventory systems, payroll programs—everyone wanted them yesterday. Tiny software houses sprang up overnight. Enthusiasm replaced process; intuition replaced testing. Programs shipped half-baked, and “Run” was often a leap of faith. Month-ends could be thrilling or terrifying. Yet businesses kept buying, because even imperfect software finished work in minutes that had once stolen days. It was the Wild West of code, and a little chaos felt like the cost of progress.

When Accountants Became Coders

For those with an accounting background, the PC felt both alien and familiar. The interface was new, but the logic of debit and credit was eternal. Installing early accounting software was like walking into a half-built factory—no manuals, few guardrails, and endless potential. If you understood ledgers, you could decode the machine’s prompts: behind every “trial balance” menu lurked a cash book waiting to be automated. I was one of those accountants. At first, I thought programming would simply total columns faster. Then two moments changed everything.

University — Discovering the Meaning of Data

Computer Science didn’t just teach me to code; it taught me structure. Data had anatomy. It could be normalised, indexed, retrieved—and designed for meaning. The machine was logical, not magical. Once that clicked, accounting transformed from a record of what happened to a medium for discovering why it happened. Reconciliation remained essential, but the discipline expanded: you could now interrogate information instead of merely totalling it.

SCALA — The “Baby SAP” and a Big Idea

Then came SCALA, a professional accounting package marketed as the “baby SAP.” SCALA treated data as a living organism. It let you pivot information—sales, stock, margins—from angles no paper ledger could manage. Patterns appeared. Inefficiencies blinked into view. For the first time, I understood the difference between storage and insight. That shift—from keeping data to learning from it—became the north star for everything I would later write in Tales from the Trenches: a story about how human judgement and machine logic collaborate, and sometimes collide.

The Price of Speed: Fragmentation

Speed arrived; coherence left by the side door. Early PC systems did one thing brilliantly: they printed. Statements, invoices, debtor lists—spat out in minutes. But identity and structure lagged behind. A customer could appear as “J. Doe” today and “Joe Doe” tomorrow, and the software dutifully treated them as two different people. Addresses were free-text, clerks invented their own codes, abbreviations shifted by taste. Monthly statements ran on time, but ageing reports drifted away from the ledger’s truth. We bought velocity and mortgaged consistency.

How Dirty Data Was Born

Duplicate entities. Slight spelling changes created “new” debtors and suppliers.
Free-text chaos. Inconsistent address lines, narrations, and item descriptions.
Key drift. Different clerks invented different codes for the same thing.
Islands of data. Sales, purchasing, inventory—all on separate programs with no single source of truth.

The paper era was slow but traceable. The digital era was fast but fragmented. Without shared definitions and validation, speed amplified noise.

Spreadsheets: Freedom and Fragility

Spreadsheets felt like liberation wrapped in gridlines. Practitioners could model cash flow, discounts, and pricing without waiting for IT. But each new version held a quiet menace: which file was the master? Which cell should have been absolute but went relative? One broken formula could turn good data into confident nonsense. The spreadsheet democratized modelling—and also democratized mistakes.

Floppies, Backups, and Single Points of Failure

Another new risk lived in the hardware. Data perched on floppies and early hard drives that failed without warning. A corrupted disk could erase weeks of work. Paper had been cumbersome but resilient; PCs were quick but brittle. Backup discipline became the difference between a hiccup and a disaster.

What “Good” Already Looked Like (Even Then)

Early governance playbook:

  • Data dictionary. Shared definitions for names, addresses, item codes, and GL accounts.
  • Validation rules. Mandatory fields and standard formats for phone, VAT, and addresses.
  • Master data stewardship. One accountable person for deduplication and correctness.
  • Regular reconciliations. Tie AR/AP to control accounts every month (no exceptions).

Tales from the Trenches: The Rise and Lessons of Budget Auto Lease

When Budget Auto Lease first opened its doors, we had a bold idea and very little sleep. The concept of Full Maintenance Leasing (FML) — where clients rented cars long-term and the leasing company took full responsibility for maintenance — was still relatively new. Most people thought leasing was just glorified renting. But FML was different. It promised predictable costs, zero admin, and peace of mind. The client would pay a single monthly fee, and we would handle everything else — from oil changes and licensing renewals to tyre replacements and roadside assistance.

At the time, this was revolutionary. Fleet managers, used to juggling invoices from mechanics, tyre suppliers, and insurers, suddenly had one line item on their budget: “Monthly Lease – All Inclusive.” And they loved it.

How the Model Worked

Our revenue model was simple but elegant. We bought the vehicles, leased them out for 36 to 60 months, and the monthly payments covered both the bank finance and our margin. The trick was managing three moving parts — time, money, and maintenance.

