The Internet Has Gone Foul. [Part I]

“Let me tell you some things I find productive. Positive reinforcement. Negative reinforcement. Honesty. I’ll tell you some things I find unproductive: constantly worrying about where you stand based on inscrutable social clues, and then inevitably re-framing it all in a reassuring way so that you can get to sleep at night. No, I do not believe in that, at all. If I invited you to lunch, I think you’re a winner; if I didn’t, I don’t. But I just met you all. Life is long, opinions change. Winners, prove me right. Losers, prove me wrong.”

— Bob Kazamakis

Founded in 1944 as the “Winnipeg Sales and Ad Club”, the Advertising Association of Winnipeg (or AAW) has for many years – and rightly so – laid claim to being “Manitoba’s largest advertising, marketing and graphic design community”. The AAW recently hosted its 2023 Signature Awards, an annual event intended to honour and celebrate the very best work put forward for consideration by its members. That the winners each year indeed represent the “best of the best” work being done, we can be certain, for the event’s homepage assures us that “the Signature Awards entries are judged by an unbiased international panel of senior-working industry professionals”.

This year’s call for entries solicited nominations in the categories of “best individual online ad”, “best collection of online ads” (these first two categories to be judged explicitly on their ‘creative’ aspects), “best website”, “best microsite”, and “best social media campaign” (though these latter two appear to have been folded into the “Miscellaneous” category, by the time finalists were announced). You might note the absence from this list – as in every year’s call for entries preceding it – of any category of honours, recognised by the largest local “professional society” to which I might belong locally, under which any aspect of my own career as a digital marketer sensibly fits.

Nevertheless, for curiosity’s sake I recently found myself running a few checks on the winners of this year’s AAW Signature Award in the “Best Website” category. The winning agency’s client is (or had been) a rather charming-looking bed-and-breakfast in rural southwestern Manitoba. I have refrained from naming either party here, so as to save them any undue embarrassment:

• A cursory Lighthouse audit gives AAW’s “Best Website of 2023” a performance score (when rendered on mobile devices) of 39 out of 100 – which is to say, firmly entrenched in the bottom half of all known websites (when rendered in that context), according to the HTTP Archive.
• An analysis generated by whatdoesmysitecost.com found that loading the AAW’s “Best Website of 2023” on your phone, via a Canadian mobile network, would typically cost more than $1 (USD).
• The WebPageTest “Carbon Control” report estimates that with each new visit to the AAW’s “Best Website of 2023”, roughly two grams of CO2 are emitted into the atmosphere. This figure is more than triple the average estimated CO2 emissions per visit of the top 1000 sites on the Web.
• The AAW’s “Best Website of 2023” includes no privacy policy, which is a de facto infringement of the Personal Information Protection and Electronic Documents Act (PIPEDA). That is, the website (and its owners) fail to satisfy the literal bare minimums of data protection and respect for users’ privacy rights which have been required of Canada’s private sector, under law, for almost a quarter-century.

It would be unfair of me to let you go on thinking (if you had been thinking, just now) that my intent in this post is to continue “punching down” at my chosen example here. It would be similarly unfair to suggest that the kinds of criticisms one might level, given the points above, are in any way especial. That is not the case. Having now surveyed all of the AAW’s 2023 Signature Award finalists, judged either as a “website” or a “microsite”, I can tell you in full confidence that they’re not even the only members of this ‘exemplary’ subset which fails to reach basic PIPEDA compliance. In that second example, where one might normally find a link to the requisite privacy policies, one finds instead this brief note of gratitude: “Funded by: Justice Department Canada”.

There’s Something on the Wing of the Plane!

“Digital marketers”, as a rule, are not good at digital marketing. That much is obvious. What I mean by this is that the prevailing majority of those employed as “digital marketers” today – in varying capacities – could not tell you, accurately and in terms you might understand, what it is that they do, or what they are trying to do, or even what it is they’re meant to be doing (for in practice, the honest answers to these questions are often distinct). This too is obvious – or at least, I contend, it should be. It is my carefully considered opinion, and you may cite it for all that is worth.

You might assume (supposing that you’re willing to grant that it exists in the first place) that this phenomena tends to be concentrated towards the lower end of the pay scales. I harboured a similar suspicion myself, earlier in my career, that perhaps my own experience could be chalked up to “local factors”. Again, it’s just not so. The kind of systemic, pervasive ignorance I have described is endemic to virtually every level of marketing management, precisely because the conditions under which it thrives are so common to human relationships, and thus the organisations we create.

Perhaps the main benefit of what recruiters would describe as my “diverse working experience” – obvious code for “routinely unemployed” – has been that I’ve managed to get an inside look at marketers work in a wide range of contexts, both on the “buy-side” and “sell-side” of digital media, and in service to organisations in many industry sectors, of many sizes, with many varying objectives, and serving a wide range of local, national, and international markets.

