* The Business of Software

A community discussing the business of software, from the smallest shareware operation to Microsoft. A part of Joel on Software.

This community works best when people use their real names. Please register for a free account.

Links:

» Business of Software FAQ
» The Business of Software Conference (held every fall, usually in Boston)
» Forum guidelines (Please read before posting!)

Moderators:

Andy Brice
Successful Software

Doug Nebeker ("Doug")

Jonathan Matthews
Creator of DeepTrawl, CloudTrawl, and LeapDoc

Nicholas Hebb
BreezeTree Software

Bob Walsh
host, Startup Success Podcast author of The Web Startup Success Guide and Micro-ISV: From Vision To Reality

Patrick McKenzie
Bingo Card Creator

Hundred Million Dollar Projects

Has anyone got experience of working on private sector (i.e. non-Government) projects that cost hundreds of millions of dollars?

I keep reading about projects in the UK costing hundreds of millions of pounds and then either being years late, or being scrapped as soon as they go live, or both.

I'm curious as to how an IT project can get to £350,000,000, like the recent child contact database that was scrapped, even though it was never rolled out properly and was apparently going to cost another £40+ million a year to run.
Scorpio Send private email
Wednesday, August 11, 2010
 
 
What would you like to know? There have been various post-mortems on these projects. There are various factors. Scope creep is a big one, typically the biggest. Another is no central direction. Another is planning for years without writing code. Another is irrelevant requirements imposed for political reasons which create impossible to fulfill specs that are unrelated to needs.

Success is more likely when adding on to a give system in specific ways. Many of these projects though are start over from scratch, reinvent everything projects that expect a totally different workflow no one has tried and even when the software works, it is rejected by its users.

It's extremely uncommon for technical incompetence to be a factor, although in the end its the programmers who get blamed for everything.
Scott Send private email
Wednesday, August 11, 2010
 
 
"how an IT project can get to £350,000,000"

And those things, that's probably something like £100,000,000 for new computers and office chairs for every government worker in the country, £200,000,000 for conversion of handwritten paper records going back 50 years, and £50,000,000 for the actual development.
Scott Send private email
Wednesday, August 11, 2010
 
 
There was a regular here a while back who worked on a USPS project that was like that.  He posted about it a few times.

Wish I could remember his name...

Pretty much every large government IT contract seems to end like this.  Huge cost and schedule overages, and the project is either abandoned, stillborn, or just never works right.

I'm not sure I've ever heard about one of these projects that didn't go bad.
Jason Send private email
Wednesday, August 11, 2010
 
 
OK, I looked up ContactPoint, what he is asking about. It was a big brother birth to 18 database of every minor in the UK, containing unified information about every event in their life and tied into every agency they could possibly have contact with.

"The database, which would be operating in 150 local authorities and would be accessible by at least 330,000 users, was expected to be fully operational by the end of 2008; however, following the 2007 UK child benefit data scandal, the deadline was pushed back for five months to allow a security review prior to implementation."

It's interesting there that that says "prior to implementation". Sounds like they may have spent years on design, then just before actually writing the first line of code a completely new scope comes down the pipeline - now it has to actually be secure!

The 330,000 users were not to be kids or parents but authority figures monitoring their lives.

In the end it seems it was finished on time and worked properly, but people were so horrified by the privacy implementations an order was given to physically destroy the databases, throw out the hardware, and abandon the program.
Scott Send private email
Wednesday, August 11, 2010
 
 
My own experience is that the more is paid for a piece of software, the less likely it is to be used. I have seen plenty of bespoke software costing 10s or 100s of thousands of pounds that went straight on the shelf, never to be used.
Andy Brice Send private email
Wednesday, August 11, 2010
 
 
I've worked on a few of these projects, I remember one where there were forty people in the office I worked in, two of us were creating the code.
Tony Edgecombe Send private email
Wednesday, August 11, 2010
 
 
Not directly answering, but I worked for the government (nothing to do with IT) for 5 years and my spouse still works there.

Based on this experience, I firmly believe that those stereotypes about government workers being unmotivated, having way too many layers of management, etc. are 100% accurate.

I worked primarily at a facility that literally did almost nothing and had maybe 20 people working there, which included five managers, each of which made around $100K, and each of which literally spent most of their day drinking coffee and chatting with us, who also had nothing to do. Everywhere I have seen in government is more or less similar.

Throw in generally much tech savvier consultants who are motivated to maximize their billings and it's hard to see how a government IT project ever succeeds (though they clearly do, on occassion, somehow).
GregT Send private email
Wednesday, August 11, 2010
 
 
Typical ERP projects involving software from Oracle and SAP run into millions.

