Army's Huge SAP Project 'At High Risk'

itwbennett writes "The Army's $2.4 billion SAP project is delayed, over budget, and, once implemented may not even meet its original objectives, according to a recent auditors' report. For its part, the Army is less concerned with the auditors' findings about the project that will manage a $140 billion annual budget and serve nearly 80,000 users once it is complete: 'The Army believes the risks identified in this report are manageable and do not materially impact the [project's] cost and schedule,' said an official with the Assistant Secretary of the Army (Acquisition, Logistics, and Technology)."

Comments

That's what you get

[sortius_nod]

When you go with SAP.

Re:That's what you get

[Anonymous Coward]

Horseshit. That's what you get when you don't clearly define what you want, when you change requirements all the time, and when you delude yourself into thinking that SAP will work for you "out of the box."

Re:That's what you get

[PhunkySchtuff]

Exactly, the thing with a SAP rollout (or anything else of this magnitude) is that you pass the point of no return quite early into the project and then the consultants have you exactly where they want you - you can't go back now to your old system, but the new system doesn't really do what you expected it to either so as expensive as it seems, it's cheaper to keep paying more to fix the new system than it would be to migrate everything back to the old system...

Once it's all in place and working as it should, SAP can be a fantastic thing to have but getting there is _never_ as straightforward as one would be lead to believe initially.

Re:That's what you get

[amiga3D]

And when it's more important for you to set up your big contractor job after retirement than to watch out for the public's money. I've seen that time and again where officers get seduced by contractors for a big 6 figure post retirement check.

Re:That's what you get

[Anonymous Coward]

EXACTLY!

I was in the ARMY for six years and I know how it works. I have worked with (not for) SAP for six years. If you KNOW what your REAL requirements are and you define them well, implementation does not have to be hell. Unfortunately, people don't usually understand the difference between actual REQUIREMENTS and their old processes. Almost without fail the business will declare their old processes as their requirements... and try to force SAP to function almost exactly the way their old system did... which makes me wonder why they didn't just stay with their old system. I was a developer for 15 years before I worked with SAP. It is AWESOME. But no matter how awesome it is, if you don't understand you needs, implementation will be hell.

Re:That's what you get

[Anonymous Coward]

They forgot there wasn't an 'A' in front of the SAP.

Re:That's what you get

[Surt]

I'm confused why you're modded funny rather than informative.

Re:That's what you get

[sp4ni3l]

Actually there is. There is a ASAP method, not sure if it is from SAP, but i think it was. It is a method to "quickly" implement SAP. Guess they did not use it here.

Re:That's what you get

[sphealey]

> Exactly, the thing with a SAP rollout (or anything else of this
> magnitude) is that you pass the point of no return quite early
> into the project and then the consultants have you exactly
> where they want you

Here's a thought: consultants are there to do just that: train, guide, and consult (e.g. _assist_ with analyzing). The people doing the implementation should be the organization's managers (not "project managers" - managers) and the employees who will be using the system. Once a project of this type is segregated in the "IS Department" and handed over to "the consultants" it will certainly fail; the former may know what needs to be done but lack the authority to do it (hence the need for managers to be at the core) and the latter's perverse incentives and conflicts of interest are well documented.

sPh

Re:That's what you get

[Anonymous Coward]

20 competent geeks on /. could probably set up the entire system in under 1 month split between hardware and software with a budget of $30million, including paying each of them $500k for completion. Besides the savings, they'd probably have a more crypto secure system, hardware if failed would be replaced nearly off the shelf, and with

Budget $10million for overruns or as bonuses, and another $20million to hire some of them for ongoing support. That would 15% of the .4 in the 2.4 billion.

80k users? That's it? In a country of 300million people, there's probably thousands of companies with a customer base that large. Of course, the dataset for the army is probably freaking huge, but still.

Seriously, WTF. No wonder this country has gone to the shits.

Re:That's what you get

[GNious]

There is a saying, I keep hearing: "No-one ever got fired for choosing SAP"....

Re:That's what you get

[index0]

Isn't the first step some kind of feasibility study?

Re:That's what you get

[trevelyon]

Yep, never been involved in any SAP implementations but I've seen several. Not one was completed anywhere near the deadline or original budget. Additionally, none of the companies got what they thought they were going to from it (always decreased deliverables). Mind you this is a relatively small implementation pool (5 companies) but zero successes is not a good sign.

Re:That's what you get

[Anonymous Coward]

80k users but I can't tell you what they need, or how they will be using the system or what the system is really supposed to do. What now?

I'll give you all the money in the world if you can make me a black box that does exactly what I want without me telling you want it is I want it to do.

Re:That's what you get

[Tablizer]

Almost without fail the business will declare their old processes as their requirements... and try to force SAP to function almost exactly the way their old system did

I've seen this with a system in which it was decided to mirror the paper forms in structure, which also meant keeping the duplication of the paper forms. Complex algorithms were needed to find the latest version of things like addresses. It would be like asking Henry Ford to make his cars shit like a horse and gallop away twice a month when a rope broke.

Re:That's what you get

[JAlexoi]

Egh... You have no idea what an SAP consultant title constitutes, do you? Hint: There's not much actual consulting going on there.

Re:That's what you get

[sphealey]

I dunno - after my first 10 ERP implementations I sort of lost track of which system I worked on for which project/employer. Clearly you know far more than I.

sPh

Re:That's what you get

[Anonymous Coward]

Isn't "SAP project" + "High Risk" redundant?

I never hear about how SAP made a corporation so intensely profitable that they fired all the foreign contractors and replaced them with domestic workers and gave them 10% annual raises.

Instead it's things like "[Fortune 500 company]" completely shuts down for 5 days due to SAP problems.

I'd say that surely such an expensive product must provide some sort of benefits, but then again, I read Dilbert.

Re:That's what you get

[Evets]

This should be modded insightful.

Re:That's what you get

[Hognoxious]

Agreed. Very often the "requirement" is written about how they do it, rather than what the end result should be - which you can get by a different way.

Re:That's what you get

[HungryHobo]

Story I heard second hand where jeff is a friend of a family member:

There was once a small startup where the founders noticed that most of the software for handling a particular task was needlessly complex and stupidly hard to use.

So they created their own version which was apparently good and very easy to use. the few customers they did pull in were very happy with it but they couldn't seem to break into the big leagues.

They discovered that some such attempts to sell their software to big corps had been shot down by the local consultants.

They tried contacting the big consulting firms to try to find out what they considered to be wrong with their software so that they could fix whatever problem was putting off the consultants but got back useless boilerplate replies just stating that they didn't think it was suitable.

Then one night in the bar at an exhibition one of the founders (lets call him jeff) got chatting candidly to someone from one of the big consulting firms(lets call him carl).

So jeff asked carl if he'd seen their software and what he thought of it.
carl said that it was quite excellent.
jeff asked why then did carls firm recommend against their clients using it.
Carls reply was that it was simply too easy to use and too easy to set up.
If Carls company recommended a worse piece of software that was hellish to set up then they were guaranteed many many billable hours as the client would be almost guaranteed to need consulting services.

So jeff went home and his startup set to work adding a myriad of essentially useless options and made the software vastly harder to configure.

it was still the same software once it was running but now the manual was a tome rather than a pamphlet and the setup took an expert rather than an amateur.

Like magic sales went up as consultants were suddenly willing to recommend it to their clients.

Re:That's what you get

[dwhitman]

There is a saying, I keep hearing: "No-one ever got fired for choosing SAP"....

Not a developer and certainly not a SAP consultant. But I did live through implementation of SAP at a Fortune 500 manufacturing company. For all practical purposes, the CEO bet the company by committing us to implementing SAP. The project cost was something like $3e8 in a company with sales of about $3e9. You spend that kind of cash, you bloody well better recover it in improved profitability.

Politically, with those kinds of stakes, the project is going to succeed. If it isn't a success, reality will be adjusted to make it a success. The only people getting fired in an SAP implementation are those stupid enough to say that it isn't succeeding.

Re:That's what you get

[Hognoxious]

Story I heard second hand

You appear to have mistyped "totally made up".

I'll be charitable and assume it's true that such a system exists and does all you say it can.

All it would take is one small, hungry consultancy - even a bunch of mates - to start pushing it and they could undercut their competitors by a huge margin, while still coining it in.

Prove you're not a liar, on this one at least: provide a citation. Name of company and system. Real names for jeff[sic] and carl[sic].

Re:That's what you get

[HungryHobo]

consultant?