I’ll spare you the effort: it’s turtles, all the way down.

This can all get to being – to put things mildly – a bit of a problem, especially when contemplated on the scale of an enterprise, or an industry, or the digital economy writ large. Indeed, as someone whose day-to-day work often involves grappling with the many problems exacerbated by this more fundamental one, I’ve often found myself at a loss to accurately convey its “bigness” (though I will make another attempt here, a bit later on).

After much dedicated effort – plus a heap of trial-and-error – for a while I came to rely on a particular favoured metaphor to describe one such “big problem”. I won’t recount it here in its entirety here, but it involved a commercial flight taking off from one airport, bound for another, when suddenly a major natural disaster causes catastrophic damage to the runways at their destination (with various stakeholders stood in for the flight crew, the passengers, air traffic controllers, etc.).

In retrospect, it is perhaps more telling to simply note that this was the clearest metaphor I’d yet to arrive at, even with non-technical audiences in mind. But the point of this imagined scenario was this: We can no longer follow our earlier plans, because real-world conditions are irrevocably changed. Nevertheless, the aircraft must still land. It has a finite supply of fuel, which constrains its remaining flight-time, and thus the places which might be reached in order to attempt a landing. Some landing sites are preferable to others; most landing sites would likely cause catastrophic losses. So, what should everybody do now?

Beer Ain’t Drinkin’

More recently, I’ve begun to incorporate a great many concepts drawn from “systems thinking” in general, and “statistical process control” in specific, into my work. Among these most recent projects, I will tell you in full frankness, are some of the finest works of my career. My employer (and our clientele) seem pleased as well. But before any of these things could happen, or indeed can keep happening on an ongoing basis, a new (yet by now familiar) conversation must always be had.

At the outset of each new project, some new grouping of people (that is, the “stakeholders”) must agree on what their problems are, and how they intend to go about fixing them, before they may commit, collectively, to having a(nother) go at solving them. One becomes acutely aware of this process when, for example, they begin to advocate for a specific approach (like SPC), within a novel context (such as digital marketing management), wherein most of the stakeholders involved can safely be assumed to consider most of its methods and precepts to fall somwhere between “counterintuitive” and “career-threatening” – or at least, at first blush.

Imagine my happy surprise, then, in recently having arrived (by way of Deming) at the works of Stafford Beer, and in particular, his 1973 book Designing Freedom. A brisk, fifty-page read (many of which are given over to the author’s own hand-drawn notes) and first presented as a six-part series of lectures intended as a popular introduction to “the science of cybernetics”, Designing Freedom was commissioned by the Canadian Broadcasting Corporation, and subsequently presented, published and broadcast by the CBC as the 1973 Massey Lectures.

I will call my personal discovery a “happy surprise”, for that is the kind thing to say. There is also, I freely admit, a keen sense of intellectual jealousy in my recognising so superior a distillation of so many concepts which I’ve struggled and meandered my way towards expressing, over many years and by various sources, all being presented so clearly, and in far more suasive terms.

[Incidentally, I believe the most recent Massey Lectures, Astra Taylor’s “The Age of Insecurity: Coming Together as Things Fall Apart“, engage with many of the same themes and topics addressed by Stafford Beer in his own contribution to that venerable series, fifty years prior. I’ve not yet gotten ’round to the more recent lectures, but I aim to before this series of posts is done.]

Here, let me show you a little more of what I mean. It’ll be good for me.

How It Started, How It’s Going

A year or two ago, I was invited to give a lecture on the topic of digital advertising to a class of students enrolled in the “Advertising” stream of the local Red River College Polytechnic’s “Creative Communications” diploma program. It was, to me, a very exciting prospect, for a great many reasons. Indeed, how often are we “working people” presented with an opportunity to speak directly to those now seeking to enter our profession, and to tell them (a bit of) what it is (we think) they’re really getting themselves into?

I doubt that I’ll be invited to give another. Not that there was any trouble “on the day”, mind, for my little talk seemed well-received – at least, as far as one can tell these things through emails and video-calls. It was a fine time, and I greatly enjoyed the experience. No, my doubt stems from having neglected to return some of the attendant paperwork, related to a token $50 honourarium, and by now this has no doubt caused some annoyance both for the instructor who so kindly invited me, and foir a few accounts payable clerks, too.

