On a plane flying back from Boston (Mass.), eaten second breakfast of the day, watched a bit of "Where the Wild Things Are", annoying, fractious kid who needs therapy (or a sharp slap) and a bunch of needy, fractious rather dopey creatures, disappointed, switched it off, didn't even care to see if he was reconciled with his poor benighted mother, bored, listening to Muse, need a coffee, some battery life in laptop, here goes...
Recently, I was working with a colleague who exclaimed "You've got to be a professor to understand that page" when looking at a consulting 2x2. Indeed there are some great pages in the world that capture some key thoughts or concepts so concisely that they can be expressed just on one page, but need a voice-over to talk through the layers of meaning embedded, maybe like one of those pointillist paintings or a fractal montages that is made up of pictures that are made up of pictures...(but perhaps not a Dali-esque or Picassoid other world view?).
This diagram below (not the one being commented on, I hasten to add), captures the entire eco-system of outsourced application development on both technical & commercial dimensions, ranging from the narrow individual project up to the strategic vendor relationship level
Very clever, of course, but it really deserves to be supported by 20 following pages to unpeel the layers and break out the key concepts, etc., etc.
(oohh, a quick round of orange juice...)
But it looks like this when you morph it Dali-style...
...but that is just plain silly, of course. (but an excuse to try out the Virtual Plastic Surgery Software, why don't you give it a go on one of your favourite photos, and make your self look like your favourite film star, or the Bride of Wildenstein...)
Battery dying....break to watch X-Men Origins: Wolverine, just another load of shouting and uber-angst
Down on the ground now...
However, there are some charts that are very easy to understand, but they do convey a message that is counter-intuitive, and so take a while to get your head round.
This chart is a good example
This shows the output of a model of software development productivity which paradoxically shows that total coding cost falls whilst the developer daily rate increases. This is, of course, quite counter to the expectations of typical Aggressive Sourcing gigs which tend to focus on bashing down the daily rates. The Old Wives and proverb writers of yore new about this since the principle of "Pay peanuts and get Monkeys" is well known.
This is what they used to say, but I do wonder if this phrase might be considered racist in these days of off-shoring, say maybe it should now be "Pay Peanuts, get Numpties" or something like that...
The twist in the tail on this analysis is that in the formula P x Q, where P is Daily Rate, and Q is the number of days needed to complete the project, some people (yes, them) are not aware that Q is inversely proportional to P. This is the essence of the move to Agile development methods, which favours people over process (amongst other things).
Finally, I also offer you the 2x2 I wrote all by myself one day after an afternoon's presentation by one of my erstwhile colleagues, a quite (self) important and entitled sort of chap who gave a long presentation from which I came out reeling with "Framework overload", having survived the discourse from the evolution of Sailing Ships to Dell's policy build to order policy and positive cash to cash resulting...
So I drew this...
So there you go...
add comment
| 0 trackbacks
| permalink
| related link
| 
( 2.9 / 285 )
I have been away from my desk quite a lot recently cavorting around the
motorways of England, racking up the miles on my poor hard-worked
steed, but now I have a few minutes to sit down and pass on an
interesting observation....
Anyway, my recent revelation is related to this framework below plucked from the world of transformation consulting and change management as relayed to me some years ago by one of my erstwhile consulting chums. The blobs relate to managing communication with people during significant changes on three dimensions: Rational, Political and Emotional.
The 'sweet spot' is in the centre when all communications are most
compelling as they appeal to all these three.
Coincidentally, whilst trying to be a useful parent and reviewing a Classics essay, I prodded Google about some topic to draw back the veil of my ignorance on such topics and it popped up with Aristotle's three modes of persuasion
Whilst equating Ethos to the Political dimension somewhat turns my stomach when I think of the more venal and self-aggrandising aspects of the political world, the three blobs of the R...P...E model are a pretty good match for what Aristotle laid down.
So there you go....
