I retired from personal blogging in July 2008.
But you can find me over at http://blog.xero.com.
The Ministry of Social Development is buying a client Management System for $54m dollars
MSD buys $54 million client management system
I would have thought that Government would have learned the lesson on big software projects before. Nothing costs $54m dollars. We can fly to the moon for $54m dollars.
At a developer charge out rate of $150 per hour and working 2000 hours per year that is 180 man years of software.
Software does not need to be that big.
Get 5-10 people on it for a year. Done.
How about taking the 10% contingency (5.4m) and letting one of our outstanding local development companies like Intergen have a go in parallel and see who implements first. We might even create a product that we can export.
Government procurement can be used to create an industry.
$54 million. Nuts.
blog
With the greatest of respect Rod I think you’re being a little bit simplistic here, the $54m is the total project cost not just the software, presumably it includes purchasing hardware to run it on and things like data conversion to the new system which would be a massive task.
Whilst I’m sure the people at Intergen are very clever, I don’t think they have ever attempted anything on this scale before. Writing a system which is essentially the largest most complicated payroll in the country is a bit different to developing web sites, intranets and reporting systems.
However, I agree with you about the gummint and big projects, I’ve always thought small incremental improvements are the way to go, big bang projects very rarely meet their objectives.
I also like the idea of Government procurement being used to create a local industry esp. if a product is created which can be exported.
Lets see the budget breakdown. Don’t they already have PC’s? How many servers do they need?
Well hello. Hello!
Starsoft will do it for 10%. Simplistic has got nothing to do with it. This is just another case of the taxpayer over paying on big IT projects.
50% of large IT projects over $US1 million never see the light of day. Over 50% of all large projects worldwide whether tunnels, airports, roads run over budget and over time.
If you can project manage a reasonably large project, let’s say $1 million or above, you can sure as heck do this one.
I agree with Rod, nothing costs this much infrastructure included; even nationwide communications included.
The problem is the people who make the decision on these projects don’t know their elbow from the proverbial.
hi rod
How ironic - $54 million sounds an awful lot of money when UNESCO has just come out with a confronting report on the state of children in this country - next minute the agency dealing with dysfunction, violence and child abuse is spending a lot of money on tracking and tracing and documentation which of course is important in the greater scheme of things…but is it, as rod suggests, the best use of public money when there could be alternatives?
But I think one of the problems in govt. procurement of large IT projects is that many of the senior managers in govt overall don’t have a deep understanding of the intricacies of an IT system - or the possibilities of doing things with their intranet and interface with their user base in new and innovative ways.
The IT managers know these intricacies, but sometimes the detail is lost in the way that the complexities are or are not communicated to senior management. Thee is also very little consultation with internal staff about how to actually do things differently or better with a new system.
Some govt depts even recommission failed IT projects to the same company I’ve heard so is this ‘the devil you know’ at work?
I wonder if its just the mythology of IT - keep it mystical and keep charging huge amounts to those that ‘know not what they are doing’…….
clare
Rod - I’d love to see the bugdet, can we get hold of it under the OIA or would it be considered “commerically sensitive”
Dermott - I’m sure Starsoft would like the $5.4m contract to have a go but without knowing the funcitonality of the system how do you actually know you could do it for that ?
Have you guys ever worked on a really large mission critical database application systems and know what it costs to run them alone for a year, let alone replace them ?
I’m talking systems with hundreds of thousands of customers and handling transactions worth hundreds of millions of dollars ? There are actually very few systems like this in NZ, the MSD would be one of them.
MSD handles a budget of something like $13 billion and has over a million “customers”.
They have something like 8,000 staff, training each one of them alone in the new system could wipe out your $5.4 million ($675 / head).
I liked this line from Curam’s website:
“The Curam Social Enterprise Management (SEM) solution is a comprehensive ****commercial-off-the-shelf (COTS)**** product designed specifically for human services…..”
$54m for an off the shelf product!! As an IT professional, and as a taxpayer, I’d love to see the breakdown of license fee, hardware, implementation.
These systems are hard, without a doubt. But $54m!!!
“COTS” - yeah, right.
Dan - “COTS” - I think Dick Smith’s Newmarket have a copy in stock. ;-)
Glenn, yes Starsoft have worked on mission critical system. No we have not developed systems this expensive (unfortunately). Whether we have developed systems this complex is debatable as you are saying Cost = Complex. And this does not necessarily follow.
