Slashdot Mirror
IT Trends In and Out of Downturn
An anonymous reader writes "Washington Post has an interesting article talking about how IT industry is changing its business models to survive (IBM: "Pay As You Save"; HP: universal printer driver; Consulting weak; Oursourcing booming), as well as how outsiders view the downturn (Merrill Lynch: it's just another bust after PC and mainframe, but the good thing is, "each 'wave' has so far represented a tenfold increase in the number of technology users."). I'm particularly interested in the outsourcing story. It might explain why IBM will benefit and other vendors like Sun Microsystem which don't have a strong service arm will suffer."
Nope.
The customer doesn't understand the technology well enough to best determine how it meets their needs. We get into big trouble around here when our "power users" get ideas about how things should be done, sell them to management and their peers, and then come to us to "implement".
It is vital that we understand the businees of the people we support, so we can determine how best to spend their limited resources. I find that people here sometimes understand how technology can be used, but they don't understand its limitations well enough to determine how best to use it. A recent example would be a construction manager here who tried to email a 37 MB word document to someone. He knows how to make complex documents, insert photos, and compose his message-- but he doesn't know he should compress his jpg images before he inserts them, and he doesn't know what a "reasonable" file size is.
To narrow down the list of things that are possible to solutions that are practical and efficient, you have to know both the technology and the business it supports.
This is a business technical term, it means that revenue growth slows, but in a secular pattern not just a cyclical downturn. He isn't arguing that the products we have now will be with us forever, but that the total revenue pie will decelerate and eventually decrease in size for all IT products. It all comes from a model businesspeople use to show how new companies/products/industries behave over time. They are introduced and only a few early adopters use their products. New features are added regularly. Plasma screens are a pretty good example of a product in this statge. Then growth begins to take off, as many people decide to start using the product. DVD players are in a growth stage. After enough people switch to the new product, fewer people are adopting it, and maturity sets in, new variations are introduced, but they don't usually spark the same excitement and rarely get new users, just people switching within the product. Televisions and PCs are here. Finally, decline sets in as everyone has begun using the product and they simply replace broken products. Washing machines stoves and other things are in this stage. New produducts are introduced in all of thes industry groups, but mature industries usually cannot grow as quickly because most people already have their product and don't want or need additional products.
Degaussing scares the bad magnetism out of the monitor and fills it with good karma.
IMHO knowing the business is key. I'm a scientist working in a small high-tech manufacturing company and don't do IT for a living, but I dabble. We have one in-house IT person to handle everything from help-desk to maintaining our ERP system for ~250 people. I take care of the linux boxen and generaly act as a sounding board for the IT guy when stuff ain't working...
When stuff really hits the fan, everyone thinks that their particular issue is priority #1. One of the roles of a well-trained IT staff is to sort out those priorities and get the comapny back to business efficiently. You really need to know how the pieces are interconnected in order to do that effectively.
Another example: my wife works for a large company that recently outsourced most, but not all, of their IT staff. Now, when she has issues with the PCs in her office/lab she no longer has a single point of contact to deal with recurring problems and ends up having to explain the issue from the beginning again with the tech-of-the-day (Can you close all your apps and reboot? Is it plugged in?)... So what happens? One of the people in her group who knows something about PCs ends up being the local "mr. fix it" and ends up doing that instead of his real job.
Yet another example: She ordered a PC to drive a lab instrument since the old one died a horrible death. It took over two months for it to show up. Why? The only authorized PC for that purpose came with a "free" RAM upgrade and a 19" monitor, and even though the PC showed up within days the RAM was delayed. So the PC and, the >$100,000 piece of equipment was idled because it only had 128M of RAM and could not be delivered without the RAM upgrade. Never mind that the old PC that ran the machine had only 32M to begin with. Oh yeah, and that 19" monitor? It wouldn't fit in the equipment rack, so they ended up repurposing it as someone's desktop monitor. Of course the people in the lab knew it wan't going to fit, and mentioned this up front but the package deal could not be unbundled.
The irony of the whole thing was that it later turned out that the PC was incompatible with the data aquistion board required to drive the instrument even though it met all the manufacturer's specifications.
How did they find out? The in-house IT person from a neighboring department dropped by for something else and mentioned that they had found the same problem with their instruments. She found them an old clunker that was up to the job, and had them up and running the next day....
So, if you add up all the time wasted in dealing with these consultants that don't know what the group does or how it fits into the big picture of the company's business, do they still save money over a dedicated in-house IT staff?
You be the judge...
