Slashdot Mirror
Teachers Give ERP Implementations Failing Grades
theodp writes "Nine months after the Los Angeles Unified School District launched SAP HR and Payroll as part of a larger $132M ERP rollout, LAUSD employees are still being overpaid, underpaid or going unpaid. In June, about 30,000 paychecks were issued with errors, falling somewhat short of the Mission Statement 'to effectively deliver services to meet the payroll needs of all District employees serving our students.' Meanwhile, a $17M PeopleSoft-based payroll implementation has been making life miserable for Chicago Public Schools teachers and staff since last April, including June retirees who were stiffed for more than $35M. It's been a bad computer year for CPS staff, who also had to contend with a new $60M system that wasn't up to the task of taking attendance."
It looks to me like they didn't buy enough equipment to survive peak usage, and they knew they didn't.
so it didn't work on a peak usage day, um surprise?
In my experience, this kind of thing is typical.
It's almost a rule that the more expensive the software, the more likely it is to really and truly SUCK.
It's also a rule that the bigger the company/organization/school district/whatever, the less likely it is that "technology" purchasing decisions are made by someone who actually HAS A CLUE about technology. The reason being, of course, that technology is too expensive to let the "tech" people get involved with the purchasing process.
Like I said, this is all par-for-the-course in the American corporate world. And school districts/government organizations are even WORSE.
Considering we pay her half what a BA in the business world would make because she works in education.. her quitting is not a option for my district.
"Slashdot, where telling the truth is overrated but lying is insightful."
http://csueu9.blogspot.com/2007/08/peoplesoft-no-pay-for-arizona-state.html
"The move to PeopleSoft at Arizona State has left hundreds of employees high and dry with smaller or empty paychecks. Employees are bouncing checks and having to scramble for loans to pay bills."
One of the main problems facing these ERP systems is that they try to be far too generic. Site-specific functionality is jammed into the overall framework in the form of modules. Unfortunately, business logic is often very complex, and so it doesn't always fit well into these modules. This can lead directly to hackish attempts to circumvent the limitations imposed by the ERP modules system, which often leads directly to faulty software.
Another problem affecting lower-end ERP solutions is the use of data abstraction layers like Hibernate. These layers separate the application developers from the databases that are actually storing the data being manipulated by the ERP system. Since the developers tend to now avoid writing SQL statements, they lose sight of the real relationships between the data stored within the database tables. Losing sight of these relationships means that the developers often take obtuse, roundabout ways to getting to data through the data abstraction layer, when the same data could be obtained in a few lines of SQL.
Underpaid Teachers want to fight!
Underpaid Teachers sent out Empty Wallet!
Go! Meowth!
Foe's Empty Wallet used Debit Card!
Foe's Debit Card is overdrawn!
Meowth used Payday!
Picked up $$$!
Foe's Empty Wallet's Absorb Taxes absorbed the money!
Giovanni: The teachers have been paid. Now, I shall take over the world!
PeopleSoft 'implementations' have been making life miserable for those of us at UW (University of Waterloo) for years now. I'm guessing it's mostly due to vendor lock-in. It's not surprising to me that they're doing poorly elsewhere. Their systems are used for our co-op job system and our student information system (choose classes, view grades/transcripts, etc). Finally, as I'm about to graduate, they are using the talent we have at Waterloo to hire some co-op students to write our own system, at least for the job portion of it. Sigh.
There is another kind of evil which we must fear most, and that is the indifference of good men. -- Boondock Saints
Ummmmmm, yes. And
Where is the graceful failure? We are talking MILLIONS OF DOLLARS spent on this project. And they couldn't come up with any way to overbuild it by 10% or so?
Who really gives a fuck if 10% of your system is "sitting idle for 90-plus percent of the year," if it works perfectly EVERY SINGLE DAY?
We are talking MILLIONS OF DOLLARS and they STILL cannot accomplish that.
The OSFA software model is a problem, too, I'm sure, because it often can't be configured to do exactly what's needed for your industry. I have a customer on QuickBooks that still has to manually figure time cards because of weird union rules that it can't handle. The cost of custom software, however, makes it impractical to have something written for him.
HexaByte - he's a square and a half!
It's common to think that ERP or some other big software is going to be the silver bullet for all of your company's problems. In fact, that is just throwing money at the issue.
