Volume 108: Live from the 2021 archives.
1. 1984 called. It wants its dystopia back.
Since worker surveillance software is rearing its ugly head again, I figured it might be worth dusting this one off. Microsoft ultimately backed off of its “productivity score,” but that hasn’t stopped myriad others from stepping into the breach. To me, the whole worker surveillance thing says way more about weak management than it does worker productivity. Originally published in Off Kilter 51: 1984 Called, on December 5th, 2020 (OK, I know that isn’t 2021, but it’s close.)
tl;dr: Microsoft goes all in on employee surveillance.
You may not have seen it, but just before Thanksgiving, Microsoft got into a spot of hot water for releasing a new suite of analytics attached to Microsoft 365 under the banner of a “productivity score.” The major problem being less the name and more that it’s a full-blown employer surveillance tool.
Initially promising to deliver a person-by-person dashboard of exactly how many times a named employee attended a Teams meeting, interacted with a collaborative document, or even sent emails. After a Twitter storm blew up, they backed off a little by pledging to anonymize the data, but it’s still a terrible thing for them to be rolling out for three big reasons:
The metrics they’re delivering have literally nothing to do with productivity; they’re just measures of a person’s engagement with Microsoft’s suite of tools, so what this actually measures is busyness. And employees are really, really good at gaming daft metrics like that (reducing actual productivity in the process, btw).
Productivity is also an outmoded metric for whole swathes of knowledge work, originating as it did with the time and motion studies of the early 1900s, where men with stopwatches and top hats would stand there trying to optimize production line workers to deliver the same amount of work in less time. That just isn’t how we solve problems in the 21st century, where often the problems to be solved are different rather than the same, and nor does it reflect the value of the kinds of problems we’re solving.
Finally, productivity is one of those red herring client demands. When you research what businesses are looking for, they’ll all tell you they want to improve productivity. But it’s only a single data point. They typically want to improve other things, too, like engagement, culture, morale, education, trust, and the ability of the employee and their teams to take responsibility and solve problems. All the things that a “productivity score” imposed from on-high is likely to decrease rather than increase.
And really, this is the rub. Microsoft sees itself as “the” productivity company, so it’s looking at a highly complex subject through a very narrow lens. They’re then taking that narrow lens and applying an even narrower set of behavioral metrics to it: did the employee use the tools? Rather than a performance metric: was the challenge tackled effectively? I mean, God forbid, what if an employee was so good that they only needed a single message to solve a problem and not multiple documents, meetings, group emails, sharefile folders, and all the rest? In Microsoft world, you’d be passing that person over for promotion due to their low productivity score and promoting the annoying busybody who schedules the meetings to nowhere and then faithfully shares the results with everyone over and over again instead.
No, this, unfortunately, has little or nothing to do with actual productivity improvements or anything meaningful to do with how we work or how the very idea of work progresses. No, this was only ever about one thing: Showing how much people use Microsoft products through a faux measure of the value of those products to make sure clients keep paying for them and ideally buy more of them.
And, because of this, it’ll probably just end up dying like Clippy did all over again.
2. Citi jumps every shark that ever lived to pitifully pump Bitcoin.
With the crypto meltdown of recent months, I figured this was worth dragging up by its ears, screaming, from the bowels of the archive. Once again, Citi proves there isn’t a bandwagon it won’t gleefully jump upon. Originally published in Off Kilter 61: Airbnb learns $660m brand lesson, March 4th, 2021.
tl;dr: Abysmal Bitcoin report is giant red flashing neon warning sign.
If you read the business press, there’s a good chance you stumbled across this headline about Bitcoin becoming the currency of choice for international trade. The source for the article is this 108-page report by Citi Global Perspectives and Solutions. But you don’t need to read it because it’s filled almost entirely with made-up bullshit.
Take, for example, the following nugget: “36% of small/medium businesses in America already accept Bitcoin”. Umm, no, they don’t. So, let’s do a rabbit hole check and see if we can find out where this comes from. Ah, here we have it. The source? 99bitcoins.com. Hardly an unbiased actor, quoting an insurance company that hired a research company to ask 500 small businesses if they take Bitcoin.
Unfortunately, this is just the tip of the iceberg, as the full 108 pages are riddled with basic errors, outright falsehoods, and charts that make no sense. Most worrying is how it downplays, dismisses, and falsely characterizes a significant legal case happening right now that has the potential to impact the crypto landscape profoundly.
So, what’s going on here? Well, two things worry me. First, this looks suspiciously like the kind of conflict of interest boosterism that previously blew up in Citigroup’s face after the dot-com bubble of 1999. Does anyone remember Jack Grubman? The at-the-time famous analyst whose glowing reports were partly responsible for driving up tech stocks' value and who continued to boost soon-to-be bankrupt companies like Worldcom long after others had become more cautious. In his case, it turned out that he wasn’t an independent analyst at all but more of a salesperson reaping big rewards from the investment banking division behind the scenes.
Well…this feels suspiciously similar. Somebody within Citi will benefit financially from a bit of judicious Bitcoin boosterism; truth and accuracy be damned.
The second thing that worries me is that aside from the last-remaining-great-business-newspaper, I’m the one pointing this out. Why can’t the business press do their job and call out this nonsense for what it is instead of just re-hashing the press release and calling it newsworthy?