Re:That's what you get

[BooRadley]

Since you're a member of the 4-digit ID club, then you may just be old and gray enough to have survived more than 10 of them. Are you functional or technical?

Re:That's what you get

[fyngyrz]

"The Army's $2.4 billion SAP project is delayed, over budget,

...they should have planted more maple trees, no?

Seriously, when someone writes a summary, please don't use unexplained acronyms. You know we don't RTFA.

Re:That's what you get

[AK Marc]

I was the first sale on Avaya's SAP implementation. After three sets of the wrong equipment and complete failure of accurately scheduling project member time internally, our system was installed. I'd have complained, but as far as I can tell, the billing was no better and we ended up with a credit on our bill when we should have owed $120k (they billed us once for the product we ordered, and then credited us back each time when we shipped the three sets of wrong equipment back).

Given that it took them at least two years from start to the first order (and 6 months after that order, it still wasn't complete) 10 of those would take 25+ years. But maybe they aren't all shit. Just every one I've ever seen.

Re:That's what you get

[rtb61]

Let's not forget the biggest trouble makers, politicians and lobbyists all skulking about in the background distorting all the purchasing decisions and then blaming government agencies for those faults.

So corporations via lobbyists, set it all up for making maximum profits via corrupted politicians and then market the benefits of privatisation, the root cause of most of the failures.

Instead of an in house evolving system growing over generations, digital no different to analogue solution evolution, there is the big bullshit push for the external privatised one hit solution that never exists, that funky magic box solution that will do every ones job for them. The only job it does of course is pay the corporate executives bonuses, pay the lobbyists and pay the politicians, those asshats making the decisions but taking zero responsibility for them (hint - any politician saying government is bad, is the shit head making it bad).

Re:That's what you get

[kestasjk]

How do you determine your needs in an organization where the managers don't want to be involved in the move to SAP, the processes vary quite a bit across the business, no-one really understands what SAP can do (and they don't want to learn), and the current processes depend on individual Excel/Access/etc apps which can't be stamped out if people don't buy into SAP?

Can you give me a typical success story for a large, complex, diverse, IP driven enterprise?

Re:That's what you get

[Anonymous Coward]

It is fun to see a joke turning out to being the reality. Now that the name came up, I also remember reading about it years ago..

Re:That's what you get

[xouumalperxe]

Sure, and we'll also start spelling out International Business Machines. Except that International Business Machines at least is recognizable as IBM. I doubt that most people familiar with SAP would recognize Systemanalyse und Programmentwicklung.

Not surprised

[edgedmurasame]

The only people who will get something out of SAP are the consultants who get paid to "fix" it.

Re:Not surprised

[gander666]

ah, my kingdom for mod points.. As to the earlier post that requirements weren't locked down, and changing needs lead to this. I doubt that there has EVER been a SAP project that didn't have significant scope creep and redefinition midpoint. Also, two or three different phases with different consultants (early, mid, and closer) to get to a mostly functional system.

Ah, the joys of enterprise software.

Re:Not surprised

[Anonymous Coward]

I think this is standard for any Government project (Military for sure) and any SAP install.

Re:Not surprised

[halowolf]

If requirements et all aren't locked down, then frankly don't start a massive scale project. Instead, start small get something delivered and work to build it up over time when they have a better understanding of what they actually need. A bit of discipline in the programming side can work wonders in making small things big. I've done it, it isn't impossible in the least.

Re:Not surprised

[St.Creed]

If the scope of a project is big enough, it is actually impossible to nail the requirements because they will never get consistent. This is also the reason that there is a direct correlation between the size of the project and the successrate. To my knowledge, no single IT-projects over 100 million dollars has *ever* been completed within a reasonable amount of time and in range of the budget, with most of the desired features intact.

Ofcourse, even with a gazillion users there's no need to have a really complex set of requirements. Look at facebook. The problem is the amount of processes this thing has to support, which would be better served by splitting everything up and defining open interfaces so you could build things project by project.

Re:Not surprised

[Anonymous Coward]

This is the purpose of a relational database system. The system itself maintains accounting and other data which can be modified by the tool the system gives you, and also by many other tools that you or some other company writes and then connects to the database. 90% of the problem is that after you discover that the database supports a certain kind of functionality, you inevitably want to implement that functionality and plug it into this new doodad that you found online but that would save your company huge amounts of money.

Re:Not surprised

[Evets]

That makes sense for everything. But when the consultants don't know the product, the client doesn't know the product, and the sales rep doesn't know the product, it's impossible to know what the requirements should be. Mix in the fact that the consultants need projects to move forward because they need the money, and sales people need projects to grow because they need the money, and the client needs projects to move forward because their current state of operations is overwhelming and obviously inefficient at all levels of the chain from the lowest worker level to the highest executive and you get a special set of wrong.

Throw in offshore development of the base product that goes untested, requirement and development priorities in the product base that are mismanaged, and a support structure that is designed for profit and CYA instead of support and you have an SAP project.

Another money sink...

[spaceplanesfan]

Why that isn't cancelled, but Webbs telescope is? Ah, its thats the Army....
RIP US space program

Re:Another money sink...

[Anonymous Coward]

What! What is this bullshit!? I did not know that. Fuck America now, they have lost their way entirely, and reduced me to incoherent rage. That was the only Nasa thing left that I was still genuinely excited about.
I propose a campaign of civil disobedience by astronomers and their sympathisers.

Re:Another money sink...

[jank1887]

http://www.jwst.nasa.gov/

I just visited NASA Goddard (great visitor's center, take the kids) and according to the web site they're working toward a 2018 launch of the James Webb Space Telescope. Quite impressive if you can also wrangle a lab tour.

Re:Another money sink...

[filthpickle]

The latest proposed budget for NASA cancels it.

Too bad we spend so much money on wars that don't make us any safer that we can't afford science. I am so utterly disappointed in Obama...I didn't fall for any of that "change the world" stuff, but I honestly thought he would be better than this. I can't remember who said it, but this isn't my quote, 'As a President, Barack Obama makes a great Senator'.

Re:Another money sink...

[stinkbomb]

Just because NASA isn't launching anything in the next few years, doesn't mean that the US government isn't launching anything. The US space program continues.

Re:Another money sink...

[doublebackslash]

To be fair Obama asked for a raise in their budget and Congress cut it and asked for a small cut in the military budget and Congress upped it.

President don't have as much control as the talking heads may imply. On the other hand I'm sure he could have done more. Just thought it was a good thing to know, though

Government IT projects

[Stormthirst]

What is it about government IT projects that makes them go so disastrously wrong? The UK government are no better at getting it right. The MOD* procurement system was a similar mess - over budget, and didn't do what it was set out to do.

* Ministry of Defense

Re:Government IT projects

[Joce640k]

Because:

a) They always employ people with the right connections instead of the right competence

b) Because the consultants they hire know the real money comes from doing it wrong? Why make an effort to deliver on schedule and under budget when you can take your time over it and earn twice as much money in the process?

You might think I'm joking but I've sat in some of the meetings. When I arrived I was under the delusion that I was there to do some work but I was completely wrong, we were only there to kill time before going off to a nice little French restaurant somebody had discovered. My bad.

Re:Government IT projects

[MichaelSmith]

the real money comes from doing it wrong

The longer I stay in the software industry the more this fact depresses me.

Re:Government IT projects

[Anonymous Coward]

c) They still believe in Waterfall development methodology. They also believe in "fixed-price" contracts. It's the change requests that kill you. The consultants gladly build what you asked for. Then when you realize that you really didn't know what you wanted, they have you.

Re:Government IT projects

[Anonymous Coward]

Read "Catch-22" and understand that little has changed.

I am posting this as a civilian, working in IT, for the US DoD.

Re:Government IT projects

[16K Ram Pack]

Because it's not their money, and they're not spending it upon themselves.

Milton Friedman identified 4 types of spending:

• Spending your money on yourself. You'll get what you perceive is the best value.
• Spending your money on someone else (like a present). You'll be quite careful how much you spend, but perhaps less careful about what it goes on
• Spending other people's money on yourself (you're given a budget to buy a PC). You'll buy the best thing you can, but not care too much about value
• Spending other people's money on someone else (most government spending). You don't care about how much is spent, nor do you care too much about what its spent on

I've done work on government projects and seen money thrown at projects that made absolutely no sense at all. Projects that just fizzled out or never got implemented. It doesn't happen in the competitive private sector. People do a project because it makes/saves money, and then make it work.

Re:Government IT projects

[amiga3D]

Oh man. You nailed that dead center. Too bad you posted anon 'cause that was a plus 5 informative if ever I saw one.