ERP implementations are meant to mirror existing business processes. If your business processes are ass to begin with, and there is no change before an ERP roll-out, your business will still experience the same issues.
All this "blame the ERP vendor" stuff is crap. I blame the people who are running the system and poorly implemented it.
The game.
My undergrad college rolled out a Peoplesoft based system with (IIRC) the objective of avoiding having to deal with fixing the old mainframe based setup. After a very large amount of money (which included fixing the old system anyway since the new system wasn't ready in time) we got a new system that (at least from the student side) was less attractive than the old one. I don't know all that much about the admin/teaching sides of things, but from what I saw Postgresql + PHP + better initial design considerations + a few good coders would have done wonders for a fraction of the $$. At that time we also had wind of other schools having similar trouble with Peoplesoft.
Perhaps the system got better over time, but I can't help wondering why Peoplesoft is so dominant in such situations - do people have better experiences with them they can report? My experience with it was admittedly very light (in the form of rather useless and highly non-intuitive grade reports) but if that was a sample of their standard work quality the market should be begging for competition.
"I object to doing things that computers can do." -- Olin Shivers, lispers.org
Payroll won't pay you if they have a choice.
Our school system recently made a transition from individual electronic gradebook servers per school to centralized gradebook servers serving the district. The troubles they didn't foresee in testing came from not having actual teachers around to place a realistic load on the system. Not just in the number of concurrent users, but the varying operating systems in place at schools, the varying age of equipment from room-to-room, and other factors have popped up. I'm responsible at my school site for handling people's issues with the system, but I had no part in the decision to move to a centralized server. It makes sense though, I just wish it had been set up in parallel for a while last year so that we wouldn't have all this failure to deal with that could have been anticipated.
The worst case with our gradebooks is that we get a little behind putting scores into the computer. No one's livelihood is at stake. I would hope that with something like payroll they could have tried it in parallel for a while to catch issues like the ones they're having now.
When the axe came to the forest, the trees said, "Look out - the handle was once one of us."
But it is VERY difficult to "mirror existing business processes" because of TWO things:
#1. Duplicating a human decision process is difficult in software - the human may have 50+ years of experience that s/he cannot articulate to the analyst. Which leads to
#2. The process and business logic that is actually implemented is what the ANALYST believes should be implemented and how it should be implemented. (and then how it is written by the coder and whether it passes the test cases (and whether any test cases were written and tested)).
It's all about the edge cases. Depending upon your market, your "edge cases" could be almost all of your business (and profits).
Chalk this up as yet another failed project by the IT folks. IT is more than just blindly installing software and pulling cable, and installing server. It's about load testing, performance analysis, and finding out just what your requirements are.
But then again, that's real work, and doesn't involve pulling cables, or configuring servers. IT people just don't have the discipline or training to solve these problems.
PeopleShaft not working right? Thats unpossible!
Seriously, PeopleSoft sucks fiercely unless you have an army of people spending thousands of manhours on it to make it work right. At the university I attended, when they rolled out PeopleSoft to do EVERYTHING (including tuition, enrollment, etc.) all kinds of random errors would screw up what you were trying to do, and the university's stance was "oops, sorry." This was their stance even if it meant you couldn't enroll in a class (or couldn't drop a class), or pay your tuition on time.
This is one the biggest scam in business history. You get some company to buy into a huge software package, hire armies of consultants, schedule months of meetings, and they end up with something worse than what they had before, only poorer.
From what I can tell, peoplesoft is an unorganized pile of crap. Because of peoplesoft, I'm listed in my company's payroll as the "Vice President of Communications". Hefty title, eh? They can't find where the problem is, though. For four weeks, they were giving me a full-time paycheck instead of the part-time that I was doing. It took them about a week to find and fix that problem. It's always a nice conversation starter, though.
Results ... of about 387,000 for: peoplesoft failure.
Some of those hits are irrelevant, but many lead to Peoplesoft horror stories.
I get so tired of this lame meme.
Neither SAP nor Peoplesoft suck.
Suck is sort of a generic term but when it comes to specific customer installations go I've never seen one go smoothly...ever. Never seen one come in on budget, either. The best thing I can say for either one of them is they're better than Seibel.
