top of page
Search
Cat Hicks

Corporate Attribution Error

I made up this term the other day and frankly, I just liked it so much that I wanted to document it on my blog so I would remember it. The term is Corporate Attribution Error. 


Corporate Attribution Error is what happens when a business (or anyone channeling the voice of The Business Narrative–so this could be a leader, a consultant, a spokesperson, a manager, an IC, your own mind talking to you inside of your own skull, it doesn’t really matter who it is, this describes an argument not a person, and I do not like the We Are So Smart And They Are So Stupid Exceptionalism that dominates many conversations about humanity in the workplace, so I won’t indulge it here) illogically and systematically attributes a large observed outcome in the world to the intentional and planned initiative of a business. 


What makes this an error is the illogical and systematic part: a claim that’s asserted with confidence unwarranted by the evidence, and a default to always align it with what the business wants. What also makes this an error is the exclusion of other factors. 


An obvious example: the macroenvironment around us is constantly changing. As it changes, it produces effects, and we only really see or have measurement of some of them – people want to buy less software, let’s say, but it’s not really an effect of “wanting to buy less software,” it’s an effect of “having less money.” Still, such things produce behavior changes (perhaps RTO mandates, because “having less money” also produces effects like “no longer feels ok to not really be using that lease”) that translate into numbers at a certain point in our accounting of the world that often ends up feeling like a gut punch. So perhaps a business sees a massive drop in renewals across the business, at the same time as their competitors. 


It is typically pretty obvious to people that many large things happen around them, e.g., “we all have less money right now.” In the first moments of big changes we feel free to make many explanations, see many factors. Yet my guess is that as the causal story evolves inside of a business, we frequently lose that early, clearer analysis that comes to us as a free gift for being aware participants in the world. Or perhaps we lose permission to say it. Either way the story winnows down, often not quite at once but as a casualty of the emerging narrative that gets fertilized by sources of authority that aren't interested in counter explanations. For good (at some point in this overwhelming world we have to focus on what we can do) and bad (you fill in the blank) reasons, many business narratives seem to become fixated on the aspects of the story where we feel we “control” the features involved: the sales strategy, extremely small product decisions we know a lot about but users don't even notice, any number of bullet points that we named as our key initiatives. The sensible larger story, where a massive change in the market is a multivariate function of both macro and micro factors, gets reduced to just this one micro factor, this one time. Another example: “so-and-so oversaw a growth in user engagement of over x%.” “Oversaw” is a typical weasel word when causality is implied but hardly evidenced; I oversaw the rise in my neighbor's garden but I hardly caused it.


Corporate Attribution Error can calcify into something really ugly and stupid. The economy can, for instance, pick back up for reasons that have zero to do with our little initiatives inside our one little company. So-and-so might actually be a deeply harmful leader who is ruining people's careers and taking all narrative credit for their work. But because we’re all packed into this place like sardines, because as Vicki Boykis says we are in workplaces in a terroir, it is easy to stop seeing the macro that we all know is there and instead affirm wild Corporate Attribution Errors simply by co-locating two little stories in the right places on the timeline. We lost sales, then we changed our deck, then the sales started happening. This deck must be magic.


***  

 

Causal thinking is hard for everybody. If it weren’t, we would have solved a lot of things in the world like cancer and climate change. Causal thinking is difficult for people who literally spend their entire careers trying to do science about it. So I write this not to castigate, but to contemplate. We live inside of attributions and our errors, always. I am curious most of the time about which ones are beating out the others, like invasive species crawling up our mental telephone poles.


Corporate Attribution Error as metaphor is, of course, me blatantly restating the Fundamental Attribution Error. The Fundamental Attribution Error is one of those effects that you might learn about in a Psychology 101 class. In its original incarnation it refers to the cognitive bias that many people have (how many people and in what regard, is disputed; originally this was claimed to be a general bias across most people) to attribute causes to individuals and not to the situations, circumstances and environments around people. “You’re late because you’re lazy,” instead of “you’re late because of traffic.” 


(Yet more parenthetical complexity. It’s important to know that “effects” in psychology are always being disputed, reconstituted, and reexamined; for instance counterweights to the original claims of the fundamental attribution error say that over-attributing to situations should be seen as just as much of a “fundamental” error than over-attributing to an individual. It’s also important to know that an ‘error’ in one context might be an adaptive, efficient strategy in another context. AKA, don’t get too tripped up by the terms bias and error when we’re talking about cognition; we are unlikely to ever be able to experience our own thinking without these cognitive mechanisms, we need to be able to ‘think quickly’ and in doing so balance the cost of thinking against the need for efficiency, but it is important to ask what our errors are leading us to do, and what other, better uses we could put them to).


