What If We Could Stop Unwanted Behaviors Before They Started?

Chapter 12 in my ongoing story about workplace surveillance

< Previous Chapter | First Chapter | Plot Summary | Next Chapter >

Keith was pretty pumped. Only 60 days into his job and he was the lead presenter on this month’s Data Deep Dive, which was still a fairly new component of the Human Dynamics team’s regular reporting schedule. He’d been to the last one, 30 days ago, but he had just sat in the virtual meeting with his camera off, since his manager just asked him to observe. Today, he was in the room at Building 11, and the whole Seattle team was there.

It was an exciting day for their team, and Keith was on point to show off the correlations he had been working on between the cyber stuff (which had all started before he got there) and the DEI stuff that he had been so immersed in. It turns out that there were some pretty powerful correlations between impulse control around things like phishing and data sharing and the tendency to “misalign” with Amazon’s DEI goals. Apparently the division’s Head of Strategy, Kate Stamper, had caught wind of what he had been reporting to his manager and wanted to be there when it was reported out.

Photo by Headway on Unsplash

It’s not that Keith was nervous, really. After all, his job today was purely to identify the intriguing cross-behavioral affinities between actions related to their cyber and DEI behavioral triggers. This stuff was pretty solid and he knew he could present it well. 

He knew his data was impeccable: after all, they had thousands and thousands of incidents (that’s what they called the behaviors that violated protocol), taps, and re-taps; a ton of tap validations, or TVs, which was where the behavior was corrected after the first tap; a smaller pool of TEs (tap escalations, where second and third taps were required to get to long-term tap validation); and, thankfully, a really small set of FTVs, or failure to validate, which meant that repeated taps—they’d gone to T4 on a couple people—still weren’t bringing about corrected behavior.

It’s what lay “beneath” the data that lingered in Keith’s mind, causing him a shadow of doubt. He had dug down into the commentary, and he was seeing something that troubled him. He’d heard that Stamper was pretty smart. Would she be able to sniff out some of the troubling undercurrents that lay beneath these intriguing data correlations? He guessed that today he would find out.

---

Kate Stamper occupied the room unlike any executive he had ever seen. Most of them walked in at the last minute, usually alone or paired up with another high-level person who got all their attention. They acted like they were super busy, and expected the whole room to quiet down the moment they signaled they were ready to go—even if that was five minutes late and everybody had been waiting.

But Stamper was different. She’d gotten there early—though not as early as Keith, who wanted to make sure he was ready to present. She entered quietly and immediately sat down next to one of his colleagues and introduced herself. As others started filling the room, she went around, introduced herself (as if she needed an introduction), asked people their names, how long they had been there, just regular stuff. She seemed to be seeking a connection with as many people as possible, and her affable demeanor set everybody at ease. But she was watching the clock, too, and just before the top of the hour she sat down next to Keith at the table in front.

“Hi, I’m Kate Stamper,” she said, offering Keith a vigorous handshake. “Most people call me Stamper.” 

“Hi, uh, Stamper,” said Keith. “I’m Keith, Keith Conn.”

“I know who you are, silly,” she grinned at Keith. “Your brother has told me all about you. I’m so glad we finally get to meet.”

“Oh, that’s funny, I didn’t know Dan had mentioned me to you,” said Keith. He was generally kind of formal with his “superiors,” but Stamper was making that hard. She had eye contact that just wouldn’t quit and a kind of vibrating energy that Dan had told him about ... but there was no way to understand it until you were in her presence. I guess this is what they called charisma.

“Are you kidding me?” said Stamper. “He says you got all the number smarts in the family, but that he’s better looking. I’m not sure I believe that.” She smiled and looked him right in the eye. 

Holy cow, Keith thought, is she flirting with me? But then she pivoted to her other side and turned that same charm on the other presenter and he realized that this was just Stamper. No wonder everybody liked her so much: she just lavished her undivided attention on whoever she was speaking to.

And then, at one minute past the hour, Stamper started the meeting:

“Hi gang,” she started out. “Gosh, I’m just so happy to see all of you here. I’ve heard that these Data Deep Dives are where the action is in Human Dynamics, so I’m so tickled to be here. I’ve met a bunch of you already, but for those who I haven’t met, I’m Kate Stamper and I’m the head of strategy here at Human Dynamics. Now, if you think ‘head of strategy’ means you can’t talk to me, you’re wrong! I really want to get to know all of you, so please, if you ever see me around, just come up and say hi. I’m just like you—I’m psyched to be here at the world’s greatest company and, I know you’ve heard it a million times, but it feels like Day One to me every day.”

Stamper had this way of saying the standard “Amazon exec” things that made them feel personal and real. No wonder Human Dynamics was getting such a great reputation as a cool team to be on, thought Keith. 

“For me it’s Day One because every day I get to work on something I’m really passionate about, which is helping employees improve their performance and keeping them from wasting their time. I joined because I wanted to end training as we know it—you’ve probably heard that already, right?—but once I got here and saw what we could do, it just really opened my mind. I wanted to help us all be the best Amazonians and the best people we can be. I’m so happy that so many of you share that goal with me, and even though he hasn’t said so directly, I bet our first presenter shares that goal as well. And with that, let me introduce you to Keith Conn.”

And that was it, bam, she turned toward Keith. It had happened so quickly he was almost taken by surprise ... but Keith was the master of preparation. He was ready to go.

“Thanks Kate ...”

