Learning how to learn in information and records management
How many of us would consider our policies an error?
Chris Argyris wrote my favorite ever piece on learning. He describes learning as the ‘detection and correction of error.’
He defines errors as ‘any mismatch between intention and outcome’, and goes on to say that the evidence for learning is that we can implement what we have learned - we can correct the error.
Through that lens, how many of our policies are an error?
How many of them produce the intended outcome?
How many of our practices should we consider errors?
Functional classification is (as many of you know) my favorite example.
The intention of functional classification is that it will result in structures that don’t have to change over time.
The outcome is generally a system that captures less than 10% of what was intended because it doesn’t support the way people organise and manage their work.
What about our lifecycles?
We generally state the intention of lifecycle management as reducing legal risk and cost - and it’s a very good story.
Common sense says that those things will occur if we operate a lifecycle.
But have we measured the outcome?
Every time I do it, I find that the cost reductions achieved are dwarfed by the investment needed to achieve them.
The general design principle for retention schedules is also that we destroy the records once our legal liability has ceased - so we argue that the practice will remove a risk at the point where the risk is already gone.
Chris Argyris also spent decades looking at defensive routines, whose core function is that it blocks either the detection of the error, or prevents its correction.
For at least three decades, the error rate of core practices has been rising, and we’ve been engaging in defensive routines.
It’s entirely understandable, and a major factor in the underlying problem actually has it’s own book - the innovators dilemma describes exactly how incentive structures prevent established organisations from responding to change.
What’s clear though, is that we do need to change.
The way to change is to focus on the outcome, and then ask ourselves whether the existing practices are delivering it, and whether there might be other ways.
Policies - does restating the best of ISO15489 and the legislation produce high quality records and compliant outcomes? No? What could?
Classification - is functional classification producing high capture rates and transactional aggregations that are easy to manage? No? What other options are available? What are the trade-offs of those other options?
Is lifecycle management the way we do it now reducing cost and risk? No? How else could we deliver those outcomes?
We do have the innovators dilemma problem - the incentives are against us. While we deploy best practices, our approach is defensible because it’s best practice, even when it fails. The nice thing is that we have a management standard now, and ISO30300 is based on the quality management cycle - Plan, Do, Check, Act. Fundamentally, the quality management cycle embeds learning - plan how we’ll achieve the outcome, operate the plan, check whether it achieved the plan, act to embed the plan or change it.
I think our organisations will all support us doing this if we present it the right way - as an opportunity to improve performance. They both want improved performance, and they genuinely want what we promise. The first step though, is all on us - detecting the error. What do we intend when we implement a practice, is that the outcome?