There is software license cost
There is implementation cost from third party vendors or software partners.
There is training cost
Legacy conversion cost
and the list goes on

The software even though run into millions ends up being used effectively.

So "More pricey the software...less used" is not true.
nj Send private email
Wednesday, August 11, 2010
 
 
"I've worked on a few of these projects, I remember one where there were forty people in the office I worked in, two of us were creating the code."

That reminds me, there was a guy that used to post here on the Joel board that had worked on some kind of submarine software. I'm probably imagining half this, but if I recall at all correctly there was like an entire defense contractor getting fat off this top secret contract and under all the onion layers of corporations and subsidiaries and subcontractors there was this one guy in the end that didn't have any clearance at all writing the actual software and he was the only person actually doing anything among thousands and thousands of people making a living here. He might have even been a college student that was making $5 an hour or something. Hm, sounds too good to be true, but it sure makes a lot of sense. Would make a good movie that no one would believe.
Scott Send private email
Wednesday, August 11, 2010
 
 
Hey I found it. Guy was Scorpio. Designed top secret nuclear submarine control systems for beer money in college.

http://discuss.joelonsoftware.com/default.asp?biz.5.701626.10

Second hit for "submarine" on this site.
Scott Send private email
Wednesday, August 11, 2010
 
 
And he retells the same story using the alias "Odysseus" in this other thread:

http://discuss.joelonsoftware.com/default.asp?joel.3.601128

Both stories are consistent so I bet it happened like that.
Scott Send private email
Wednesday, August 11, 2010
 
 
Ha ha, Ok "steam submarine" catches all four of his tellings of the story, adding in these threads:

http://discuss.joelonsoftware.com/default.asp?joel.3.543881.12
http://discuss.joelonsoftware.com/default.asp?joel.3.530076.11

Bow to my powerful Google-Fu, puny hu-mans!
Scott Send private email
Wednesday, August 11, 2010
 
 
"The military contracted to their main supplier, who sub-contracted to the guys I worked for, who sub-contracted to me, as a part-time student, working for beer money."

Wow, I actually remembered it all right. I guess he told it so many times that that reinforced it.
Scott Send private email
Wednesday, August 11, 2010
 
 
Ha, and in the last thread there there is a "me too post" from "Student Worker" giving some dollar figures from his own experience:

> That happened to me in the US. Like you, I was a student at the time, and I was hired by a professor who was doing some side work (without university clearance) for a contractor working across the country, who was working for another shop, who was working for a defense contractor, who was working for the pentagon on a top secret system for the stealth bomber.

> And no, I didn't have any clearance. And yes, I was paid $6.50/hr at the time for work that the government was paying $375/hr for. I even had to fill out timecards.
Scott Send private email
Wednesday, August 11, 2010
 
 
Classic. I have a decade of government contracting experience. This, is absofreakinglutely how it works. The contracting process is so broken that there have been an infinite number of ways to "go around" the bottleneck that is contracting. This includes using 8(a) contractors, disadvantaged contractors (SDVOB), sole source requirements, and big multi-award contracts that limit competition to 3-4 big vendors, anything to avoid having to do something competitively. The big contractor's squeeze everyone else out of the business by either buying them up, or gaming the system to keep everyone else out. The other problem is that the government never knows what they want until they see it, try evaluating proposals in accordance with FAR for something that has no face. It's rediculous and painful. Oh and, once you make a contract award, it costs $.42 to for a competing vendor to protest. This can delay a contract by 6-9 months, or kill it completely. Working under this system is insane and inefficient, but I know of things like this going on all the time. It's yet another government system that is need of dire overall. Add it too the list, SEC, Fanny/Freddie, Medicare, Medicaid, IRS, Congress...
Glade Send private email
Wednesday, August 11, 2010
 
 
Add to this the fact that most of the companies doing business with the government are X-governement employees and politicians...and you have one mell of a hess.
Glade Send private email
Wednesday, August 11, 2010
 
 
There is a blindingly simple solution: stop giving them your money to spend and pass laws that makes it illegal to have a deficit.

Maybe people will actually earn a living instead of cracking the backs of the productive part of the economy.
Bring back anon Send private email
Wednesday, August 11, 2010
 
 
@Scott
Yes, that submarine thing was me, back in the day when I was s student. It is absolutely true, although I apologise for posting the story so many times. It's a good story, but I didn't realise it'd been four times.

The ContactPoint database is just one of many, many government IT projects in the UK that go bad. It was never rolled out properly. Just 15,000 of the 330,000 users got it, then it was abandoned a few weeks after it went live.