BalamThis may sound a bit trollish, but I mean it more in a Devil's Advocate sort of way... Could the fact that this 'downturn' is lasting so long have anything to do with the majority of tech writers/industry pundits/whatever continuously writing articles of doom for the tech industry? I mean come on, anyone that is reading a column like that on a regular basis already KNOWS it's going to take a while for IT spending to pick back up, for companies to rediscover the value of their IT dept, from the PC techs to the UNIX security admins. It's almost like these writers are practically trolling for hits themselves, isn't it? Let's face it, reading the tech news online is just as monotonous as watching eMpTyVee or CNN.
"See, we plan ahead! That way, we never have to do anything now."
The trick is getting competent outsourcers with sound business principles.
Not easy, but they are out there.
We outsourced some stuff. Then, when it was all successful and all the powers that be decided to insource it. Here's the result...
It was outsourced for $1 million to develop, and exactly $580K/year to run. There were 2 in-house project managers assigned for 20-40% of their time. The few outages we suffered were all corrected 100%, far, far, within the timeframe terms of the contract. Upgrades and changes were T&M, a phone call away, and done on a 2 month upgrade cycle. Ongoing change budget was about $20-40K/year. The contract was renewed each year, and the project was live 5 years.
It was insourced for $4 million to develop, and has an ongoing group budged of $1.2 million/yr. It drives, we think, from about $200K/yr in Corporate (shared) infrastructure expenses. We've had full day outages, about once a month. We don't dare ask for many changes. The smallest change request takes months, is ALWAYS met with a request for $20K or more in funding, and the rollout ends up with the sites offline far in excess of anything we've ever experienced with the outsourcers.
Bottom Line
Outsourcers SHOULD be cheaper than any in-house effort. They should be dedicated to the task you seek to have done, their people should have far more experience than you can ever hope for, and have the advantage of scale. Your outsources should be "on topic" and they should be able to show you existing production along the lines of what you're trying to do.
When it comes to specifications, they should be able to help you understand and tailor their existing systems and tools to fit your needs. Anyone that demands you send them a XXX page "spec" for them to implement to the letter should be PROMPTLY discarded.
Why discarded? Because ground zero development is always best done in-house. Outsourcers can have no prior experience in such, no scale. But, they surely do bring another copy of management overhead expenses to the table.
But even if they offer an existing solution set to what you need, if they are running their business just as poorly as you are, then you get screwed. Many, like EDS et. al., burn all ends of the candle. The seek the cheapest employees (low level, high turnover) and thus are probably less bright than just pulling people off the street. They have no shame in "locking you in" then squeezing you for all you're worth. Etc.
Bad business? Yes. Do they lose customers? Yes, but there's another being born every minute. The sad part is they're poisioning the market's view of an entire industry.
Does your outsourcer have a long term or short term view? Does it serve it's employees well enough that they stay and continue to grow on topic?
Last, and not least... Are you a big enough customer to maintain influence? Are you funding the effort fairly? If you'd pay $1 million to do something in-house, you shouldn't feel all too secure in suckering someone in to do it for far less.
Very good point. I currently work at a 13 person company.
1) I am the only IT they have, in addition to being a front-level programmer. I take care of about 8 Sun/Linux boxes, and about 4 test Win2K boxes.
2) We currently oursource our laptops/mail/network support, which I think is absolutely great. That company does what I wouldn't want to touch, nor have the expertise for (firewall, main file server, exchange, remote backups, laptop hardware support, recovering crashed hard drives, stuff like that).
While they cost us about $6000/month, their role is clearly defined, they are very responsive, and do what we would need at least a two-three people here full time to do. (and we didn't even have to buy all that equipment, the $6000 amount covers leasing the hardware, and pretty much unlimited support).
For a company like ours, there just isn't the need to hire two more people who are just going to sit around. As the parent mod suggested, the economy of scale is certainly there.
"If you could only see what I've seen with your eyes..." - Roy Batty
Not at all. He started off with a really dumb idea, but managed to avoid shooting himself in the foot. For the record, the agency I worked for was outside the jurisdiction of the executive branch, so this little adventure had no impact on me personally.
The outsourcing deal was ridiculed from day one. By the time they had it narrowed down to a single bidder, nearly half of all IT in the state had managed to exempt itself. Had Rowland (a Republican) gone through with the deal, the Democrats would have never stopped talking about the long-term commitment that never produced any savings. The same smoke-and-mirrors technology that the Republicans would have used to "prove" savings would work even better in reverse, as the Democrats "prove" cost overruns.
Instead, the decision was made to outsource a few projects where the scope and risk were more manageable. As usual, common sense will prevail, as soon as all other possibilities have been exhausted.
As it happens, Rowland is up for re-election. I would never have considered myself a Rowland fan, but the Democratic opponent is worse.