Helping IT Save Money ... and Jobs?

An anonymous reader asks: "I work in a small, overworked and understaffed IT department at a profitable business. We recently got the news that we needed to cut costs. While every penny counts, simply turning off the computers at night and saving pennies on processor cycles isn't exactly a noticeable savings. I'm curious what measures other Slashdot readers have taken to save money within their IT departments."

Comments

Every Penny Does Count

[fembots]

Yes it does. And it's even more important that at time of cost-cutting, you show the initiatives to help the company cutting costs whenever, wherever and however you can - So that your head is not on the chopping board.

If you're in a small, overworked and understaffed IT depatment, are you sure there's anything left to be cut besides offshoring? Does it always have to be cutting costs in IT? How about, for once, in other departments?

My company recently merged 3 production servers and 2 test servers into 1P and 1T, and saved 3 SQL2000 licenses (yeah, ex ex ex developers just set up their own "independant self sustain" web+data servers whenever they needed one).

Also, how about cutting the 'net costs/time spent on Slashdot?

Re:Every Penny Does Count

[Jhon]

Aside from your cute "offshoring" slam, there are other ways to save...

Keep overtime to a minimum. Do a cost analysis of overtime paid vs. off-hour staffing and consider the addition of rotating on-call time for your employees.

Keep your hardware CLEAN and read your logs! You can ID many hardware problems well before they cause downtime. Remember, when an office of 100 cant work, every hour of downtime translates to 100 hours of lost productivity.

Change from Cells to Pagers.

Don't let inkjet printers in the office AT ALL. They are a constant headache and steal more in support costs than ink.

Need new workstations? Most software packages will run fine on older (say -- 5 year old) hardware. Buy off-lease Compaq, Dell, Gateway, etc... You can get 5x the hardware for the same money with win2k licences included. It will cost you in setup time -- but if you can manage identical hardware profiles (not that difficult), set up a single machine and clone it.

Re:Every Penny Does Count

[OECD]

My company recently merged 3 production servers and 2 test servers

You might also be able to move users onto a thin client setup (like LTSP.)

Also, consider using free alternatives to licensed products, like OpenOffice.org. (Bleedin' obvious around here, but I haven't seen anybody mention it yet.)

Re:Every Penny Does Count

[Anonymous Coward]

Paid overtime? You kidding, right?

Re:Every Penny Does Count

[Anonymous Coward]

Nearly all IT workers are defined as "managemet" to get around that rule.

Re:Every Penny Does Count

[SharpNose]

When you're looking at this sort of thing, you have to make sure you know the difference between eliminating things that represent sunk cost and eliminating things that are not sunk costs yet. It's the latter that you care about; you only care about the former to the extent that they're firmly attached to the latter.

Fembots here talks about saving three SQL Server 2000 licenses; well, you don't get to cash those licenses back in or resell them, so that's a rather empty gesture, although he/she'll avoid any renewals that might be associated with the three licenses.

Some costs are per-user: desktop operating system licenses, desktop app software licenses, desktop machines, MS client access licenses. If your company has expansion plans, get rid of those costs by using Linux, Firefox, OpenOffice, etc. and inexpensive beige-box semi-disposable PCs instead of paying so much just for the letters D, E, L, and L. If you're real good at setting up application servers under Linux, you can use junkers (down to P/90) as desktop systems and your users won't know the difference. If this is a company in trouble and being able to scale up operations is one way the biz managers could solve the problem, DON'T sabotage the effort by adding on so much of your own expansion costs.

If you needed DBMS software, you were being irresponsible with your company's money if you didn't evaluate PostgreSQL to see if it would do what you needed and went with MS SQL Server or Oracle just on the basis of the name.

If I were your IT manager, I'd already be doing these things, but I'm not, so what I think you should do is listen carefully to any discusions about how the line-of-business managers might want to fix things and do your damndest to help them succeed.

Re:Every Penny Does Count

[Anonymous Coward]

Unfortunately, over here in the United States where most of the slashdot audience is from, it's not like that. If the company wants to cut costs they pay you a low salary and work you until you quit. Any small companies that pay hourly are the ones that aren't on cost cutting crusades, unless they're breaking the law.

Re:Every Penny Does Count

[General Fault]

We just cut our IT department. I would not recomend this as it makes our development staff our IT department. Paying a programmer to do IT stuff is like paying your contractor (as in home building contractor) to clean your house. Not only is it going to cost you $200 per hour to have a clean house but it is really going to piss off your contractor.