“”Stamper!!,” Stamper insisted. “Really, I like my nickname!” 

“Okay, thanks Stamper,” Keith smiled. “I’m happy to get the chance to talk to you all today” — Keith was reading off his slide notes — “and to follow up on the data that Matt shared last month. As you all know, I’ve only been here a couple months ...”

“We’ve all only been here a couple months,” laughed Stamper.

“Right, you’re right,” Keith agreed. It was uncomfortable for him to be interrupted, but he pressed on. “Anyway, I’m ready to share with you some of my findings related to the correlation between the cyber behavior response matrix and our new DEI response matrix.”

And in his careful, overly corporate way, Keith shared his data with his team: he laid out the correlations between sub-optimal behaviors in cyber and sub-optimal behaviors in DEI. Both cyber and DEI had a list of desired behavioral targets for employees, and it turns out that employees who performed sub-optimally (that was the word Keith preferred) on the cyber stuff also showed a tendency to perform sub-optimally when it came to measures of diversity, equity, and inclusion. For each of his assertions, Keith was careful to show a slide and reference the data tables that he had provided links for in his handouts.

“Are you telling us that people who click on phishing links are also prone to being assholes?” Stamper joked when Keith paused for questions. She was one of those people who could get away with swearing in big meetings—no one held it against her. 

“Ah, well, that’s not ...” Keith was flustered. He was NOT one of those people who would ever swear in a big meeting. “That’s not a leap I’d take, but I would say that people who score below average on our cyber matrix also score below average on DEI, and that the worse the score is on one the worse it is on the other. So, you know, definite correlation.”

“Keith, that’s just fantastic—I’d hug you, but isn’t that on our safe workplace matrix?” Stamper joked, and the whole room laughed, not least because they all knew that unwanted physical attention was on the safe workplace matrix that they had started working on. “I wonder how that would correlate with phishing?”

“But seriously,” Stamper continued, “I love your tracking of correlations! I’d tell you why—but I’d rather hear what you guys think. Like, can you all see why tracking the relationships between these behaviors can be so important?”

Amazon wasn’t a place where a leader asked a question and everybody sat there quietly. These were ambitious, smart-as-hell people, for one thing, and they all saw having the best ideas as a great path to getting ahead. So Stamper’s question ignited the room:

“Correlations can help us design better taps.” 

“Correlations can give us insight into invisible drivers of behavior.”

“If we can track these behaviors into the hiring process, we can make better hires.” 

“Uncorrected behaviors probably also correlate, so we can predict how a failure to respond to a tap in one domain might predict behavior in another.” 

Stamper let it burble for a bit, nodding and affirming here, drawing out a point there, but not for long before she laid down her trump card: “Guys, these are all correct, but let’s not forget the big one: efficiency. If we can predict, based on one behavioral clue or violation, what other interventions a person may need, we can get them their behavioral corrections before they even get a tap! We don’t have to wait for someone to get something wrong before we show them how to do it right!”

And then Keith—who had been quiet through this last part—spoke up: “So you’re saying we might deliver a bit of training, or even a small intervention, based entirely on what we predict you’ll do?”

“Yes! Exactly!” gushed Stamper.

“So I could get a Kickstarter on inappropriate physical contact even though I’d never touched anyone?” Keith continued, referring to the little 5-minute learning snippets they were piloting with people before they were allowed to log on to the network.

“That’s the basic idea,” replied Stamper, her tone shifting to match Keith’s seriousness. “Wouldn’t it be great if we could stop unwanted behaviors before they ever started?”

Keith paused. On the one hand, he did think that at the pace they were collecting data, it wouldn’t be long before they could predict behavior in one domain based on what they were gathering in another, and they could intervene accordingly. At the same time, Keith couldn’t help but recognize that not all corrections were working, and to think about some of the feedback they’d gotten, comments that made him wonder if anyone was considering the downside to this little enterprise. Was the value they’d gain from these predictive interventions worth the cost—the cost of alienating some of their employees, or of dampening their overall performance due to their paranoia about the level of behavioral monitoring?

But he couldn’t just come out with that worry, could he? He didn’t have the data to say anything substantive, and he knew where he’d get if he voiced the “feeling” he had that this could backfire. Was this time to have backbone and speak up? Or was this the right time to disagree and commit? Keith decided right then and there that this wasn’t the time to counter Stamper’s enthusiasm. Perhaps he’d raise his concerns with her some other way, ideally when they had more data.

“Keith?” said Stamper. “Are you going to answer my question?”

Keith snapped out of his reverie. 

“Do I think it’s possible to stop unwanted behavior before it starts? Yeah, I think it’s absolutely possible,” he said with as much enthusiasm as he could muster. He just didn’t know if it was a very good idea. But how did he say that?

“I do too!” exclaimed Stamper. “So let’s end it on that note. Folks, here’s my challenge to you: as we expand the number of behavioral domains we’re scanning, let’s lean in on identifying correlations between seemingly unrelated behaviors. The more we know, the better we are able to nudge employees to be the best Amazonians they can be!” A couple people clapped, and there was an energetic murmur as Stamper left the room, smiling at people as she went.

Keith sat back and sighed. This could get complicated, he thought.

< Previous Chapter | First Chapter | Next Chapter >

Disclaimer: This is a work of fiction. I’ve made up the story and the characters in it. While certain businesses, places, and events are used to orient the reader in the real world, the characters and actions described are wholly imaginary and any resemblance to reality is purely coincidental.