I’ve been thinking about the utility of certain change models, and why even when you use them, change is still really hard to do! Some-where in my brain I recalled an academic paper from a while ago. It struck me it explained the conundrum well. Let me run it by you.
Karl Weick was the academic who wrote the piece. Weick’s was one of my absolute favourites to study. He’s an organisational theorist who writes on sense-making (among other things). He is super storyteller – and his academic writing is incredibly easy to read. If you get the chance check out his paper The Collapse of Sensemaking in Organizations: The Mann Gulch Disaster , in Administrative Science Quarterly, Vol. 38. At the risk of sounding very nerdy 14 years later when I catch up with friends from uni days we still discuss the concepts at dinners and BBQs. Yep, OK. A bit too nerdy?
Anyhoo, the article I am thinking of tells of the trade-off in writing good theory. (“Conclusion: Theory Construction as Disciplined Reflexivity: Tradeoffs in the 90s” in The Academy of Management Review, Vol. 24, No. 4 , Oct., 1999, pp. 797-806). In it Weick argues there is a trade-off in theory development of three attributes: generalizability, accuracy and simplicity. He suggests you can only ever get two of the three.
So for instance, a theory that is simple and accurate can’t be generalisable to the greater population. A theory that is simple and general will never be accurate. A theory that is accurate and general is going to be very complex. Of course, change models and change frameworks are simply extension of theory right?
Simplicity in change models is easy to align – any of the 3 steps, 8 steps, 5 stages are representative of simple models.
Accuracy is less easy to define. For me it speaks of comprehensive, complicated and nuanced change models – Taking into account multiple contextual factors. I don’t know I know many change models that do that to be honest.
And general suggests one size fits all. Something that can apply to most change.
So it stands to reason that a change model that is simple and general won’t be accurate (or comprehensive). A change model that is accurate (comprehensive) and generalized to many situations is going to be very complicated. And the change model that is simple and accurate – well it might just be a one-off right? Industry specific? Magic happens…
Food for thought…what has your experience been? I’ve found that anything that is simple and general has needed a lot of tweaking and adapting to make it work for the circumstances. Which returns me to the original pondering…what do you think?
I think models help us make sense of the reality we’re currently being faced with as change agents. I’m a novice in complexity science, but what I’ve learned so far is that the human brain is the most complex “thing” on the planet. People can’t design any model or process to manage the complexity of change because we can’t design something more complex than what our brains can process.
I think the tactics that change agents employ widely differ within which model they choose to use. Most of my change experience is with bringing Agile practices into organizations. I use Kotter as a checklist to make sure I’m doing something to meet each of his 8 steps (which I don’t believe were intended to be done in a linear way). I use the ideas in ADKAR to decide on tactics for gaining early adopters of Agile processes by finding people with high Desire. I use ideas from BJ Fogg’s behaviour model to make certain changes easier when I detect the motivation is high to do a new practice but the ability is low.
I prefer to keep my use of models as simple as possible and let the brain, communication and collaboration deal with the complexity. I think when people find that a ‘change model’ isn’t accurate or complex enough they may be focusing too much on following a plan and a process and not paying attention to what’s happening in the organization.
My specialisation is not ‘change’ so can’t make a considered response there but I think your thinking applies to models generally.
I come more from a lit theory background, and the general approach is that any model must be based on accuracy first. Problem is that the model become arcane and complicated. That is not so much an issue if your audience are peers or have the same kind of understanding.
In a business environment it is a requirement to simplify so a process of reduction is necessary. My thinking has been around (1) identification of conventions and constituents, (2) process flow, and (3) measurement and feedback.
So can you generalise from there and prove Weick wrong? My thought and experience is that generalisation is pointless. First, it is highly unlikely that a general model will outperform a tailored model. Second, no-one that counts is going to appreciate the additional work. Third, generalised content seems to be fine for internet articles but in a business environment it (rightly or wrongly) implies a lack of thought or finesse.
