Where's My Jetpack?

Author: 
John Sheridan - CIO & CISO
Category: 
The Department of Finance Archive

The content on this page and other Finance archive pages is provided to assist research and may contain references to activities or policies that have no current application. See the full archive disclaimer.

I gave this speech at the Australian Government Leadership Network Annual Conference (New South Wales) on 7 September, 2012. The audio quality in this case isn't great so I recommend you read the transcript.

Audio (MP3)

Transcript

No-one one born after 1935 has walked on the moon, the promise that was made to people of my generation memorably in Mr Gould’s 6A class in Lindfield East Public School on an afternoon in July 1969, has not been kept. And indeed it hasn’t been kept not for one, not for two, not for three, but potentially not for four generations, all of whom expect that I think realistically that we would see the promise made that July reflected in our lifetimes, that many of us would walk on other planets, would fly in rocket ships, would shoot above the earth regularly all the time, we’d whip between continents at a moment’s notice, we’d do all sorts of wonderful things.

Now I guess we did get Teflon and Velcro and those sorts of things out of that promise of technology, but famously as Leo McGarry said in that wonderful series West Wing, “Where’s my jetpack, where are my colonies on the moon”.  Today I’m going to talk to you about the failures that mean government sometimes doesn’t get the sort of technology outcomes that it expects, and what you as leaders in government can do to address those things now and in the future.  Not because you’re working in I.T., although some of you maybe, but because you’re in the business of running parts of the government, and that gives you a role in ensuring that you also run parts of I.T.

I’m going to talk about four things; jetpacks, why we didn’t get to ride in jetpacks or on jetpacks or strap them on; the coming C.I.O. apocalypse as I call it; the sort of things that are coming together to make technology even more frightening for those of us in management; and then what we can do about it collectively about governing in government.  Dick Tracey’s famous watch was first mooted in 1946 that was about when those astronauts were nine I think or something like that, no 11.  The precursor of the mobile phone as we all know it today, has been around for a very long time.  The first contract for a jetpack was lead by government in 1959.  Now jetpacks have been produced before that unsuccessfully, and indeed still not many of us commute that way even in Sydney. The robot from Lost In Space, which by the way was set on other planets which we weren’t walking on. 

The robot in Lost In Space was supposed to have been around in 1997, that was the year that the Lost In Space accident was supposed to occur when the story was written and filmed, clearly that didn’t occur.  In 1968 when Arthur C. Clarke’s wonderful story 2001: A Space Odyssey was turned into a movie; it was about 2001, 11 years ago.  And indeed you can see there vaguely the very first cover of Isaac Asimov’s I, Robot series, about the sentient humanoid robots that existed in this future world, were set only about 25 years into the future from now.  Not sorry yeah, only 25 years into the future from now, not very far at all.  While my children will certainly be working, and I hope supporting me on my trips to other planets.  The idea being of course that all this promise hasn’t been achieved.

Now the challenge is of course to say, well what has been achieved in technology in robots in that period of time, and of course the most popular robot in 2012 is a vacuum cleaner.  Now I’m all for someone else doing the vacuuming, as my wife would attest, but it does seem to suggest that maybe we set our short term sights a little high.  And what we have to do is do something about making sure that both our short term goals and our long term goals to use technology, are put in a way that make things work a lot better.

In 2008 the government contracted Sir Peter Gershon to do a review of the use of I.C.T. in government.  Sir Peter’s review had seven recommendations, it looked at a whole range of things, some of which people in government liked, some of which they didn’t and some of which I was in charge of, some of which I wasn’t.  The challenge of those things though covered a range of areas, but one of them was this idea that sometimes I.T. projects didn’t work.  So Sir Peter’s survey looked at 193 I.T. projects, and he asked the agencies making them to self report on those projects.  Now I think all of us know in our heart of hearts, that maybe when we self report we don’t take it quite as hard on ourselves as an external auditor would.

And nevertheless about 23%, a quarter of those projects were self reported as running over budget.  Another quarter of the projects didn’t provide, or their agencies didn’t provide information about their budget status.  Thirty three per cent, a third of those projects reported running late, again about a quarter didn’t report how they ran according to schedule.  So what we can see is that there’s clearly an issue in achieving the sorts of promises that we want out of technology.  It’s interesting to say, well we’ve been doing this for a while now, what is it that we aren’t getting right?