The vehicles were financed through a bank facility — a revolving credit line that allowed us to buy cars in bulk. The bank held title to the vehicles as security, and we repaid them through fixed instalments. Sometimes, to make repayments more manageable, we negotiated residual values — balloon payments left for the end of the term, roughly equal to the expected resale value of the cars. This reduced monthly obligations and improved cash flow.

We, the FML company, were effectively a bridge between the banks and the drivers. The bank owned the metal; the client paid for the mobility; and we made our profit in the space between. It was a beautiful model — until Excel got in the way.

The Team That Kept the Wheels Turning

We started small — just four accountants and a secretary doubling as the CEO’s PA. One accountant oversaw the operation and liaised with the bank, while the others handled the maintenance department, reconciliations, and payments.

It was an efficient setup for the first few months. Each new lease meant adding another line to our master spreadsheet — an Excel workbook that became our lifeblood. Invoices, statements, and vehicle details were all generated using carefully designed macros. It was magical at first. Click a button, and the system spat out customer statements like a seasoned accountant on autopilot.

But as with all magic, there was a price.

The Spreadsheet That Ate Our Nights

At first, running statements took about 30 minutes. Then, as the business grew and the number of vehicles ballooned, it took an hour… then two… and eventually, an entire night. What had started as a clever spreadsheet had turned into a fragile monster, buckling under the weight of thousands of transactions.

The problem with spreadsheets — as I’ve often told younger accountants — is that they’re like trying to move a 10-ton load with a 1-ton truck. It works beautifully at the start, but it wasn’t designed for the strain of a growing enterprise. Every new formula, every lookup, and every linked sheet added another thread to a tangled web. One wrong edit and the whole thing collapsed.

We were living in the age of manual backups and floppy disks. When the statement run failed at midnight, someone would be sleeping under their desk, waiting to restart it at 3 a.m. The growing data load also meant that a single mis-typed cell could misstate a client’s balance by thousands.

Our invoice system had no audit trail — no way to confirm that a statement had actually gone out. It was a lesson in why control systems are not just accounting theory, but survival mechanisms.

The Porsche Incident

Then came the Porsche 911 — the car that taught us humility.

It was a stunning machine, leased to a client who had an impeccable credit record. But one small oversight in our Excel-driven process caused chaos. Because there was no automated audit trail, his monthly statements never went out. For an entire year, no one noticed. The car remained on our books, but no payments were coming in.

When we finally discovered the error, we sent the arrears statement. The client, somewhat amused, responded with a stroke of brilliance — he sold the Porsche at a profit, settled his arrears, and still walked away smiling. We, on the other hand, were left with red faces and a hard-earned lesson in internal controls.

That incident became the stuff of office legend. Whenever someone proposed a “quick fix” in Excel, someone else would mutter, “Remember the Porsche.”

The Dirty Data Dilemma

As the business expanded, so did the complexity. We were now managing hundreds of vehicles, each with its own account number (based on the registration plate) and payment history. Capturing vehicle details was simple enough; matching the payments to the right accounts was where chaos lived.

When the parent company — a much larger organisation with its own mainframe — offered to migrate our operations onto their system, we jumped at the chance. It seemed like salvation.

But what we didn’t account for was dirty data. Years of manual entries, typos, duplicate vehicle codes, and incomplete information had created a digital swamp. Data that looks neat in Excel often doesn’t meet the rigid structure of a database.

We learned that lesson the hard way.

The Great Migration Project

The migration took eight months, and it nearly broke us. Because our business couldn’t stop for system upgrades, we had to migrate data while continuing daily operations. Every invoice, every payment, every repair order had to be captured in both systems until the new one went live.

This was what IT professionals politely call a “big bang approach.” In plain English, it means switching from the old system to the new one in a single dramatic moment — no gradual phase-in, no fallback. It’s like changing the engine of a plane while it’s flying.

We worked late nights, hired temporary data clerks, and burned through enough coffee to keep a small café afloat. Every morning brought a new round of reconciliation reports and error logs.

By the time we flipped the switch, it felt less like a launch and more like survival. The relief was immense, but so was the exhaustion.

Looking back, the irony wasn’t lost on us: we had spent eight months trying to make sense of messy Excel data when a proper leasing management system — designed specifically for FML operations — would have done the job in half the time, and probably at half the cost.

