3 reasons why I hate pretty graphs

People prefer the pretty guy over the uglier guy cos they like pretty things more than they like things that are right.

For over a decade I’ve been working in an area where the obsession has been about finding just the right type of pretty so that managers would pay attention to performance reporting and things would just all click.

Every year or two a new format of performance reporting would go through the approval mill and be approved as being officially prettier than the last one. The last framework is now thought of like an ex boyfriend, considered unattractive, thinking with a shudder “how could we have, this one is so much prettier

The word “pretty” is never used when people look at a new arrangement of graphs or tables on a screen or page. It is smarter, easier to read, cleaner, more modern. Pretty, basically but unsayable as an adjective.

Why the urge to prettify? I’ve noticed the thinking is about trying to be SO pretty that managers pay attention, that the graphs and tables would sashay into the room and BAZINGA all eyes go on the graphs.

The assumption is that managers don’t use performance measures etc because they find measures and graphs etc reeeeallly boooorrrriiing and the only way to get them to pay attention and do something worthwhile  is to gussy up measures and data, and slip knowledge in with something pretty. Sugaring the pill.

If the assumption is that managers aren’t using measures and data because it is boring and irrelevant, then it is entirely logical to make your communication and presentation of data and measures more user friendly. To follow solid design principles is not a bad thing. But the reason why i hate pretty graphs, as the blog title says, is because….

Whilst I’m a big fan of doing communication right.

I’m a MUCH bigger fan of doing the right communication.

The assumption that managers just aren’t using or thinking about data and measures cos they’re ugly, and therefore will when they are pretty is probably wrong. This is the first reason i hate pretty graphs…

1 There’s no use for graphs full stop

They just aren’t using them cos nobody uses them.

If you are trying to get managers to pay attention to data and measures then you don’t work in an organisation that does data based decision-making.

 

 

You work in a normal ordinary command and control organisation where decision making is separate from the work, where the further you are from work, the higher the pay and the more powerful and ignorant (in the literal sense of the word) is the decision making. There’s no call to use measures and data because that isn’t how management happens.

Whether it is day to day management of work, or changes to the design of work, if you ALREADY worked in an organisation that needed data to make decisions then they would ALREADY be paying attention and using measures and data.

Instead of managers disinterestedly or uncomprehendingly staring at performance reports they would be simply desperate to find out what was happening, tearing them out your hands.

There would be pull, in other words.

Pretty graphs are a cunning attempt at pushing data, when the opposite is the case, you need to create pull.

Managers need to pull data and measures because without it decisions cannot be made. If they can be made without data cos they’re being made cos the highest paid man wants it done, there’s no pull for data. The data is literally useless, because it cannot be used.

The problem with focussing on prettying up performance reports is that no attention is paid to WHAT is being prettified. When you’re prettying up what you’ve already got, then….

2 It’s trying to solve the wrong problem

This also means that you are not solving the right problem. If the wrong problem is “we’ve got ugly graphs” and that’s what you’re focussing on, then the right problem isn’t being addressed. The right problem is management that doesn’t need data to make decisions cos it uses other things to make decisions. Pretty graphs aren’t going to work cos graphs arent going to work, no matter what.

The temptation is to think that what’s needed is more TECHY graphs, like control charts and run charts with all the bells whistles and systemsy add-ons.

Things like THIS….

With all the added analytical tools like this

Stuff that could just hammer home rational decision making. Who could fail to see the light when presented with stuff like this?

This is something that won’t work because those silly wrong measures exist for a reason. Red/green scorecards and binary comparisons weren’t magicked up by foolish performance people because they didn’t know any better, but because bad performance measurement is  a symptom of bad management thinking.

Red/green, up/down, higher/lower, all forms of binary comparisons fit in with the thinking.

A scorecard that shows if a target has been hit or missed with a silly up or down thumb next to it or a smiley face is the answer to the question “have we hit our targets?”.