Just a momentary tangent before we head into the main meat, so to speak, there is another blog post that I have been meaning to write about Broadband Britain, Cloud Computing, the Innovators Dilemma, passing by the new statistic that the number of of old people in the UK now exceeds the number of young, and arriving finally at some as yet unthought pithy comment about Silver [read, Grey] Surfers. However, it is really just an excuse to create a comic juxtaposition alluding to the alleged practice of North American ethnic peoples (no longer Eskimo) to abandon their old folk on ice floes, whereas I have observed over the long miles I have travelled in the last few months that we British seem to abandon them at Cherwell Valley Services on the M40...so lets move on
Anyway, my recent revelation is related to this framework below plucked from the world of transformation consulting and change management as relayed to me some years ago by one of my erstwhile consulting chums. The blobs relate to managing communication with people during significant changes on three dimensions: Rational, Political and Emotional.
Coincidentally, whilst trying to be a useful parent and reviewing a Classics essay, I prodded Google about some topic to draw back the veil of my ignorance on such topics and it popped up with Aristotle's three modes of persuasion
- ήθος - Ethos
- λόγος - Logos
- πάθος - Pathos
Whilst equating Ethos to the Political dimension somewhat turns my stomach when I think of the more venal and self-aggrandising aspects of the political world, the three blobs of the R...P...E model are a pretty good match for what Aristotle laid down.
So there you go....
Recently our esteemed Green Knight, Sir Jonathan Porritt was attributed
with saying
"Overweight
people are 'damaging the planet'". Of course it turns out
that he
said something like this in about 2007, in fact building on a
comment by the then
Secretary of State for Health, Alan Johnson. But somebody
else unearthed it again for some typically
twisted reason - nothing can be more topical than mixing global warming
with a bit of "fatty slapping".
The hypothesis behind the hype is that fat people use more resources because they eat more food, but why not then include teenage boys (unfillable, as empty fridges around the country can testify), people with very high metabolic rate, and other some such big eaters. Ah, well, the logic goes that fat people also drive everywhere and so contribute more CO2 than thin people who, of course, walk or cycle everywhere. Well, maybe it applies in towns, but it is certainly not true in the countryside, so drawing a different intersection in the Venn diagram I am sketching out here in hyperspace, maybe the headline should have read "Teenage boys and country people with very high metabolic rates are 'damaging the planet''" - not quite so catchy, or right-on, eh?
But, of course, there is a secondary thesis which is that obese people can be "cured", especially if they all got out of their cars, walked and cycled, and stopped scarfing all the pies, whence their weight would magically drop away and they would join all the normal people in the happy mean.
When you look at whole populations analytically then of course you usually see some sort of distribution (Normal or otherwise) of whatever factor (weight, in this case) that you might be measuring. So the theory is that by thinning down the fatties, the shape of the distribution will be changed. However, there are flies in this particular ointment, and if you look around you can find suggestions that obesity is actually a structural feature of a/the/any human population, that everybody has got fatter and that you need to treat the population as a whole, not just focus on the upper tail.
All in all, an example of woolly loose thinking gussying up to a political agenda.
BMI is one of the weapons in the "fatty slapping" armoury, a metric with some very well documented short-comings, yet standard (mis-)guidance would label people like Lawrence Dilaglio, Jonah Lomu & Mel Gibson as over-weight or obese. Whilst BMI might have some trivial diagnostic uses, some lard-brained, fat-heads try to use it as a decision-making metric, vide 'Too fat' to donate bone marrow - the 18-stone 5'10" sports teacher with a technical BMI of 36.1 who was ejected from the National Bone Marrow Register. To make a proper health assessment, you need to have a more detailed look at structural features, like waist size, percentage of body fat and so on, before pronouncing.
Thus BMI is a prime example of a benchmark ratio or KPI that is NOT a good basis for making decisions, as it fails to take account of significant structural factors.
This parable provides an important lesson for practitioners in the world of Information Technology Economics, where many a ratio is measured and analysed by pundits including Gartner et al, a classic being "IT Costs as percentage of Revenue", one of their IT Key Metrics.
It is defined quite simply as:
If you dig into the typical drivers of the top and bottom parts of this formula as below, say,
then you might surmise that it is quite possible that the Revenue numerator has significant elements that are certainly outside the direct control of the IT organisation, and indeed outside the control of the company, whereas the IT Costs are defined largely by the structure of the organisation, its distribution channels, and internal policies and practices. The top line is also, I conjecture, more volatile than the denominator, and being mostly outside the control of the IT so a very unfair stick to beat the IT donkey with. So in qualitative logical terms this metric is certainly appears to be a very poor 'apples and oranges' comparator.