So, multiple accurate, simple models tailored for each occasion seems to be the go for effective usage and execution.
TL:DR I agree with you.
I see change models as something that help derive order from chaos, and support the successful journey. Even with the best model change is not easy – the right model makes it easier, the wrong model adds no value at all. The problems really come about when people apply an innappropriate model to their change journey and then blame the model. A bit like using a screw driver to try and hammer a nail and then blaming the tool when it does not work.
Intuitively Weick’s trade-off’s feel valid, and actually provide a pretty good approach to consider the applicability and effectiveness of the model/s being considered. The trick in selecting the right model is to have a good handle on availability of situational awareness/implications (part of sense making in the Mann Gulch case) and innate expertise available (wisdom in the Mann Gulch case). One needs to understand the degree to which high quality situational awareness is available during the journey and the degree of wise people available and empowered to support the journey. With this one can then assess the weakness & strengths of the available models on Weicks theory trade-off scale of simple, general and accurate scale. E.g.:
A. High situational awareness & wise people available = a simple and general model is fine to add that little bit of extra support (easy and low overhead to maintain)
B. Low situational awareness and inexperienced business people available = an accurate and specific model is needed (will cost more but is needed to drive both clarity of context and detailed actions needed for a successful journey)
C. Low situation awareness and wise people available = a specific and simple model is appropriate (model asks the specific questions to obtain greater situational awareness, however the available wisdom can apply this to the change journey with a little help).
D. High situational awareness and inexperienced people = a general and accurate model is appropriate (situational awareness is OK, model focus is needed to tightly guide the change activities to support the journey)
The above is not the only selection criteria I would recommend – still also need to look at cultural fit, etc. But it is a bit of a start.
Jennifer – generalizability, accuracy and simplicity… Oh what it brought to mind related to some of the recent dialogs regarding ACMP “Standards”. It’s not just “models” but methods and processes that can be challenged on multiple dimensions. I’m still noodling on it, but to me, my gut is saying there is a 4th dimension at play. I haven’t yet articulated the one simple word to summarize it. Something around applicability. How easily it can be executed against. This could roll up into generalizability, depending on how that is defined, but to me there is an element of interpretation and then application which makes it different. Still pondering. Thanks for stoking the brain today.
Jason – Yes, our brain is very, very complex and ever evolving. We map, rebuild, and rewire it as we learn. At times we also lose pieces and parts of it.
I’m with you in picking and choosing what makes sense in the specific context. My sense is that as people evolve in their “change experience” they need the methods and models less and less as they fundamentally understand things at an intuitive level. Earlier paractitioners often have a greater need for them in order to help them “make sense” of a situation and plot a forward course of action.
Scott – You elegantly point out why it is never “one size fits all” with your A, B, C, D. Thanks for laying that out.
Why is change always hard even if you have “good models”? Because it is people changing and people are complex organisms. Often it is people, processes, and technology all changing at once which introduces even more complexity and a dynamic system. Even for the most basic activities, such as getting dressed each day, individuals have their own unique approaches and sequences. How and when do you select your cloths? What influences your choice for that day? What room do you get dressed in? What sequence do you put your cloths on? Where are your shoes kept? How and when do you put them on? What type of outer wear do you put on?
“Getting Dressed” can be defined as a 3 step process or a 20 step process (or even more) depending on the level of granularity. In writing this, it struck me that the “generalizable, accurate, and simple” trifecta applies to many processes, not just methods, models, and frameworks…
I’m always concerned when people use change models as a rule or a truth. I’ve always thought that they were useful in helping you get your head around aspects of change and in certain cases to bring some sort of order to decision making. But change is a complex thing and anyone who tries to model that complexity does run the risk of making the model unusable. So treat a model as a guide, an insight, an aid but never a black and white truth.
This! I like the recipe analogy. And if you saw me in the kitchen you would understand!