Now, back when I saw a student, I worked on this submarine project, with no security clearance and... Oh wait, you already know that ;-)
Scorpio Send private email
Thursday, August 12, 2010
 
 
I've worked on a few of these sorts of projects, and I've had contact with a bunch of others. Particularly in the IT market, the disaster goes something like this;

1. There's a problem.

2. There's a solution.

3. Assume that 2 and 1 go together and decide that we're doing the solution. Announce it publicly so everyone knows that a solution is being done and when it will be done by.

4. It will be big and comprehensive and do a whole set of stuff. It will need to handle all the special cases and the places where data doesn't exist yet and all the rules about all of these things and it will have, clearly, MILLIONS of users. So it will be expensive.

5. It must not be expensive -- this is the taxpayers money we are spending and we must spend it **responsibly**.

6. Hire the cheapest people we can find. It doesn't matter if they're any good, having more of them will make up for that. Hire people who don't speak the computer or human languages used if necessary. They'll pick it up as we go along; we'll make our OWN experts and they'll still be cheap!

7. Throw them in a building and change stuff about what they're doing every week; even better if you can fail to actually understand what you're doing to start with. The UK specialises in this initial failure to understand processes, starting with exciting advantage of having one of the world's most complex tax systems and one of the world's most complex benefits systems and working out from there.

8. As they get further beyond the initial guesstimated delivery date, hire even more people.

9. Boooooooommmm!
Katie Lucas Send private email
Thursday, August 12, 2010
 
 
The large consultancies that run these projects make their fortunes on change requests. The initial contract goes out to tender and so has an element of competition, although due to the cost of bidding each stage will reduce the number of bidders. However once a single consultancy is running the show out of plan changes or requirements get negotiated separately and end up costing a fortune.

Governments tend not have the best business analysts so end up realising the requirements don't do everything they need - cost.

But in the UK far worse was the previous governments constant reorganisations. New minister or quangocrat wants to make their mark, reorganises something, either deliberately tinkering with the project or with the organisation that causes a project change. That is where lots of the massive price inflation comes from.

The new governments plans to end these massive projects should be applauded. Some form of UNIXy mentality of software doing one thing well working together might also help.

But the key part should be transparancy, big projects should have some form of public change log - short description of change, reason, who authorised and estimated cost.
ChrisH Send private email
Thursday, August 12, 2010
 
 
A number of years ago the company I worked for was selected as a supplier (subcontractor) to one of the huge IT companies working on a huge (billions) military IT project.

We were suppose to help supply the bit of the project related to our company's specialist field of expertise.

The first meeting I went to with a sales guy who had got us the deal.

In the room were similar groups of 2 or 3 people from around 20 other similar vendors. Some of whom were our company's fierce competitors. Some of whom were in related fields to our company. And some of whom had absolutely nothing to do with our field, just had something of a similar sounding name/description (similar to this: if you were having a meeting on spreadsheets, and as well as inviting spreadsheet software companies, you invited a couple of "bed sheet" and "aluminium sheeting" suppliers along as well).

We all sat there for about 2 hours while the guy from the huge IT company waffled vaguely. He invented whole new dimensions of vagueness.  The actual amount of information from the meeting was less than in the 2 line summary of the project you might read in a newspaper.  And what's more, I understand similar meetings were taking place all over the huge IT company's building.

Afterwards, the sales guy said to me that the meeting was a waste of my time, and that I needn't come again at this stage. This was from a guy who was never reluctant to use developer's time to his own ends.  The sales guy of course kept going to subsequent meetings probably for at least another year. And I'm not sure if our company was paid for our contribution (or non-contribution) as a subcontractor by the huge IT company,  but I know the huge IT company was paid handsomely by the government (more on this later).

The huge IT company is BTW the way a very successful company with tons of sharp people - and can produce stuff - but in this case, it seemed like an exercise in looking busy.  And I didn't get the impression the huge IT company had assigned their sharpest tools in the box to the project.

Coincidentally a couple of years later I worked on a different project with one of the guys from another vendor who attended that first meeting. He asked me what impression I had got at the meeting... yes we both had the same feelings, and a good laugh about it.

Over company's involvement dragged on for a year or more, until the sales guy couldn't take any more.  I don't think a requirement for anything deliverable was ever made to our company.

BTW, the project was eventually cancelled several years later by the government (I learned this from the press).  Several billion had been sent on the project by the government.  Not one line of code, and not one piece of hardware had ever been delivered AFAIK - it had all been sent on "defining requirements".
S. Tanna Send private email
Thursday, August 12, 2010
 
 
BTW the thing about ContactPoint is that it was triggered by the murder of Victoria Climbié, an 8 year old girl.