Re:Government IT projects

[Anonymous Coward]

*Ministry of Defence

Re:Government IT projects

[owlstead]

Gosh, that's going to be the most insightful comment. The only thing I can do is ammend it a bit.

Currently, as least in Europe, every high priced contract needs to be tendered in the open. This however means that you need to know beforehand what you want. So basically they bring in the consultants at an early stage to create the business case. Then they tender the thing (after a Q&A of the possible participants). And of course price will be a big decider for who wins the tender, so each and every participant will have little flexibility in their business plan. And then they will over charge when change requests are made.

The problem with this kind of business practice is that while the practice of tendering to the lowest bidder is - in itself - a good thing, it completely breaks down if you want to do any kind of modern software development techniques. The use cases are already cast in stone when the contract is done (and even much of the design will have to be present for the participants to create any kind of realistic price). IMHO, they would be better off doing it inhouse, with a company that supplies the software people and software development practices, and an external consultancy firm that specializes in project management (and gets payed for getting that part right).

Re:Government IT projects

[cyber-vandal]

I've worked for both the UK government and the private sector and the failure of large IT projects has one thing in common. Shite external contractors who promise the earth without knowing the first thing about what's actually required. It would be far better to have people who know the area doing the work, but for some reason senior managers all seem to believe their staff are less competent than any of the external companies who all have a well-documented record of uselessness. Private Eye should be required reading for all senior executives and senior civil servants, so that they can't claim ignorance when someone like Capita lets them down.

Re:Government IT projects

[Kagetsuki]

The Waterfall model works very very well in *cultures* where it fits. In Japan for example time and time again they have used the Waterfall model successfully. Americans with a government contract? Yeah, they don't culturally fit the model and, as it's been pointed out, a government contract will make you money for as long as you can extend it in America.

There's nothing wrong with the Waterfall model - but perhaps the American government should learn no American company bidding for a contract with a Waterfall model based project is going to finish on time, under budget, or even get anything working.

Re:Government IT projects

[Anonymous Coward]

What is it about government IT projects that makes them go so disastrously wrong? The UK government are no better at getting it right. The MOD* procurement system was a similar mess - over budget, and didn't do what it was set out to do.

* Ministry of Defense

Posting as AC... the reason why these projects go so badly is due to a combination of one (or more, mostly all) of the following:

1. - They don't know what the fuck they are doing. Seriously, there is people there from the Sumerian clay tablets era, who have been in the same companies for decades never bothering to learn how to build systems. As long as they get a paycheck and claim to work for the government or a supposedly uber-tech contractor, they are happy.
2. - They watch each other backs. When they fuck up, they just get moved somewhere else. Unlike the commercial sector, in the government, there are little consequences for fucking up a project in such a colossal manner. I mean, they can't demand accountability since it will imply that the major techno-clusterfuck that exploded in their own faces is in great part their opus magnum, the pinnacle of their technical achievement.
3. - Contractors never face competition, only during the proposal phase. Once the project is granted to them, they have a near monopoly, so there are no market pressures to make shit work correctly and in time.
4. - They are for the milking. Yes, they are. They stretch the projects for years on end. Even worse, sometimes they decide to burn all the money during the initial phases (when the contractor has very little expenses) just to get the project cancelled by the government before entering the manufacturing phases (which is where contractor's costs rise up.)

All the while we are told by management that we are responsible for the quality of what we do, for the soldier, the government blah blah blah. It's a repugnant business.

Re:Government IT projects

[cyber-vandal]

It doesn't happen in the competitive private sector.

Yes it does. You just don't get to hear about it either because it's confidential or because private sector waste isn't a good story.

People do a project because it makes/saves money, and then make it work.

I have worked on many projects in the private sector and heard about plenty more where the IT director has believed what a salesman told them and ended up with an absolute disaster. What you say might be true for SMBs but big organisations are not too different to the public sector.

Re:Government IT projects

[Plugh]

This is not a fixable problem. Governments are simply not properly incentivized to be able to cope efficiently with large, complex projects. This is part of David Friedman's Machinery of Freedom, essentially an argument for stateless anarco-capitalism. The full text of the book is available as a link off the wikipedia article. Worth a read.

Re:Government IT projects

[lucm]

Huge projects usually fail because they are deadline-oriented. From there everything goes down the drain because every bump on the road will cause the project managers to either :

1) compromise on the quality of the implementation, which leads to resistance from the people in charge of maintenance because they feel that problems are offloaded to their department. This actually initiates a downward spiral in quality and collaboration areas.

2) compromise on the number of features that are delivered, which leads to resistance from the project owner and will usually cause significant side effects as the impact of missing components is underevaluated.

Both alternatives lead to more bumps in the road, and the cycle starts again.

A deadline-oriented project also places consultants in a perfect position to do whatever they want, which is usually either repeating a solution that worked in another organization (without really knowing if this is a good fit in the new one) or taking the opportunity to learn something new that they can put on their resume. Whenever someone complains they have this magic wand called "deadline" which they can use to make all the restrictions (such as design patterns or IT best practices) go away.

When the deadline becomes a goal instead of a guideline, and when an organization is ready to let consultants call the shots instead of placing them in a controlled sandbox where they can bring value without creating chaos, then the project is doomed.

Re:Government IT projects

[jd2112]

the real money comes from doing it wrong

The longer I stay in the software industry the more this fact depresses me.

If I were a little better at programming and were a LOT less honest I could get stinking rich writing enterprise software.

I take that back, my programming skills, though weak, are better than a lot of the enterprise software developers I've encountered.

Re:Government IT projects

[Bert64]

Because the system is grossly corrupt...
There are only a very small number of very large and highly bureaucratic consultancies who ever get picked to manage these projects, and they tend to have very little in the way of technical skills and a corporate culture that scares such people away.
They massively over charge, deliver extremely poor quality work safe in the knowledge that there are very few competitors all of which are equally incompetent so there's no danger of losing out.

Re:Government IT projects

[darjen]

Right on, great book. These people essentially have unlimited funds, thanks to the taxpayers. Why bother getting it right when you can just take more and more money anyway?

Re:Government IT projects

[16K Ram Pack]

I have worked on many projects in the private sector and heard about plenty more where the IT director has believed what a salesman told them and ended up with an absolute disaster.

Yeah, but the difference is that they make it work. They hold the company to the contract, have clauses about non-delivery, that sort of thing. The government will just let stuff die.

Re:Government IT projects

[SunTzuWarmaster]

I believe that it is really because the software industry is young. Most industries went though this phase for a while before discovering the truth: The REAL money comes from the next project.

Re:Government IT projects

[Anonymous Coward]

What is it about government IT projects that makes them go so disastrously wrong? The UK government are no better at getting it right. The MOD* procurement system was a similar mess - over budget, and didn't do what it was set out to do.

* Ministry of Defense

The problem is that it's a GOVERNMENT IT project. Can't wait til that obama care kicks in...

Re:Government IT projects

[SunTzuWarmaster]

Amen. "COMPANY wastes money on software that doesn't work!" is a terrible story, to which people respond "who cares". "GOVERNMENT wastes YOUR MONEY on software that doesn't work" is much better story.

Re:Government IT projects

[Anonymous Coward]

Taxes are but a very small piece of the booty, the lion's share is from inflation (that is, Treasury bills for fiat currency, which are artificially valuable dollars backed by nothing but the paper on which it is printed).

Re:Government IT projects

[JamesP]

Waterfall works if you have:

- a very clear definition of the problem, hence, the problem shouldn't be too big (think a Wordpad sized project)
- a very thorough knowledge of the technology and its limitations (project that works in several web browsers? forget it)
- almost no variation in deployment and use conditions

Not many projects fall under those categories

Re:Government IT projects

[bertok]

Well, I'm a consultant working primarily for government, and I can honestly say that those aren't the reasons why these projects fail.

Corruption in hiring is surprisingly rare in western countries, and usually involves only a small subset of the people on a large project. I've heard it's a serious problem in developing countries, but I've only seen it a few times here, and and I've only ever seen it lead to a project failure when the project team was only a handful of people.

You'd be surprised about the work ethic of consultants.

For some consulting organizations, there's so much work that they prefer to have their consultants finish projects quickly so that they can go on to other projects and hence satisfy all of their customers, not just some of them. Not turning up at all is a surefire way of losing a customer to your competition, which can go from nonexistent to serious in very little time if they suddenly start landing big projects.