If you stretch the analysis further, you can ask the question "what does it mean?" Is the ratio intended to show the importance of IT? or IT leverage/gearing (bang for the buck)?
Well, if it is some level of importance we are trying to assess, then we should analyse the relationship between this benchmark ratio and true measures of business value, such as, Operating Margin. Looking across a range of industries the curve looks like this:
OK, is is a deliberately silly chart, just to make the point that this is clearly a wobbly relationship.
If you do a linear regression analysis of the relationship between Operating Margin% and the IT Cost/Revenue ratio and a sibling ratio "IT Cost as a %age of Total Operating Costs" (or "Systems Intensity" to its friends), then you get these results for R2
What this shows is that there is no particularly significant linear relationship between these two key metrics and Operating Margin, so quantitatively, the ratios do not really tell you anything about how IT costs/investment drive overall business performance at all.
Even within an industry ratio comparisons are fairly meaningless. For example, in the past UK Banks had an average Systems Intensity around 20%. If you were to calculate the Systems Intensity for Egg, the Internet bank, at its height, you would come out with a number ranging from about 17% to 25% depending on how you treat the IT cost component of outsourced product processing and some other structural factors. And I do recall having a conversation with one Investment Bank CIO who declared, "Yes, of course, we do spend 20% of our operating costs on IT, it's how we set the budget!"
The whole averaging process loses information too. Look at the four distributions below, they all have the same mean (i.e., average) but are wildly different in shape.
Without further detail on their parameters than just the mean value of the curves, you cannot make a sensible comparison at all.
So all these ratios give is some rather weak macro illumination of the differing levels of IT spending between industries, like saying to a Bank "Did you know that, on average, Banks spend 7.3 times more on IT than Energy companies" to which the appropriate response is "YEAH, SO WHAT?"...
...Oh, and maybe, some vague diagnostic indication that there may (or may not) be something worth looking at with a more detailed structural review. So, why not just go straight there, and dig out the real gold!
And so the morals of this story, O, Best Beloved, are that just because you can divide two numbers, it doesn't mean that you should, and be prepared to dig into the detail to truly understand how cost and performance could be improved.
Just so.
The hypothesis behind the hype is that fat people use more resources because they eat more food, but why not then include teenage boys (unfillable, as empty fridges around the country can testify), people with very high metabolic rate, and other some such big eaters. Ah, well, the logic goes that fat people also drive everywhere and so contribute more CO2 than thin people who, of course, walk or cycle everywhere. Well, maybe it applies in towns, but it is certainly not true in the countryside, so drawing a different intersection in the Venn diagram I am sketching out here in hyperspace, maybe the headline should have read "Teenage boys and country people with very high metabolic rates are 'damaging the planet''" - not quite so catchy, or right-on, eh?
But, of course, there is a secondary thesis which is that obese people can be "cured", especially if they all got out of their cars, walked and cycled, and stopped scarfing all the pies, whence their weight would magically drop away and they would join all the normal people in the happy mean.
When you look at whole populations analytically then of course you usually see some sort of distribution (Normal or otherwise) of whatever factor (weight, in this case) that you might be measuring. So the theory is that by thinning down the fatties, the shape of the distribution will be changed. However, there are flies in this particular ointment, and if you look around you can find suggestions that obesity is actually a structural feature of a/the/any human population, that everybody has got fatter and that you need to treat the population as a whole, not just focus on the upper tail.
All in all, an example of woolly loose thinking gussying up to a political agenda.
BMI is one of the weapons in the "fatty slapping" armoury, a metric with some very well documented short-comings, yet standard (mis-)guidance would label people like Lawrence Dilaglio, Jonah Lomu & Mel Gibson as over-weight or obese. Whilst BMI might have some trivial diagnostic uses, some lard-brained, fat-heads try to use it as a decision-making metric, vide 'Too fat' to donate bone marrow - the 18-stone 5'10" sports teacher with a technical BMI of 36.1 who was ejected from the National Bone Marrow Register. To make a proper health assessment, you need to have a more detailed look at structural features, like waist size, percentage of body fat and so on, before pronouncing.