Reflections from the Trenches

The experience taught me lessons I still carry today:

  • Excel is not an ERP. Spreadsheets are incredible tools for analysis, but disastrous as operational systems. They don’t enforce data integrity, audit trails, or user controls. Once you’re managing millions in assets, the margin for error disappears.
  • Data cleanliness is not optional. “Dirty data” is not just an IT problem — it’s a financial risk. Every duplicate, typo, or missing field represents a potential mismatch, misbilling, or compliance breach. Cleaning up bad data after the fact costs ten times more than doing it right upfront.
  • Control systems protect you from yourself. The Porsche story wasn’t about a dishonest client — it was about our overconfidence. A well-designed system doesn’t rely on trust or memory. It enforces discipline automatically.
  • Never underestimate the cost of transition. Migration projects drain morale and money. The promise of automation often hides the mountain of manual work needed to get there. The lesson: plan for the pain, not just the payoff.
  • Right tools, right time. Growth exposes weaknesses. What works for ten vehicles fails at a hundred. The sooner you adopt purpose-built software, the fewer fires you’ll fight later.

The Human Side of the Story

Despite the chaos, there was a deep camaraderie among us. Long nights blurred into early mornings, punctuated by laughter, frustration, and takeaway dinners. There was pride in the grind — a sense that we were building something real.

One night, around 2 a.m., I found our youngest accountant asleep at his desk, his cheek pressed against the keyboard, fingers still on the number pad. I nudged him awake and said, “Go home.” He blinked, smiled, and muttered, “Just one more reconciliation.”

That, to me, summed up the spirit of Budget Auto Lease. We weren’t just processing numbers; we were powering mobility, keeping businesses running, and learning the limits of what spreadsheets — and humans — could endure.

In Hindsight

If I had to do it again, I’d start differently. I’d invest in proper leasing software from day one, design for scale, and enforce data standards religiously. But the truth is, those early mistakes were our greatest teachers.

Budget Auto Lease wasn’t just a business story; it was a case study in digital transformation before the term even existed. We learned how fragile systems become under pressure, how easy it is to lose control of data, and how technology without structure creates chaos faster than it creates efficiency.

Yet, for all the sleepless nights and costly lessons, we came out stronger. The systems improved, the teams grew, and the next generation of accountants inherited a cleaner, faster, more resilient operation.

And whenever I see a Porsche 911 today, I can’t help but smile — a reminder that behind every spreadsheet, there’s always a story.

When the PC Met the Ledger

The PC didn’t just change how accountants worked; it rewired what “working” meant. The ritual of month-end close—the clatter of posting machines and the smell of carbon paper—faded into the soft whirr of a printer and the blink of a cursor. What once took teams now fell to one person. Efficiency soared; so did isolation. Paper errors were communal; digital errors were private, hiding in cells and codes. Transparency gave way to opacity. Yet the exhilaration was real: a machine could compress days into seconds.

South Africa’s Quiet Digital Pioneers

While Silicon Valley minted legends, South Africa built quietly. Before browsers and Wi-Fi, a national online network called Beltel connected people to email, banking, directories, and news via ordinary telephone lines. Inspired by France’s Minitel and launched through the South African Post Office and Telkom, Beltel offered text-based menus navigated by keystrokes. Each subscriber had a Beltel number—an account ID with a password—and access to capabilities that prefigured the web: person-to-person messages, selected bank balance enquiries and payments, news and weather feeds, directory look-ups, and early digital classifieds.

Picture a dawn routine in Johannesburg: the modem hisses, the screen glows, overnight messages arrive. Quotes go out without a fax. Journalists skim headlines; universities experiment with tutorials. It demanded patience—characters blinked one line at a time—but the magic was unmistakable: information moved without paper.

Beltel faded in the mid-1990s. Graphical browsers brought colour and a global ecosystem. Text menus felt clunky; per-minute phone tariffs and infrastructure costs hurt; institutional shifts after the Post Office and Telkom split dulled momentum. By the early 2000s, Beltel was decommissioned. Yet its legacy is everywhere: in online banking, digital classifieds, government portals, and the culture of “being online.” History rarely celebrates prototypes, but breakthroughs stand on their shoulders. Beltel was one of those shoulders.

Speed vs Sense — The Permanent Lesson

The PC era taught a truth that still holds in the age of AI: if we don’t govern master data and logic, acceleration multiplies errors. The reconciliation, narration, and cross-checks once done by hand must be rebuilt in software—intentionally, not assumed. The difference between a ledger and a database isn’t the machine; it’s the discipline of the people who design and use it.

Epilogue — From Code to Meaning

Looking back, 1980 wasn’t just a date; it was a threshold. We moved from paper to pixels, from ink to insight. We learned that automation without understanding breeds chaos—and that data without context is noise. Out of the Wild West of code came a new craft—not programming, not accounting, but something in between: data thinking. That’s where the real revolution began.