A control chart showing how much a service varies does NOT answer that question. It is an answer to a question that has not been asked.

This is not the problem you should be trying to solve. Giving managers control charts that answer systemsy questions is falling into the trap of giving things from ONE worldview, yours the systemsy performance type, to someone from another worldview, a manager in a normal ordinary command and control world.

They still have the questions and assumptions from their own worldview, the control chart will be understood through that worldview. The one that’s causing all the mischief that you don’t like or want.

So a pretty stupid graph or an ugly clever control chart aren’t the thing, so what IS then clever clogs? I don’t like pretty graphs because…

3 Pretty means finished means done

Producing pretty charts, either dumb-ass ones or clever ones, is a bad thing because if its pretty it can’t be messed about with. It is a finished product. Signed, sealed and delivered.

“Et, VOILA! Your monthly performance reports Sir, transcribed on vellum as you requested”

Pretty graphs are often the product of people like me, performance types, possibly the worst people to “report performance”. Cos reporting performance implies THREE types of people where there should only be one.

  1. the worker who does the work
  2. the performance type who measures it and
  3. the managerial type who is distant from the work hence needing a special performance report to tell them all about it.They get the pretty graph. When they get it, it is finished and done.

There is a way out of this, by removing the need to report TO someone, it removes the performance type and the managerial type. Now there’s just the worker who does the work, measuring the work, so they can improve the work.

They do [drum roll] THIS!…


This is a measure, in time series, that identifies true change using objective methods, IN THE WORK….

Graphs on a wall in the work. Not signed and sealed and delivered, but written on and annotated. Somebody marked the shift with marker pen. Arrows pointing to when things happened that made an impact, linking data and efforts to improve, creating learning about the work.

This is not a signed, sealed and delivered pretty graph sent away from the work.

This is a living thing used for a purpose. Not pretty, much better than that, useful and used.


These are the 3 reasons I don’t like pretty graphs.

  1. There’s no use for graphs full stop
  2. It’s trying to solve the wrong problem
  3. Pretty means finished means done.

Pretty graphs are an attempt to make an information radiator without anything to radiate.

Start UGLY. First measure the right thing, in the right way. Then build your information radiator.

Advertisements
Posted in data, measures, statistics, systems thinking, thinking, vanguard method | Tagged , , , | 7 Comments

How to make the world seem REEAAALLLY BOORRRIIIING

How to be bored

  1. Take a performance report.
  2. Get the data out of those damn tables and silly bar graphs.
  3. Stick it in a run chart, or a control chart, s’up to you.
  4. Look for any actual change over the course of the whole time period.
  5. Yawn.

Whenever I get my hands on data, a rare occurrence for a performance person, I’ve found 90% of the time there has not been any real change.  That over the whole time period nothing happened. Booger all. Sweet FA.

You might find something like this…

These are the standard run chart rules to check for one consistent unchanged process…

Applying these tests to the chart above what do we find?

YAWN!

Nuffinks changed!

No exciting shifts or trends.

Nowt!

Over 3 years!

 Zero!

Nada!

 

Managers and strategic leaders DO do stuff over this time period, but because they’re human beings, they react to things that are special and new and distinct, or at least LOOK special new and distinct. They see signal where there is just noise.

If this week’s performance number is lower than last weeks performance number, then look at this week and find out what’s different about it. And probably this will result in some kind of shouting at somebody. Either way, the focus will be on the special, the one-off, the peak, the low-point, the very latest. The focus will be on….


Or they might focus on the outstanding features to the eye….

 

 

Either way, they’re looking at where their eye drags them. They are eyeballing data.

Using the eye for analysing data is as bad as using the gut, another mis-used body part often used for feeling or reacting instead of it’s evolved function of producing faeces. The irony.

Instead there is another useful body part just dying to get involved in decision making and data analysis, the brain. Once equipped with the more accurate mental models of how the world works and the necessary tools to apply them, the brain will look at something else…

This means instead of looking at the latest data point or the sharpest looking peak, you look at what the actual whole process is doing. Cos, as the cliche goes, it’s perfectly designed to get the results it is getting. Especially if all there is to look at IS the boring old up and down of common cause variation.