I have seen the reps leapfrog over the technical department to pitch the executives, glossing over the implementation and cost issues. Seen them give out customer testimonials that didn't hold up to investigation, low ball hardware requirements and suggest that the IT people were well-meaning but out of their depth.
I also disagree that it's something that couldn't be custom built for less money and deliver longer and more reliable service. Now if you mean having EDS or Dell Consulting build it for you then, yes, you're completely correct in that context.
That's our life, the big wheel of shit. - The Fat Man, Blue Tango Salvage
I don't know much about the innards of Peoplesoft, but speaking as a faculty end user, it is a steaming pile of crap. The current implementation of Peoplesoft running across all 23 campuses of the California State University system is estimated to have cost over $700 million at this point.
.pdf). After skimming through it, I couldn't believe that no CSU executives were not indicted on insider trading and corruption charges.
Just as one example, this fall students were being booted out of classes they legitimately enrolled in, because the financial aid module could not talk to the enrollment module properly, leading the system to think that these students did not pay tuition. Our department office spent the better part of the last 3 weeks manually re-enrolling everyone.
There is a state auditor's report on the CSU selection and implementation of Peoplesoft, which began back in 1997 (too lazy to link to it but Google will find you the
NO CARRIER
great fucking summary...
The headline is cool. You (are supposed to!) explain the fucking abbreviation in the first sentence. Or at fucking least somewhere in the summary...
this is fucking 101 stuff!
BURN IN HELL!
My wife is a teacher in LAUSD. Her paycheck has been screwed up on a number of occasions. She no longer knows how much she is supposed to be paid, because her salary is now different every month. The worst case was when the district deposited her check (direct deposit into the checking account), then withdrew every penny of it the next day with no warning. Why did they do this? No one has been able to explain it. The following day, they deposited the exact same amount back into the account. Even when we have the money in the account now, we feel like we can't touch it.
Since this has affected us personally, and since I'm an I.T. professional, I've been following this pretty closely. Here is some more information that wasn't talked about in the article:
"Any fool can make a rule, and any fool will mind it."
--Henry David Thoreau
I am a SAP Integrator (not on the functional side, but on technology - SAP Basis).
1. There is nothing wrong with the software or architecture design.
2. SAP is highly customizable to customer's requirements.
3. Projects are normally rushed thru without proper planning.
4. Lack of quality SAP specialists. These days, SAP consultants are commodotized.
It is difficult to identify a good consultant. It appears consultants without relevant
industry experience were deployed (SAP+Government+Education+HR background)
5. Testing, testing and testing !! I think corners were cut here.
Go identify the culprits.
I wonder how much of these pay mistakes may be directly related to the recent increase in minimum wage. It has been nearly a decade since U.S. Federal mimimum wage has increased, an eternitiy in software design. How many of these pay mistakes were for people whose pay rate is in some way tied to minimum wage ($X over minimum, etc.) I work for a small business with less than 100 employees, we use a simple comercial payroll software product, and generaly get quaterly software upgrades to deal with changing tax rates, min wage, etc. Often the update CD's will arrive only a couple of days before the end of each quarter causing a rush to install before the next pay period. Are the big players any better, and what about when IT is backlogged installing other updates.
Ike
No one in the LAUSD speaks English anymore, and the company forgot to code in a Spanish language option. Viva La Rasa! Return California to Mexico.
Some of these problems have little to do with the software itself. Rather, the process was a load of crap from day one with a zillion stupid little rules. Naturally, such rules can turn a fundamentally dead simple process into a giant hairball in an instant.
The give and take in business software should be at the boundary the old process and the new software. That is, developers should have input into the business process side of things. After all, most business programming uses procedural languages, so it's developers just might have a clue or two about the appropriate design of procedures. That doesn't mean making people work for the computer, it just means things like recognizing that a set of 20 rules can be compressed down to 3 if you don't mind the result being up to a penny off (meanwhile, the cost of implementing all the rules will cost way more than those pennys over the life of the system).
Then, of course, rather than implementing reasonable simple systems that preform related functions, they try to implement one big swiss army knife that performs all functions badly instead.
Now, add in the PHBs who want to specify languages, toolkits and technologies to be used because they heard they're magic bullets that let 4th rate code monkeys build perfect systems for petty cash and you're almost certain to have a failed project.