The Victorian Ombudsman did a study of ten major I.T. projects across the Victorian Government, to see why all of them had failed in one way or another.  Now what I’ve done is taken his report and I’ve pulled out the things that aren’t specific to the Victorian Government.  Now I’m not picking on the Victorian Government’s I.T. staff in this regard as I go through these things not at all, these are problems that are common in I.T. projects I think across the commonwealth and indeed in I.T. projects across industry as well.  Let’s dwell for a moment and just see what they are.

Firstly the difficulty of getting people to take responsibility for the project, this is a really interesting challenge.  You would think as an I.T. deliverer like I am, that I’d be able to find someone in business who wanted a project, who could take responsibility for things, who would be the lead, who would be happy to sit at the table and tell us what to do in business sense, I don’t mean which wires to solder or anything else like that, but to tell us what they want in a business sense.  But actually it’s really hard to get that, it’s really hard to set up governance decisions about projects, so that you know they’re working the right way.  It’s amazing how many times senior officers have been reluctant to make critical decisions in projects.

Indeed whenever I see a committee set up to run a project, I almost the sort of hair raises on the back of my neck, because what I can see is that too often this is a mechanism for not taking decisions, or alternatively for sharing the blame at the subsequent board of enquiry.  It’s the challenge about what it is we do to make sure that people can make those decisions, but it’s important to recognise that it’s not that we just sort of get a whole bunch of new people.  But rather we should make sure the circumstances are in place, so that they’re willing to take those decisions and make them.

Now as a sort of opposite of that, one of the big problems of project management in I.T. is a bias for optimism, this is the notion that says, someone will look carefully at a program schedule and they’ll say, well OK, it’ll take us 18 months to deliver this project in the best case, and in the worst case it’ll take two and a half years.  Now what they do in those circumstances or at least what happens when that report is put forward, do people say, well OK, let’s assume that it’s a two year project plus or minus six months, which would be an average sort of view.  Now the project deliverer in me says no, it’s a two and a half year project that I might finish early if you’re lucky, but what actually gets reported is that it’s an 18 month project, because the expectation is that everything will go well. 

Interestingly in the face of all evidence to the contrary, you see this happening time and time again.  I ask people and I’ve had this experience in a range of areas, I ask them what sort of contingency do you have in the project, and they’ll say, oh I think we’ve got about 10%.  And I said, well what normally happens, oh well often we run about 30% late or something, well let’s have 30% contingency, or indeed the big trap for large departments is when the project ceiling that an executive can decide is as it was in Defence when I worked there $5 million, and people would bring a project to me for approval at $4.9 million. 

And at $5 million it had to go to the Minister, and I would say, what’s the contingency in this project, and they would say, oh its 10%, and I said, well that’s interesting ‘cause usually we have 30%.  But if we have 30% it’ll have to go to the Minister, I said yes, it will.  But if we don’t have 30% if we don’t do that, if we sign it at ten and I sign it and it runs late, then I’ll have to go to the Minister and explain why I didn’t take it to him in the first place, which is a much more difficult conversation to have about project management.  And this is again this challenge that it’ll be right, we’ll get through, I.T. projects don’t get through, most projects indeed run late in those circumstances, not because the original work wasn’t good but because of this optimism bias in deciding what might happen.

Business cases, if you’ve got a good project it should start with a business case, and the business case should say, we’re going to do this project for these reasons, it’s going to have these benefits.  Now you can’t just write that at the beginning of the project, particularly a decent sized project and file that and say OK, that’ll be good, we’ve got it started, we’ve got the money now, let’s get stuck in, that’s not what can occur.  Indeed though it is what occurs, what you need to do is go back in these projects at every stage and see how they’re performing against the business case, as we’ll talk about later structure the projects, so there are milestones so you can stop if something’s going wrong. 

Would you continue on your holiday if you ran out of money after the first two stops hoping that somehow more money would arrive? No I don’t think so, but that’s the mechanism that we have without referring back to those business cases.  But jumping through a few inaccurate costing is a challenge, the notion of being able to work out exactly what something costs.  Now regularly at my place we watch... well we’re not watching shows about robots, which I’d rather do, we watch a show on Foxtel about lifestyle. 

Now I know more about building houses in the U.K. than anyone needs to know, particularly when I have no intention of ever living there or indeed ever building a house like that.  But the interesting challenge that you see on that, is the question at the end, how much did you expect to save, how much did you expect to spend, 500,000 pounds, god knows where they get that money, and how much did you spend at the end, oh I think about 900,000 pounds or something like that.  I got the house building project, I got that about 50%, about 100% wrong in price.  Now relate that to the sorts of things that we see that we’re not familiar with, technology things, new things, things that might be on the edge, and think what the risk about costing those is, it’s a big challenge to get right. 