But regardless of where we locate the error, error happens. We attribute credit, connection, recognition, belonging, ownership. We attribute cause. I personally guess that the direction of this error goes in the direction of a desired corporate narrative, and I would guess that it goes in the direction of privileged initiatives rather than environmental causes, based on my pretty educated experience that we always underweight environmental causes in my particular (US, tech) societal context.


As a social scientist in tech who primarily studies the environments and cultures around people, the affordances that make us feel like we can or can’t execute on certain types of strategies (e.g., what I’ve termed contest cultures vs thriving cultures, and the psychological affordances that either constrain or free developer experience), I do think the original definition of the FAE maps on to what I hear people ruminate about when they really open up to me about their past experiences in a business. When we share our work-related postmortems with each other, our “you know what that time was really like…?” I hear a lot of frustration at the causal stories that became unquestionable inside of a business. If the business is the "person," in this case, we seem to lean awfully heavily on explanations that reify the business and endow its “decisions” with both existence in the first place and with sole responsibility for outcomes in the second place (sometimes these stories are totally post hoc, and an “initiative” that was never clearly defined at all at the time, later becomes defined post hoc and talked about in the corporate mythology as The Big Change!).  


Actions that we care about analyzing in a business are of course made more complicated by the fact that they are usually not inherently one “thing,” but a word or phrase we are using to refer to a whole bunch of things. Particularly in large organizations, we are often forced into a kind of Executive Communication land in order to make information travel far enough to do something. This can make our definitions contextless, our accounts of cause more about tapping into as many associative ‘good things’ as we can in a short amount of space, than about actually operationalizing and testing a relationship. For instance, “customer satisfaction” is a succinct phrase that everyone can agree they want as an outcome. But what is satisfaction and what causes it? Undoubtedly it is multivariate. This is probably the subject for another blogpost. More likely, a book.



***


What the hell, Cat, should I do about it? You might be wondering. I mean, I don’t know. I just like naming things I’m thinking about. 


But some good news is that in my experience you can make a lot of progress by simply recognizing that a unique, important sphere of work IS trying to gather rightful evidence about what causes what and trying to make good decisions based on that evidence. That we won’t just automagically know this, and that there are many ways it can go wrong. If you want to not entirely fool yourself as you go about your business and presumably impact lives with it, a more scientific approach to causality is needed. If you’re wading through a corporate narrative terroir trying to figure out how to make a decision, a more scientific approach might be your only option. 


I’m not in a lab, Cat! No one said you had to be. Start thinking about this from the perspective of scientists who have had to work in the real world, with large, real-world problems and large, real-world data. We think about how to observe and crucially, how to set ourselves up for the claims we want to make.


Perhaps you can design small structured experiments with efficacy measures. Before you make a change, 


Perhaps you can empower the experts inside of your business who have thought and worked on evidence building. 


Perhaps every large initiative can have an evaluation plan. Perhaps features can have an efficacy measurement plan. Perhaps you can hire some good quantitative researchers? 


I’m not going to be able to drum up all the solutions in this one blogpost, because mostly, I just wanted to document that I found a name for this thing that bothers me. Corporate Attribution Error. If you name it, as per Madeleine L'Engle, perhaps you can wrestle with the chaos.


***


Here are some questions you can ask the Corporate Attribution Error demon in your head, the one that always wants to say WE DID IT

  • What are the other, plausible reasons that we might see this increase/decrease? What exist outside of our actions, and what exist inside of our actions? For the latter, are there large, unmeasured things that we also did at the same time as the thing we’re talking about? (for instance, we redesigned the user interface, but we also made a change in pricing that caused a bunch of our users to change) 

  • Does the size of the increase/decrease feel reasonable given the size of the change we made? Does the timeframe make any sense? (for example, it is reasonable that a single newsletter going out to a small group for three weeks could be associated with an increase in a brand awareness measure that sampled from a large market research pool?)

  • Did we set out to measure a change we wanted, and also measure counter evidence? (for example, we see an increase in the buyers of our product feeling happy about it, but our users are different, and they’re not actually using it any more than before, nor do they seem to be attaining the outcomes we’re promising to those buyers)

  • Did we exercise care in measuring multiple outcomes that could serve to illuminate each other, together? (for example, we thought a new feature would increase user engagement, and we do see users take longer on the site, but not on the features we wanted, and not taking a conversion action, therefore we aren’t reaching the type of engagement we thought would happen from more time on site) 

  • Are we able to replicate the change in similar circumstances? (for example, we rolled out a new feature to a small group of users at one point, and then tried it a second time with a different small group doing the same kind of work, and saw a relatively comparable change) 

  • Are we able to replicate the change in different circumstances? (for example, we rolled out a feature for our teams when the economy was thriving and everyone was jubilant, but it still helped achieve good outcomes when everyone was experiencing friction and pressure a year later)

bottom of page