Go-Live

First, understand the bet.

Replacing a company's main system is not buying software. It is open-heart surgery on a company that must keep running.

Every order, every invoice, every pallet and every salary payment has to keep flowing — while the organs that process them are exchanged.

These projects run for years. They cost from tens of millions into the billions. A whole global industry of consultants exists just to do them.

Boards approve them for a real prize. One system, one set of numbers, one version of the truth. The alternative is a tangle of old systems, held together by three specialists who are close to retirement.

And roughly half of these projects, in survey after survey, decade after decade, end up seriously over budget, behind schedule, or worse.

This lesson is about the "worse".

The founding legend is Hershey, 1999. America's biggest chocolate maker was replacing its core systems. It was running late, and it made two decisions that now appear in textbooks with red circles around them.

It cut the testing to protect the date. And it chose to switch everything on at once, in July — exactly when the orders arrive for its two biggest seasons, Halloween and Christmas.

The new system started. The flow of orders through it did not.

The warehouses held over a hundred million dollars of chocolate that the system could not put onto trucks. Shelves were empty in the most chocolate-important quarter of the year. Sales fell. The share price dropped sharply.

And business schools got their favourite opening slide: the company that had the chocolate, had the customers, and lost Halloween in the software between them.

Here is the part the legend usually leaves out, and this series keeps it. Hershey survived, fixed it, and later completed further upgrades smoothly. The same company became an example of doing it right.

So the lesson was never "systems don't work". It was that a go-live date had been treated as a promise to the calendar, instead of a conclusion from the testing. And the calendar, unlike the testing, cannot tell you anything.

The pattern repeats with the reliability of physics. The list is worth reading slowly.

FoxMeyer, mid-1990s, is the extreme case. A five-billion-dollar medicine distributor bet on a new system and an automated warehouse at the same time. It also signed aggressive new contracts. Those contracts were priced on savings that the unbuilt system was supposed to deliver. Orders misfired. Stock disappeared into the automation. Costs exploded. In 1996 the company went bankrupt and was sold for a small fraction of its size. It remains the reference answer to the question: can a software project actually kill a company?

A US utility company went live on schedule in 2012, in the same week that a hurricane destroyed much of the area it served. Contract penalties and project exhaustion had made the date impossible to move. The payroll broke immediately. Workers were overpaid, underpaid or unpaid through the storm recovery. The repair bill ran towards a billion dollars.

A famous cosmetics company jammed its main distribution centre in 2018. It could not supply its biggest customers. And it learned the modern ending: shareholders now sue over a bad go-live, arguing that the risk was known and played down.

A large British city council ran its finance project from tens of millions towards ten times that, and it became one contributing factor in the council's effective bankruptcy.

And then there is the industry's most discussed anti-disaster.

A giant European discount retailer simply stopped. Seven years and roughly half a billion euros into its programme, it cancelled the whole thing and went back to its old system.

The reported reason is beautiful. The software assumed one particular way of valuing stock. The company refused to give up its own way of doing it.

People laughed at the time and called it half a billion euros wasted. The decision has aged into grudging respect. It is the only entry in the whole catalogue where somebody pulled the plug before the go-live pulled it for them.

Read as one long text, the catalogue keeps confessing the same three sins. Testing squeezed to protect a date. Complexity multiplied by changing the software to match old habits. And everything switched on at once, when it could have been done in pieces.

The companies that succeed do the opposite, and their success is boring and invisible. They go live in stages. They clean their data before they move it. And they treat the whole thing as business change with software attached, rather than the other way round.

The game never ends, either. Whole generations of these systems are now reaching the end of their support, which is pushing a worldwide wave of moves to newer cloud versions on the suppliers' deadlines. Thousands of companies are queuing for their own go-live day this decade, with consultants rarer and more expensive than ever.

And think about the people who carry these projects. The analysts, consultants and administrators translate between the company's habits and the system's rules. They have quietly become some of the most important people in any business. They are the interpreters at the collision.

So the case ends with the question every project in the catalogue answered too late.

When the company's way and the system's way disagree, who decides which one bends — and is that person in the room?

Companies that can answer before the go-live get the boring, invisible success.

Companies that cannot find out on the day the switch is flipped — with the trucks waiting, the payroll due, and Halloween coming.