Many concerning things are happening right now in the world’s financial system, so having a megabank like Citi put its reputation on the line (again) to boost a spurious narrative about a cryptocurrency that’s risen in value by 2,000% in the past year should be a giant red flashing neon warning sign to all of us.
3. What gets measured gets manipulated.
The likes of Google and Facebook have spent considerable time, energy, and money persuading us that what’s on their dashboards is all that matters to marketers. But as effectiveness declines and earnings calls highlight the rising costs of online advertising, this seemed an apropos reminder that what gets measured gets manipulated—originally published in Off Kilter 67: A day late and a dollar short, April 16th, 2021.
tl;dr: An update for the oft-used management cliche.
One of the more common cliches’ in business is the statement that “what gets measured gets managed.” On its face, it makes sense. If you can measure something, you can baseline it objectively and see whether your management actions have a positive or negative impact. This is why measurement and business metrics are like sports scores. They mark your fitness to perform and thus your likelihood to advance and win.
However, I’ve observed over the years that “what gets measured gets managed” isn’t particularly accurate. The long-winded reality should be “what gets easily measured gets manipulated, sometimes falsified, and what’s hard to measure gets ignored.” I’ve yet to meet a business that, in some way or another, doesn’t manipulate easily measurable metrics to make it look like they’re winning the game. Why? Well, if your career depends on climbing the rungs of any large corporation, the metrics and measures you’re judged by don’t just indicate whether you’ll progress but whether you’ll even have a job.
When looked at through this lens, you see that the temptation to measure and manipulate can quickly become oppositional to good management. Here are three examples from large to small:
Many years ago, I consulted with GE, which at the time was the world’s largest corporation by market capitalization. Its primary strength was its scale and scope - it had discovered flywheel effects long before Amazon popularized the concept - but it didn’t compete that way. Instead, GE fragmented into thousands of mini-businesses that often competed harder against each other than they did the competition. Why? Because former CEO Jack Welch had said, GE would “be number 1 or 2 in any market or get out,” which led the organization to drastically narrow its definition of what a market was to ensure it was always “no 1 or 2” no matter how small or insignificant the market might be. Measure meet manipulated.
The second comes from a friend who used to work in the automotive industry. The measure to be manipulated was sales volume, which meant working diligently at the end of each quarter to ensure cars were “sold” and then just as diligently, but a lot more quietly, to ensure they were “unsold” at the beginning of the next. How else to ensure you remain the volume leader in car sales?
And finally, think about any agency business where you’ve been told to fill in your timesheets only with the hours approved in the budget, irrespective of how long it takes to do the work. This makes the people judged on their ability to manage a budget look great but wreaks havoc on the business's ability to accurately price future work and understand how busy the agency is. (As a side note, if you’ve ever been asked to do this, you’ve almost certainly found yourself working crazy hours because the people in charge of resourcing are working from manipulated data and think the agency is a lot less busy than it is).
It isn’t just easily measured and manipulated metrics that are the problem. It’s the deliberate ignoring of difficult-to-measure metrics too. Beautifully illustrated by Tim O’Reilly when he talks about the clothesline paradox: As a society, we only measure the energy used by dryers and entirely ignore the washing that gets hung on the line.
The reason this matters is that as marketing has become vastly more technology-driven and embraced an unprecedented level of quantification, it’s also become ground zero for manipulated, downright falsified, and largely ignored measures, which have led to all sorts of unintended consequences, such as the effectiveness of marketing going into a ten-year decline.
This is why performance marketing (easily measured, manipulated, and falsified) often gets budget preference ahead of brand marketing (hard to measure, typically ignored), even when the evidence points to a requirement for both.
It’s why we create bizarre and arbitrary attributions across the customer buying journey through surveillance that gives us a tremendous understanding of correlation and almost zero understanding of causation, excellently illustrated through this story of attributing a % of sales in a physical store to the door you had to open to enter. Or the observation that a significant % of digital advertising doesn't have an advertising effect at all but at best acts like wayfinding signage (thus suggesting the cost of things like Google AdWords are vastly greater than they should be based on actual commercial impact).
This neatly brings me to the dominant advertising platforms themselves. Google & THE FACEBOOK now control the vast majority of US digital advertising and a good chunk of total advertising. Both provide powerful black box measurement tools that drive marketing management toward measures that reinforce their platform dominance rather than what’s right for the marketer and have proven extremely difficult to audit independently. Concerningly, both seem to have problems telling the truth, as multiple lawsuits continue to demonstrate.
My final point, inspired by Ben Evans, is that no amount of technology, digitalization, and quantification of marketing changes the fact that marketing has and always will have marketing problems to solve. When technology points itself at a market, it moves in, destabilizes it, and then moves on. And when it moves on, it becomes apparent that those things that were being framed as technology problems often were not. In retail, we are left with retail problems. In banking, banking problems. Entertainment, entertainment problems, and so on.
And this is where we are at. For the past ten years, technology has destabilized the market for marketing, showering us with black box insights, surveillance ecosystems, rampant fraud, hucksterism masquerading as intelligence, and a broad swathe of automatically measured, easily manipulated, often falsified, and sometimes downright dangerous metrics combined with the hard to measure things that we completely ignore because they don’t fuel ad-tech or mar-tech monetization.
But, as technology moves on from marketing to bigger and more profitable arenas, I believe marketers will begin rediscovering the fundamentals of marketing again, which will mean less time spent worrying about click metrics and more spent creating customers, delivering on their needs, and creating value for them.