More commonly, consulting is only a part of what a company does. Large vendors like SAP sell licenses, support contracts, and consulting separately. If one branch of the company starts annoying the customers (too expensive, slow, incompetent, etc...) then this drags down the results of other branches too, and then their executives become very angry and complain directly to the CEO. God help the consultant working for an organization that makes most of their profit from licenses if the fuck up a sales deal!

Lastly, consulting firms tend to hire better-than-average people, and those tend to be high achievers and motivated professionals. Delivering projects on time and on budget looks good on a CV, and can lead to even more lucrative positions.

I've seen enough projects that I've figured out that government contracts go over budget or time for several inter-related reasons:

- Ridiculous levels of risk aversion -- if there's no bonus or profit to be had, then no risk is worth it. This leads to some very stupid decisions, over-engineering, etc...
- Management overhead -- big bureaucracies ignore the cost of management overhead, because the only way to reduce it is to fire a bunch of managers, but management makes hiring and firing decisions! Almost nobody would ever fire themselves. Instead, managers rationalize the need for management. There's no arguing with people about useless processes, when the existence of that process, useful or not, keeps them employed.
- Conservative approach to IT -- a big project is hard enough, but when you also have to deal with decades old software and sometimes even hardware, the difficulty becomes astronomical. In quite recent times, I've come across all sorts of fun things in the core infrastructure of large organizations. For example, OS/2 is still in use. Novell NetWare refuses to die. I've seen Windows 95 as a server in a data center just recently. I did a lot of work on an enterprise DOS application just a couple of years ago. It's not just systems, but processes to. Why change anything, just because the software is completely different, and the hardware is six orders of magnitude bigger or faster?

So imagine being the consultant hired to rip up and replace 10s of millions of lines of code across hundreds of undocumented systems, most of which should have been cleansed with purifying fire decades ago, but you're not allowed to. Instead, you have to sit patiently through a never ending series of pointless meetings that serve only to prevent any bureaucrat from ever having to make a decision, or take any blame for anything.

Re:Government IT projects

[Hognoxious]

For suitably small values of "work", and at a cost of a never ending death-march of fixes, workarounds and bodges.

Been there and seen it. I haven't actually done it, but I've cleared up the mess (or at least tried to) after other people have.

Re:Government IT projects

[Anonymous Coward]

"I've worked for both the UK government and the private sector and the failure of large IT projects has one thing in common. Shite external contractors who promise the earth without knowing the first thing about what's actually required. It would be far better to have people who know the area doing the work, but for some reason senior managers all seem to believe their staff are less competent than any of the external companies who all have a well-documented record of uselessness."

One problem is that managers don't get the essential difference between things like custom IT and things like say, printing the business cards and the car fleet. You can outsource running the car fleet. You find a company doing the same things for dozens of other companies, they all want the same things, which means you have scalability, and therefore, it's cheaper to push your car fleet management out.

IT is like hairdressing. Most hairdressing companies are small, because it doesn't have economies of scale. If you create a hairdressing business that does 10000 heads a day, it would still cost as much per head as a hairdressing business that does 100 per day.

I've worked for one of the large cuntsultancies and worked with at least 3 of them. And the thing is that in terms of process, they don't scale either. You work on a site with a manager and a cocksucker account manager comes along occassionally. They don't scale their operation to do things like create common code libraries or processes or whatever. So when people see them as this "large company", they're getting fed bull. You might as well hire a small company.

Worst of all, they will just put out work to wherever it's cheapest, regardless of quality control. I've seen schoolboy errors in stuff that cost £600/day and not just the odd one, but lots of them, stuff that should never have gotten through.

Really, if you are big enough, and have the work for 2 or 3 developers, you should just hire them, because all that Accenture/Crap Gemini/IBM/Crapita will do is hire those same people and add on a large wedge

Re:Government IT projects

[St.Creed]

There have been succesfull projects with waterfall methods. There have been a lot of succesfull projects with fixed price contracts: for the last 15 years I've never done business on any other basis (both as buyer and as supplier) - if you know what you are doing it's not a problem at all.

Even competence of the people involved isn't an issue. In a project that big, there's bound to be a lot of nitwits but the competent people can usually work around them.

No, what kills this thing is that even with the Gods of IT themselves on this project, once the scope exceeds a given size (you can measure that in the budget, anything over 100 million will never succeed) the amount of time needed to build and confirm that requirements have been met, exceeds the time before those requirements change. Also, the more people on the project, the larger the amount of overhead and internal communication.

Now, what agile development says is exactly what Fred Brooks says: more people on a project makes it later.

What they need to do is to cut down the scope to something manageable. They can waterfall or Scrum it to their hearts content then, but size the scope first.

Re:Government IT projects

[sgt scrub]

Because this math "a $140 billion annual budget and serve nearly 80,000 users" makes sense to the government.

Re:Government IT projects

[Anonymous Coward]

The use cases are already cast in stone when the contract is done (and even much of the design will have to be present for the participants to create any kind of realistic price).

It depends of the field. In the construction and building industry, for example, the specifications corresponding to use cases are sometimes defined so ambiguously and widely that the differences between the bids are not simply measurable with money. The government, particularly a local administration, tends to have little experience and know-how in dealing with the bidding processes, the less the more technical the bid is. The relation between the initial offering and the final price with hourly charges and such, extensions and other additional work is sometimes less than 1:10.

Re:Government IT projects

[sphealey]

> - a very clear definition of the problem, hence, the problem shouldn't be
> too big (think a Wordpad sized project)
> - a very thorough knowledge of the technology and its limitations (project
> that works in several web browsers? forget it)
> - almost no variation in deployment and use conditions
>
> Not many projects fall under those categories

And more generally: that is knowledge and wisdom that is typically only obtained by /doing the project/ - hands-on.

sPh

Re:Government IT projects

[cyber-vandal]

You really have no idea do you? Plenty of private sector projects fail too after many years of incompetence and it may shock you to know but governments hold companies to contracts, have clauses about non-delivery, incentives for early completion that sort of thing. Come back when you've actually worked somewhere large rather than talking utter bollocks.

Re:Government IT projects

[Anonymous Coward]

The German model of problem solving is said to be different from the Anglo-Saxon model particularly in the way the Germans like to design for any possible outcome and the English and by extension, the Americans like to react to the unforeseen circumstances. SAP is an item of German culture, rather than that of the English one and the Waterfall is similarly more likely suited for the Germans rather that the Americans.

There's nothing wrong with the Waterfall model

The waterfall model was represented as a failed model for software development. I think the Kaizen, or continuous improvement should be closer to an iterative software process than a waterfall one. Therefore, I'm not sure how strong the influence of the waterfall model is in modern Japan. It depends of the field I would guess.

Re:Government IT projects

[Kagetsuki]

I've worked on a lot of projects here in Japan and I can tell you with absolute confidence you are wrong. Most console games are developed under a Waterfall model for example. A lot of embedded software as well. The Waterfall model works if you have developers who actually implement according to the designs and do it well. The Waterfall works in getting products to market. I'm convinced this has a big part to do with social and developer culture in Japan.

Perhaps another factor is how the Waterfall model is implemented here. We have planners for example - they do all the planning and they deal with inconsistencies that arise during implementation. Planner work -under- designers, but planners work -for- programmers. Furthermore, the Waterfall model is just used for the core application - peripheral features and post-release enhancements [kakuchou - kaizen] are rarely handled using the Waterfall model. In addition the implementation step is usually broken up between departments with a set of joining critical-passes - basically becoming a separate process in and of itself.

Re:Government IT projects

[JamesP]

I've worked on a lot of projects here in Japan and I can tell you with absolute confidence you are wrong. Most console games are developed under a Waterfall model for example. A lot of embedded software as well. The Waterfall model works if you have developers who actually implement according to the designs and do it well.

Except you just proved me right.

Games: 20% code, 80% art. And that's true even for the first Super Mario games. And game engines are reused often.
Fixed platform (same thing for embedded sw). Working in a restricted platform is much easier in some aspects than, for example, a desktop program.

Reading the Mythical Man Month may explain it better.

The Waterfall works in getting products to market. I'm convinced this has a big part to do with social and developer culture in Japan.

Perhaps another factor is how the Waterfall model is implemented here. We have planners for example - they do all the planning and they deal with inconsistencies that arise during implementation. Planner work -under- designers, but planners work -for- programmers. Furthermore, the Waterfall model is just used for the core application - peripheral features and post-release enhancements [kakuchou - kaizen] are rarely handled using the Waterfall model. In addition the implementation step is usually broken up between departments with a set of joining critical-passes - basically becoming a separate process in and of itself.

That's very interesting. Yes, for core features Waterfall works better.

