originally posted on one of my several now defunct blogs, called On Engineering, on 7th July 2013
Who reads procedures?
What, you don’t? Alright then, another question - who writes procedures? Ah, well, on that front, I have a confession to make: I do.
But hold on - if you’re not reading, why bother writing?
Let’s go back to the start, see what we’re dealing with. Procedures are company-internal documents that describe how things should be done. In that sense, the closest analogy I can offer is that, like the laws of the land (which I presume you, like me, haven’t read in their fullness), procedures form part of a distinct yet only subconsciously perceived scenery - barely noticed until you crash into them, or if a newcomer starts asking for directions.
That newcomer point is important: the bigger the company, the less likely it is that someone can naturally pick up and run with the workflows and networks of people and systems that combine to get things done. A start-up’s first instinct is not to write its procedures. Only when the company is established and has grown do such things begin to matter and procedures - instruction sets for beginners, for those lost in a maelstrom of work, or for those about to be audited - are written. (Rands in Repose has written a lot about transitioning from start up to big company, including and very eloquently about procedures and process in the software world)
So, I’ve written procedures. I’m not a prolific proceduralist, thankfully, and the ones that I’ve been involved in are hardly internal best-sellers, but at least I can say that whilst novels they certainly are not, I have at least experienced some satisfaction at having the procedures approved and published.
Or perhaps journalism would be a better analogy than novels - when I was writing them, there was a real sense of urgency surrounding them. There was a sense of deadline, of needing to fix something that was broken.
But with a procedure? Which nobody will read? Hmm, that doesn’t quite add up. Unless people do read them. Some few, key people who then tell others how things should work - and those others more or less understand the intent and get on with it.
Until teams drift apart, new people come in and things get broken again.
Procedures - the making of
The key to remember is that procedures are principally a work in the written language, which means they always need some knocking back and forth and into shape - editing - at best within a small, motivated and expert working group. They are official documents, so they’ll need to be given their ID numbers and released into whatever system your company uses.
Procedures are also visual documents, so a certain design aesthetic is also worth investing some time in. Even if the only graphic in there is the workflow, it’s always worth trying to make it look and feel right by considering format, colours and alignment. If it looks shitty now, it’ll look shitty in three years’ time, and it will still have your name on it!
Procedures are structured documents, too, like technical reports: Introduction, Scope, Definitions, Resources involved, the step by step guide and the workflow should all be in there.
Structure, word and graphic need to be sound in logic and clear enough to read without frustrating or confusing a reader by descending into ever diminishing circles of definition (or spiralling up its own arse in grandiose sweeps of meaningless bluster. Ah, speaking of which, I’d better be moving on here…)
What is a good procedure?
A procedure is like a process flow, but for people. Do this, then that, inform this person, file that data there, fill out this form, review with that team, finish and get on with your lives. Like a good process flow, a good procedure should capture the steps, decisions and loops that result in a final product. A process results in a part, a procedure results in information. Both should maximise output and minimise waste.
Do procedures define or describe?
Companies generally muddle by, if we’re honest about it. There are things that need to be done and, buoyed by our honest and human sense of duty usually overriding any sense of “oh, let’s just forget it and do something fun instead”, we talk and email and chivvy and stress and eventually get things to a sufficiently complete state that we can move on to the next disaster.
Procedures formalise and idealise those inner workings of a company. Going back to the process analogy above, a procedure should be designed to maximise efficiency - in once sense, it is an engineered document (hopefully with decent spelling), in the sense that defines how things work in their optimum way.
So the question above is misleading: procedures do describe, but really they should define.
Are procedures a cure or a symptom?
One key reason for writing procedures is to correct course on some workflow that is sailing onto the rocks. Often, the rocks weren’t formally recognised before, so the new (or updated) procedure is effectively charting dangerous waters.
One of the biggest balances that a company has to manage, however, is resources. Even if we were all able to live and work on one project only, and all of our team members could do the same, tensions and bottlenecks would arise. With teams being involved in the multiple carousels of projects, individuals start think-feeling for themselves, taking on tasks that aren’t strictly theirs “to keep things moving”, while others try to keep low and out of the firing line. These tensions generate ever greater emotional potential.
This is where procedures form the “laws” of the company. He should be doing this, she should be doing that, but only when given this signal or that information. They should serve to de-escalate things. If they can’t, then it’s a clear sign that the respective teams aren’t set up correctly for that combination of tasks.
Ghost protocols
Eventually, all of these new definitions and workflows become the new day-to-day and procedures are no longer referred to. Or or those new spangly ways of working fade into disuse, yet their respective procedures remain, sitting in (often a bewildering array of) libraries somewhere, gathering digital dust, crumbling as entropy attacks their bits and bytes.
The dust is kicked up by everybody’s favourite bugbears, audits. As audits suddenly loom (which they tend to do), links are checked and if they still work, are distributed, some conscientious colleagues bookmark them and may even scroll through the procedures themselves. They generally won’t read them, though.
Occasionally, an auditor will prudently validate his job description and note a discrepancy between a job done and a procedure. Hands will be wrung, failure to comply noted and efforts made to be better in the future.
By that stage, though, procedures are usually dead. They lie in state - Lenin’s waxy visage springs to mind, though without his enduring popularity. Perhaps three people in my company know where to find them. Everybody else who needs to know anything about how things should run simply ask me directly - or kind of muddle by.
Anything, of course, being preferable to reading and understanding procedures. Although: at least procedures don’t get irritable if you consult them.