Over customisation, this is the challenge of saying, I’m going to buy something off the shelf and I’m going to use it and that’s going to be good, and yet I’d really like to twig just this little bit here, and it would be better if it just did those things there and it had that.  Imagine if you bought cars like that, imagine if you rocked up to the Ford place and said, I’d really like to buy a Falcon, but I’d like it to have Commodore wheels, a Volkswagen engine, that boot from the Toyota I saw downstairs, and the ski racks on the Porsche that just went past.  Is it any wonder that such a car wouldn’t be built properly, that wouldn’t run properly, indeed bits of it might not even go together, and yet that’s the sort of thing that we see happening in I.T. projects regularly.

Now there’s also an interesting challenge about the way that we recruit public servants.  I’m a really great fan of the public service, I’ve worked in for the government for more than 30 years, and you couldn’t get a stronger fan of the public service than me, and yet I sit in these meetings and what I see is public servants are lateral thinkers.  They think how could this affect me, how could this be expanded to do my bit of work as well.  So we’ll start with a project that looks at doing a particular thing.  And someone will say, gee it would be good if we did that for this thing too, we could just expand it to do that, we could just expand it to do that, and then we could use it in that division too, maybe we’ll need to speak about that branch to that branch as well, and those guys in that other department they’re looking as well. 

And next thing you know you’ve got a project four times the size of the one that you started to do, something that was achievable, something that was little and could be done, is now an enormous cross-departmental project, it needs interdepartmental committees to run, you need to attract funding through new budget proposals, all sorts of things have changed.  What we need to do in these areas is concentrate on one thing and build it quickly and build it well, and I’ll talk to you in a moment about how you as part of the business can make sure I.T. does that.  There’s a challenge of course of getting inexperienced staff to work in this area.

Now I actually think one of the core skills of being a public servant is to be a project manager, as project management discipline doesn’t need to just be used in technical fields, it can be used in a whole range of delivery of papers, of briefs, a whole range of things can be delivered with a solid basis in project management.  Indeed next time you’re going to come to a conference, if you’ve got a couple of days don’t, go to a project management course, learn about the foundations of project management, because I think you’ll find... present company excluded of course, that in most circumstances, two or three days spent doing that, will be far better than two or three days listening to old fat guys tell you about how to do things better.

The other challenge I think in running these projects is the reticence that people have about saying where they’re going when things are going wrong.  Now there’s a mechanism in project management about measuring the tolerance of the project, how far can things go out before I actually say something’s wrong.  Now the idea is in good project management that if you know it’s going to go out of tolerance, that’s when you’re out of tolerance.  So you report it immediately, you do something about it, you think of a new plan, you work out how you’re going to adjust it.  But actually that’s not what happens regularly.

What happens regularly is people don’t admit that they’re in that situation, they don’t tell their bosses.  Now this isn’t unusual behaviour, I rarely tell my wife when we’re lost, I wait till it’s absolutely obvious, and indeed the G.P.S. is going off and saying no, you should do a U turn at that next mechanism or something like that.  But the challenge though is to not apply that behaviour at work, to recognise that the time to escalate that project is when it is out of tolerance.  What you want to do is not surprise your boss with the project was going really well last week, but you wouldn’t believe what happened on Monday, that’s not the notion.  The idea is to go out, make sure you can see where the project is, make sure you understand it, and then do something about it.

Now the next thing I’m going to talk about is what I describe as the four horsemen of the C.I.O. apocalypse.  This is the challenge that’s coming to address I.T. in government, but indeed I.T. across the world at the moment.  And it’s got four things that create problems for it, and I’m going to go through each of those in turn.  Hype, social media, some of which you’ve heard about in the last couple of days, mobility and the cloud.  This is the Gartner hype cycle, Gartner’s a big I.T. research company, does a lot of work across the world about I.T.  What they identified was the way that things are reported, now you can’t necessarily I suspect see all of that text, but when a new idea comes around it gets hyped up, you start reading about it everywhere.  If you’re in I.T. your boss comes back from the Qantas Club and says, I just read about this, how come we haven’t got one.  The notion that you can say, well actually ‘cause it’s imaginary, doesn’t necessarily wash very well with them, you know, just stick to the sports pages boss. 

