12 Steps to Beat Your Service-Provider Addiction

eastbayted writes "It starts off simply enough: Your company signs on an outside firm to help you finish an important app dev project on deadline. But then they convince you they can be of service in getting other work done at your company, and you agree. Before you know it, your organization has become far too dependent on this team of outsiders on whom you're wasting a ton of money and perhaps not getting much in the way of a return. InfoWorld has devised a 12-step program 'that can help wean you off unhealthy dependencies on service providers, consultants, and outsourcers — without having to check into the Betty Ford Clinic or make a tearful confession on Oprah.'"

Comments

Printer (and reader) friendly:

[Anonymous Coward]

http://www.infoworld.com/archives/emailPrint.jsp?R =printThis&A=/article/06/08/28/35FEservices_1.html

Use strategic open sourcing

[pieterh]

The key problem is (apart from the fact that inviting a large consultancy firm into your organisation is like inviting Tom Cruise into your marriage) that closed applications depend on a small skill pool that can be easily turned against you.

For many larger organisations, a straight-forward way to create a competitive market for services is to either open-source major systems, use existing open source applications (which is still difficult), or mandate that any new custom software must be open sourced.

For government departments, especially, this policy would improve quality and cut costs significantly, simply because anyone wishing to offer their skills would have access to the information they need.

Re:Use strategic open sourcing

[arivanov]

This is not necessary. In fact it is not enough. Both open source and close source applications can be a quagmire where developers simply go and get lost for ages. In fact making the application open source does nothing and proves nothing as far as long term maintainability. Similarly, an open source application can store its data in an absolutely nightmarish format understandable only to itself.

What matters is splitting projects and applications into small understandable modules which well defined and well documented API and make them operate on well defined data flows which are as open and easy to understand as possible.

From there on a module can be thrown out, replaced and modified at will at any particular moment with minimal fuss. Similarly, any vendor which has become too pushy can be shown the door and replaced with an alternative one.

Further onto this the first person to manage "easy" object persistence (like the Open Source Prevayler) should be quartered, skinned, boiled and the remains hanged at down. It is essential for the long term health of a module for it to store the data in a format that is understandable and accessible by third parties and not just itself. Prevayler (and the similar commercial frameworks) break this to bits. In fact it is possibly the best example for an Open Source lock-in tool I can think of.

Re:Use strategic open sourcing

[legoburner]

This is spot on correct. There have been far too many occasions where companies get so rushed to deliver they bring in outside help but do not bother getting full documentation sorted out as they did not have enough manpower to begin with, and they then move the outside help on to other projects without getting the sufficient docmentation written, potentially leaving the system in a state which is unknown to the actual employees and will require more input from the outside help if modification is required in the future.

Like most things, incompetence is to blame.

[mgblst]

The print version: http://www.infoworld.com/archives/emailPrint.jsp?R =printThis&A=/article/06/08/28/35FEservices_1.html

This comes through incompetence - it is too easy to hire outside help, and not setup an exit strategy (you listening, bush and blair?), when you don't understand the problem and won't ask for help. It is easier to get outside help than realise what you will need in the long term, and start hiring people. Oh, but when you need a new secretary, that gets done within the week.

Too many non-IT (and I am sure this happens in other departments) people are put in to manage IT infrastructure, and because they have in the past, feel the need to be making the important decisions. This is what happens.

And hire someone like IBM, and you will never get rid of them.

If people could do it themselves, they would

[petes_PoV]

Of the article's 12 steps, 2 are the reasons you got the SPs in, in the first place.

Step 5. Seek out expertise. Yes, that's a good reason to bring in external people. You don't have the skills in house and it's not cost/time effective to hire or train your own staff.

Step 8. Hire knowledge you need. Sounds pretty much like step 5 to me.

As for step 12: Give yourself over to a higher power -- your employees.

So, who's going to do their jobs while they "work side by side with the consultants"? Oh, I know. let's get more consultants in.