The point was she needn't have died, but for poor communication among social workers, etc.  So the there was a perceived need to improve that.

In short the problem:  About 100 (?) UK kids killed by parents/careers annually,  poor communication among social workers etc., in these serious cases where kids are at risk.

Government proposed solution: Build a database of 11,000,000 children, accessible by 400,000 people... figuring out issues like confidentially as they go.
S. Tanna Send private email
Thursday, August 12, 2010
 
 
"The point was she needn't have died, ..."

Of course, but the reason she died was not because there was a missing £350M contact database.

That money would have been better spent elsewhere, for example in cleaning up drug-infested estates, creating jobs, etc.

Stupid politics strikes again.
Scorpio Send private email
Thursday, August 12, 2010
 
 
What I'm interested in most with this topic is whether such enormous budgets exist in the private sector, or does it only apply to the government/military gravy train?

Even the most complex private sector IT projects I have been involved in have only been a few tens of millions of dollars, so I just don't get why government projects should cost tens or hundreds of times more, apart from the excuse that everyone is milking it for every dime.
Scorpio Send private email
Thursday, August 12, 2010
 
 
@bring back anon,

I wish there were a way to keep them from taking our money.  It doesn't seem to matter which party is in power, they all love to bring home the bacon for their constituents.
consultutah Send private email
Thursday, August 12, 2010
 
 
> Even the most complex private sector IT projects I have been involved in have only been a few tens of millions of dollars, so I just don't get why government projects should cost tens or hundreds of times more

They don't always. There are some very rough numbers for the development costs of Vista here. About 10 Billion USD (for payroll only)

http://seattletimes.nwsource.com/html/businesstechnology/2003460386_btview04.html

As for some of the other comments here, what I've found is that Government projects always want to include the edge case functionality that a private company would disregard because they know they won't get a return.

Also, because those bits are usually difficult, a disproportionate time and effort is spent trying to satisfy those requirements that nobody is likely to ever use. So the core functionality that everyone will use gets less effort. Not only that, but the nice clean design also gets compromised to ensure the never used functionality is incorporated.
IanH. Send private email
Thursday, August 12, 2010
 
 
"I just don't get why government projects should cost tens or hundreds of times more"

Where's the mystery here? Government IT contractors are labor providers that charge by the hour. The more people involved, and the longer it takes, the more they make. They have absolutely every incentive to just be quiet and make exactly what the government specifies, even if they know there's a better or less expensive way, because they make more with change requests when the specs have to be changed (again).
CC Send private email
Thursday, August 12, 2010
 
 
Hey Scorpio, I didn't realize you were still here. I have always wondered about that (and it's OK if you just have a random estimate that's not accurate because I only intend to repeat it a million times but it's all anonymous so no problem), how many people do you estimate were employed and sitting on top of you in the chain of solving this steam pipe problem, and how much did the government spend on people other than you, the only guy in the system that was actually doing any work at all.

Obviously I am asking this so I can say, "Well this dude got paid $5 an hour and the entire project involved 75 companies, 1500 people and $150 million, all of it to produce this steam pipe that he made."
Scott Send private email
Thursday, August 12, 2010
 
 
"The more people involved, and the longer it takes, the more they make."

Yes. Couple further comments.

When I graduated college, one of the main employers looking for grads at the time was government defense contractors. So I'd get calls from them, and at school job fairs they'd sometimes grab my resume out of my hands even though I wasn't interested in their jobs. I went on two or three interviews for well known outfits though out of curiosity and because I wanted to see what went on with our tax dollars in these places.

What was interesting was these were some of the only interviews I went on in my life that did not result in any job offer.

At the third interview, when the big guy met with me and it was obvious I was a no hire I asked him about this, because I had graduated Summa Cum Laude, already held a patent, etc, and I for a fact knew I was the best candidate he had ever seen in his life. He closed the door and said (I am paraphrasing), "Look, we both know you would not want to work here. We work defense contracts. That means we don't want to deliver early. We justify our rates by the number of engineers required. It is in our interests to hire the less capable graduates, while not quite utterly incompetent. That's how this industry works. Obviously if you ever repeat this I will deny it."

It was interesting because the industry was not always like that. I have older relatives that worked in defense from the 1920s-1960s and back then you needed to be good and the best were sought out.
Scott Send private email
Thursday, August 12, 2010
 
 
Oh sheesh, I was totally oblivious that the original poster in this thread was the same guy whose ancient post I then dug up in response. LOL. Doh!
Scott Send private email
Thursday, August 12, 2010
 
 
"Of course, but the reason she died was not because there was a missing £350M contact database."