Anyways: this being a rather expansive subject (and me being the sort of person I am), I eagerly set about slapping up the dozen-or-so presentation slides I felt I’d need, and combing through some of the “prior art” to help tie concepts back to real-world examples. And so, following some perfunctory definitions of “AdOps” and of some relevant industry jargon, and a brief enumeration of the kinds of tasks which are common (in my experience) to this line of work, I arrived at the thesis of my speech that day, which I’ll restate here:

• Digital advertisers work within, and are tasked with administering, systems of enormous complexity. One need not go on at length recounting all of the terrible and wondrous things that online advertising can do (or purports to do) in order to appreciate that these depend upon numerous, complex, interconnected systems.
• Operating within (and upon) these complex systems, digital marketers are tasked with performing good work on behalf of other stakeholders, and with communicating the results of this work to them.
  • This is a key detail, since (with rare exception) the essential functions and goals of “marketing” are subordinate to those of a larger institution.
  • To put a finer point on it, few organisations exist with the “animating purpose” of communicating this existence to the outside world.
• It is useful, therefore, to briefly consider how it is people (and organisations) actually go about “knowing things” – and here, I submit that one will find three strategies are prevalent, in general practice, in the context of online marketing:
  1. Inspection: Careful examination and analysis of the products of one’s work, by subjecting these to (at times, destructive) testing. This method is analogous to (and to great extent modeled after) the “intuitive” ways by which children have been observed to go about learning about the world around them.
     • Pros of Inspection: Capable of identifying (perceived) errors, defects and/or strengths, in a given process, or its outputs.
     • Cons of Inspection: Labourious; time- and resource-intensive, and typically occurs “after the fact” (AKA “post hoc analysis”).
  2. Fraud: Fabricating aspects of one’s work, in order to satisfy (in the immediate, surface-level sense) the requirements of the larger organisation. This is, I must emphasise, the most prevalent strategy one finds adopted in practice by digital marketers.
     • Pros of Committing Fraud: It is often less difficult (in many contexts, trivially so) to produce correct-sounding responses to informational requests. “External stakeholders” will also tend to lack the domain knowledge and/or resources necessary to challenge answers which deviate from reality.
     • Cons of Committing Fraud: Moral hazard; reputational risk; it tends not to feel very good.
       • I try not to belabour the point, since most folks don’t need to have “fraud is bad” explained to them. People know that it’s bad. Nevertheless, fraud is inarguably a strategy, so it’s useful for us to consider the conditions under which one might choose to adopt it.
  3. Statistical Process Control (SPC): This was the subject of the remainder of my lecture; in the words of one of its pioneering minds, “the statistical control of quality is the broadest term possible for the problems of economic production“.
     • Pros of Statistical Process Control: Near-real-time monitoring of the “outputs” of our work; an emphasis on early identification and prevention of faults in a process, rather than relying on “post hoc” review and/or analysis to understand (and iterate on) the finished results.
     • Cons of Statistical Process Control: Less intuitive than either of the preceding two strategies; some degree of training and education is often required for all project stakeholders, in order for such methods to prove successful/effective.

From there, I went on that day to describe the role of “variation”; of “normal distributions”, and Shewhart’s invention of the “control chart“; of the need to distinguish between the “stability” of a process and its “capability”, and so on *sniff*. In the intervening years, I’ve also added a section which deals with “tampering” (aka “overcontrol” or “management-by-numbers”), and another on the basic natures both of “quality” and of “loss”. These most recent versions, you see, have come to form an hour-long “crash course” of sorts, which I’ve made it a habit to offer (some might say “threaten”) to present to most colleagues and/or clients, at some point or another.

There. Now, did you follow along with all of that? If you had, then great! I’m so pleased! And, if you hadn’t, that’s fine too – you wouldn’t be the first, and the fault is mine. But here, now just have a look at these four pages, from the author’s lecture notes as reprinted in the published version of Designing Freedom. Know a little better my seething intellectual jealousy:

If the ideas (and doodles) above have intrigued you at all, but there is some other part of you shouting that yes, this is all fascinating, no really, but that it’s just not feeling up for three-or-so hours of heady, far-ranging lectures on fundamental cybernetics, that is perfectly fair. We’ll pause here, for now. But if that does happen to be the case for you, then might I suggest that you consider saving it for a summertime beach-read?

No, really; here are just the first three sentences, from Beer’s first lecture:

The little house where I have come to live alone for a few weeks sits on the edge of a steep hill in a quiet village on the western coast of Chile. Huge majestic waves roll into the bay and crash magnificently over the rocks, sparkling white against the green sea under a winter sun. It is for me a time of peace, a time to clear the head, a time to treasure.

In Part II (coming soon!), I plan to return to those “big problems” we spoke of earlier, while relating a few of the ideas Stafford Beer articulates in the first three lectures of Designing Freedom. After that, I hope to focus on Beer’s fourth lecture in Part III, and maybe sort out why his notion of an “Electronic Mafia” feels in many ways, to me, a far more apt metaphor than the more recent (and fashionable) framing of “surveillance capitalism“.

Happy New Year,

– R.