This article looks like it was written by the very people you're trying to get rid of. They can give you pretty prsentations and high-level bullet points. However, when you look under the covers at the substance. it all disappears.

Use consultants when you have an extraordinary need, if you really have to.

Better to have them do the mundane stuff, and train you own people to do the cutting edge, interesting, high-value work....... Assuming they're good enough.

Re:If people could do it themselves, they would

[Anonymous Coward]

As a consultant, I see it all the time - companies have a really poor opinion of their employees and a slick consulting firm can easily appear to be attractive (regardless of their competence). So organisations really need to understand what consultants can do for them and why they cannot do it in house. A good consulting firm and service provider is worth their expertise and experience.

In general organisations that have trouble getting rid of consultants are really really bad.
I have been working on a six month project for two years now and we (as the consultants) are trying to find an acceptable exit strategy for ourselves. But due to staff turnover and limitations of our scope to design (XP/Win2003 infrastructure implementation) work only , we cannot get them over the edge in terms of operations procedures so that they can run with the deployment to 1000 sites.

It is frustrating because we know we don't need to be there and we are losing good personnel because they are not being challenged as they continually need to hand hold the existing and new company staff.

Understand why

[James Youngman]

Interestingly the article doesn't point out to the reader that they also need to pay attention to the reasons why the service provider got called in in the first place, any why they needed to stay so long. There's an underlying issue there (be it manpower, organisational ability, wrong executive sponsorship of projects, skills, poor control of scope creep, etc.) The underlying issue needs to be addressed or you will be back in the same situation before you know it.

Re:Understand why

[UKRevenant]

The reason they stay so long is often simply down to inertia. The "We've started so we'll finish." coupled nicely to the "Why did we pay them so much if we are not going to listen to them?".

I used to work for a company that was taken over by an asset stripping industrial conglomerate, to make sure they got the best return for their money they sent in consultants for every department. Sadly, in the engineering department we used lots of fancy computers running non-industry standard programs. So, when the consultant came to look he had no idea what the system actually was doing or what it was capable of. His recommendation was to shift to industry standards, including AutoCad, which the company did despite my best efforts. The company lost its competitive edge as soon as the standard software was put in. I had 2 of the existing suppliers (one the main application provider, the other the hardware/os support company) they both submitted very similar suggestions for the way forward. The application provider, obviously, had a vested interest. The support company had no vested interest as they expected to continue to provide support regardless of the direction chosen.

I presented my own findings and pointed out several flaws in the consultants report, I also presented the reports from the 2 other companies. The only thing I was asked about the reports was who had authorised the spend on the additional reports and when I said they had been provided for no charge, I was told the company had spent thousands on the consultant and for that reason would be going ahead with his suggestions. I resigned at that point.

In case you are interested, the consultants suggestions were implimented in full at great cost and since the old systems were decommissioned the productivity of the company has dropped, which has had the obvious outcome of reducing it in size to about 10% the size it was when I left.

On a plus note, the consultant made me move into a much more enjoyable and profitable job.

Re:Understand why

[gusmao]

For what I've observed, high management just doesn't see any underlying issue. I once have prooved to my old manager that it would be much cheaper and less riksy for the company to hire permanent employees and pay them fairly than keep spending tons of money on consulting and outsourcing for long periods of times. He replied that new employees would be seen as an increase in headcount (and consequently in expenses) as opposed to hiring a consulting firm, which is considered an investment. In other words, it would look good in the balance sheet for the stackeholders, even though the company would be losing more money.
As long stakeholders are happy and high management is getting their bonuses, there's no issue at all.

Re:Understand why

[petes_PoV]

I was told the company had spent thousands on the consultant and for that reason would be going ahead with his suggestions.

This is one of the most common reasons I have heard for going ahead with a consultants recommendations. "if we don't, the money's been wasted".

Never mind that the consultant cost (maybe) 5 grand, and the recommendations cost five hundred to implement. One of the perverse outcomes is that the more your consultant charges, the more likely their recommendations are to be accepted.

