A Vexing Problem, Part 5, Post 2

This is a continuation of a response to:

Joseph Carrabis, A Vexing Problem, Part 4, Post 2

My previous entry dealt with language issues and why an analytics argument won’t work with any audience.

Surely, it must work with pure marketing scientists? Or no dice?

My previous entry dealt with language issues and why an analytics argument won’t work with any audience. Here I’ll mix in cultural learning and how working memory affects how we interact with clients. It turns out that most people are not only prisoners to language, most people are also prisoners to what they’ve learned, how they learned it and the last time they used it.

(And at some point (tha mi duil na dhia) I plan on actually getting back to what Christopher wrote in his post)

The folks in charge of teaching and designing web analytics courses might benefit from knowing that how they present their information in class is going to deeply influence how that discipline will be presented and performed in the real world. This is true of everything, not just web analytics.

Stephane Hamel, for example, teaches some of UBC’s online web analytics courses and shared that there’s little being done to mentor students in real world practice (“…there are a couple of assignments but the feedback are sometimes not very positive (too disconnected with the course content, applied to specific sites instead of their own employer’s site, etc.) they don’t have access to a live tool, or real data, etc.”).

Stephane is currently preparing some courses for Laval University and is working “…to create a text book (maybe even just a structured list of articles to read). The information is all out there, the challenge is to structure it, give it a sense of continuity and think about real situations and cases.”

I’ve known Stephane for a while and have a great deal of respect for him. I’ll also be quite curious about how his future Laval students do applying their knowledge in the real world because (as I’ve pointed out in previous posts and Stephane has admitted to) his thought processes aren’t typical to typical web analysts. His metaphysic is different enough that his differences will be demonstrated to his students via what is called cultural transmission.

Cultural transmission is an anthropologist’s way of saying “structured, standardized learning”. Learning (not to be confused with “education”) occurs in two basic ways — social transmission and cultural transmission. Social transmission is learning that’s done by all animals on the planet, cultural transmission is specific to humans (so far as we know) and the difference between the two has to do with humans’ ability to store information internally (in our memories) but also externally (in print, electronically, etc.). Cultural transmission also tends to be both person and topic specific. For example, a professor will culturally transmit knowledge of quantum physics to students while parents will teach their kids how to cook. As the language in the previous sentence demonstrates, the former is specific knowledge transmitted between specific individuals, the latter borders so greatly on social transmission that I could just as easily have written “…dolphins teach their calves which fish to eat.”

So, long story short, how web analytics is done and how it is presented to clients has a lot to do with how it is taught and by whom.

And this is probably why people who take my or NextStage trainings or who’ve been in my presentations tend to reference them as “experiences” rather than “classes”.

Now pepper how something is taught and by whom with the fact that our behaviors are most strongly influenced by our most recent experiences (unless you’ve had lots of training. This is becoming my standard caveat, me thinks). More exactingly, our behaviors are most strongly influenced by our most recent experiences rather than the sum of our experiences. Example: someone burns their hand on a stove. They come back the next day, the stove is obviously off and has been for a long time, there’s nothing boiling, baking, broiling, braising, burning and they’ll still think twice before touching the stove.

Thus, not only is how someone does web analytics and interacts with clients (even in-house clients) going to be strongly influenced by how and who taught them, it’s also going to be strongly influenced by their last web analytics experience.

And if that last experience was less than positive? That less than positive experience will be re-enacted (totally non-consciously) in their present experience (again, unless they’ve had lots of training). Heaven forbid if the sum of their experience matches their last experience, both practitioner and client are doomed to a painfully non-positive experience.

Meanwhile, back in Canada…

So we have web analysts that are being trained to think in a way that doesn’t match the cultural metaphysic or identity, then sell into (make best use cases for) a mindset in ways that don’t demonstrate what that culture recognizes as having value.

And because their training doesn’t include (I admit I’m guessing their training doesn’t include) the tools necessary to adapt facilely to new situations (they are pioneers who have been trained by colonists, primarily) the less than positive experiences become the sum of their experience and their personal metaphysic becomes one of self-fulfilling prophecy (as was demonstrated by the Toronto WAW question, “Why are Canadians so reluctant to embrace data driven strategy?“, that led to these missals).