Re:Government IT projects

[Anonymous Coward]

hahaha, so you're solution is to bring it yet another consultant group? I work as an enterprise software consultant, and let me just say we try very hard to meet deadlines, budgets, and scope. Let me be the first to say that it's never easy to build systems for people that have no idea what they want but sure know what they don't want. Most people don't have a good idea of if they like something until they see functioning code, and therin lies the power of Agile development. Here's the catch though, it's extremely difficult to have a client that likes the Agile process. Everyone wants to know what they're getting for their money. Saying the equivalent of, pay us time and materials for this amount of time and we can't tell you exactly what you're getting until a few days before we start building the first demo you've agreed to some requirements drives people crazy. So you go back to Waterfall. Now as a consultant I need to get paid, and I estimate against the best requirements and wireframes I can get my hands on, and if the client changes their mind on a feature they don't usually get charged unless it adds at least 3 days of effort.

I've also worked on governement projects. The reason they go so far over budget is the same reason some private sector projects go over budget, no one wants to make firm decisions and so everything is designed by committee and therefore every last feature under the sun is included, and there is no sense of good enough. You can buy third party software, integrate it quickly and usually meet a huge list of requirements, if project owners are willing to let go of additional functionality they could save tons of money. The real kicker is that the highly custom functionality is usually totally unused because it's insisted on by someone that has no idea about user interface experience.

So you can blame the consultants if you want, and that might even make you feel good. I'm not trying to defend all of them, I've met some piss poor consultants, but I've also been called at 3 am in the morning by angry clients saying our software didn't work properly when the delivered code worked just fine and I find out the internal IT team only added "logging statements" but clearly had no idea what they were doing.

Re:Government IT projects

[jrminter]

Pretty good description of the environment in a large corporation too...

Re:Government IT projects

[JAlexoi]

a) They always employ people with the right connections instead of the right competence

And even when they do, they believe that certificates and training takes precedence to experience. In short people in government are taught to be ignorant of experience, but guess who's teaching them that? Yep, the genera populace’s distrust of people breeds such ignorance. In all fairness, the state employees are not dumb they just have to make sure that they follow the rules and generally know that experience is worth much more than the paper.

Re:Government IT projects

[JAlexoi]

Lastly, consulting firms tend to hire better-than-average people, and those tend to be high achievers and motivated professionals. Delivering projects on time and on budget looks good on a CV, and can lead to even more lucrative positions.

Yeah.... No.... Do you happen to work for IBM or Accenture? Because they definitely don't hire 100% of high achievers and they definitely don't tend to. Although IBM is the best "supplier" of freelance SAP consultants on the market(EU) at the moment.

Re:Government IT projects

[JAlexoi]

Ever heard of a budget? You know, the thing that all government agencies have. A bureaucrat most definitely cares about how much is spent, because he can't spend more.
Government agencies do IT projects because it saves the people their time. So that the average government employee can play some games or read news. In the enlightened world, it's not the IT department that orders a system to be upgraded or created, it's the actual department that needs it. IT departments is usually the one that oversees the implementation.

Re:Government IT projects

[TarPitt]

I do similar work and I'd like to add that sometimes projects are, by nature of the customer itself, impossible to accomplish successfully. Requirements may be impossible to pin down or shift randomly. Resources necessary to finish the project are unavailable. Key users may be hostile to the project and unwilling to use a system that they feel (right or wrong) is a political hammer against them.

What I have NEVER seen is a consultant or system integrator willing to tell the client blunty that they have created a management environment that guarantees project failure.

I have also NEVER seen a consultant or system integrator willing to back up this assessment by walking out of a multi-year contract, knowing the project will fail and that the client will be unhappy at the end (resulting in litigation and widespread bad publicity for both parties).

The business economics of these projects create very strong incentives for consultants to white-wash client failures, to not risk the wrath of the emporer by telling him that he is in fact naked, and to turn down what looks like a large multi-year income stream for the consultant's firm simply because everyone working on the project knows it is a disaster waiting to happen

Re:Government IT projects

[Anonymous Coward]

A bureaucrat most definitely cares about how much is spent, because he can't spend more.

And what incentive does he have to spend less?

Re:Government IT projects

[olau]

There's nothing wrong with the Waterfall model

Yes, there is. Go read up on Design of Design by Fred Brooks (known for the mythical man month and no silver bullet).

The problem with the waterfall model is that it does not allow for learning. With many IT projects, both the client and also the developers need to learn about the problem space and each other.

Re:Government IT projects

[turbidostato]

"What is it about government IT projects that makes them go so disastrously wrong?"

Who said it's only the government?

Want to know one of the most (in)famous SAP fiasco back in the day? The one from the company that wanted to put itself as *the* reference consultancy for big SAP projects on its own implementation. The verily non-government HP.

Re:Government IT projects

[turbidostato]

"Spending other people's money on someone else (most government spending)."

Well, just exactly like corporations. You don't think a general manager or a CEO is expending *his* money, do you? He is expending *your* money via your retirement funds that go to the stock market.

"Because it's not their money, and they're not spending it upon themselves."

Yes, quite exactly like corporations. No wonder their success rate for big projects is more or less that of the government.

"I've done work on government projects and seen money thrown at projects that made absolutely no sense at all. Projects that just fizzled out or never got implemented."

Yeah, exactly the same as corporations.

Re:Government IT projects

[Anonymous Coward]

Sorry buddy, Fixed-Price contracts represent the _minimum_ risk to the buyer out of all contract-line item types. Consequently, it results in higher risk for the seller - they can only operate with the amount of money obligated by the government. You're thinking of Time & Materials type work which is where the contractor is able to do work and charge for whatever time and materials they used to do the work. The latter (T&M) makes tracking spending vs deliverable deployment much more difficult. You raise good points, however!

Re:Government IT projects

[MichaelSmith]

Say you are a builder. If you make a good building for a client they will come back to you for the next one. Software is different. They will install the next one from your distribution, unless you foresaw this problem and broke the software so this couldn't happen.

Re:Government IT projects

[scamper_22]

The other big problem is government's are run by bureaucrats. When people get fed up with the bureaucrats, the only other option is to 'run' the government like a business... which means the finance people and auditors come in.

Their job of course is to quantify things in long detailed documents and then open up a bidding process.

Anyone involved in a complex project in either software or systems administration knows this is going to lead to collapse.

Now in tech companies and even many large companies, they still have IT people and developers to push back and basically say 'trust us, we'll get it done'. As silly as it sounds, that's basically what we say. That's what we're saying when we push back against endless requirements specs and endless conformance matrices... And magically, things get done.

This layer is totally missing in government. Governments demands bureaucracy and accountability. 'Trust us' doesn't cut it.

It affects every area the government touches. The most recent is healthcare. Instead of you 'trusting' your doctor as there would be trust in any other business transaction, like getting your car repaired, the government demands accountability.

Now, granted your mechanic can screw you over... and people can rip off the government, but the amount of paper work in healthcare is insane.

Doctors too have learned to game it. Just like IT consultants.

Click the right boxes, maximize your tests, say you checked for 50 things when they just came in for a cold... equal more money.

As the government is a single entity, the 'trust us' attitude just won't work. 'Trust us' only works in the private sector where failure is an option. You can have different companies provide the same service. They each trust their own staff. If their staff don't deliver, the company fails. The others are still there to do the job.

Re:Government IT projects

[Kagetsuki]

It's obvious you know absolutely nothing of what goes into game programming. Console game engines are often copied and modified, but as console game engines are highly tailored to their games the term "reused" is massively inappropriate. And it's not like we have tons of available libraries or easy memory management or anything either - things are often built from the ground up. Game programming for console systems requires an extremely high familiarity with the hardware, patches are usually out of the question so you need to get it right the first time, and are often built from the ground up. In many ways it is significantly harder to develop for a restricted platform than on the desktop. It's obvious you have absolutely no idea what you're talking about.

Re:Government IT projects

[Kagetsuki]

I didn't mean for my statement to be all encompassing. Obviously different models will be better suited for different projects and different people. That's the reason we have different models in the first place, isn't it? Choose what works for you, but don't claim a method like the Waterfall is sure to fail just because it hasn't worked for you.

Re:Government IT projects

[JamesP]

I've worked with several embedded software platforms, both with and without an OS, both with a C compiler and without a C compiler available. So, yeah, I guess I know what I'm talking about.