Worse, as the costs go up the harder it becomes to say "hang on, this project's not delivering the benefits". There's so much investment, that it becomes politically impossible to lose face and cancel it. In the end, the money runs out. The consultants leave. Everyone agrees that the project was a success.

If ever there was a need for "the emperors new clothes", the IT industry is it.

Oprah = Good time

[Lord Fury]

"tearful confession on Oprah."

But on a positive note, a trip to Oprah could result in an iPod, a digital camera, or a sighting of a crazed Tom Cruise.

Dr. Phil on the otherhand is just an ass.

Hey!!

[AnomalousTurd]

I work for one of these service providers. Reliance on IT outsourcing is pretty common these days for major companies. Even technology based companies. I feel in many cases that paradoxically, it is a way for the companies to get back in control of their IT which may otherwise have become a self-serving money-eating monster.

Incidentally I would think a large percentage of slashdot readers are in outsourcing.

EDS & CSC are gonna LOVE this /. article (not)

[Anonymous Coward]

South Australia seems to have an addiction to its EDS contract,
but there are more students studying IT, hopefully to take jobs
in SA Gov't, to help position itself in an EDS-free place, "any
day now"... ;-)

A EDS-story has been cirulating in recent years:

  The Adelaide Crows (Aussie Footy Team) needed a web site, &
  EDS (reportedly) won the contract, after submitting a bid
  which estimated it would take 4+ weeks and cost Au$ 32,000.

  In fact, the project took just 2 weeks... Too bad a local
  South Aussie web making business couldn't have been the
  winner, in this case.

(SA also has a "whole-of-gov't" contract with Microsoft,
that calls for penalties whenever a non-Microsoft server
is added to the gov't N/W in contract's scope, ie, for
the first time (replacing an -old- UNIX server by the
same -old- version of UNIX may not lead to a penalty).)

How do such contracts get written or won?

There are very few palms to be greased & a company like
EDS has a lot of "grease" to offer, or so we suppose...

Outsourcing Done Wrong.......Sigh

[segedunum]

Many companies get outsourcing wrong, particularly in IT, because they have have managers who don't know how things work, they've never written a line of code or put in a server in their lives and don't seem to have a learning gene inside them. I work in a company who has a contract with a financial company for a major system, and honestly, I think we're the only ones who care sometimes. Of course we're the only ones who knows how the app works. We're certainly the only ones who seems to know what's going on, and if something happens even remotely related in the general area of our system we have all sorts of people in the company instantly hanging on to our apron strings and phoning us up. I suppose it's because many boring companies, like financial ones, just can't attract the right people who can think for themselves - hence they become even more dependant on outsourcers and consultants than they otherwise would be. When you see how many of these companies operate, you can see why.

This part of the article I always worry about:

You must roll out a major enterprise app on a tight deadline and you don't have the bodies to pull it off. So you borrow some money from next year's budget and hire a global services firm to help.

This never works - ever. Managers of IT projects who don't know much about IT seem to have this incredibly bizarre idea that IT people, programmers and analysts are all interchangeable. You can drop someone from a project two months away from the deadline, bring someone else in who knows nothing about what's going on and the new person will instantly hit the ground running. They also do it again, and again, and again and again. They also equate getting bodies on the project directly with getting it done faster. If something is late and obviously a complete mess it instantly becomes a resource problem. Not that I like calling 'people' 'resources'.

I've seen it time and again. Company gets an outsourcing company and consultants in to develop a system because they don't have the people or the expertise. Said company has no real idea what the requirements are in terms that they can get over to the consultants, they have no real idea exactly what they want these consultants to do and the whole thing becomes a mess with the outsourcing company, quite rightly, creaming off whatever money they can because of the ignorance and lack of clarity from the main company. The company then starts to bitch and whine about the 'leech' outsourcer and the relationship deteriorates. Rinse and repeat the process for the next outsourcing company.