Re:Every Penny Does Count

[KungFuPenguine]

Some other advantages the OP didn't point out by merging multiple servers into one. * centralised management which translates to time saving and time is money. * energy saving. * you can sell the old servers on ebay if you don't need them and you company allows. * and as the OP noted, cost saving on software licenses (unless your are already using Linux or some free alternatives). note that although you have to fork out money to buy a more powerful server to do the job of three, it will save you in the long run. Off the top of my head, some other ways to cut cost are * buy CRT monitor instead of LCD * buy AMD CPUs instead of Intel ones * use IM instead of phones * when calling overseas, use international phone cards which are IP based instead of going through your telco. * educate your users about virus, spyware etc, teach them how to do simple things by themselves. There are always some lazy ones who prefer you to do things for them. Usually these are the managers who probably rationalise by thinking they have better things to do with their time. * run a terminal server instead of having a workstation for everyone. (e.g. Linux Terminal Server Project, www.tlsp.org) * justify every expenditure, make sure the money you spend will increase productivity. This is very important, if you gotta spend, you gotta spend, you gotta give people the right tools for the job to increase productivity, by saving a couple of hundred bucks, you might be losing thousands of dollars in man hours and productivity.

Re:Every Penny Does Count

[Cramer]

Salaried employees don't get overtime. It's rare (from all the surveys I've seen) for IT staff to be hourly workers -- for this very reason. As an "IT guy" and knowing many more in IT, it's rare for anyone to care that they're working more than 40hr/wk. (spouses, on the other hand, complain a lot.)

Pagers are less reliable with worse coverage. And in many cases, a cell phone is simply cheaper. With a cell phone, you are talking with the person (or can be), so you instantly know if they are aware of a problem and when they'll be in a position to fix it.

Having recently repaired a 5 year old computer -- a K6-2/350 running windows 98, there's no damned way you could get any productive work done with that thing. Just browsing the web is horribly slow. God help you if you have to run any real office applications (word, outlook, access, etc.) 2-3 year old (1GHz+ processor speeds) machines might be passable if your company is flat broke, but those machines are costing the company some employee productivity.

Re:Every Penny Does Count

[Jaysyn]

Or you could do what I do, get a couple of overpaid, underworked management types fired. You'd be surprised how much overhead that can clear up. Usually doesn't take much more than a quick trip to their browser cache either.

Jaysyn

Re:Every Penny Does Count

[Syncdata]

Don't let inkjet printers in the office AT ALL. They are a constant headache and steal more in support costs than ink.

While we're talking about printing, check some of the software to see what's being printed, and how.

Where I worked, the software package by default printed a light grey background along with whatever actual data was being printed. Changing the background to white was a seemingly trivial change, but since the organization prints reams worth of paper every day, the drop in toner use/cost was extremely noticible.

Re:Every Penny Does Count

[Blakey Rat]

Pagers are less reliable with worse coverage. And in many cases, a cell phone is simply cheaper. With a cell phone, you are talking with the person (or can be), so you instantly know if they are aware of a problem and when they'll be in a position to fix it.

I don't know where you live (or what pager company you use), but here (in Western Washington State) pagers have much, much, much better coverage than cells. Not only that, but they keep working even while in parking garages or in the center of large buildings where cellphones almost always lose signal.

Pagers also have much better battery life, lasting 3-4 weeks on a single AA battery. You'll rarely miss a page because the battery is dead-- but a cell battery won't even last a full day.

Not to mention that no doubt the vast majority of your staff already carries around a cell phone. Carrying around a cell and a pager is not that weird, but carrying around two cells would be very strange.

I agree with your other points... overtime watching doesn't help with salaried employees, and 5-year-old computers are older than you think, but pagers are definately a better idea than cell phones.

Re:Every Penny Does Count

[PitaBred]

Piss him off? If I were the contractor, I'd sub-contract it to some high school kids for $8/hr, and sit on my ass laughing.

Re:Every Penny Does Count

[TWX]

"Having recently repaired a 5 year old computer -- a K6-2/350 running windows 98, there's no damned way you could get any productive work done with that thing. Just browsing the web is horribly slow. God help you if you have to run any real office applications (word, outlook, access, etc.) 2-3 year old (1GHz+ processor speeds) machines might be passable if your company is flat broke, but those machines are costing the company some employee productivity."

Okay, then go get a machine off a two year or three year lease. These can run Windows 2000 *gasp* Windows XP if it's your fancy.

By the way, running a business-normal 350MHz machine puts you squarely into Intel land. I loved the K6/2 line, but that was because for the same money I could buy a 550MHz processor when I could only buy the 350MHz or 400MHz Intel P-II. MHz for MHz the Intels were ahead. My employer has 25,000 PCs on the desktops of users and only around fifteen people to do field work on them across a hundred sites in a metro area. We have machines still out there as slow as 75MHz that are officially off the supported list, but we still support machines down to 300MHz. Take that 300MHz Intel P-II and put 512MB RAM in it and it's capable of doing all required tasks in a reasonable time. I know this because our accounting department is still using them because they're the last PCs we bought in desktop form factor cases, and they don't want towers.