Game programming for console systems requires an extremely high familiarity with the hardware, patches are usually out of the question so you need to get it right the first time, and are often built from the ground up. In many ways it is significantly harder to develop for a restricted platform than on the desktop

Of course I know that. But the hardware and environment is the same always (unless of course you switch platforms). All SNESs are very similar.

It is hard, but once you get it right, it's done. And you know the hardware, and you know its limitations, and you don't have an OS to 'get in the way'. DOS programming was almost like that.

And it's obvious you haven't read the Mythical Man Month. Or understand agile.

I'm sure now the PSN was done in a waterfall way.

Re:Government IT projects

[jimnorcal]

It sounds like we need some laws that prosecute software developers and project managers for stealing from the people of the united states. Start prosecuting people for this crap and it wouldn't happen so much.

Re:Government IT projects

[Anonymous Coward]

There are many reasons why a large SAP implementation or other project may fail - I have seen this from all aspects, permanent employee, consultant, manager, team leader and buyer.
First ,and foremost, project failures almost always tend result from management failures:
It is management responsibility to ensure that any contractors working on their project have the right experience and whose CV's and references demonstrate their capabilities.
It is management responsibility to ensure they have competent project management and budget for the project.
It is management responsibility to ensure employees 'buy into' the project from early on in its life to ensure thorough scoping, budget analysis and full cooperation (lack of employee cooperation can kill a project)
It is management responsibility to track a project and to tie down the scope to ensure scope-creep doesn't drift beyond the realms of reality with regards to project delivery times and budgets.
It is senior management responsibility to ensure politicking and bickering amongst the management team itself does not lead to conflicting departments with different agendas.

ITIOEWPA?

[Anonymous Coward]

Clearly the Information Technology Investment Oversight Enhancement and Waste Prevention Act is defective. I mean, the initials don't even spell a word! ITIOEWPA? What the hell is that? Come on congress, put in some overtime and brainstorm 'til you come up with an legislative acronym that means something, and sort out this mess.

Re:ITIOEWPA?

[Hognoxious]

It's a word, but not in English. It's Greek for "sit on your arses all day and hope Germany pays your bills".

Extremely difficult

[Anonymous Coward]

This was to be expected. This is one of the first accounting systems ever where time and cost had to be exponential functions rather than fixed amounts.

Another SAP disaster

[CuteSteveJobs]

In Australia a State Government used a ridiculously expensive "off the shelf" SAP payroll solution that turned into a complete disaster. A year later and staff still aren't being paid properly. Lots of finger pointing between IBM, SAP and Corptech who is the State Government's IT corporation. They paid $40M for software that didn't work, and still doesn't work.

Take that number in. $40M. Ridiculously overpriced even if it did work, but this doesn't even do that. Payroll isn't rocket science. A few competent programmers locked away for 6 months could do better. Far too much money is thrown at so-called 'enterprise software'.
http://www.itnews.com.au/News/218348,ibm-under-fire-for-qld-health-bungle.aspx
http://www.arnnet.com.au/article/351650/ibm_says_queensland_health_sap_failure_its_fault/
http://www.zdnet.com.au/qld-health-sap-woes-lead-to-cash-advances-339302381.htm
http://www.goldcoast.com.au/article/2010/05/07/215335_gold-coast-news.html
http://www.theaustralian.com.au/australian-it/qld-health-pays-hefty-price-for-sick-payroll-system/story-e6frgakx-1225813063057
http://www.computerworld.com.au/article/351608/updated_qld_govt_blames_ibm_health_payroll_bungle/

Re:Another SAP disaster

[JamesP]

What the government should do is sue all the contractors. Return the software and ask for damages. Lawyers are good at coming with bullshit numbers.

Sue them for 100x the value they got from the gov.

Not a big surprise - Implementing SAP is hard

[mvfranz]

Many SAP projects go over budget. It is a large complicated piece of software - which many do not understand. Like any vendor, the ease of installation is over sold by SAP and confirmed by implementors wanting the work.

The biggest issue that any implementation has is under estimating how much the business process will change.

I worked on a few implementations, the biggest failure was also a success. The system did exactly what the company wanted. However, the company did not understand it's own business and thus asked for the wrong features. Once implemented the company's operations came to a stand still as they could not follow the business process as implemented and was eventually bought by their biggest competitor.

I no-longer work with SAP, but I think it is a great product - just be sure you understand what your business process is and how that process will change once implemented in SAP. Rushing a project and taking shortcuts is a surefire way to end up with a failed project. Oh, and any project that lasts as many years of this, will have had 300% turnover of the implementation staff, so plenty of time will have been lost rolling consultants on/off the project.

Re:Not a big surprise - Implementing SAP is hard

[White Flame]

But have you ever actually seen a business successfully happy with what resulted from getting in with SAP?

Probably...

[Anonymous Coward]

They probably hired a big, fat consulting company like Accenture to implement SAP for them, who's in it to get even fatter by putting as many unqualified warm bodies on a contract as possible, rather than hiring a someone who will actually run it like a project and get out... someone who is actually concerned about the customer's best interests. One would hope that the Army would know better.

SAP is a huge piece of shit

[Nicolas MONNET]

1. It's written in fucking COBOL

2. It's the vilest user interface I've ever seen. I have no idea how anyone could come up with something that bad.

3. C. O. B. O. L.

Re:SAP is a huge piece of shit

[cyber-vandal]

Gasp it's written in something that has a proven track record of working and being scalable. For fuck's sake rewrite it in Ruby on Rails immediately!

Re:SAP is a huge piece of shit

[Anonymous Coward]

And you've never done a comparison against its competitors to see TRUE crap in action. (My work picked one of those other ones)

Re:SAP is a huge piece of shit

[Hognoxious]

There actually is a French competitor. After twenty years of development and a lot of cigarettes, wine and coffee they've made a lot of progress - they've got the shortlist for the name down to "Camus", "Descartes" and "Nefaitrien".

Re:SAP is a huge piece of shit

[KingAlanI]

My experience with SAP is that some of my classes tried to introduce us to it from a user/manager perspective, so I don't know about the technical backend, but I agree about the epically shitty UI.

It's not gov't, it's SAP

[Nicolas MONNET]

It fails just as often in the private sector, the difference being that there, the client usually goes bankrupt before you hear about it.

Re:It's not gov't, it's SAP

[rtaylor]

Indeed. Worked one of SBCs failed one-bill attempts. I have no idea what the total loss was but they lost $2B buying and later selling consulting firms who were pretty much dedicated to doing that piece of work.

Re:It's not gov't, it's SAP

[JamesP]

It happens with other corporate crap, like ClearCase.

If ClearCrap wasn't crap google would use it, simple as that.

Re:It's not gov't, it's SAP

[16K Ram Pack]

all "enterprise" tools or software that you can buy are overpriced shit. That's not an anti paid-for software comment. Stuff like Visual Studio is great, but is also used in smaller companies.

The most hateful examples are enterprise helpdesk systems. It's like someone designed them that just hated users. Bugzilla's less nasty (and still nasty), but at least it's free.

Re:It's not gov't, it's SAP

[JamesP]

You're 100% right.

Talking about 'enterprise helpdesk systems', I had the displeasure of using one that made Clearcase look like the best sw in the world.

Re:It's not gov't, it's SAP

[Anonymous Coward]

True enough... Ask the lovely people who worked on the "roll out" of the SAP project for Caca-Cola at their hindquarters in Atlanta. It was an open secret, within the bowels of this multi-national juggernaut, that they were behind schedule (3 years after the start), well over budget, and had no end in-sight (or insight, if you will).

Lesson: Beware the candidate who wants your government to be run more like business.

Note: There's a reason it's called roll-out... sounds smooth, but the end is really dependent on the dough.

SAP is worse than anything M$

[Anonymous Coward]

My company recently switched to SAP... it more than doubled the time it takes us to do anything. that software is garbage, yet people keep buying into it. they have great salespeople.

Now the US army is switching to it? The military will shut down. We joke that the Germans (it is german software) are getting the world back for beating them in WWII.

You can't understand how bad this software is until you actually see it.

Re:SAP is worse than anything M$

[16K Ram Pack]

MS stuff is like swimming in a pool of golden unicorn tears compared to most "enterprise software". The reason is quite simple: a lot of small companies use MS stuff, and if it was as hideous as most "enterprise" software, people wouldn't buy it or upgrade.

In enterprises, the purchasing decisions are made by people who don't use it day-in-day-out. They look at things like the reporting module, see that's great and buy it. I worked in an organisation that dumped a working in-house change control system for a hideous IBM system.

Great product?

[Nicolas MONNET]

- The user interface is the worst of any software product I've ever used, and I'm not exaggerating.