But what if it was? Then the blood would be on your hands.</sarcasm>
Scott Send private email
Thursday, August 12, 2010
 
 
"as well as inviting spreadsheet software companies, you invited a couple of "bed sheet" and "aluminium sheeting" suppliers along as well"

Ah dear Sunil, that one had me rolling on the floor.
Scott Send private email
Thursday, August 12, 2010
 
 
Wow, Katie Lucas's recipe is really good.
Scott Send private email
Thursday, August 12, 2010
 
 
"The UK specialises in this initial failure to understand processes, starting with exciting advantage of having one of the world's most complex tax systems and one of the world's most complex benefits systems."

Oh! So it goes beyond the addresses then.

UK addresses are the current worst case scenario in my DIY envelope addressing code, a result of hundreds of years of different jurisdictions splitting and merging in new shapes presumably. Maybe Chinese is worse, but for those we just allow free-form entry.
Scott Send private email
Thursday, August 12, 2010
 
 
>> The large consultancies that run these projects make their fortunes on change requests.

Too true! I can't count the number of times I heard "we'll make it up on the change orders" while I was working in Big Time Consulting.
a former big-fiver Send private email
Thursday, August 12, 2010
 
 
Scott: "Hey Scorpio, I didn't realize you were still here.
...
how many people do you estimate were employed and sitting on top of you in the chain of solving this steam pipe problem, and how much did the government spend on people other than you, the only guy in the system that was actually doing any work at all."

Yeah, still here. Been here since the beginning, back in 2000, apart from a short break when Joel said I should check myself into a mental hospital because I disagreed with him (part of his charm, I suppose).

The story is absolutely true and serves as a great example of what is wrong with government procurement, especially military projects.

It is hard to know how many people were involved, as there are so many layers of management. My bit was part of a major re-fit of the "boat". We weren't allowed to mention the "S" word (submarine) and definitely not the "N" word (nuclear), even though we all knew what it was. Maybe they were worried about the Russians bugging the place.

I did the work, sub-contracted to a small business of four people. The two directors signed the OSA and has security clearance, but weren't involved in the project in any way (I had no clearance and no OSA, of course).

That business sub-contracted to an engineering company who specialised in pipelines and was part of a ship-building group. I believe that there was at least one more layer involved between them and the MoD who would have been running the project.

I was paid £7 per hour at the time, which was better than the £5 a hour my student peers got for working in bars, etc. The company I sub-contracted for did the project for £50,000. The engineering company contact laughed at them and assumed it was a typo, as the other bids were in the high hundreds of thousands, so of course their figure would probably have been in the low millions and the eventual charge to the government could easily have been several million pounds.

There was about £2000 worth of electronics in the system (sensors, communications equipment and computer hardware), with my hardware and software value-add, maybe £10,000 of costs, which mysteriously turned into several million pounds charge to the tax payer.

For perspective, this happened about twenty years ago.
Scorpio Send private email
Friday, August 13, 2010
 
 
@Katie Lucas
Your recipe looks about right, apart from step two. I'm not convinced that they always have a solution in mind when they announce that the "problem" will be solved.
Scorpio Send private email
Friday, August 13, 2010
 
 
There has been talk of the UK government canning its £12.7 billion IT project:
http://business.timesonline.co.uk/tol/business/economics/pbr/article6946336.ece

As far as I can tell it is a glorified CRUD database. I have a sneaking feeling that a team of smart guys from Google could have written something that fulfilled 80% of the software requirements in a year for < £1 million. Heart breaking.
Andy Brice Send private email
Saturday, August 14, 2010
 
 
The project I talked about was also a glorified CRUD database (albeit spread over multiple locations).  It was for what you might call a rear/support activity, rather than an actual weapons system.

In other project, I also encountered an IT system used in an actual weapon system project (a major weapons project). You'd be surprised, probably worried, if you knew how low-tech and cobbled together that was, especially in terms of hardware  - I know this system is still in service,  I wonder if they're still using the same laptop in it, and will be for the next 20 years - hope the hard disk doesn't crash.
S. Tanna Send private email
Saturday, August 14, 2010
 
 
> I can confirm that a lot of what my organisation is trying to do could very easily be achieved by 10 smart guys

Take for example the task of standardizing provincial/national medical records, and making patient data portable, so that it's available on demand and in an emergency to all service providers (hospitals etc).