The article can be summed up thus. Fire the useless people in your company and employ good people who can define requirements well, and consequently, can lay it on the line to outside consultants exactly what they want. The consultants will then actually be much happier, because they will know what it is they've got to do - something they probably haven't had much of ;-).

Re:Outsourcing Done Wrong.......Sigh

[swillden]

This never works - ever.

I disagree. I've seen it work. The reason it hardly ever works is because (a) managers wait too long before adding staff or (b) they don't add enough staff.

If you have a project that's running behind schedule and you want to add people to catch up, you first have to realize that the new people will be a huge drain on your existing staff's time. Very high-quality people will be less of a drain, and they'll get up to speed faster, but they'll also cost more, generally. Even the best will take time to educate, and they'll be a distraction in the meantime. You also have to realize that no matter how you go about bringing them on, not all of the people you bring in are going to be good.

So, if you want to add staff to a late project in order to finish it on time, you first have to plan on the fact that doubling your staff is going to halve the team's productivity for the first few weeks, and even after the 'ramping up' period is over, your team is not going to be twice as productuve as it was. Maybe 50% more productive. So you have to add far more new staff than would be predicted by looking at the tasks remaining to be done.

A good rule of thumb is to assume that each person you add will accomplish nothing useful during the first month of work, and will cut an existing team member's volume of useful work in half. After accounting for that productivity hit, look at the time and tasks remaining, and make sure you're adding 50-100% more new developers than the project plan says you need. Unless all of the new people you get are excellent, lean toward the higher end of that range. Finally, if the planning process shows you're going to end up with more people than discrete, separable tasks, just forget it.

My current project did this successfully. We had a very good, aggressive PM and a couple of excellent team leads who realized about three months before the deadline that the current team of six was inadequate, and that all reasonable analysis showed that they were going to miss the deadline by about two weeks. A naive manager would assume that since there was 12 man-weeks more work than time, and 12 weeks remaining, he needed to add one person to make up the difference. Our PM added six more, and all of them senior, experienced developers. It took two weeks to get the new people on board, slowing the project some because existing devs had to spend time with the interviewing process. Then it took a month before the new people were really productive, during which time they slowed progress further. Six weeks after the decision was made to add staff, the team was five weeks behind schedule (they had been two weeks behind, remember). But at that point we had ~66 man-weeks of work remaining, and 72 man-weeks of developer time available (12 people times six weeks). Even that wasn't quite enough, but with the addition of a few weekends (all billable hours, increasing costs further), we finished development a week ahead of schedule, and had that time for bug fixing and cleanup before user acceptance testing started.

I've seen it work in other cases as well. The key, though, is to add lots more staff than you think you need, make sure they're competent engineers and quick learners, and do it early. It's expensive, but it can work.

From the consultant's side of things

[bhmit1]

I am one of those hated consultants, and I see things pan out three different ways:

1. Never ending project. This one usually seems pretty straight forward and then management keeps extending. When those extensions are because they see the value of me doing more, that's fine. But more often than not, it's because they can't get their own staff to pick up the new challenge. Typically that's a result of under staffing.

2. Scope creep. Essentially I'm brought in for something small, and groups are constantly adding on more tasks. When this is combined with the "never ending project" above, I basically become entrenched. I don't mind if it's interesting work, but all too often, after the first few months, I'm doing things that won't apply to any other customer and have stop growing. When I'm on a project like this for 3 times the original duration, I tend to get antsy and weight the cost to the relationship of not signing the next contract to extend. If the work stays interesting, I'm happy to be paid consulting rates for full time employment.

3. The right way. Not many people successfully do this. The thing these customers have had in common is that the staff wasn't overworked and were truly interesting in learning what I was doing and taking over. Also, the work I'm doing typically involves drawing from experience at my previous clients and vendor training. Any extensions are usually to do something above and beyond the original contract, and not to maintain what I've developed.