- The user documentation is even worse

- I'm told the developer documentation is worser still, esp. if you don't speak German.

- COBOL is so fucking awesome.

- It costs a leg, an arm, your first born and your left nuts. Oh and your soul.

Re:Great product?

[mvfranz]

- The user interface is the worst of any software product I've ever used, and I'm not exaggerating.

- The user documentation is even worse

- I'm told the developer documentation is worser still, esp. if you don't speak German.

- COBOL is so fucking awesome.

- It costs a leg, an arm, your first born and your left nuts. Oh and your soul.

- I have seen worse UIs - a proprietary X11 interface for one.
- the variable names are mostly in German, but once you have programmed in it, you learn what they are - very consistent usage within and across modules. There is English documentation provided by SAP.
- SAP is NOT written in COBOL. As someone else stated, it is C kernel with ABAP for the business language.
- it is expensive, but it is cheaper now than it was 5 years ago. Most of the cost is in paying people to do the implementation, and then support contracts. Much like every other enterprise software.

Re:Great product?

[St.Creed]

- I've seen worse UI's but only in free software. Never in something paid for.
- User documentation sucks
- Database field names are a joke (5 positions? come ON!)
- If the metadata in SAP doesn't match the actual use of the field, too bad.
- Modules are by and large crap, except for the ones like FiCo that have been developed over decades.

I've just had the misfortune of working on the SLcM module, which is the bastard child of the HR module, which itself is a demon-spawned object oriented abomination, created by what must be an automated implementation of a logical model designed by someone who hasn't a clue about relational modelling.

Also, the fact that this module only recognizes SAP technical keys and treats business keys as attributes, that can be valid for several technical keys at the same time, and each having their own timeline, creates a horrendous mess if you ever want any usefull management info out of it.

Another feature is where SAP hides the fact the database contains multiple overlapping timelines for attributes: they don't show them. You can even lose student grades this way. Joy.

No, SAP implementations should stick to FiCo, without customization. That's where it is really good and reliable. For everything else I'd highly recommend looking at alternatives.

$30K per user

[Anonymous Coward]

and still overbudget! - ah to be in government...

Re:$30K per user

[cyber-vandal]

Or a big bank, or a big insurance company, or a big oil company.

Re:$30K per user

[Surt]

Well, at least in the insurance industry it looks like they're all wising up and getting off of SAP/Accenture these days.

Editors, there's no need to repeat yourselves

[Anonymous Coward]

"The Army's $2.4 billion SAP project is delayed, over budget, and, once implemented may not even meet its original objectives"

Surely "The Army's $2.4 billion SAP project is a SAP project" would have been sufficient, guys. ;)

C and ABAP

[kleinesRaedchen]

The SAP R/3 kernel is written in C. The application layer is written in ABAP and can be extended in ABAP or Java. So, the the claim with COBOL is BS.

Re:C and ABAP

[Anonymous Coward]

That said, his second point is very true. I work for McKesson and we use SAP. Evidently when we first went live on it a few years back, all the error messages that would pop up were still in GERMAN. The UI is incredibly unintuitive. I have a tech background and even I found the UI cumbersome and difficult to use. I've seen less tech-savvy coworkers learn SAP and it's a complete nightmare.

The worst part about SAP (at my company) is they rolled out a CRM piece for it. We had previously been using Vantive/Sage which, although kind of old and not the best, at least worked reasonably well. I was in the first group of people getting training for the new SAP CRM tool. I literally spent two days in the training room breaking the program and submitting error reports back to the coders of this POS. I repeated over and over to the SAP trainers, my coworkers, and my managers that this thing was NOT ready to go live and we would suffer serious productivity issues should they continue with the implementation. I work for a corporation, so obviously I was ignored, the CRM tool was rolled out, and productivity dived. Even now, two years later, the thing still suffers from glitches and is much slower than our old system.

I honestly think that the only reason SAP is implemented is because company executives see those ads for it in airports and think, "well if company X uses SAP, then is MUST be good! I need to get that for my company too. I'll be a hero and get a nice, fat bonus!"

Posted anonymously to protect my job. I can't find the link, but McKesson is the same company that slashdot reported a couple of years ago about implementing a policy on their internal social network program (that you had to accept) indicating that we, as employees, agree to not post anything about them online in any forum, social network, blog, etc.

Re:C and ABAP

[FreelanceWizard]

ABAP is sufficiently similar to COBOL that I think it'd be fair to call it a relative in the same language family.

And if you think the SAP user interface is bad, may I introduce you to BAAN or Daly & Wolcott, both of which make SAP look like god's own gift to UIs?

Re:C and ABAP

[Hognoxious]

Evidently when we first went live on it a few years back, all the error messages that would pop up were still in GERMAN.

All? Really? Either somebody had installed it incorrectly or this is a corporate legend. Or someone had set his logon defaults wrong.

The UI is incredibly unintuitive. I have a tech background and even I found the UI cumbersome and difficult to use.

It's a bit clunky, but I think incredibly unintuitive is going a bit far. Would you rather type in a command line? Do you think end users would?

It does tend to work in a "fill it in, hit send, get a response which is probably an error message" kind of way, but that betrays its origins as a mainframe application. In fact, most websites used to work like that in the pre Ajax days, but I can see why generation F might think it's not shiny enough.

Re:C and ABAP

[Anonymous Coward]

ABAP is a superset of COBOL

Re:C and ABAP

[john_chr]

Plus 1 to the parent - The back end of SAP is in fact written in it's own proprietary language: ABAP, although when you look at the code it has more than a passing resemblance to COBOL. I've called it "mutant COBOL" for years now. It grew out of a reporting script in the early years of SAP and even today every program starts with a "REPORT" keyword. Whilst SAP can be extended in JAVA and they have a massive J2EE front end for a lot of their web based applications they are moving away from Java (something to do with ORACLE maybe?) and can seerve html straight from their backend SAP servers running ABAP using their "Webdynpro for ABAP" products.

Haha

[Zedrick]

Haha, SAP.... :-/

(mod me +5 insightful, I used SAP for years and promise that I deserve it based on the short but very insightful comment/review above)

Re:Haha

[PirateSmiley]

HA SAP indeed, they deserve what they get. I have never come across software that makes work twice as complicated, slower, aggravating, frustrating and the UI has to be designed by... well, 'designed' is being generous.. and we just eventually end up doing everything again in a spreadsheet anyway.... Their sales teams must be amazing.

No surprise..

[Bert64]

Has there ever been an implementation of SAP that didn't go massively budget and fail to meet its initial goals?

Re:No surprise..

[FreelanceWizard]

My company runs SAP as its ERP system, and the project was only a little late -- but on budget and met its initial goals. We were migrating from Daly & Wolcott on an AS/400. Then again, we only have about 260 employees, and we did a fair amount of the work using our own people. We didn't just foist the whole thing off on consultants, as is most often the way.

As someone who writes integration code with ERP systems, I can say that for all the problems SAP has, it's not nearly as terrible as others. I've worked with CORRIDOR, BAAN, and Quantum Control MaxDB, and all of them are terrible, horrible monstrosities that barely work, are wildly oversold, have terrible user interfaces, are mostly undocumented or improperly documented, and are apparently designed to be as difficult to interact with as possible. Add to that stupid programming decisions (CORRIDOR uses materialized views for all DB work as opposed to stored procedures; Quantum Control loads DLLs by reading them into memory as data then jumping to their entry points, causing massive issues with DEP and weird crashes periodically) and it's amazing anyone buys these pieces of crap. By comparison, SAP is a thing of pure beauty, with its (usually) correct documentation, rock-solid stability, and actual supported interface points (RFC and IDOC).

The problem is that ERP systems all stink. SAP just happens to stink the least.

Re:No surprise..

[JAlexoi]

SAP has this brilliant option for medium sized companies: If you adjust your business process to what SAP deems appropriate then SAP implementation in your business will succeed.
And AMEN to comparing SAP to other ERP systems. SAP is somewhat good...

Re:No surprise..

[sapped]

I have been implementing SAP for 13 years (look at my nickname) now and the majority of our implementations went live at the right time and within the original budget. Of course occasionally you get a client that hires you and then proceeds to do the complete opposite of everything you tell them to do and eventually it turns into a disaster. That said, it's not unique to SAP, I have seen this problem when I was developing other kinds of software as well.

Re:No surprise..

[KingAlanI]

Yes, that's part of the idea - to try and whip a company into shape w/r/t best practices.

What's there to worry about?