As for people not getting paid, that's just a beureaucratic disgrace. If you know a person should be paid, surely it's better to pay them what they got last month and adjust later than to pay them nothing. Considering that problems with a new payroll system are hardly surprising, they should be prepared for that when they cut over.
Considering we pay her half what a BA in the business world would make because she works in education.. her quitting is not a option for my district.
Yep. But why don't we say it like it really is. Gross management incompetence. I have consulted for a major college and could not believe the lack of depth in the head of I/T. Totally freaking clueless to to I/T and industry best practices. Not one molecule in his head was into I/T and being irrational and political type no hope too either. A freshman in CompSci in the same institution would be better qualified to run the department.
For they are political institutions run by politicians that set aside all the money for pet projects, or slush it to "more important needs". The head of I/T is often just a "yes" man/woman patronage appointment that never says what should be done because they don't know. I/T has no representation to the heads of the institution, as if they would care even if they had. I could crack a joke about sex and I/T "yes" but it is exemplified in government and educational institutions.
Since then I have talked to others. And similar experiences can be told from almost every educational institution around. A personification of Dilbert at best.
But they care once they have been hacked an payroll is down.
I think the approach to complex systems with lots of site-specific business logic is to provide kits and modules to write semi-custom code instead of take an existing package and bend it with extensions and attribute settings. Custom code tends to fit better, but requires a lot of nitty-gritty stuff to be written from scratch. In other words, we need kits to surround and help the custom code process, not nail custom extensions onto a somewhat rigid framework.
One could use a given framework's or library's parts for a given task, but modify them or ignore them if they're not a good fit. The framework should be a helper, not a dictator. The problem is that it is hard to sell something like this because it is hard to measure and bill for ownership of a given piece of code.
Table-ized A.I.
This brings up an important point: organizations don't bother to try to simplify their business rules. Complex business rules make life harder even IF the computer does work because people still have to verify the results and answer questions from users (paycheck receivers). Beurocrats build up layers of messy rules like a desk or fridge that nobody ever cleans. Until real AI is invented, it may be unrealistic for a computer to magically fix it all. If such a system is too complex for regular payroll clerks to understand and navigate, then it is probably beyond automation also.
Table-ized A.I.
Rush Limbough's influence on society's perceptions is amazing. He may be a manipulative ahole, but he is a clever salesmen of political philosophy unmatched by the other side.
Table-ized A.I.
I suspect they tried solving their problems with lots of cash, relied on outside consultation, didn't consult with internal key staff, and got hit with reality.
I wonder how many meetings were made with the schools' operational staff to analyze the payroll system that was in place to take care of all the factors and how much was planning was done to make sure the transition was smooth?
"Enjoy what you're doing! If it becomes drudgery, you're doing it wrong!" - Jim Butterfield
With something as complex as payroll, it may make sense to create two parallel systems, one with simplified rules. Then you do a trial period run on each, and sort by the size of the differences between them and inspect the biggest deviants. After finding the reason for the differences, make adjustments and run a trial again. Repeat this until the differences are either documented (legitimate reason found) or below the threshold. This will not eliminate all problems, but can reduce the bigger ones.
Table-ized A.I.
PFFFFFT! $132M... $17M... $60M... Bah! Nickels and dimes! Come see me and bitch when your school system's people soft implementation has cost you $800M+.
And yes we bitch that the state doesn't fund our university well enough. That we should be given more funding. When, in fact, we are given enough money. Our administrators, chancellors and trustees just choose to waste it in the most inefficient ways possible.
And don't get me started on the lack of business case. That's just S.O.P.
I will never live for sake of another man, nor ask another man to live for mine.
Now imagine if you had to work with the crap systems from Peoplesoft, Lawson, etc?
I once worked for a consulting firm that though there was going to be big bucks with Siebel. Nearly became a Siebel consultant. Fortunately the company went under before I got into that too deep.
I've been overpaid, not sure how much. Could be 900, could be twice that. Either way, the money is spent. I did spend almost 4 hours one day at the district offices only to be told they had no answer and couldn't help me. Lame! I was also told that if I was overpaid they will request it back. However I got mixed answers to how and when I'd have to repay them. SOme say they will deduct it from one full paycheck. Others say it will be taken out in smaller increments. We'll see what happens. What really sucks is those of us who have been overpaid have also paid taxes on that money.