But I do know what is involved in developing and hosting here at Starsoft web based financial systems that have to work and be available to a high degree of availability. And I understand the infrastructure that this type of system has to run on.
And as Rod said, don’t they have servers and PC’s? Even if they don’t, unless they are buying the stuff at 3 times retail, it’s still a lot of money.
A IT system could cost $50,000 and be more complex than this. Sure it’s large but that does not necessarily mean complex or difficult.
I have also been underwhelmed by large IT vendors in NZ and elsewhere who time and again, despite screw up after screw up keep getting these type of jobs.
Case in point the UK National Health Service IT disaster. See below.
The £6.2bn National Programme for IT will henceforward be known as the £12.4bn National Programme for IT, after a long-awaited National Audit Office report into the ambitious NHS IT scheme revealed the full extent of its costs to date.
But the Department of Health always knew it was going to cost as much as £12.5bn, it said today, even in those days when it said the programme was going to cost half as much.
Moreover, it insisted today, that the doubled price tag did not mean that the Programme has gone over-budget.
Speaking
Never underestimate the amount of money the vendor will be adding to the bill just to deal with the amount of nonsensical political back and forth that will go on for years while trying to deliver the project. Probably half of that budget is for meetings with over 15 people in them :) Gotta love trying to deliver software by committee of ill-equipped plebs.
When you start spending that sort of money it basically goes to create big structures of people - there’ll probably be 20 developers working for 3 years + a supporting cast of project managers, program managers, test designers, etc.
Since it’s impossible for any one person to understand what all those people are doing, you instantly create a game of chinese whispers - and then grind everything to a halt in meetings.
NZD5 million would seem reasonable = 3,000 man days = 6 people over 2 years.
But on the other hand, it does piss me off how a lot of NZ clients come up with unreasonable requirements and expect them to be implemented at ridiculously low cost.
Hear hear! I’d love to see a break down of “project costs”. Rich - I think you’re totally right re: the chinese whispers. I’d also like to know whether they plan on “designing” it before “building” it. Of course, I might be a little biased here (I’m an interaction designer), but I think too much time and money is wasted on projects like these because building starts too early, when stakeholders and developers aren’t clear (or on the same page) about what’s actually going to be built - and why.
Build it from scratch or pick an suitable OSS system, open source the bugger and build a services industry in round supporting it. The model has been proven here already in education, let’s do it again. This is fundamental infrastructure we are talking about.
Yesterday Laurence Millar (e-govt God for SSC) gave a talk and mentioned that he didn’t think NZ IT had the capability to deliver to Government’s wonderful e-govt “vision”.
Well, no shit buddy. With decisions like the one above, what do you expect?
Oh well, at least we have cornered the world markets in shrink wrapped cheese, right?
And lets not forget Incis
http://en.wikipedia.org/wiki/INCIS
Funny, I was discussing government procurement just today … over lunch.
Our educated and very lunchtime thesis was that rather than carefully specifying, developing and deploying one $54M IT project … that it would be better to approximately specify and deploy half a dozen or so for $100-200k each and see which actually gets used. As a method of eliciting _real_ client requirements it has to be a winner; six small local companies would get what would quite possibly be their first real gig; and you never know, we may find that $54M is completely ludicrous and we could realistically save the taxpayer … say … fifty millon bucks?
A friend of mine needed a good plasterer. He didn’t want to be bothered with getting a plasterer and finding out some way down the road that it wasn’t the plasterer for him.
So he asked five plasterers to plaster three square metres of wall and a metre of square-stop for him. Paid them each $500, and asked them to just go at their usual pace and explain what they were doing.
Out of the five, two stood out. He chose the cheaper of the two. Ended up with a first-class job.
If he was working for the government, there would be no need for first-class work, the selection process would have been seen as wasteful, he might not have used all of his budget that year and lost out next year, and probably would have had to take the cheapest option, or spend a lot of time and effort justifying why he didn’t. There aren’t that many people in the public sector who have the motivation to do anything other than the safe thing for them, their career, their party, etc.
Robin - are you suggesting the govt selects 5 vendors and pays them all say $100k to develop a part of the system as part of the RFP process ?
I have been involved in a govt RFP response where the short listed vendors were asked to do exactly that.. except for free
In the end they pay though because the cost of sale process is always passed on in the project one way or the other.
I think it is important that the govt has procurement processes, the $$$ involved would make it too susceptible to dodgy practises otherwise. It’s bad enough responding to RFP’s to the private sector where the playing field isn’t level.
