A Spectrum of Actions, Part III
In part 1 of this series on the organizational “Will we or won’t we?” of action items, I made vague reference to “higher cost” action items on the action item magic quadrant requiring “more” (and promised more on it later!) And in part 2, proposing the action item spectrum, there’s a weird blue line pointing to the right:
What do I mean by ‘more?’ And what is that blue line about?
For the third, and final, part in this series: let’s talk about that.
In the two previously presented models, the X axis represents “cost” (conceptually) on the quadrant and “probability of completion” on the spectrum. In day-to-day discussions, “cost” is a rough, though acceptable proxy in most economic systems for “likelihood of completion.”
But while they are distinct concepts, they act as proxies for each other because they generally are (inversely) related: higher cost action items have less probability of being completed. But they also share a number of characteristics and effects, all of which are important to acknowledge as we honestly examine whether or not any specific action item will find its way to fruition.
When we talk about “costlier” action items, there’s obviously a sense of ‘moreness’; but more what?
- “Costlier” action items usually involve a larger, often broader, set of tasks. By the very nature of large and/or broad remediation tasks, this usually translates into involving “more functional areas,” which means…
- More people are involved. More people doing the work, more people managing the people doing the work, more… “socio-” of that socio-technical system rearing its head. Ignoring those dynamics (as many would prefer) creates additional complexity (which can dwarf the increased technical complexities).
- More people and more teams involved? You’ll definitely have more communication, more communication paths, and more coordination between groups to get the item completed.
- More people and more teams means… more buy-in is required! The action item spectrum is littered with action items that lacked sufficient buy-in, and therefore never got implemented.
(It’s worth noting: buy-in, conceptually, is a two-way street: in an optimal world, we would seek buy-in to complete an “expensive” action item, just as others would seek buy-in to eschew that task. In practice, it seldom works this way: declining to do an “expensive” action item is often a fiat declaration; it’s those arguing to implement an “expensive” action item who bear the burden of cultivating buy-in and promoting an AI’s benefits.) - “Buy-in” is really about building alignment and common ground for the particular task (or, as items become “costlier,” projects). For these types of AIs, someone must sign-up to carry the torch to maintain group alignment and common-ground and repair it when it starts to break down. Failure to do so is a common contributing factor to half-implemented action items.
- Because of the ‘moreness’ of the people involved, the complexity of the solution, and the associated complexity of the remediation project, as we move to the right of the spectrum, we are more likely to observe organizational power dynamics. More “permission” is required to spend precious capital and time to implement the fix. (As the interplay of these dynamics becomes more prevalent in the planning and execution of the remediation item, declarations of purely “rational” or “data-driven” decision making also become more suspect, precisely because organizational power dynamics have nothing to do with rationality or data.)
Jabe Bloom has done a lot of work for his PhD on design, our perceptions of time, frames, and storytelling. It turns out: these are all critical aspects of incidents.
One aspect of the “cost” of those “costlier” action items is time horizon: they take longer. And because they take longer, the frames in which both individuals and organizations reason about them becomes wider. And because of the wider frame, there are more potential futures, which business typically translates to ‘more’ risk. The following slide from one of Jabe’s talks illustrates this perfectly:
These longer time horizons for “more expensive” action items also impacts the “organizational story” around both the action item itself: in the same talk, Jabe explores the phenomena of executives and engineers having trouble communicating during an incident, due to the time horizons of their respective frames:
Why is this relevant? The actual framing — the story — of any particular action item requires translation as it moves “up” the organization. It’s more than just the ‘moreness’ of communication and coordination described above: it’s an active translation process. And someone has to sign up to translate if these action items are to get implemented. In the real world, organizations have varying capacity and capability to actually perform these translations in an effective way.
The last ‘moreness’ of “costly” action items worth exploring is their impact on the story of the incident that was their genesis.
Every incident has a story. In fact, incidents have multiple stories, some of them “official” and some of them recorded only in the recesses of the memories of the engineers who worked them, only to be retold in the hallway tracks of conferences.
But stories need to have endings. Our brains are hardwired to expect a conclusion to a story, even if it’s abrupt, vague, or unsatisfying. (The Sopranos fans know what I’m talking about.) Large, “costly,” “complex” action items extend an incident’s story. They challenge the establishment of an “ending” to an incident, its containment. This, in turn, forces the organization as a whole — whose leaders just want to know “Is it fixed and it won’t ever happen again, right? — to “keep the story going.” It also challenges our personal cognition to “keep paying attention” to the story, because… “it’s not over yet.”
(Ever wonder why “large” action items so often morph into their own beast of a “month’s-long remediation project?” It’s because it allows the story to conclude, so the organization can view “the incident” as “done and solved,” and “move” “on.” It also creates a new story… and this story isn’t in the ‘storybook of incidents.’ It’s in a different ‘book,’ a book of ‘normal’ work that doesn’t hold the calamity and uncertainty of ‘an incident.’ The metamorphosis into a ‘new story’ makes it is socio[-technic]ally acceptable for the organization to react to it differently.)
These are just a few aspects of ‘larger,’ ‘costlier,’ ‘more complex’ remediation items. Surely there are others aspects which increase (or decrease) as we move to the right of this spectrum. (And I would love to hear the ones you observe in your company!)
So… that’s, in a (three-part) nutshell is how I think about incident action items: as being on a spectrum that if we pay more attention to, we can see… more.
When teams use it, revisited over time, it serves as an important source of information on how an organization, in the aggregate, thinks about, reacts to, and actually executes on the action items that fall out of incidents.