What do unions provide anymore? Are Americans still subjected to sweatshop labor practices anymore? There are PLENTY of laws protecting the worker.
They both SUCK! I have used both and implemented both at 2 fortune 500 companies. Both were multi-million dollar efforts.
Did I mention they both SUCK? They are both so bloated and slow. The PeopleSoft HR and portal systems I worked on were just crap. They took _millions_ in hardware just to run OK. Oh, and then there is this "watchdog" service for PS that has one purpose (not according to PS), to restart the crappy processes that keep dieing in PS or.. even worse, the processes that just go crazy and start to suck up memory. We had to write our own watchdog that just killed the whole PS process tree and restarted everything. Oh, and talk about sloooow. To startup the PS portal/HR systems was just a bitch. start them, go get some coffee, get lunch, take a dump if you want, it is real slow.
To everyone out there, save your money and use some other system(s). SAP and PS just suck. PS has this crappy, bastardized VB-script-like language. It just sucks. SAP, yeah, not any better.
Brrr... bad memories... bad times....
I never understand why these folks insist on doing it themselves. There are many companies who will do payroll for them and habitually "get it right the first time." Here in Rochester, one of our fastest growing companies (Paychex) comes to mind.
These types of things can no longer happen. Why? Because projects and HR types now insist on only hiring certified project managers with PMP after their names.... mmmmmm errrrrrr.... never mind.... forget I said anything. Sorry. Carry on.
-- I ignore anonymous replies to my comments and postings.
I worked on a PeopleSoft ERP project for a large midwestern University back in the late 90s and can corroborate most of the stories here.
The project was an unmitigated disaster, yet somehow I attended these gala release events at which the managers, consultants et al. pronounced great successes (while the programmers and developers looked sheepishly around at one another because we knew the whole system was running on the digital equivalent of tin-cans and rubber-bands). The rumor-mill had it that the University had gone to the people responsible for the existing mainframe system to ask their advice. After moths of ERP due dilligence, the University's own top geeks said they'd be better off rolling their own than attempting to bend an ERP to handle the University's business logic. Nevertheless, the University promptly signed-on with PeopleSoft and hired a major consultancy to help implement the ERP.
The University was publicly funded and I waited for the impending multi-million dollar lawsuits . . . which never came.
And what does ERP mean? I checked wikipedia, but none of it's possible meanings of ERP seemed appropriate. I'm just going to pick one that sounds intestering, though.
I don't know why these schools can't pick a decent Erotic Roleplaying system. There are so many good ones to choose from, and it's very important that we get these students involved with one of them, so they can practice and know what to expect in the real world.
My mom works as a secretary at a high school in my hometown and her school system uses something like what they're describing and every time i see her she's bitching about it. Usually i just tell her about my college experiences with peoplesoft (which is a beautifully exicuted system btw) to shut her up, i think i'll give one or more of these articles to do it.
Coming to you live from another dimension.
I worked on a PeopleSoft implementation for a large US University on payroll, specifically. I cannot believe they would push a payroll implementation into production without being pretty damn sure it's going to work correctly. Payroll is not the thing you want to get wrong! I know, first hand, about the convoluted business rules and complexity that make these implementation so difficult and risky, but that's precisely why you do testing (and plenty of it). As someone else pointed out, they should have run test payroll cycles in parallel until any differences were resolved or documented as acceptable. I find it particularly appauling that Deloitte let this happen. They deserve more of the blame than the school administrators. They should know better. It is there business to know better. Presumably, that's that they're being paid for. After my University's PeopleSoft implementation they had a change of heart regarding ERP. To replace their current financial system they have adopted the "Community Source" model. Basically, it's the open-source model, but "the community includes some organizations or institutions that are committing their resources to the community, in the form of human resources or other financial elements." -- http://en.wikipedia.org/wiki/Community_source. This model seems more appropriate for all public sector enterprises versus each shelling out tax payer money to high price consultants and vendors. Take, for example, the states' Medicaid Management Information Systems. Every state has their own system, some outsourced to folks like EDS for millions. As a tax payer, I would much rather see my state pool their resources with other states, all of which have the same need, to create their own solution. That doesn't mean every state has to have the same business rules. Ideally, the "kernel" would be modular and flexible enough to meet their unique needs, but if not they can fork it. This would probably cost the states less up-front, but certainly there would be substantial savings in the long run as well.