Just pausing a moment to dissect BMI further, it has units of kg/m2 which is not unlike the metric used to define paper thickness.Many organisations these days used 80gsm printer paper which is more environmentally friendly than the more sumptuous 100 paper of oldAnd even less rich feeling than the 120gsm paper that Tier 1 consultants use to create a table-thumping report - the dollars are in the loudness of the thump.As Marshall McLuhan told us, the medium is indeed the message, thickness = quality, and just feel that silky china clay high white finish. Oooohhh...
Sorry, started to get rather indented there, must coach self, control tangents...
So a person who has a BMI of, say, yeah, like 25, is like a piece of 25000gsm paper, no really...equally a piece of A4 paper might have a BMI of about 0.08...
Thus BMI is a prime example of a benchmark ratio or KPI that is NOT a good basis for making decisions, as it fails to take account of significant structural factors.
This parable provides an important lesson for practitioners in the world of Information Technology Economics, where many a ratio is measured and analysed by pundits including Gartner et al, a classic being "IT Costs as percentage of Revenue", one of their IT Key Metrics.
It is defined quite simply as:
If you dig into the typical drivers of the top and bottom parts of this formula as below, say,
| MicroEconomic Drivers - Typical Examples | |
| IT Costs | Revenue |
|
|
then you might surmise that it is quite possible that the Revenue numerator has significant elements that are certainly outside the direct control of the IT organisation, and indeed outside the control of the company, whereas the IT Costs are defined largely by the structure of the organisation, its distribution channels, and internal policies and practices. The top line is also, I conjecture, more volatile than the denominator, and being mostly outside the control of the IT so a very unfair stick to beat the IT donkey with. So in qualitative logical terms this metric is certainly appears to be a very poor 'apples and oranges' comparator.
If you stretch the analysis further, you can ask the question "what does it mean?" Is the ratio intended to show the importance of IT? or IT leverage/gearing (bang for the buck)?
Well, if it is some level of importance we are trying to assess, then we should analyse the relationship between this benchmark ratio and true measures of business value, such as, Operating Margin. Looking across a range of industries the curve looks like this:
OK, is is a deliberately silly chart, just to make the point that this is clearly a wobbly relationship.
If you do a linear regression analysis of the relationship between Operating Margin% and the IT Cost/Revenue ratio and a sibling ratio "IT Cost as a %age of Total Operating Costs" (or "Systems Intensity" to its friends), then you get these results for R2
|
|
R2 |
|
IT Costs as %age of Revenue vs Operating Margin% |
0.175 |
|
IT Costs as %age of Op. Costs vs Operating Margin% |
0.330 |
What this shows is that there is no particularly significant linear relationship between these two key metrics and Operating Margin, so quantitatively, the ratios do not really tell you anything about how IT costs/investment drive overall business performance at all.
Even within an industry ratio comparisons are fairly meaningless. For example, in the past UK Banks had an average Systems Intensity around 20%. If you were to calculate the Systems Intensity for Egg, the Internet bank, at its height, you would come out with a number ranging from about 17% to 25% depending on how you treat the IT cost component of outsourced product processing and some other structural factors. And I do recall having a conversation with one Investment Bank CIO who declared, "Yes, of course, we do spend 20% of our operating costs on IT, it's how we set the budget!"
The whole averaging process loses information too. Look at the four distributions below, they all have the same mean (i.e., average) but are wildly different in shape.
Without further detail on their parameters than just the mean value of the curves, you cannot make a sensible comparison at all.
So all these ratios give is some rather weak macro illumination of the differing levels of IT spending between industries, like saying to a Bank "Did you know that, on average, Banks spend 7.3 times more on IT than Energy companies" to which the appropriate response is "YEAH, SO WHAT?"...
...Oh, and maybe, some vague diagnostic indication that there may (or may not) be something worth looking at with a more detailed structural review. So, why not just go straight there, and dig out the real gold!
And so the morals of this story, O, Best Beloved, are that just because you can divide two numbers, it doesn't mean that you should, and be prepared to dig into the detail to truly understand how cost and performance could be improved.
Just so.
