A third decade, and some things just don’t seem to change. The following stories have been lightly fictionalised. The names have been changed…
One: The Nineties
Pete didn’t have a clue what he was doing.
He’d made the jump about eighteen months earlier, using his masters to move to a consultancy. Things had looked bright, but after its initial promise, the Next-Big-Thing had turned out to be Nothing-Much-At-All: the market now looked like a coal mine après-Thatcher. So much for his big break in consulting.
He had a young and growing family, had bills to pay, but was determined to keep a mortgage over their heads. He’d been terrified about being laid off and having narrowly dodged the bullet at the last annual RIF he figured his best strategy was making himself useful. The bench was not a healthy place to be. He’d fired up the internal jobs board and scanned the list of open assignments, desperately searching for something, anything, that he might be able to do.
Test Manager, he’d thought, how hard can that be?
The first couple of days hadn’t been too bad. He’d been introduced to some people and sat in a bunch of meetings, eyes glazing over as the team debated the finer points of implementing Soothsayer Enterprise.
It wasn’t until the third day that things got real. “So Pete,” the PM had said, “what about the testing?”
“Urghh”, he’d replied, grasping unconvincingly. THINK, he’d thought, don’t look an idiot in front of the client. What would testing this involve?
“Scripts”, he’d finally blurted out. You need scripts for testing, right? I’m sure I heard that somewhere. “We’ll need scripts – so who’s going to write them, who’s going to approve them, and who’s going to do them?”.
After that, things had settled down. He turned his hand to something he excelled at: spreadsheets. He had spreadsheets of scripts, spreadsheets to track the plans to perform the scripts, spreadsheets to report on the scripts. He was in his element.
Unfortunately, Pete didn’t have a clue what he was doing. Worse: he didn’t know it yet.
Two: The Aughts
Angela was having a bad day, probably the worst since she had joined Delanorme. Things had started out well: first class honours, snapped up by one of the Big Four, aced her GlobalMethodsTM Boot Camp in record time and parachuted into a high-profile presales gig with a leading bank.
Today though, it’ll all gone wrong. Worst of all, she couldn’t figure out why. She’d just heard the news: her bid had been beaten out by TLA Consulting. TLA Consulting! With pretty much no name in the testing business, and no recognized methodology to speak of, TLA had sent some old guy in to counterbid at the last minute. This guy, must be in his late thirties at least, didn’t seem to know anything about testing at all. He’d refused point blank to talk about the number of test scripts needed, rolled his eyes during a discussion about how many test phases would be involved, and seemed to spend most of his time talking with the client about what they were trying to do and what problems they’d had on the last release. Finally, he’d rocked up with a bid twice as high as Angela’s and talked some nonsense about “high volume mechanized checking supplemented by exploration”. Angela had been convinced she had this in the bag.
She checked the time. Twenty minutes: less than half an hour before one of the partners showed up wanting to post mortem. She desperately needed to check her facts, make sure she hadn’t screwed up. Angela fired up her screen, double-clicking on the TestQuoter icon.
Right she thought, Data Testing tab, let’s see…
OK, number of tables, check.
Average number of attributes…let me see, that looks OK.
Complexity factor, medium, seems reasonable.
Expected number of iterations; default. Was that the problem? No, she thought. She remembered her training: always leave that as the default three. No one ever took it seriously if you quoted fewer, or more, than three test cycles.
Her eyes scanned down the screen, her heart beating in her ears as the scrolled to the results box.
“Phew”, she muttered. I’m going to be OK. No matter what went wrong, it wasn’t me, I got the estimate right.
The results box read: 2000 test cases = $1 Million.
Three: The Teens
Watson sat in Zhang’s office, watching skyscrapers slowly emerge from London’s tungsten mists.
“Morning”, said Zhang, walking in, dropping her backpack on a spare seat.
“Hey”. No need to be formal: Watson and Zhang went way back. Whilst Zhang had moved on and up, riding her reputation as someone who could get things done, Watson was sticking it out at the old gig: he’d mentally committed to getting his current project over the line before even considering his options.
“I need your help”, Zhang continued, “your BU’s lit up red like a De Wallen window. Needs fixing fast or Krige’s going to get involved.”
Krige: Avasarala’s fixer-in-chief. Watson knew Zhang well enough to know that this wasn’t idle name dropping or an appeal to authority, if Krige was taking an interest they’d need a solve ASAP. “Go on?”
“It’s the Global Reg Programme: no-one upstairs has a clue what’s going on. And now the Test Governance Group are calling you out as red. Say you haven’t submitted any test reports yet. What’s up?”
Test reports. It didn’t seem to make any difference how many advances were made at the shop floor; this always came up on the big programmes. There was always an information vacuum at the top of the house, always a need to know where the flash points were so the top execs could muster support where it was needed. It was a legitimate need, but it always went the same way: by the time the need had been expressed through multiple layers of organization, it always degenerated into a data collection exercise based on the same assumptions: discrete test phases, script based testing, the same old meaningless metrics. Maybe it was the army of consultants, armed with their methods du jour, drafted in to prop up what was, in effect, a transitory endeavour. Maybe it was the inevitable consequence of miscommunication in how these asks percolated, like a game of Chinese Whispers, through the hierarchy. Maybe these things were just the default assumptions of people who’d never tested anything more than Watson’s patience.
Watson shrugged, mind already working on who would be the appropriate adult. “You know I’m not just going to fall in line for the sake of it, right? If I’m going to fix this, then I’m going to fix it. Do something meaningful.”
Zhang smiled. “I wouldn’t expect anything less”.
“I’ll get on it. Expect more noise before it gets better.”