[Anonymous Coward]

I'll admit that I don't know the ins and outs of government contracts, but it seems reasonable to assume that a thorough investigation will take place. If the consultants/contractors are found to be at fault then the cost overruns will come out of their pocket. Obviously the goal of all civil servants is to avoid unduly burdening tax payers. If the fault lies on the federal side of things then I'm sure that the people responsible will be dealt with accordingly.

There, nothing to worry about. It'll all be sorted out before teatime.

100% correct

[mikein08]

I spent 30+ years in IT doing administrative programming. I saw this sort of thing happen constantly. Almost all the users I ever dealt with were of the "How do I know what I want until I see what I get" persuasion. So we gave them what we thought they needed and told them to live with it. They did. If you tried to force users to define their needs as completely as possible, you'd never get out of the requirements-definition phase of a project. Never. Users have neither time nor inclination to define their needs that thoroughly. And user management assuredly isn't competent to do it, at least not at any place I ever worked. And as far as SAP goes, it's a steaming dung heap (one way to assure your continued employment is to understand your employer's SAP implementation thoroughly, because no outsider will ever be able to do so).

You think they could have learned from Navy

[Anonymous Coward]

Navy is sort of making it work after years, but everyone still hates it. I attended an acquisition class and SAP was used repeatedly as an example of what not to do, and how not to write requirements. The requirements were written along the lines of "Vendor A will integrate software B to interface with Army Supply system C" instead of "Guy in Afghanistan orders a wheel, and the order is processed and he gets his part".

At least it wasn't Salesforce.com

[Anonymous Coward]

If it was then nothing would have been done even after all this time.

Back to Sap. It has its own way of doing things. If you don't follow that methodology then you will fail miserably.
Then RFC Calls to a BAPI are handled differently to accessing it from an interactive session.
Don't talk to me about BAPI Caching sessions.
SAP BW Licensing? If you thought MS-Office & Windows was money for old rope, think again. They ain't got nothing on that. $100K per connection good enough for you?

SAP? A total piece of crap.

Business process more important than IDE

[Anonymous Coward]

If a fortune 500 company (not your web startup) can implement and maintain it's ERP processes faster with another software package I'd like to see them try.

As for everything outside of core ERP, SAP is not the best.

Wow

[brilanon]

What the fuck is SAP

Re:Wow

[sphealey]

There's this thing out there called "business", which tends to use software to assist with running its day-to-day activities. You might want to Google a bit along those lines - I hear these "business" thingies might hire some people from time to time.

sPh

Re:Wow

[brilanon]

I'm serious, what does it stand for, and why does everyone seem to assume everyone else already knows. I'm a software developer but I work for myself and I've never heard of this

Re:Wow

[brilanon]

OK, is it -this- SAP

http://www.erptips.com/Learn-SAP/What-is-SAP/SAP-Defined.asp

Or what

Re:Wow

[brilanon]

(My three replies here are in reverse order for some reason)

Re:Wow

[sphealey]

SAP develops and sells various forms of business process management software (which generally includes subsystems such as financials {General Ledger, A/P, A/R, etc}, customer order processing, manufacturing management, purchasing, warehouse management, shipping, receiving, payroll, etc), often known as ERP (originally Enterprise Resource Planning). SAP AG generally trades places from year to year with IBM, Microsoft, and Oracle as the world's largest software supplier.

sPh

Re:Wow

[sphealey]

Or you could go here:

http://www.sap.com/solutions/business-suite/erp/index.epx

sPh

By the way, my posts in no way constitute an endorsement or anti-endorsement of SAP; purely informational ;-)

Re:Wow

[brilanon]

ty

Re:Wow

[brilanon]

ty .

Too verbose...

[telekon]

The Army's $2.4 billion SAP project is delayed, over budget, and, once implemented may not even meet its original objectives, according to a recent auditors' report.

You could have just said, "The Army is engaged in deploying SAP."

The rest can be inferred by anyone familiar with an SAP roll-out.

why am I not surprised?

[KingAlanI]

SAP implementation is a mess even in the civilian world - and then throw military procurement bureaucratic issues into the mix?

SAP is German for "Our Hand In Your Wallet"

[boogahboogah]

I had the wonderful experience of seeing SAP, which was German manufacturing software at the time (1998), sold to an extremely large travel tour operator. SAP kept trying to shoehorn the tour operator into their software. Everything was an 'operation'. Why ? Because that was the only SAP function that would come close to providing the tour operator the functionality they needed. Tour pricing ? The morons ended up reading most of the pricing table each and every time they had to calculate a price.

SAP is smart, they sell upper management only, frequently in bed with consulting organizations like Accenture & such (the culprits that I saw in action at the tour company). After the years go by management is forced to continue so they don't have to acknowledge how big their f#&kup was. Don't believe me? I was there. I watched it all happen. I have talked to the culprits and the victims and the grunts.

After six (!) years this travel company finally had something which would run, pretty much, and the tour pricing was right, most of the time. About the time I left their system uptime was around 85%. All of this is after having a team of 16 programmer/analysts (and spending and spending for OJT), also after spending millions for more and more hardware because SAP is such a resource hog, They went from one big server and a few disks to a super SAN and 60 dual processor blades and lots of other hardware I can't remember at the moment. They still have a crew of 16 programmers to keep this crap system patched & running but it is not pretty by any means. Now that they have customized their system to make it actually work, they have to do partial re-writes every time SAP comes out with a new version (big upgrade $$ for SAP to be sure). I used to feel bad for them, but if management is stupid enough to believe the salesman and not do due diligence before the purchase, it's their own darn fault.

On second thought, SAP is German for "Our Shit Doesn't Stink, To Be Sure!"

why do this projects had to be so big?

[martiniturbide]

Came on, a 2.4billion dollar contract must be something almost unmanageable. Sure the salesman must be very happy by signing a contract like this, but he does not do the delivery. I'm sure that if any company signs a 2.4billion contract does not have the required resources at hand to execute it.... because you are not expecting to sign a 2.4b contract every year. ... so you had to go out and hire... so, does the people that you are hiring has the expected skill that was calculated on the 2.4b dollar business case in the contract??.. uhm...

$1.75M per soldier?

[Anonymous Coward]

That's inconceivable!

Open source

[unsolicited]

Congress should pass a regulation to use only Open source in Govt

beijing massage-beijing outcall massage

[cuiyongxia0122]

My name is cath, 26 years old, 170cm tall, 53 kg, sexy and good curve, a passionate [url=http://www.elitemassageclub.com]Chinese girl[url] with charming personality. I'm a highly trained Chinese masseuse who has 6 years experience in massage therapy,Offering best outcall [url=http://www.elitemassageclub.com]massage in Beijing[/url], sensual & tantric massage, deep tissue massagesexy stocking massage, etc.that is very therapeutic yet sensual call me:+86-13611087267 ervice detail:[url=http://www.elitemassageclub.com]beijing massage[/url] ,beijing escort,beijing erotic http://www.elitemassageclub.com/ http://www.qingdaoescortmassage.com/ http://www.beijingvirginpark.com/

ABAP=COBOL

[Anonymous Coward]

ABAP is a COBOL specific to SAP. At least that says the wikipedia page on SAP. Check.

...please only speak from direct experience.

[RogerRamjet98]

I think most of the people commenting here have no experience on the leadership side of a large, complex software project (from either side of the table, consultant or customer). Let alone, what's involved in an RFP or vendor/consultant selection process (again, from either side).

A project like this will stress the capabilities of even the most competent management and execution teams, and are simply beyond the capabilities of any one person (yes, even the armchair experts posting here).

Technology is rarely the issue (these problems get solved). Things like the internal politics and culture are among the real problems - they are hard to manage, and are always unique in each organization.

To boil this down to "SAP Sucks" or "Consultants Suck" is ridiculous and unhelpful.

Same old

[Anonymous Coward]

Requirements, requirements, requirements. The problem is assumptions are made that are wrong and then it is found out too late. The wrong people drive the process: they do not understand systems management and design. That said there should be a little chaos for discovery but have it be controllable.

Infor SXE

[Anonymous Coward]

I have never used SAP just only heard about very hard implementations in my business channel. The databases of ERP systems I have looked at are scary. One of the worst is Infor's SXE.

ABAP is not Cobol

[Nicolas MONNET]

It's a proprietary dialect of Cobol.

So much fucking better.

ABAP is Cobol

[Nicolas MONNET]

Well, it's Cobol++, but it's Cobol nonetheless.

You make a good case for coal power plants.

[Nicolas MONNET]

Also the track records you speak of includes many spectacular failures, including bankruptcies.