Being out of work now for a while I applied for an "SAP HR business analysis" even though I have very little HR or SAP training. Much to my surprise the University of Texas called me in for an interview. They had 'reciently implemented the HR SAP' and are now looking for people to 'fine tune' it to make it run correctly. They could not find any trained HR types with this knowledge so they interviewed me - an IT type. One comment they made was 'not to make any interface programming changes because after they update it gets overwritten'. There exists a great lack of SAP trained HR and ERP and payroll type of IT support people. The training and education to get this knowledge is very expensive. An artificial shortage designed to keep profits up for SAP.
zenray
The problem with ERP (software), like any other big-money big-market app, is that every single commercial ERP solution is a steaming pile of shit. Massive development teams of varying (lack of) skill, textbook project management that meets deadlines at the expense of code quality, sales and training staff that earn way more than the developers, but most importantly the absolute ignorance of clients who buy into these systems without even knowning what they want out of it.
I've been involved in ERP systems big and small, and the one thing that's constant about them is they do very little and cost very much. How hard is it, really, in 2007 to manage a few lists of assets, costs and schedules ? It often takes more people to install and maintain the app, than it did for humans to do all the work on paper before the ERP got implemented. The prime difference is that once the app is launched, you give all your PHBs a login and they can run zillions of misinterpreted reports at whim. You don't gain efficiency, you gain the illusion thereof.
-Billco, Fnarg.com
As a certified bookkeeper in Belgium AND a graduated Engineer I can tell you some things...
There is no country in the world with more complex pay laws as Belgium. We're the only country in the world, to my knowlege, that still uses the archaic Napolian based distinction between those who work primarily with their hands, and those that work primarily with their brains. And that's just to get started.
I can tell you that pay calculations are really quite easy, when you compare it to most subjects related to accounting and bookkeeping. As a certified bookkeeper I've always been a "Do it right the first time" type of person. When you spend 48 hours looking for a 0,05 (or 5 euro CENTS) difference in a yearaccount, you learn to do all that is humanly possible to do it right the first time.
As I said, I'm a certified bookkeeper AND a graduated engineer, so I just wonder how they can fuck things up THAT badly. Seriously, you can make all nessecary distinctions in pay regulations easily by adding a few fields to the database that keeps track of a employee's data, and adding for example the raises linked to inflation, seniority raises, etc...
I've worked with my share of accounting and pay software, and even created some quick fix databases and spreadsheats for use in departments and smaller companies under the umbrella of the company I worked in (think 15k+ employees spread over 60+ companies under one main organisation, but working independently). I worked at the personnel department, mostly involved in pay calculations and ROI and all that basic accounting. Most of the time we just used good old AS400 and crystal reports on top of that to get proper printouts of the pay info to pass on to the finance department who actually sent the pay to the employees.
It cost almost nothing, compared to all this ERP and such and it did the job, plus we had HUMANS with proper education in charge of every company, and every department. Strangely the humans caught the errors, surprisingly (cough). When you have properly educated people and a KISS (Keep It simple and PeopleSoft, oops Stupid) type of IT system you get the best results.
Low amount of mistakes thanks to humans checking the results, (I'm just wondering why nobody blowed an alarm when they saw teachers getting no pay at all according to the system) and thanks to using KISS principles in the IT part, you get an easily adapted and properly managed IT system.
End of story, the first 2 things ANY engineer should learn are:
A; if ain't broken, don't fix it!
B; you know you're done NOT when there's nothing left to add, but when there's nothing left to take away.
So the lesson here really is that you should NEVER have pay done solely by IT sollutions, especially external IT sollutions. Get a few people who know bookkeeping/accounting, and let them check the plausibility of the paychecks. That way you cut 95% of all errors before they are sent out to employees, and you save yourself a lot of legal and financial heartache, not to mention ruining your companies image of doing things right.
And to end things in beauty, if I ever see a salesrep for any ERP or Peoplesoft or whatever, I'm putting itching powder on my "straight kick up their arse" boots.
"What really sucks is those of us who have been overpaid have also paid taxes on that money"
NOT! You are a state bureaucrat. You have never paid taxes. You spend others taxes.
Score & Karma: SASA: Slashdot Approval Seekers Anonymous