There are some technical difficulties: e.g., how to standardize the content of the records.

There are already some standards though, and, open-source implementations of those standards.

The biggest problem I think is conflicting, mutually-contradictory requirements:

* All information available, to authorized users
* Many, many 1000s of authorized users (doctors, medics, receptionists, paramedics, nurses), all over the country (in hospitals, doctor's clinics, nursing homes, ambulances)
* Information mustn't leak
Christopher Wells Send private email
Sunday, August 15, 2010
 
 
Yes, this is a very good point. The requirements are not achievable as specified. You can't always have everything you want. No system with thousands of users much less hundreds of thousands can be made secure. What will happen is the information will leak and then the programmers will be blamed. When in fact the system never should have been built as it intrinsically conflicts with a right to privacy.
Scott Send private email
Sunday, August 15, 2010
 
 
That said, the ultimate failure of the system could have been achieved more quickly, at lower cost and higher quality if done with a small autonomous agile team than with a giant bureaucracy. And then open source and small teams would be blamed and the next iteration would demand a much larger system and time frame to achieve the same failure.
Scott Send private email
Sunday, August 15, 2010
 
 
"I can confirm that a lot of what my organisation is trying to do could very easily be achieved by 10 smart guys"

Every project I have ever worked on could have been done better by just me, or me and one or two of the excellent people I know.

That isn't the issue. The real problem is convincing management to give you the chance, when "conventional wisdom" suggests that teams of 50 and several years and many millions of dollars are required.
Scorpio Send private email
Sunday, August 15, 2010
 
 
@Andy Brice
You are closer than you think with the Google suggestion, as I believe Google are moving into the electronic patient record space in a big way.

The project is a national disgrace and has been since day one.. It is the canonical example for this thread.

It has a budget of £12.7 billion, although I've heard alternative estimates up to £30 billion for total life-cycle costs.

And yet we have "ministers had previously defended the system, insisting that it could save the NHS £1.14 billion by the revised deadline of 2014."

Hmm, no wonder this country is in such a financial mess. Spend twelve (or 30) billion and several years later you "save" 1 billion.
Scorpio Send private email
Sunday, August 15, 2010
 
 
Unfortunately the government doesn't seem to have learnt the lessons that businesses have - start with a small system and then iterate.
Andy Brice Send private email
Sunday, August 15, 2010
 
 
Some friends and I were discussing this last night: why don't some groups of people ever do things the smart way? And the answer seems to be natural selection. There is absolutely no reason to do things differently.

If a government could not borrow money in your name, it could not survive and I believe that would be sufficient to guarantee a smart way of doing things.
Bring back anon Send private email
Sunday, August 15, 2010
 
 
In the car today - listening to the radio - this was basically the discussion:

1. IDS (Conservative MP) wants to reform the UK welfare system

2. But one of the things holding him back is it takes investment to make improvements.

3. One of the major investments required is a new computer system to calculate welfare benefit payments, because the current one is not good enough

4. And the estimated cost of developing a new system for this......  7 billion pounds!


WTF I thought. Then I thought of this thread.

Even if you have to buy new PCs for all the welfare offices in the UK (which actually I think is unlikely to be included in the 7 billion)... WTF

Even if there are 100 different possible welfare benefits how hard is it possible to calculate how much to pay each person, and store what has previously been paid in some kind of database..
S. Tanna Send private email
Sunday, August 15, 2010
 
 
"There is absolutely no reason to do things differently."

This is the crux of the matter. There's absolutely no motivation for politicians to do the right thing. Even just before election time, they just have to present an illusion of doing the right thing.
Scorpio Send private email
Monday, August 16, 2010
 
 
@S. Tanna

This is exactly my frustration.

The "system" IDS was talking about is just a glorified spreadsheet. However complicated and ridiculous the UK benefits system is, there's no way you can justify spending £7bn on it.

I bet I could write a system to solve this problem in a year.

The obvious thing to do is merge the benefits system and the tax system together, then everything only needs to be calculated once for most people.

I'd happily do the whole thing for £5bn, saving the taxpayer £2bn in these difficult times ;-)

Actually, I would do it for nothing, just to prove to these idiot politicians that there is another way. However, even if I stopped IDS in the street and told him that, he wouldn't be able to accept my offer.

Governments seem to be so inured to spending billions and getting nothing useful back. I can only assume these outsourcing companies do really good golf days and other freebies.
Scorpio Send private email
Monday, August 16, 2010
 
 
S. Tanna: "In the car today - listening to the radio - this was basically the discussion..."

Do you know which radio programme/station that was? I'd like to hear it, if it is on the BBC iPlayer.
Scorpio Send private email
Monday, August 16, 2010
 
 
Radio 4, about 10pm, Sunday night. I think it was the Westminster hour.