“Anchor and adjust” springs to mind, and the implications of the model are important, because it all folds in.

Ouch! (and did I mention NextStage is available for trainings?)

Plugging it. Plugging it.

And now a return to Christopher Berry’s “A Vexing Problem, Part 4“

You liken my approach to Jim Novo‘s. I’m flattered. Don’t know if Jim is or not.

My use of “if” in “if any of you are interested…” has more to do with my methodology than an intentional structuring of the language, me thinks, although I believe the outcomes are the same even if the motivations that lead to those behaviors is different (a clear case where understanding the {C,B/e,M} matrix is vitally important to solving the problem). I tend to present suggestions backed up with lots of data and research (both NextStage’s and others’) then let clients make their own decisions regarding what suggestions to act upon. You can see this methodology echoed in NextStage’s Principles as applied to cultural transmission — people learn best and most rapidly from their own mistakes (even though that learning will probably be anecdotal in nature).

Skepticism and credibility are topics I dealt with in yesterday’s post, Responding to Christopher Berry’s “A Vexing Problem, Part 4” Post, Part 1.

Your statement “…understand how the buyer of your product is thinking” I agree with completely. I would edit your “…we’re not quite enunciating the real value of action-oriented analytics” to “…we’re not enunciating the client-based value of action-oriented analytics”, ie, what would the client recognize as the real value of action-oriented analytics?

You also supplied the root of the answer, me thinks, in “Good old fashioned Canadian pragmatism at work.”

I disagree with “Front load the argument with a bottom line value statement” and apologize if I’ve mislead or confused. Or more likely I misunderstood what is written. My rephrasing would be “Front load the argument with a the client’s bottom line value statement, not necessarily a ‘web analytics’ bottom line value statement“.

Two things fall from this, me thinks, one a question and the other a suggestion (and did you notice how I gave you all the data and research, am offering some suggestions and leaving it up to you to decide?):

• What would a pioneer use web analytics for?
• Find more pioneers who’ve found a use for web analytics and invite them to teach it to other pioneers.

Whoosh! And great conversations, folks! I love these kinds of backs&forths.

(and I hope Dr. Geertz is pleased with my use of multi-syllabics in this post. More on The Good Dr. Geertz soon)

So, here goes an attempt to solve the vexing problem:

First, I’m going to neatly arrange all the facts on little pieces of paper.

Canadians are by and large pioneers. They expect to encounter unusual situations, and think first, act second. It’s a country of pragmatists and flexible instrumentation. It’s hell of a lot easier to sell a Sham-Wow that’s a towel, sponge, and pseudo-vacuum cleaner than it is to sell a towel, a sponge, and a wet-vac separately.

Front loading the “what’s in it for you – the client”, directly into any presentation, isn’t as effective as presenting an argument, carefully selecting data, and then leaving the client to decide for themselves if they’ll do anything.

Humans, not just Canadians, have an ‘anchor and adjust’ tendency. Reaching back to my public policy roots, the theory mirrors the entire “path dependence” of organizational behavior almost exactly. We’re cumulative products of our experiences. Unfortunately for web analysts, people tend to be a species of story tellers and analogies. Worse, most Canadians had the curiosity beat right out of their asses by the end of the “Terrible twos”.

But, there’s more! The tendency to use analogies, which is another way of saying “the comparative method” or “case studies”, is that it makes it very easy for logical flaws to go completely unnoticed. This method is also particularly susceptible to “seeing what you want to see”. (There are multiple interpretations to the Three Little Pigs case study – one of which is “blow harder”…not necessarily the right takeaway.)

EVERYBODY LOVES A NUGGET (TM) because everybody seems to love telling short, snappy, stories.

“I’m trying to think of a word that describes what you’re wearing…………….AFFORDABLE!”

Worse, business schools (and comparative political scientists!) tend to reinforce such usage of case studies.

Beware of how you socialize analytics into an organization: the impact will long far longer than you will.

My instinct, to solve the vexing problem, is to exploit the weakness in the case study methodology, however, there’s that good old arrogance bone that comes in that says, “no, I don’t have to co-opt that method, the web analytics method is better.”

I’ve really enjoyed the dialog 🙂 There’s layer after layer here.

Like a web analytics lasagna. Only meatier. And with better cheese.