The challenge around this is getting, because people like to write, journalists like to write about things that are new.  They like to write about interesting stories; they like to write about things that might happen soon.  We over estimate how fast technology can change things in the short term.  Now the next interesting thing that happens is that what journalists really like to write about is when things go wrong, and when things go wrong you fall off that peak of expectations into the trough of disillusionment.  Things don’t get delivered when you say they’re going to be delivered, your projects start to run late, it turns out that the vendors in Australia don’t have that technology yet, they don’t have staff ready to do that, it’s hard to buy those things that you might want. 

We see this happen time and time again, and all of this means though that by the time this has occurred a lot of attention of bosses and leaders and other people has moved on, ‘cause they fly a lot, there are new magazines, new things to bring back to talk about.  And it’s this challenge of how do we get round delivering what it is when those things.  Now things have to go wrong in order to get things right.  One of the really interesting discussions I often have with people outside the public service is they say, the public service doesn’t have enough innovation, we need lots more innovation in the public service. Good I say, good.  Could you just tell me isn’t it true that one of the things you need in order to innovate is to allow things to go wrong.  Yes, yes that’s right, things have to go wrong in order to, you know, that’s good.

Now you’re a taxpayer aren’t you, yes they say.  Could you just tell me what percentage of your taxes you’d like us to devote to failure? ‘Cause if I knew that it would be very easy for me to plan how much innovation I could do.  And of course the answer is, we can’t do that realistically.  What we need to do in government in I.T., but in a range of other technical things, what we need to do is be the first to be second, not the first to be first.  So there’s going to be a lag of expectations as we build these technology projects, it’s going to take time to get things right, and we have to pull up to what is described there as the plateau of productivity over time.

Now my one for those of you with a technical bit, I won’t bitch about Gartner, in this regard is they don’t show the cliff of legacy which occurs just after the plateau of productivity, where I.T. has to keep things going when the rest of the world has moved on.  And if you don’t make a wrong move you just fall into the pit forever.  There’s a bit of a challenge with that, but this is one of the issues we have to deal with.

Now, as I suspect, you’ve heard a fair bit about social media recently, these are the top ten social media sites in Australia measured in August 2012. 11.5 million Australians use Facebook.  Now one of the interesting things when you’ve been married as long as I have, almost 30 years, my wife and I sit in the study with our backs to each other ‘cause that’s the way the architect designed it, and I work away on my computer on very important government projects with my two finger typing speed.  And behind me I can hear my wife typing, ‘cause if there’s one thing they teach you at Port Augusta High if you’re a girl, or they did in the 60s, it’s how to type. 

And my wife can type up a storm, and I know she’s just sending rubbish on Facebook, I know that’s all that’s going on, and it’s incredibly frustrating.  Because one of the interesting things though is social media isn’t about young people, the biggest growth in social media is the people sort of women 35 through to 55, they’re the ones who are using social media an awful lot, 11.5 million Australians, half the population on Facebook.  What it means is that this is, as you’ll see, a sort of almost sneaky creep up on technology, that somehow C.I.O.’s weren’t expecting. 11 million people use YouTube.  This is unique Australian visitors, and it’s a useful thing that it is, because if it actually counted my three sons visiting YouTube, there’d be so many things that I wouldn’t be able to get them on the graph.  But there’s a challenge though in what it is that we’re doing around this, how we’re using that technology to good effect.

LinkedIn, 2.2 million Australians on LinkedIn.  I don’t know if you’ve used LinkedIn, it’s a site that people put their C.V.’s on, essentially they comment on each other’s C.V.’s, they add qualifications to each other, they build themselves up, and conference organisers ring you up and ask you if you want to talk at conferences.  That’s the main use of it that I’ve seen.  But the thing is that people are increasingly comfortable with using these things.  And as I said, it’s not just young people, I really think one of the problems with technology is we make a mistake about thinking we should manage people on the basis of their age, and I think that’s a really interesting error that we continually make.  There’s a lot of people who can use technology in clever ways, who aren’t in their 20s.

This is the hashtag from Twitter, the gov2au hash tag.  Now the reason I bring that up is you would have heard in the last couple of days, I’ve said a lot about the use of social media in government.  The gov2au hashtag is a Twitter hash tag that grew out of the work of gov2 that Martin and Lindsay Tanner were talking about yesterday.  I actually started this archive online in August 2010, so it’s just over two years old, in that time there’ve been about 24,000 tweets on Twitter about gov2au.  If you’re looking at the top users, because that shows who’s using Twitter.  The interesting thing about this new technology is that first big slice of the pie that you see, Simon A. Roberts, that’s actually a robot, a bot if you like that re-tweets everything that’s with the gov2au hash tag.

