Useful fictions

The more I learn about product management, the greater the array of models, frameworks, methodologies, processes, maps, tools and tricks I become aware of. This proliferation isn’t unique to the discipline; everyone runs their own representations of reality to some extent. But the intensity of this proliferation is particularly high within product management.

Here’s why: product management sits at the intersection of software engineering and development, business analysis and strategy, UI and UX design, and project management, and involves knitting these fiefdoms together with an ever-evolving blend of communication (what I deem “interfacing and integrating”). The fluid positioning of product management quintuples the available pool of what I now term “useful fictions”.

I think I picked the term itself up while reading Matt Lemay’s Product Management in Practice, and it’s stuck ever since. It’s short-hand for a component of fictionalism, “the view in philosophy according to which statements that appear to be descriptions of the world should not be construed as such, but should instead be understood as cases of “make believe”, of pretending to treat something as literally true (a “useful fiction”).”

As I discovered first-hand two years ago, each of the five above realms has their own, vast network of yarns. As a newb looking to parse the constituent elements of product management in preparation for a career pivot, it was easy enough to identify over five hundred useful fictions. The number of available useful fictions expands even more when product management itself is taken as its own, independent domain to fiction-craft in; I have no doubt I could find another hundred.

Some more informal characteristics of useful fictions? They are grounded in reality, and are neither baseless, relative concepts nor absolute, all-encompassing ideologies. They tend to be wider and/or deeper than a single metaphor or analogy, and narrower and/or shallower than a cohesive theory or philosophy. They’re often domain-sensitive and typically rely on context for their efficacy. They are lenses, models, perspectives and conceptual toys of varying scale, rigour and complexity that we use to explore questions and posit answers. And most importantly, there are many of them, some of which we keep to ourselves, some of which we share with others, and the majority of which are abandoned to the compost heap of history.

Bucketing these things as “useful fictions” has helped me navigate the often bewildering story-scape. But it’s also raised two questions: what sort of useful fictions are out there, and how does one evaluate their usefulness?

The first question—concerning the taxonomy of useful fictions—isn’t easy. After a little thought, I settled on two dimensions (and thus, four quadrants) for cataloguing useful fictions:

  • Fidelity (either high or low)
  • Constitution (mostly real, mostly unreal)

“Fidelity” refers to the resolution at which a representation of an object describes the object itself. A kid’s treasure map is lower fidelity than an atlas, for example. A sketch is lower fidelity than a wireframe which itself is lower fidelity than a prototype.

“Constitution” is a little trickier to define. I’m assuming that no thing is ever absolutely real or absolutely unreal, so our useful fictions will tend towards one pole. They’ll be mostly real, or mostly unreal. Take the class of useful fictions associated with understanding and empathising with users. Some of these useful fictions will be less real than others—traditional user stories versus profiles of revealed user behaviour created by processing high volumes of analytics data, for example.

The axes of fidelity and constitution give us four taxonomy quadrants. Useful fictions that are:

  • High fidelity, mostly real
  • High fidelity, mostly unreal
  • Low fidelity, mostly real
  • Low fidelity, mostly unreal

You’ll notice that I haven’t listed any examples for the given quadrants. You may have also noticed that I said I “settled” on these axes. They don’t yield any obvious answers. So the game is up. I’ve been found out. I couldn’t find a more satisfying way to slice up the vast array of useful fictions known and unknown. The above is what I’ve got. Feel free to fork the question and generate a more suitable, pleasing answer—and leave me to consider the second issue: evaluating the utility of a useful fiction. Here, I feel like I’m on firmer ground.

Every Friday, a bunch of Yaks convene for a study group. There are a couple low-key, recurring study groups/workshops/sessions going on right now—distributed system studies, the Fermi Gym, an AI workshop, the Mars rover—but the one I’m a part of is focused on “online governance”. That is, how distributed collectives form, coordinate and evolve. A couple weeks back we discussed Venkatesh Rao’s talk at Devcon, titled There Are Many Alternatives: Unlocking civilisational hypercomplexity with Ethereum, during which he juxtaposed two worldviews:

  1. TINA: there is no alternative
  2. TAMA: there are many alternatives

The difference between TINA and TAMA worldviews is whether they deny and attempt to crush hypercomplexity or acknowledge and attempt to nurture it. Venkat provides a definition of hypercomplexity:

“Hypercomplexity: property of a system that allows it to sustain many mutually incommensurate, divergent narrative futures at the same time”

A TINA worldview argues for convergence towards a singular, inevitable future state, and often comes with/relies on/aggressively advocates for finite game, coercive dynamics. A TAMA worldview accepts the plurality of present states, allows for and encourages divergence to a wider plurality, and is embodied via infinite game, participatory dynamics. (Venkat goes into more detail with more coherence; view the talk.)

This TINA-TAMA dichotomy provides one axis for evaluating the usefulness of a particular useful fiction. Let’s call the axis dimensionality. Is the useful fiction reductive or generative? More practically, does the useful fiction help me come to a decision (reductive) or does it broaden my perspective and/or the accompanying possibility space (generative)?

There’s an interesting asymmetry in that last question, too. I’m assuming that no-one wants to make bad decisions, so a more apt questions is: does the useful fiction help me come to a good decision? There’s no such constraint when thinking of a useful fiction’s generativity, though. The road to good ideas trundles right through bad ideas, so we shouldn’t disincentivise the outputs of ideation based on their quality. The useful fiction should be evaluated on the additional perspectives and possibilities it yields, inclusive of the perspectives and possibilities that could be considered insert negative adjective.

Other axes for evaluating the usefulness of a useful fiction are, as in the first question, difficult. I considered impact—high versus low, subtle versus obvious, immediate versus higher order. But that just feels like evaluating a useful fiction with another convenient useful fiction; a fool’s game. I’ll admit I’m out of options, for now.

However, progress has still been made. Naming things is a fundamental, hard and human problem. I now have a label for that vast, unruly library of things I’ve been bumping into and conceptually experimenting with for the last couple years: useful fictions. I’ve also got a primitive way to taxonimise them (low fidelity versus high fidelity) and a simple way to evaluate their utility (their yield of good decisions or their generative effects).

Not a bad outcome for several days of idle speculation.