This kind of outcome is a direct result of myopic project management. People assume that only big vendors wearing flash suits, using wizzy powerpoint presentations and “enterprise grade” pricing can do the job. It’s bollocks.
The “big boys” throw a pile of money at the project, often over-engineering it to ensure it doesn’t fail. I’ve seen it happen. I’ve seen RFPs written this way. Most of the money is for contingencies - they assume that you don’t really know what you need, so better to build a lot of slack in so when they find out, sooner or later, they can still deliver something.
You’ve got to run a really clean RFP process, and ignore the smoke and mirrors. And have a decent RFP too.
And building it from scratch…don’t even start me on that. If MSD think that their needs are soooo different that they DO require a custom built system, it suggests that their internal business systems are not in great shape.
Spending $54M on a system to support a bad business process? I’ll watching this one with interest.
Richard - seriously, one of the considerations by govt depts. when selecting a vendor is “are they big enough to sue”. ie: if it all turns to custard can we get something out of them. Small local outfits would just fold and disappear.
A $54 million contract to some of the big boys like EDS, Accenture, CSC etc. is small change.
I think the MSD is actually attempting to buy a package here, not a custom development, although the level of modifications required is anyone’s guess.
Glenn, the “are they big enough to sue†comment is valid even if the companies are reasonably small relative to the companies you mention, because even reasonably sized development companies have professional liability insurance; as do large companies.
The problem with large development companies is that they turn it into a Big as Ben Hur type production.
Small and nimble actually does work.
Glenn - good points. I’d not thought of “belt and braces and room to sue”. I guess with so many IT projects ending in tears (over budget, over time, objectives not met) you’d want to have some fallback ;-)
“are they big enough to sueâ€
…if they are, they are probably big enough to sue right back at you. The PL issue was dealt with but there are two other points with that approach
1. it mitigates against just about *any* NZ company being able to get that sort of work.
2. when was the last time you heard of a supplier being sued by government for non-delivery?
Another beef I have these days with contracts being bandied around the Wellington sector at least, are the draconian IP clauses. Liability is assumed by the supplier for IP infringment in *any* jurisdiction in the world.
Don -
“1. it mitigates against just about *any* NZ company being able to get that sort of work.”
yep, also means a few jaunts overseas to visit reference sites of the supplier, fair more interesting than visiting NZ reference sites. ;-)
“2. when was the last time you heard of a supplier being sued by government for non-delivery? ”
perhaps “big enough to sue” is not the right phrase, more like “are they big enough to negotiate to some sort of settlement with”. You’re right they don’t actually get to court very often but the threats are thrown around. Of course it’s not good for either party for it to see the light of day in court. The things I have seen for example are the multi-national suppliers agreeing to bring all their “international resources” to bear at “cost price” to resolve the issue at hand. Small local vendors may not have the wearwithall or contacts to do that
Regading IP, I think this is an area where the govt. gets it wrong, they’d be better of letting the supplier keep the IP because then they’ve have a vested interest in creating a good product that they may be able to sell later and may be more inclined to do the work for a lower price. Realistaclly why does a govt. dept. need to keep the IP for a software development ? It’s hardly like they have any competitors and need to maintain competitive advantage is it ?
Dermott - Professional Liability Insurance is usually capped, I personally have some for $1m NZ or the value of the contract, whichever is the lesser. Agree about the “be hur” productions for sure.
Small and nimble vendors can be better I agree depending on the job. Small/medium size development outfits sometimes struggle with the “governence” issues at very large customers though, issues like rigorous change control, formal user traing, separation of duties etc.
I think that breaking a project down into “bite size” chunks is the better approach if at all possible. Small achievable projects that are easily justified, understood by everyone involved. For example perhaps replacing one part of an existing legacy system at a time. It might take longer overall and cost a bit more but the risk is much less and more likely to succeed.
All - don’t get the wrong idea, I’m not trying to defend govt. procurement methods here, just trying to see their side of the argument and provide a bit of insight from my experiences.
Glenn
Curam is being considered or is implemented for other Govt Social development Public Sector organizations around the world Ireland, Canada, North America so perhaps a big part of the selection process was Domain Knowledge and experience of the vendor, expectation of continual improvements, stable and reusable business processes, the ability to share and leverage innovations from other members of the user group.