So a quarter of the tweets on that hash tag have no new value, they’ve been tweeted by something – it’s a piece of technology - that hasn’t even read them, just automatically copied them.  The next green slice is Craig Tomler, The next green slice after that is Craig Tomler with capital letters, still Craig, just changed the way he recorded his name for some reason.  But he’s responsible, he’s a consultant now, he used to work in Agriculture and Fisheries or Regional Australia, he’s a consultant now he works in I.T., the grey bit’s me.  So we’ve got half of those tweets on this stream right, are actually two people doing these things. 

Now I’ve got to tell you that I don’t sit round all day at work tweeting, I’ve got too much else to do.  But what you can see is again there’s a bit of hype about how much is going on here.  Now the good news  is that there are 42 active government blogs at the moment, 130 active Twitter accounts in government, just recently the chief of the Defence Force became the fourth senior executive in the federal area to become an active Twitter user, publicly Twitter user.  I’m a bit annoyed because in about a month he’s got more followers than I have, but nevertheless he’s out there doing things.  There are 81 Facebook active pages in government and 69 when I counted them this morning YouTube channels being used by the federal government.  So we’re out there using that technology, the interesting thing is it’s not I.T. departments who are doing this, the business is doing this, because it’s easy for the business to do those things.  I’ll come back to that.

Mobile phones, 11 million mobile phones or mobile accounts across Australia in December 2011, five terabytes of data, that’s about 23 million state libraries worth, no it’s not really, but it’s an awful lot of data, I think it’s about six episodes of Underbelly, going through online in mobile phones all the time, 2.6 million tablets being used in Australia at the moment, 58% of Australian’s have used their mobile phone to access the internet.  Indeed 95% of the population has access to the internet.  All of a sudden you can get I.T. wherever you are, wherever you are, with, I have to say, some very small exceptions in terms of coverage.  An interesting exception still in terms of the digital divide, and I think it’s useful for those of us in service delivery areas to remember that that digital divide still exists for many of our customers.  But the ubiquity of mobile I.T. is a really important trend.

Finally the cloud, I bet that many of you have used the cloud even today, you probably if you’ve got an iTunes account on your iPhone, you’ve probably clicked the iMatch thing that allowed you to store your music up into the cloud, so you can get it from any device anywhere, didn’t cost very much, probably the cost of five cups of coffee a year or something like that.  A good idea there to do that you’re using the cloud.  I wrote my presentation when I started it a couple of weeks ago, and when I finished it in the last two days, I stored it in Dropbox, another cloud related thing, it’s got nothing classified in it, it’s not official government information in that sense.  I stored it in Dropbox so I can write it on a computer at work, I can use it on a computer at home, I can view it on my iPad when I’m outside rehearsing just before we start.  That ubiquity of the cloud, means you can get to the computing anywhere. 

But also for those things that aren’t classified information, there’s an enormous benefit in using the cloud, because the costs of capital, the costs of building large data centres are amortised across the world.  So data.gov.au which is a repository of information sets that government publishes as a consequence of the declaration of open government, is hosted by us on the Amazon cloud, the front end is in the data centre here in Sydney run by an outsourcer that, you know, the historical sort of outsourcer that we use, but the data is held in the public cloud.  And you would of read about a range of those things.

This now means because those clouds accounts are so easy to set up, that someone in business can set one of those things up themselves, eight cents a server an hour I think is what Amazon costs to get their cloud thing.  That means someone in one of your business areas, one of you perhaps, can ring up with the corporate credit card, set up an account and start running things on an I.T. platform, you actually don’t know where the I.T. platform is, it might be in Sydney, it might be in Seattle, it might be in Hong Kong.  Someone can run an I.T. application in the cloud provided by a provider that never went near your I.T. department.  This really is a change, because once upon a time again if you’re as old as I am, I can remember at uni taking my computer program on punch cards to the guys in the white coats, and they would run it on the big computer, and I’d get the output later on. 