The program wasn't about this issue, it just mentioned it in a brief aside.
S. Tanna Send private email
Monday, August 16, 2010
 
 
BTW I was thinking there's an obvious illustration of the ineffiency of government projects:

1. Commercial programs (sold by various software companies) to calculate how much taxes each individual has to pay:  Development cost = few million,  Development time = months, Update frequency = annual

2. Government programs (developed by various huge IT conglomerates for government) to calculate how much taxes each individual has to pay: Development cost = billion, Development time = years, Update frequency = decades

The only difference with 2, is it also incorporates a database of each tax payer, with a facility to lose track of all correspondance sent to/from them.
S. Tanna Send private email
Monday, August 16, 2010
 
 
> The only difference with 2, is it also incorporates a database

The I/O is different too. Also the project isn't just software: it's the staffing, contractors, auditing, management, and I don't know what, the front line talking with members of the public and opening envelopes and staffing call centres, and mainframe services and networks and multiple office buildings etc. etc.

Also cost doesn't scale linearly with project size.

Having said that, 7 billion pounds seems to me like about 30,000 man-years; I don't know how they spend it: what was included in that estimate.
Christopher Wells Send private email
Monday, August 16, 2010
 
 
Governments haven't always been this rubbish at managing large projects otherwise there wouldn't be any pyramids. Maybe we should need to execute a few under-performing contractors every now and again. ;0)
Andy Brice Send private email
Monday, August 16, 2010
 
 
> I don't know what, the front line talking with members of the public and opening envelopes and staffing call centres, and mainframe services and networks and multiple office buildings etc. etc.

I don't think all of that is included in the 7 billion.

Remember they already open envelopes, have call centres,  etc

The 7 billion is just the cost to develop and deploy the new IT system. Yes it probably includes hardware and training, but the big staff costs of employing thousands of people to run the system day-to-day, are not included in the 7 billion, since they exist anyway.
S. Tanna Send private email
Monday, August 16, 2010
 
 
> Governments haven't always been this rubbish at managing large projects otherwise there wouldn't be any pyramids.

There's a (funny) book about all this, called Systemantics, subtitled "How Systems Work and Especially How They Fail": http://www.amazon.com/Systemantics-Systems-Work-Especially-They/dp/0812906748

It's funny, with "horrible examples" in the same way that Code Complete has "Coding horrors".

It includes some wonderful laws like, "A complex system that works is invariably found to have evolved from a simple system that works."

On the subject of pyramids, it mentions a ruined one, which fell down; that they improved, built them bigger and stronger; and that eventually they devoted so much effort to building pyramids that Egypt itself fell down.
Christopher Wells Send private email
Monday, August 16, 2010
 
 
Pyramids are a bad example, I think. You might want to instead talk about the mission to get to the moon first. They set a goal to actually bloody do the damn thing or face humiliation and wouldn't you know, they did it within budget.

Now, they spend more on sporting events than space exploration.

How the human race gets so fucking lazy and retarded is beyond me. It's probably the taxation/welfare system that allows the unproductive part of the economy to live off the backs of the productive part of it.
Bring back anon Send private email
Monday, August 16, 2010
 
 
This is not a goverment-only issue.  There is a reasonably famous cautionary tale about the time Hershey imploded on fulfilling Halloween candy orders in 1999 due to a Siebel-Accenture-SAP ERP implementation that cost $115 million (in 1999 dollars).  I understand management was itchy to flip the switch on their new, and very expensive, replacement for legacy systems, and they ended up pulling the trigger across the enterprise, in the middle of the busy season, without a plan for rollback. 

Here's a good, succinct article on the Hershey event:
http://www.cfo.com/article.cfm/2987370
keith gerritsen Send private email
Monday, August 16, 2010
 
 
>There's a (funny) book about all this, called Systemantics, subtitled "How Systems Work and Especially How They Fail"

I've read it. It is very good!
Andy Brice Send private email
Monday, August 16, 2010
 
 
Another good read is "If We Can Put a Man on the Moon...".  It has analysis of both successful and failed government projects, and talks about the key differentiating factors between them.

Most of the projects discussed weren't specifically IT, but I think the same ideas would apply.

http://www.amazon.com/Can-Put-Man-Moon-Government/dp/1422166368
Jason Send private email
Monday, August 16, 2010
 
 
>Another good read is "If We Can Put a Man on the Moon...".  It has analysis of both successful and failed government projects, and talks about the key differentiating factors between them.