Unfortunately during implementation a lot of these expected synergies are lost, through poor requirements and GAP analysis before product selection and during implementation. Meaning that the standard processes are not leveraged and instead replaced with strange Frankstein processes that actually slow an organization down. If the MSD has done the right due diligence and not just selected because it is COTS, has all the right check boxes on the sales brochure or because everybody else has it and keeps a tight ship during implementation, they may even come in under budget and be able to free up public money from IT and for where it is actually intended Social Development :)
According to the Curam website news archive MSD have been publically talking to and about Curam since at least 2004, so the due diligence had enough time to happen….
54 million is a lot, and Curam present selling success is a very good example of why investing in IT vendors organizations in your own country is a good idea. Curam product after all appears to be based on what was implemented for the Irish Govt :)
As a taxpayer I would like to see the cost breakdown and also the return on investment from the 54 million.
What would also be interesting is the expected costs and benefits for the next 5-10 years with such a big up front investment of public funds.
Regarding IP - agree with Glenn’s comments completely. In my experience, even govt-related buyers are happy with a split between core IP with which you turn up and licence as part of a solution, and what you develop exclusively for them (eg the customised interface to your system/IP). Only the latter is theirs outright; the former is theirs by licence.
The tricky bit is in the middle, where you extend your existing IP/systems through the work you do with the client, and expect to walk away with the ability to reuse it later without any claim by the client.
Of course, it helps enormously to have an existing business model established on these lines, about which you’re extremely clear up-front. So the business must *actually* and demonstrably be dedicated to building up some core IP in a specialised area.
I can’t be bothered reading all those comments, but in regards to Glenn - Intergen might not be the best shop to implement it, but since it’s “essentially the largest most complicated payroll in the country” why not give it to the people who run the current biggest payroll in the country, Datacom (they run the M.o.Education payroll)? They’re a bespoke development house as well, as well as providing mission critical hosting and processing capacities (they manage the NZ Blood Service and banking transactions).
Adam, maybe you should have read “all those comments” because its not basically a payroll system but a client management system. So think of it as a CRM system for Work and Income, Child Youth and Family.
While there probably will be payroll involved its much more than this.
So is this mission critical? I don’t think so. Will it make any difference to the child poverty in NZ that Clare mentioned earlier? I doubt it.
Someone who worked at Work and Income (in a small town where they are the biggest business in town) said there are a lot of people who work there who are paid low wages and even get all sorts of benefits themselves.
So previous people have mentioned this could be a useless expenditure of money. Think the Police INCIS system; think the Telecom PROBE project.
I’ve written a few public sector RFP responses. Boy, they’re a lot of work. Public sector jobs often have very rigorous demands to meet in a response. If you’re a small shop, you won’t have the resources to easily complete an RFC response - so the big boys are often the only contenders for the work because they can afford to spend a lot of unbillable person-days on a tender.
And then, if you are the lone cheap bid, it looks weird. If you’re a public servant, and you get bids of 10 million, 8 million, 9 million, and 2 million, you’re going to wonder what’s wrong with the 2 million dollar bid. The risk-averse person will bin the substantially cheap bid as being unrealistic. (I don’t blame them for this. Such heuristics may be as reliable as detailed analysis in assessing complex projects. Or not. Who knows?)
This and other institutional factors inherent in the tendering process make it unlikely for small firms with novel ideas to win.
Also, many people are averse to trusting small teams, because they fear dependence on a handful of important individuals. From the outside, naive point of view, a large team seems less vulnerable to individual crises.
Fundamentally we come down to what Warren Buffet calls the “institutional imperative”, or Parkinson’s Law. And I don’t believe things will ever be otherwise, because that’s what human nature, or at least our culture, is like.
Just in regards to Adam re:Intergen’s ability to deliver vs. datacom.
Well all I can say is I hope you are enjoying working with Datacom these days. Personally, there are many companies out there that are doing amazing work. Intergen is one of many who are delivering high quality solutions to large govt. and private companies.
NZ development shops really have to pull together and team up bringing together there individual strengths to make it ‘happen’ in the IT market.
ALone NZ IT companies keep fighting over the same piece of meat. If they could come together and work as a team, we as a NZ IT sector could start vying for larger projects internationally.
In fighting over who is best never gets anywhere. But seeing the possibilites of joinging together and focusing on each individual strengths, we as a sector could deliver great projects, at a much smaller cost.
It would create the ideal soluition : An IT project, on budget at a resonable cost, on time and that works as expected - using local talent to engage and assist New Zealanders.
[...] the announcement was not entirely universally applauded with notable industry insiders questioning the size, price tag and the fact that the project went to an overseas rather than local [...]