We’ve moved so far away from that, that it isn’t funny, that now as I said, anyone in business can set up a really solid bit of I.T. work without ever going near the I.T. department.  What have we done then, we’ve created what I’d see as this nightmare for C.I.O.’s, people can do computing anywhere, they’re not tied to the agencies facilities, anybody can do this, not just I.T. people but anybody at all can access this huge amount of computing power, and they can do anything with it, because almost all computing services now can be provided externally to your I.T. department.

Now I know for some of you who’ve probably got problems with your I.T. departments, poor maligned people that they are, are probably thinking well this sounds pretty good, but there are challenges that I don’t think sometimes we appreciate, and these are why it’s a business nightmare as well as a nightmare for C.I.O.’s, the business doesn’t know anymore whose doing I.T. things.  Many of you have probably experienced the problem of having a range of Excel spreadsheets used to do accounting, or to do projects or to do something like that, and been forced to try and track after somebody has left or done something else, how they actually set things up.  Imagine if they’ve left it in the cloud somewhere in Singapore, how are you ever going to find that if you don’t have the password, or if you don’t even know it exists. 

Where is that corporate data going to be in these circumstances? Is it going to be secure, is it going to be safe? You as the business leaders are going to be responsible for that in these circumstances.  What on earth are they doing is another interesting question, that of course you don’t really have any control of in these circumstances.  When we first did the gov2 work that Lindsay spoke about yesterday, one of the big arguments that we used to have, and we had it for about a year and a half when we were doing those A.P.S.C. principles about the use of social media, was the notion as to whether or not people would be allowed to use their computers at work to do social media stuff.  And the progressive departments like Finance, the department I’m in, decided we could do that early in the piece, other departments still can’t do that.

But the interesting thing is that it’s now completely irrelevant, because everyone’s got enough computer power on their Smartphone to do that, but at their desk whenever they like, it no longer needs to be controlled by the department, indeed it’s a management problem now because managers have to be sure that your staff are actually doing their work and not tweeting about you or sending comments about you to Facebook pages.  This control thing has moved out of I.T.’s hands, how much is this going to cost, if they’re buying it on the credit card at, you know, $4,999 a time, and they do it time after time after time, when does it become a problem.  If you’re storing information like this in the cloud, and you’ve got a responsibility to keep it for seven years, you’ve got this interesting challenge.

Leasing a car’s a really good idea if every three years you throw away the old car and get a new car, but because we can’t throw government information away every three years, we have to keep it for seven years at least, military personal records for 130 years, we have to keep the old cars too.  So if people are putting it into the cloud, they are leasing information, what they’ve got is this garage out the back of their house full of old cars that they’re still paying for, because the lease model doesn’t work if you can’t get rid of the old car.  And in terms of government data, we can’t throw most of that away, so we’ve got to manage that.

Again if you’ve got different areas of your organisation doing this work, how do you know they’re not doing the same work, how do you know you’re not paying them to do the same things, how do they know what each other is doing if there’s no control of those things, are those things ever going to work together in those circumstances.  If I put all my stuff in the Google cloud and you put all your stuff in the Amazon cloud, at the moment there are no standards between those things, nothing can make them work together, what is it that we’re going to do.  The good bit about the C.I.O.’s nightmare, is that it’s not just ours, it’s also a nightmare for the business and we need to do something about it.

Now interestingly Australia has taken quite a leading position in this work, earlier on about the middle of the last decade we wrote the Australian Standard 8015, which was about the corporate governance of I.T.  And that standard swiftly became in 2008, an international standard I.S.O./I.E.C. 38500:2008, about the corporate governance of I.T.  And what it talks about it’s not about managing I.T., this isn’t a screwdriver manual, it’s not a user manual for a printer or anything else like that, it’s about how business controls I.T., about how you as leaders in the federal government, you as the head of state government in the public service generally, you as leaders in business need to understand what I.T. is doing. 

Essentially it gives you three tasks. Firstly you need to evaluate what’s going on, you need to look at the current and future use of I.T. being proposed to you by the I.T. staff, and see that it aligns with the business direction that your business is going in.  You need to direct I.T., you need as leaders to assign responsibility for I.T. and for I.T. projects to particular people, and ensure they’re accountable for those things.  You need to make sure that those people are making plans and developing policies so that you know that the I.T. situation is in control.  You need to monitor what’s going on, you need to measure performance.