Thursday, May 28, 2009, 02:00 AM - Opinion & Humour
So it has been a torrid couple of weeks for MPs outed having
been
caught with their hands in the cookie jar. Schadenfreude, Epicaricacy, aighear
millteach
and their ilk are good words to
roll
around the tongue, and savour whilst we lob cabbages and rotten
tomatoes
at those in the pillory: all the more unattractive being
that their "misfortune" was brought about by their own actions and a
display of lower moral standards than is clearly desirable in
our political representatives.Auto-Epicaricacy: a term I just made up, applying some word logic, would mean taking pleasure in your own misfortune. Definitely an unhealthy and paradoxical mental state, but I suppose optimistic, in that every cloud has a silver lining...
I was particularly fascinated and driven to ask "how does that work, then?" by the declaration of one misadventurer that "Of course I feel that my reputation is tarnished, but my integrity is intact".
Integrity: the unimpaired state of anything : uprightness : honesty : purity - Chambers 20C
What logic system do you have to apply, what set of axioms must one have, how must one deconstruct common sense to be able to make this statement? For a start, you would have to look at redefining some core words: unimpaired, anything, upright, honest, pure - take your pick.
Words are a key part of a consultant's stock-in-trade, and pictures too. One of my favourites satirical sites, now sadly defunct, was SatireWire which ran a series of bizarre and entertaining statistical charts like this...
Madrigals By Freshness
... a hearty lampoon
of opaque and
confusing "Management by Cartoon" Powerpoint presentations
(one
step up, though, from "Management by In-flight Magazine" which is
significantly more dangerous).
And of course every industry has its buzz-words and jargon, which can be useful short-hand for many forms of communication, but often quite poisonous when they leak into other places.
Note the use of the word "key" in the preceding paragraph - a consultant-y sort of word if ever there was one, it means important, significant, stands out from the crowd. Non-key things are not interesting...now go back and carry on reading here
The recent attempt by the Local Government Association to proscribe some logofluvial jargon-words was a valiant attempt to stop etymological pollution in Local Government communication with the rest of us. I am certainly a fan of Plain English, and keeping things short and sweet with some sharp Anglo-Saxon monosyllables replacing loquacious logorrheic verbal peregrinations, but equally a devotee of precision and conciseness which some longer words can bring to a sentence, by conceptual elision, perhaps.
So I was interested to see some words on the list that I have used myself and as have many of my colleagues. These are words from the consulting domain that do have proper surgically precise and correct meanings in the right hands, but indeed deadly in the wrong. Other words on the list would be posionous in any context:
- "Baseline" is a word I know well that has meaning both in project planning and also in procurement - in both areas being the datum from which you measure some sort of progress or achievement.
- "Predictors of Beaconicity", however, is never going to win any prizes for clarity....[there are 1550 hits on Google for this phrase, but I am overwhelmed by disinterest and will not clutter my brain with vacuous garbology]
And talking of words and in an interesting juxtaposition of neurons firing, I noticed that the BBC were having a Poetry Season. Being a self-professed iconoclast and fact-based sort of person, I have a completely tin-ear for poetry which is just a form of "talking funny" (in an unfunny way, unlike puns).
So to finish, I have constructed a Boston grid attempting to make some sense and classify some of the odder behaviours of my fellow human, viz....
| "Talking Funny" |
|
|
|---|---|---|
| Normal |
|
|
| Normal | "Dressing Funny" |
About a year ago, I went though one of those few moments when I thought my normal powers of memory had somehow deserted me. It was not really anything important I couldn't remember, just the word that describes the the lights you see when you squeeze your eyes tight shut. Like this...
.gif)
So not very significant in the scheme of things: not one of the words I actually use very often in conversation or in Powerpoint presentations. Just annoying, because the word was just lurking on the edge of my perception, out of reach. But something that you can get a bit obsessed about when information normally falls to hand or mind quickly...
So I Googled and Wiki'd and all those searching jobs that normally count as work, and kept finding Tom, Nicole and Stanley and their film, and other flotsam and jetsam on the endless waves of Web surf.
But, eventually, I created a mega-whiz, sharp-as-a-scalpel, spot-on search string that gave me that Eureka moment...Ding!
The word I was looking for was "Phosphene"
Mind you the Eureka moment was over quickly, as I came to that odd feeling that I had never known the word at all so how could I have semi-forgotten or demi-remembered it? But let us not confuse the story with such technical plot twists and devices.