What task using say, Microsoft Office 98 can not be done that can be done using Microsoft Office XP? Don't go to marketing literature to answer it, answer it off of the top of your head. Cop-out answers like "file versions are too new for it" don't count either. I want to know what actual features that real people use didn't yet exist in MS Office 98 that people depend on now in Office XP. If you can't think of any then running that computer from 1998 or 1999 with an OS dating back to when the hardware was reasonably new (NT, 98, 2000, hell even Millennium) properly security patched, updated, or secured behind proper firewalling, and a proper replacement web browser could do everything that the user needs as fast or faster than the user needs it.

I'm writing this on my 700MHz Celeron based laptop with 192MB RAM. I surf the web, check my email, write papers with a word processor, play DVDs with no hardware accelerator, work with spreadsheets, and work with a graphics editor. Yes, I have to be a bit careful with that last one, but it does just work to the point that I haven't really considered a need to buy the newest/latest/greatest other than because 192MB RAM is maxing out what this machine can handle.

My work computer was a 400MHz Celeron for a long time and it still let me use the workorder system (written with Access), use a word processor, a spreadsheet program, email, web browsing, and the like. The only reason that I got a better computer was that they offered us upgrades because we had some parts left over after a project.

Re:Every Penny Does Count

[TheOldFart]

Blah... If you want efficient cost cutting, fire the CEO and all the beancounters in between who can't see beyond the end of the quarter and stockholder's asses. These people are trashing corporations in name of quick profit on Wall Street even if it means destroying the company in the process. The CEOs are by then fully vested and long gone to the next demolition derby. It makes me sick.

Re:Every Penny Does Count

[mainfr4me]

I have to argue against the pagers-being-less-reliable point.

If you go in health care, the reason a lot of doctors and on call staff are carrying pagers instead of phones is because they are more reliable. I have witnessed and been attested to by health care IT who can get zero reception in the middle of a hospital on their T-Mobile/Sprint/Verizon/Cingular phone, yet are able to reliably recieve pages. Plus, the units are cheap, last a long time on a AAA or AA, and are fairly idiot proof.
<p>
Also, on the salary/overtime thing, one thing to give your full time salaried folk who would be losing overtime, give them possibly something else. What we did was if they are call, we gave them paid internet access (a number upgraded from dialup, and also reduced the questions coming to our helpdesk). Those who were not on call then saw an advantage to being in the rotation, which increased the space in between people being on call.

Re:Every Penny Does Count

[xenocide2]

You know, my old K6-2/500 ran Debian and win98 quite well up till I accidentally fried it. Its no number cruncher, but I never noticed any problems with it for writing papers and browsing the web. Of course, it helps to have plenty of RAM (384 or so in my case).

I worked for a while at an insurance company doing data entry. You'd be surprised how little they really need (or how much they purchased). The whole affair was done over a telnet terminal to their mainframe. Of course, I was running that terminal on a brand new Win2K computer, 1.5Ghz at the time, and it was total overkill for typing in claims and filling in form letters telling policy holders their claim was less than the deductable.

That said, windows 98 was a real bitch to maintain. I bought a new video card and the driver installation managed to nuke the drivers.vxd or whatever its called that stores the rest of the drivers(!). I've seen Windows ME machines crippled by software installed on it like weatherbug and comet cursor and the like. Most companies take imaging the hard drives and keeping saved work off of those PCs. That's where productivity can really be lost, by not fighting off those gremlins.

Re:Every Penny Does Count

[CAIMLAS]

A K6-2/350 is about 7 or 8 years old - but nice try. Seven years ago we were looking at 500Mhz stuff in that realm. In 1999 (6 years ago, in case you can't count), we had Athlon CPUs which (IIRC) started at 550MHz. In 2000 (that's 5 years ago, btw), we had 1+ GHz CPUs from both Intel and AMD.

If you don't believe me, look here.

On another note... I currently use two Pentium II 350MHz systems at work, exclusively. One runs linux, the other runs Win98. I have absolutely no problem being productive on it. It's combination of software (Win98, Office 97, and various other apps) are roughly as stable as Windows XP is with Office 2003. I don't program (maybe that's what you're talking about?) but I do a fair amount of SSH administration and things of that order.

Re:Every Penny Does Count

[defile]

Pagers also have much better battery life, lasting 3-4 weeks on a single AA battery. You'll rarely miss a page because the battery is dead-- but a cell battery won't even last a full day.

Not to mention that no doubt the vast majority of your staff already carries around a cell phone. Carrying around a cell and a pager is not that weird, but carrying around two cells would be very strange.