Looking at the common cause variation in the process will mean they  devise a common cause solution.

Looking at imaginary special causes leads to special cause solutions, and special cause solutions in a common cause world are just tampering. No change results, just fiddling round, wasting opportunity. At best just increasing the common cause variation. But without the mental models or tools to support them, the increase in common cause variation will go unnoticed by everybody but the customers who experience it.

This is why control charts are soooo boooorrring. They show up the futility of acting without knowledge, over and over again, world without end.

Posted in data, knowledge, statistics, systems thinking, Uncategorized | Tagged , , , | 2 Comments

The 2,500 year old lesson everybody ignores cos it’s too hard

Harold MacMillan was once asked, what is most likely to blow governments off course, he replied…

Events dear boy, events.

And events is what organisations continue to believe will change things, events meaning one off things that happen. When they want to improve the way we do things around here then it is events that they turn to.

One off things that happen, like….

  • strategic away days
  • staff development days
  • training days
  • bleedin’ rapid improvement event days
  • planning days

All with the word “day” at the end. Cos that’s how long they last. A day or two.

Then what happens? Well if you’re not in an event, you are back at work cos the very definition of event is something that begins and then ends, and once ended it’s whatever the base  level default behaviour is, ie business as usual. So what is the effect of an event?

If I look around me I see exactly what I saw last year, and the year before that ad infinitum. There are people sitting at desks typing stuff. And pretty much typing the same sort of stuff with the same sort of effect that there has been for years, i.e. not much of value to a customer.
But in these same years there have been SO MANY events that came and went. Went being the operative word. The thing that didn’t change is what was between these events. And the thing that is between events is huuuuge. Like this pic below shows.

Business as usual! That’s what I see is the way we do things round here, punctuated by brief events that don’t touch the sides.

If business as usual is what the usual is, then the usual is what you will get. Not the hoped-for things contained within the sporadic events. They don’t work. Business as usual wins.

Events don’t work. Why do we keep on doing them? We keep on doing events because we keep on doing events. You can’t get better if you keep on doing worse.

This is something that Aristotle knew thousands of years ago….

I mean this lesson is old. It is simple and has lasted cos it’s true. Getting better at anything is a repeatable every day habit, not a one off event. 

If you say this out loud, nobody would disagree. It’s a two thousand year old cliche. It’s true, but it’s safe.

What isn’t safe is the next time you’re at an event pointing out that all the other previous events didn’t work so this one won’t either, because it’s an event. You can tell them Aristotle said so.

Posted in all wrong, change, command and control, learning, systems thinking, Uncategorized | Tagged , , , , | 2 Comments

I am a police officer. 

When people ask me, “So what do you do, where’d you work?”

I reply…

I’m a Policy Officer

And inevitably people mishear that as police officer.

Oh really!” they exclaim, looking me up and down with surprise.

Then I have to correct them, “no, sorry, POLICY officer.” and then explain what that is, because you can’t just plop those words in somebody’s lap and walk off. Generally I mime typing and say “you know, reports and that” and people gladly move the conversation on.

Regular readers of this blog will know that whatever I do I don’t hold it in high regard. This is because I can see that there is no line of sight between me and a customer/client/whoever so it is highly unlikely that anything I do is of worth in helping the customers of the organisation.

Not only that, I KNOW it can’t be of value because I am a corporate cog in a normal ordinary command and control management system, so I am based on flawed theory before I walk in the door in the morning.

So, I’m a policy officer (with attendant typing-mimes)

What though if I was asked a different question…

What problem do you try to solve at work?

This is the old “what problem are we really trying to solve here?” question, but applied on a larger scale to a job, rather than a situation. I love this question, because it focusses attention on the actual thing itself, the reason why you should be getting up in the morning rather than the bells and whistles when you get to work. This question allows you to talk about something useful and relevant, like reality and its attendant concreteness.