Any chance of a quick summary Jason? ;0)
Andy Brice Send private email
Monday, August 16, 2010
 
 
Well, I didn't take any notes...

It essentially lays out 7 common traps that these projects fall into.  It's a business/management book, so they all have catchy names, of course. 

One of the first ones it goes into is the Boston Big Dig, where a woman died when part of the ceiling collapsed.  They discussed how the ceiling panel bolts had known problems, and how they kept focusing on fixing that specific problem, rather than stepping back and thinking through the root cause of the problem (sort of a 5 Whys approach).

Another big one is only planning for the best case scenario, and refusing to accurately and realistically plan for risks.

They also talk about the transition to the implementation phase, which is government terms often means that the legislature passes a law and moves on, leaving the implementation details up to someone else, without any guidance or oversight.

One success case they profile is the entry-toll for cars going into London.  Andy, I don't know if public sentiment in the UK considers that a success or not, but in their terms it represents a well executed implementation of a complex system.

It's a good read, and fairly short.  I think I read it over a weekend.
Jason Send private email
Monday, August 16, 2010
 
 
>One success case they profile is the entry-toll for cars going into London.  Andy, I don't know if public sentiment in the UK considers that a success or not

It certainly wasn't popular (tax never is) and I am not sure whether it made any real difference to the congestion. It seems to have been reasonably well implemented though.
Andy Brice Send private email
Monday, August 16, 2010
 
 
Aren't we conflating two issues here.

1. Success vs Failure

2. Fair-Priced vs Over-priced (vs Ridiculously over-priced).


The X billion government defence project that I was tangentally involved with, was both over-priced, and a failure, since it produced nothing.

The X billion government IDS benefit project that Radio 4 would be considered a success, if it *only* cost 7 billion pounds (over-priced in my view).
S. Tanna Send private email
Monday, August 16, 2010
 
 
"I am not sure whether it made any real difference to the congestion."

Congestion was the selling point but never the goal. "We want to raise taxes because we can" is not popular so a cover story is often devised, "We are here to help you reduce congestion so life will be more pleasant for all. A tax is only the means to this noble goal."

This is done worldwide. Latest murmuring in the US is that we need a federal carbon tax on parking of a minimum of $5 a day or so in remote areas if you buy tax credits in bulk, more in the cities. If you park your car you will pay this tax. The explanation is that it will reduce carbon, save the planet, reduce pollution.

In practice, to avoid what will become a $20 city parking tax, many people will have their spouse drive them to town, drop them off, then take the car back to the house, doubling the miles driven.
Scott Send private email
Monday, August 16, 2010
 
 
"One success case they profile is the entry-toll for cars going into London.  Andy, I don't know if public sentiment in the UK considers that a success or not"

It made no measurable difference to congestion and was an unpopular tax, especially when it went from £5 per day, to £8, then proposed to go to £25 per day (although that was never implemented).

That is a project that was both a success and a failure. It was successful, in that it raises money and was delivered in a reasonable timescale.

However, it was a failure, in that it made no difference to congestion, although of course, it is easy to imagine a scenario where the government knew it would "fail" in this way, but just wanted to get the extra tax revenue, as well as a nice big green tick for trying to tackle environmental issues.

Another "failure" of the system is that it is largely optional. Anyone with a disabled badge can nominate several vehicles which are supposedly used to transport the disabled person and are exempt from the congestion charge. Disabled badges are relatively easy to get and most people know someone who has one, so it is pretty easy to get on the list.

Another way to be exempt is to register as a mini-cab. This is much cheaper than paying the congestion charge and explains why there are several Lamborghini and Aston Martin mini-cabs registered in London.
Scorpio Send private email
Monday, August 16, 2010
 
 
@S. Tanna
You are correct in the conflation point.

The £7bn would be good value if it delivered what was needed on time and allowed the tax payer to make meaningful cuts to the benefits budget.

The problem really is that it is easy to fall into the trap of assuming that all government IT projects are massively over-priced and fail.
Scorpio Send private email
Monday, August 16, 2010
 
 
Scott: "In practice, to avoid what will become a $20 city parking tax, many people will have their spouse drive them to town, drop them off, then take the car back to the house, doubling the miles driven."

Of course. Most taxes are, at least in part, optional.

However, the government will still get a nice green tick for this sort of scheme, and still gain some useful revenue (as not all people can avoid the tax).
Scorpio Send private email
Monday, August 16, 2010
 
 

This topic is archived. No further replies will be accepted.

Other recent topics Other recent topics
 
Powered by FogBugz