It's not a bad thing to be at a customer forever if you are always doing something new and doing it faster and therefore cheaper than their internal staff could have done. It's bad when they keep you there to maintain their environment, and it's bad for both the customer and for the consultant, the good consultants at least.

Re:From the consultant's side of things

[Doug Coulter]

I'm also one such. We've been darned lucky, it seems, as we've not had this sort of problem, or not much. Our longest running customer for whom we developed quite a number of products allowed us to train and affect hiring practice, in effect making us unnecessary, but still desirable == we still get calls for fiddly maintenance jobs to customize this or that for some big customer of his. Having written the code in question, we can just do it cleaner and faster than his guys, even with the training. Other customers (and some of the names would surprise you) just took the design and said "thanks, see ya" and that was that -- they needed no more. Maybe our worst customer (a nice people, however) had a government grant to develop something he had no clue about, so hired us. We thought the thing was a good thing, so we took the job. It was a visio tactile aid for deaf infants. But the customer gave us NO feedback, it seemed everything we did was perfect (even though we knew better). It might sound heavenly to get nothing but unspecific praise, but we wanted more help on specs than we got. He was just burning the grant money and taking a cut, so the more he spent on us, the bigger his cut was. When the grant ran out, he of course dumped us, not saving enough to put this nice thing into anyone's actual hands though we built circa 30 prototypes for him. BTW my consulting firm never does fixed price or contract work, period. If you can't trust the guy, what good is a contract? No fixed price job is ever bid accurately, so someone is getting ripped, and we don't want the karma from that either way. And there ARE legitimate spec changes once a project becomes better understood. That long running customer had all his engineers in complete fear, as he would let them almost finish a job, then come in and stir the pot, forcing a redesign. Dumb like a fox, this guy, as we all know Rev2 of anything is the first good one. We only work directly for CEOs and owners, maybe down the line some chief of engineering handles the day to day communications and spec writing. That way there's no doubt when it comes to check writing and so on, no BS, or what the real vision for a product was. We've also gotten praise for showing someone a product they had in mind was a waste of time, and getting it cancelled. There's always something else to do, so this is no biggie. Often the worker bees at the place are afraid to do things like this, but we get to point out stupid stuff without getting fired. Saving a customer a few million dollars on something bound to fail is considered a favor by anyone we'd want to work with anyway.

Re:Service Provider Addiction?

[m0nstr42]

That was, like, the point and stuff. All in good fun.

Re:EDS & CSC are gonna LOVE this /. article (n

[dubl-u]

How do such contracts get written or won? There are very few palms to be greased & a company like
EDS has a lot of "grease" to offer, or so we suppose...

It may not even be like that. Consider your average consumer, who is boldly manipulated by any marketing agency who can buy air time. Now consider what happens when you take the top people from that agency and put them in the room with an executive. It's like a pack of dogs on fresh meat.

A friend of a friend mine worked for a few years hot-shot company that was negotiating giant contracts with the California government. As the salesmen slipped thing after thing past, he felt an overwhelming urge to get up and move to the other side of the table because it was so unfair. Government bureaucrats had no defense against these sharks. Of course, he sat still until the urge passed. He had payments to make on his Porche, and you can't do that on a government salary...

Re:government consulting

[Marxist Hacker 42]

Some organizations are mainly management staff running pools of contractors (for example NASA). Is this the best use of Tax Payer Money?

Oregon Department of Transportation is begining to resemble that remark. Luckily we're trying not to in IT- and failing greatly, we had a 26% turnover when you include the consultants, for 2005.

Having said that- ODOT came under fire in the 1990s as the largest state agency. It's people who don't want to pay taxes who are driving this boondoggle, under the assumption that "Private Industry Can Always Do It Better", a mantra of small-government conservatives and libertarians that is a complete and utter myth to somebody like me who has worked both public and private sector. It's the stupidest thing imaginable- but I'm one of those contractors who was so valuable that I was brought in house for my third year with ODOT.