People with very real jobs like Mr Bun the Baker might reply…

I’m trying to make the tastiest bread and buns I can at the lowest cost

If I were asked this question though, my answer would be similarly opaque and round the houses cos I’d try to answer honestly in a POSIWID manner…

how to strengthen the appearance of data based decision making at a strategic level

Or more honestly…

how to balance creating the appearance of  being valuable, or at least polite, with the expenditure of least energy

The tricky thing is how to describe the problem you are trying to solve, without talking about the actual silly things getting in the way of you solving the real problem. At any moment in time the work I have to do has no impact on helping a customer solve a problem.  It is getting in the way of me solving a real problem, if one were to actually present itself. So the “problem I try to solve at work” is more along the lines of “the problem I try to solve cos there’s no real problem to solve at work“. A pseudo-problem.

As an aide memoire, the problem these people are trying to solve below is the bit on the right….

If they thought the problem they were trying to solve was the bit on the left, they’d be obsessed with the bearings and the trucks etc and wouldn’t be looking at the other bits of the problem they should be trying to solve, ie the bit on the right. How does the skateboard work in the context of what the customer really wants?

Looking at the bit on the right the skateboard is only one small component. The guy is wearing clothes, probably clothing that would be suitable for leaping over things

 

So the things that “they really want” could include trainers, a cap etc

These are things that fit into and complete the thing “they really want” cos who wants to be throwing sick moves in a pair of pants that your mum’s ironed with a crease down the front?

 

Or perhaps the thing “they really want” includes the obstacle on the ground, and the smooth ground itself. Cos if where they are it’s all grass, then there’s no skateboarding.

The thing “they really want” is a much bigger thing than a skateboard, so knowing that you make skateboards is one thing, but knowing that the problem you are really trying to solve is providing the customer with the thing that they really want helps you know find the right problem, rather than the problem you have right now.

Me? I am a policy officer trying to solve the problem of strengthening the appearance of data based decision making at a strategic level.

Hard at it.

###############

This blog post very much inspired by Tobias Mayer  a thing he did called “What do you do?”. You could say it’s a blatant rip off. Think more homage, with added musings.

 

 

Posted in clarity of purpose, systems thinking | Tagged , , , | 5 Comments

When is a team meeting NOT a team meeting?

Q: Does your manager cancel it if they cannot attend?
A: It’s not a team meeting. It’s the manager’s meeting.Q: Does your manager set and send out the agenda?
A: It’s not a team meeting. It’s the manager’s meeting.
Q: Are you expected to attend by default?
A: It’s not a team meeting. It’s the manager’s meeting.

This is not a feature purely of generic team meetings, it applies to any regularly timed meeting in a normal ordinary command and control organisation.

Think of people who need to meet to solve some kind of mutual problem. They self organise and decide individually whether they should or shouldn’t go. The things that need to be talked or decided grow out of the shared problem. If somebody can’t come at that time or date, then the nature of the problem and the degree and nature of the missing person’s involvement in it would be the criteria that decide whether the meeting is rearranged so they can come later or at a different location.

Essentially the problem itself that causes the meeting, dictates the meeting.

On the other end of the scale from problem-driven meetings are hierarchical meetings. These are total crud. They happen because. They happen because of the features of a command and control hierarchy, such as the existence of managers, decisions made separate from the work dropping vertically from above, performance monitoring, “planning” in the absence of knowledge etc

The hierarchy dictates the meeting, who attends, the content, the form and how decisions are made within it.

Here’s the test of a meeting, whether it is a manager’s meeting or a problem solving meetings….

Does the law of 2-feet apply?

The law is simple…

If you are in a meeting and are not contributing anything of value or getting something valuable, you have the responsibility to use your 2 feet and walk away.[link]

If you cannot use the law of 2-feet, because you’re expected to be in the meeting due to because, then you’re in  a manager’s meeting.

