All change has inherent risk, regardless of how “insignificant” the change appears to be. Small change does not mean low risk.
Most of the organisations I have worked for over the years have had some way of categorising change initiatives, based on their scale.
In some, for example, “scale” was simply a function of its estimated cost, measured in (for US companies anyway) US Dollars. In others, it was based on estimated duration – from initial approval to, say, the “go-live” date.
Why did they need to differentiate change initiatives in this way?
It’s because they felt comfortable that “small” change initiatives shouldn’t need to subject themselves to the full rigour of the governance process. They could be fast-tracked through a “governance light“ process. After all, there would be no point in spending more effort on the governance process than there would on actually delivering the change itself. Right?
Well, there is some sense in this, but there is danger too.
Inherent in change is risk, regardless of scale. The failure of a small project may be as impactful as the failure of a large project. Scale is not necessarily in the equation.
I once worked for a company in which a small “insignificant” change caused a catastrophic and expensive failure, all because the initiative was fast-tracked, through the “Governance Light” process.
The company in question had a customer base of around 500,000. Each of these customers had one or more insurance policies of one sort or another, all of which were held on the mainframe, which was the size of a small town.
On opening a new policy, or changing an existing one, the company would send a copy of the policy to the customer, in the post, with a covering letter. Also, on the annual anniversary of the policy (and at the end of each tax year) the company would write to the customer enclosing a statement and their warmest thanks.
There was a huge amount of printing to do and consequently, the print shop was truly epic.
The atmosphere was controlled in order to maintain a constant temperature, and huge vacuum ducts ran along ceiling above the three printers to suck out paper dust to reduce the risk of fire.
The printers themselves were housed in their own inner suite, with their own fire suppression system. The room was separated from the rest of the print shop by special doors which sealed shut in the event of fire to reduce the risk of the fire spreading.
If you worked in the printer suite and you heard the fire alarm sound, you wouldn’t want to hang around. Once it had been set off, there would be a 90-second delay before the doors would seal shut. The air in the room (and hence the oxygen) would then be rapidly replaced by Halon (an inert gas which is heavier than air) which was pumped into the room under huge pressure.
The good news is that since the fire would be deprived of oxygen, it would quickly die. The bad news is that if anyone didn’t get out in time, they would too.
The room was nicknamed The Bunker.
These printers were truly huge. 15 feet long, 4 feet wide and hungry for paper. In fact, they were ravenous. Each printer, at normal running speed, could get through 100 meters of paper a minute.
With each policy comprising an average of 7 pages, and the cover letter comprising 3, each one of the 500,000 customers would receive 10 pages.
That’s 5,000,000 in all.
Although each page would be required to carry the company’s branding, logo, address, company number, etc., these details were not pre-printed on the paper, as it was a continuous roll. The branding and other accoutrements were printed as part of the overall job itself. Printed on the fly, on every page.
On this particular occasion, they were preparing for the next print run. There had been some changes to the printer software recently, so it seemed sensible to run a test, to see if things were all running as expected.
The test produced a single sided sheet of A4 adorned with the company’s world famous branding. All looked well, until some bright spark noticed that on the footer there was a lower case m, where an upper case M should have been. Branding is important.
“No problem,” said bright spark, “It’s a quick change…”
Famous last words.
The change was made, the printers were locked and loaded, and the job was run.
Each page was also printed with a series of small black lines in the margin – like a barcode. These were used by the scanners in the next stage of the process which was to collate together the 10 or so pages, fold them, and stuff them into envelopes. These would then shoot off at terrific speed along what looked like a mini roller coaster for franking and sorting into hoppers ready for the postal services to do their bit.
Some time into the print run, a technician decided to take a sample print – just to check everything was ok. You can’t stop the printers, because it would take way too much time to reset them. But, at the press of a button, the printer can isolate a single after-cut sheet and feed it into a hopper for inspection – all at 100 meters of paper per minute.
The technician pressed the button, and out popped a sheet of A4. When he looked at it, it didn’t seem quite right. There was no footer. He pressed the button for another sample and found the same thing. No footer. And again. No footer.
At this stage of the run, approximately 50,000 envelopes had already been stuffed, franked and sorted. And each one contained an average of 10 sheets of A4, with no footer.
At the post mortem, they worked out what had happened. Bright Spark was right, changing an m into an M was a quick change. Unfortunately, though, he didn’t check it. He didn’t test it. It was a simple change after all. A small change.
The company used its own proprietary type face, in which an upper case M is wider than a lower case m. Although there was enough space for a lower case m, there wasn’t enough room in the footer for the new upper case M which replaced it. The software had to repaginate, and boom.
It’s true, the change was small. So small, in fact, that it wasn’t even called an “enhancement”. It didn’t even register as a “change”. Nevertheless, there was inherent risk – as there is with all change – which they failed to address.