Now what’s going to happen when you start to do that when you’re in a position to do so, is your C.I.O.’s going to come to you with a huge bunch of tables and graphs and things like that, and say, look there the outages level here is down a lot on what it was before,  in the meantime between stop that.  Talk to you about the five things he thinks or she thinks are most important, and ask them to show you how those things are changing over time, make sure they’re explained in a way that you can understand those, ‘cause none of you are silly, none of you don’t actually understand a whole bunch of technology now. 

You don’t see in homes in Australia now the blinking clocks on video players that we used to see 20 years ago.  We know how to run that level of technology, and you know how to make sure that people explain to you what are essentially business functions in a way that you can understand them.  Don’t be put off by the fact that there are 30, ask them to come back in half an hour with the top five, and develop the looking at those things over time.  What you need to do then is measure that performance against the plans you were promised in the direct part of the discussion, so that you ensure that it can be demonstrated that what was promised is what was happening.

You need, when setting up an I.T. project, you need to work with the C.I.O. to make sure that you’re getting the sort of project that you want.  Firstly you need to make sure that you’re addressing the right problem.  Now a whole bunch of these steps need to take place before anybody buys any computers, anybody cuts any codes, anybody downloads any software.  Indeed most of this work is all about the things that you would do before you get anywhere near those things.  Make sure that you’ve got the right objectives that you’ve sat down and said yes, this is what I want this project to do.  We’re not talking about any technology now, make sure that you’ve seen what the realistic alternatives to doing it various ways are, have them explain to you. 

I’d point out to you as I do to my staff, that if you come to me with no options, if you say to me, we can only do this one way, I say, well why am I in the discussion, if there’s only one way to do it, you decide it right.  You need to be presented with options, so you understand them. What are the alternatives?- because there are always different ways to do I.T. things.  Make sure that you understand what are the consequences of each of those alternatives, that you understand if I go down this bit, I won’t be able to do these things but I will be able to do those things, they’re more important to me because they meet my objectives, they’re the more important things to do.

Balance the tradeoffs  that occur in those discussions, these are the sorts of things most of us do every day in decisions that we make, and there’s no reason that they can’t be explained in that way to you when you’re making decisions about what I.T. projects to undertake.  Having done all that, having seen what you think is the recommended way ahead, make the decision and then stick to the decision, don’t let people add other bits of scope to it.  When someone comes to you and says, gee it’d be really good if we expanded that into another department, say yes, I think it would be and that’ll be in phase two, it won’t be in phase one, we’re not stopping what we’re doing now.

Make sure that when you’re implementing these plans you do look back at those milestone stages and see that they meet what they were promised to do and that they deliver what was promised.  And if they don’t stop the project, stop the project.  It’s better to fail early and cheaply, than it is to drag something on forever if it’s not going to work.  The best I think the best I.T. projects are short I.T. projects that deliver regularly, small short projects that each time they make a delivery it’s something that you want, and I think that’s the way to manage that.  That I guess summarises how I think we would go about getting jetpacks in the future.  I still don’t expect to ride in one of those jetpacks, frankly because I don’t think the lift would match the load in my particular situation, but I’m happy for others of you to try once we get this sort of good control of I.T. by leaders, by you into the future.  Thanks very much.

 

Comments on this blog are now closed. Please let us know if you would like to discuss this post or have any general comments.

Comments (7)

That was nice.

It's so true isn't it? We overestimate what technology might be used for in the short term and underestimate it's acceptance in the long term. The best example i know of is the greek who invented steam power 2,000 years before the industrial revolution. But hey, who needs a Hero when you've got slaves.

Thanks Nick, I've just downloaded that book and am looking forward to checking it out.

I'm not sure about the correlation between large projects and failure. I do think there is a correlation between the time between milestones and failure: the longer between deliveries, the greater the likelihood of divergence from the plan. I think regular deliveries to customers of something they value keeps them happy and the project on track. It also means that if cancellation occurs, some useful things have been delivered.

Cheers

John

John raises a whole bunch of interesting issues. No Baby Boomers have walked on the moon. What an underachieving bunch.

I’ve been IT in Federal Government in all sorts of roles since the XTs were old school and 80286 was the hottest thing in town. Mainframes ruled. The rapid spread of PC’s into business (rapid was slower back then) had Central Data Processing management very concerned about how costs were to be understood and controlled and how data was going to be protected if everyone was going to have their own PC and business areas could just buy them.

The tension between control and flexibility is not new.