I wrote "normal powers of memory" at the top of this piece, though we Jungian Is "enjoy" the physical aspects of memory that are imposed by our brain chemitstry, being the dominant long acetylcholine pathway, compared the the short dopamine pathway of Es out there.
If you looked inside my head, it might look something like this...
...but brighter and probably in colour.
So I worked out many years ago that I should not waste my time remembering stuff, when a notebook works much better.
And so on into the Wonderful World of the Web, I have always found it useful to clip bits out and paste them into my digital scrapbook for longevity and to act as my long-term cyber memory. I gave up on browser Favourites early, as they quickly became useless signposts to where information was no more.
In my Adobe period, I printed bits of the Web to PDF files and stored them in a byzantine filing structure. But, eventually I settled on Onfolio and paid some brass for a real product...and then Microsoft bought it and gave me back my money because they were giving it away free in the Windows Live toolbar...then to become a zombie, twilighting product. The death knell was when they switched off the licensing servers last September.
So I had to indulge in one of those distress-driven searches to find a new digital brain. I tried Ultra-Recall which can import Onfolio collections, but has the user experience of a broken lift. I tried TopicScape but that felt like I was in Castle Wolfenstein or Jurassic Park (the " 'I know this, it's UNIX' whilst looking at a mad graphical computerscape " moment), and a host of other paraphernalia and arcana.
So I have ended up with MacroPool's Web Research, which feels a bit like Onfolio...but German...so hopefully it will be most efficient. We'll see...
So not very significant in the scheme of things: not one of the words I actually use very often in conversation or in Powerpoint presentations. Just annoying, because the word was just lurking on the edge of my perception, out of reach. But something that you can get a bit obsessed about when information normally falls to hand or mind quickly...
So I Googled and Wiki'd and all those searching jobs that normally count as work, and kept finding Tom, Nicole and Stanley and their film, and other flotsam and jetsam on the endless waves of Web surf.
But, eventually, I created a mega-whiz, sharp-as-a-scalpel, spot-on search string that gave me that Eureka moment...Ding!
The word I was looking for was "Phosphene"
Mind you the Eureka moment was over quickly, as I came to that odd feeling that I had never known the word at all so how could I have semi-forgotten or demi-remembered it? But let us not confuse the story with such technical plot twists and devices.
Palimpsest is another word a bit like Phosphene, but in reverse, I know what the letters say, but the meaning slips my mind (a reused bit of parchment, in fact). It is however a word that I have read many times but never ever had the need to write down - until today. It is definitely a clever Stephen Fry sort of a word, or maybe a Will Self word
I wrote "normal powers of memory" at the top of this piece, though we Jungian Is "enjoy" the physical aspects of memory that are imposed by our brain chemitstry, being the dominant long acetylcholine pathway, compared the the short dopamine pathway of Es out there.
If you looked inside my head, it might look something like this...
...but brighter and probably in colour.
So I worked out many years ago that I should not waste my time remembering stuff, when a notebook works much better.
And so on into the Wonderful World of the Web, I have always found it useful to clip bits out and paste them into my digital scrapbook for longevity and to act as my long-term cyber memory. I gave up on browser Favourites early, as they quickly became useless signposts to where information was no more.
In my Adobe period, I printed bits of the Web to PDF files and stored them in a byzantine filing structure. But, eventually I settled on Onfolio and paid some brass for a real product...and then Microsoft bought it and gave me back my money because they were giving it away free in the Windows Live toolbar...then to become a zombie, twilighting product. The death knell was when they switched off the licensing servers last September.
RIP, Onfolio, you served me well
So I had to indulge in one of those distress-driven searches to find a new digital brain. I tried Ultra-Recall which can import Onfolio collections, but has the user experience of a broken lift. I tried TopicScape but that felt like I was in Castle Wolfenstein or Jurassic Park (the " 'I know this, it's UNIX' whilst looking at a mad graphical computerscape " moment), and a host of other paraphernalia and arcana.
So I have ended up with MacroPool's Web Research, which feels a bit like Onfolio...but German...so hopefully it will be most efficient. We'll see...
Next
Avatar