If you can, congratulations. It’s a real meeting. Enjoy!

Posted in command and control, systems thinking | Tagged , , , | 1 Comment

Folk management

How do you think organisations work?
This is how Jeremy thinks his boiler works….

This is what a fair few people including me think a boiler works like. Not in as obviously ludicrous way, but  comments I’ve sourced from nearby people are similar…

  • I’d turn it high, so it gets really warm quicker, then turn it down to the temperature I want”
  • “If you whack it up high, it makes it hotter, which is what you want. Then turn it down”
  • “When it’s cold, turn it much hotter to reverse the cold then put it to the heat you want”

This isn’t how thermostats work, but it’s how people think they work.

Most people don’t give it much thought, instead they use the nearest metaphor they have to imagine how it works. Probably something like putting your foot down on an accelerator hard, so you speed up faster, until you reach the desired speed then you take your foot off the pedal. Or closer to the actual situation, turning on a gas fire full until it heats up a room, then turning it down again.

When people don’t know how something works, they imagine how it works. But crucially, they’ll not know that there’s a difference between imagining and knowing. There’s something called folk physics, which is the study of how people imagine the world works when they don’t know how it does.

For example people think that when water is piped through a coiled spiral of pipe, that when the water emerges from the end it will continue to spiral, in a coil of water. It doesn’t, it comes out in a straight line, but surprisingly large numbers of people think it comes out all wiggly.

What are trees made of? Where does all that wood come from? Lots of people think it comes from the soil, water and something to do with leaves and sunlight.

Trees come from fresh air. 95% of a tree come from carbon dioxide, the air around it turns into tons of wood. Sounds quite weird and unlikely, but only if you don’t know how photosynthesis works. 

This is why normal ordinary command and control management is so dumb. Because it actually is dumb. Or rather, it’s the equivalent level of dumbness as thinking that a cannonball falls faster than a marble. Just an incorrect model of how reality works, easily tested by empirical investigation. But startlingly dumb, if intelligence is the ability to acquire knowledge, then the failure to test assumptions of normal models of work is dumb cos no new knowledge is acquired.

Do targets work to make services better? No, you can test that and find out.

Do appraisals work to make services better? No, you can test that and find out.

Folk management is the result of the ongoing inability to generate new knowledge by testing the underlying theory that it relies on. Most organisations don’t know the theory that they operate under, only the visible manifestations of it. They know they do annual plans, budget monitoring, have service standards, but not that this is a choice driven by an assumption about how the world works.

As the Deming quote goes, without theory there is no learning. Without knowing the theory of how boilers and thermostats work, you can’t use it properly to keep you at a pleasant temperature. Without knowing how work works, or at least what your own theory of work actually is, then you can’t learn how to get better at it.

Posted in Uncategorized | 9 Comments

Double Trouble

Mr SquireToTheGiants did a nice thing on why organisations can be like the politics and game playing of Game Of Thrones, but without the deaths or dragons, explaining double loop learning in the process. Here it is and you should click to read more.

Squire to the Giants

Double troubleThere’s a lovely idea which I’ve known about for some time but which I haven’t yet written about.

The reason for my sluggishness is that the idea sounds so simple…but (as is often the case) there’s a lot more to it. It’s going to ‘mess with my head’ trying to explain – but here goes:

[‘Heads up’: This is one of my long posts]

Learning through feedback

We learn when we (properly) test out a theory, and (appropriately) reflect on what the application of the theory is telling us i.e. we need to test our beliefs against data.

“Theory by itself teaches nothing. Application by itself teaches nothing. Learning is the result of dynamic interplay between the two.” (Scholtes)

Great. So far, so good.

Single-loop learning vs. Double-loop learning

Chris Argyris (1923 – 2013) clarified that there are two levels to this learning, which he explained through the…

View original post 2,383 more words

Posted in Uncategorized | 1 Comment