Control was achieved because CDP areas evolved into IT Branches and provided networked PC’s, share drives, backed up the data and supported the tools. It became easier and cheaper for business to let IT do the job. Let this be so with Cloud. A WoG approach could be implemented where AGIMO provisions Cloud services for the Federal Public Service. It could and should be outsourced – AGIMO and lead agencies set the requirements and service levels and then provide the management interface for the FGCloud. If we build it they will come and data can be protected and costs contained.

John’s other point about project management and control is really the same tension – between being responsive to business needs and the time and cost of governance. My experience is that formal Governance process can absorb 30% to 50% of a small projects budget and the approval process is glacial. My business clients despair. IT Governance process is absolutely necessary but it is very poorly designed. Sadly ironic because systems analysis and design is a core disciple in IT and we would never design a systems that required our clients to enter data in three systems, reformat in word for approval, transfer it to a PowerPoint presentation for the next step in the business process and then have 3 manual workflow processes that took two months before the transition was committed. In one Agency I work in I had to record key dates for my three projects in 13 different places, systems and documents (Steering Committees, reports to AGIMO, Project Boards, Change Control Meetings, Program Office reports, PMPs, Planning Committees, ITGC’s and Executive Board Reports etc etc).This was my task most of every Friday. Seriously.

Quality business process and IT design is needed to streamline and reuse. Apply our professional skills to our own business. Make Governance easier, add some value. One example; standard estimation models, with real cost data for compute, storage, network, licences etc. . Have these generate standard project schedule WBS with work effort based on previous projects. Gateway could use the same schedules and cost models as the project actually uses to manage the project and report to their management. One fact in one place.

Put this in the FG Cloud. Better result, standard approach, better control, WoG view, easier for PMs, faster, cheaper. Just like networked PC’s back when Money for Nothing was a hit and not a consultant.

One last comment; one of the reasons that IT projects are seen to fail so often is that the cost and time are decided and agreed by people who are not going to deliver it and the solution design is unknown. Requirements cannot be accurately costed – only design solutions.

And I’m very worried that by the time I get my jetpack I won’t remember why I wanted one.

Hi John,

Another good article, thanks for sharing.

I think one of the many reasons for scope creep is the lack of initial thought when the project is first put together. It's a couple of people who write the first cut of the business case. It gets sent around to their peers and too many people comment on format or typos rather than adding future value to it. Then, when the solution is being built, these same people see the gaps and suggest them. They are good ideas but cannot be introduced because funding wasn't sought or the timeline doesn't allow it.

Cheers,
Neil

A fascinating talk, full of astute observations.

My comment refers to John's references to business people not taking responsibility for decisions and projects run by committee.

Anyone working in government has seen this time and time again. It seems to be a risk averse strategy where the responsibility is shared by many and accepted by none. The business doesn't know enough, even about what it wants, to feel comfortable taking responsibility and the project people don't want to make business decisions.

I'm not sure that the answer is to encourage, or even force these two groups to make decisions, they are usually not equipped to do so.

IMHO, what is missing is technical leadership. Someone should be appointed who will act as the bridge between the business and the project. That person would work with both groups and ensure that both sides of the equation - the business problem and the technical solution - match appropriately.

In engineering projects that would be the Chief Engineer. In IT projects there isn't even a commonly accepted title. Chief Architect might cover it but architecture has been devalued to little more than design or standards and frameworks.

My view, as an IT architect, is that Project Managers should stick to administering projects, steering committees should endorse decisions, not make them, and the responsibility for delivering a solution to a business problem should lie with someone who is fully trained and capable of leading such projects.

Or to use a phrase I've seen around in AGIMO offices - "one throat to choke"

And just to finish up, that person and their throat should be appointed to run the business case, which they should "own", and should see the whole thing through from start to finish. That would also go a long way to mitigating one of John's other points about poor project management. As I said, project managers should administer projects, not make decisions about the deliverable or how it is being implemented.

John,

Thanks for posting the transcript of this speech.
I enjoyed the refreshing absence of guarded and self-serving vapour.
You might be interested in this book by a bunch of project manager grey-beards if you haven't seen it already: Adrenalin Junkies to Template Zombies: http://goo.gl/UZ2rA
It has names for all the patterns of project behaviour you discuss:
"The sun will come out tomorrow"
"Telling the truth slowly"
"Dead fish"
"Feature soup"
"One throat to choke"
"News improvement"

A question: what do you think is the correlation between project size and project failure?

Thanks for sharing this John.
Provides some good ideas for topics at this year's Canberra GovCamp
Look forward to seeing you there again.

David Williams

Last updated: 05 August 2016