You mention all of these things like they're faults, but I consider them to be features of cell phones. Sporadic coverage? Bad reception inside buildings? Low battery life? Cell phones offer all of the political capital of being 24/7 reachable while still offering a million excuses for why you never answer or call back.

Re:Every Penny Does Count

[zeath]

Salaried employees don't get overtime. It's rare (from all the surveys I've seen) for IT staff to be hourly workers -- for this very reason. As an "IT guy" and knowing many more in IT, it's rare for anyone to care that they're working more than 40hr/wk. (spouses, on the other hand, complain a lot.)

I can be added to your list of people who get overtime. Yearly I get paid a little less than average salary for my experience level around here, but since I work 50+ hours a week they stay very competitive. I work at a printing company, and I am one of two people that manage the IT operations of the company as well as do typesetting and health care printing. The other guy is IT by education, and I'm a programmer by education, and between the two of us we can handle pretty much everything they throw at us. Well, except that we're slowly getting behind in our work despite the 10+ hours overtime we put in every week. God help us if this trend continues and we need to find a third person with the credentials to do everything we need.

Back on topic, being on the ground floor of a rapidly expanding company, we have the good fortune of basically have an unlimited IT budget. For example, about 2 months ago (before I was hired) they bought a brand new XServe and RAID array just to be a domain controller and do some file and print sharing. To go off on a tangent, they contracted its setup to some momo who broke it horribly (set /etc/hostname to the company's web address and until I found it completely borked my attempts to set up Samba) and actually left the first time saying he had to "go home and research" how to do what he was hired to do. When he called back a month later saying he found (read: subcontracted) someone with the knowledge, I told them not to bring him back and that I could do it in less time and certainly less money (inevitebly on overtime) since I know exactly what we need.

We can pay $1500 for a color laser jet printer and after we got it all hooked up (just a few days ago) all our boss says is "wow that looks great hey don't show anyone this they'll want to come in here and start using it". As far as management and finances goes, it's really the most absurd (and the laxest) place I've ever worked.

(Side note: 'Laxest' is a strange word. I would have assumed 'more lax' if I hadn't just looked it up)

Re:Every Penny Does Count

[kmhebert]

Well... I'VE gotten a million JILLION DAYS!

Re:Every Penny Does Count

[Cramer]

For the record, this PC (dual PIII 800) has Office 97 installed on it. And it's running XP now. Wordpad can eat just about any .doc I've been handed. However, discounting incompatible file versions is just WRONG. When a vendor, customer, or whatever sends you documents that your 8 year old software cannot open, you cannot get your work done. Finding converters, viewers, alternate applications, or asking for it in a different format takes time, often destroys content, and makes you look like an idiot.

The first problem with running software that old is simply that it isn't supported or patched anymore. Office XP includes the business contact manager, something some people seem to be unable to live without. Server based rules (outlook w/ exchange) with expanded functions. Improved security features -- like being able to turn off certain anoying crap like HTML rendering, hard blocking of certain file types (extensions) that can be a domain wide policy... Better junk mail filtering. IMAP support. Support for multiple POP3 accounts. SSL/TLS support for STMP/IMAP/POP. Support for more than one outbound SMTP server. SMTP AUTH that doesn't require M$ f'ing authentication protocol.

And, of course, version compatibility problems when sharing files... open an access db with access 2000 and then try to open it with access 97. They cannot both have the db open at the same time unless 97 opened it first. God help you if some nuts clicks OK to convert the db to access 2000; older versions will never be able to open it again. Open the Outlook rules editor from 97 when 2000 lasted updated the rules and you'll fuck up the rules.

But, you obviously don't care. Or you've not used the newer versions and realized what's been added or changed. And you've never worked anywhere where version compatibility and interoperability were an issue.

Re:Every Penny Does Count

[Cramer]

First off. every hospital I've been in has it's own FCC licensed pager system. It'll work for some distance outside the hospital. And certainly works in every little nook and dark corner inside the building. There's a real good reason your cell phone doesn't work in the middle of the hospital -- aside from the fact that you're not supposed to have one on in there... look around next time, see all that shit that's interfering with it? And the further inside you go, the more steal and concrete is hiding you from the tower(s). See, cell phone companies don't hang boosters inside hospitals like they do in malls.

Pagers are unreliable. They are passive, low-power devices. The paging tower has no way of knowing a) the pager even exists, b) is on, c) is in range, or d) received the message intact. Cell phones are bi-directional. When you call one, you know instantly if the call was answered or not and if your message was understood or not. Even the fancy bi-directional pagers give no idication to the originator that their message was received, even if the paging terminal has confirmation from the pager.

I could see the reasoning for carrying both, but that adds to the budget. The cell phone will do everything the pager will (and much more), gives you a phone to answer any "page" which you'll need anyway, and provides immediate indication of receiving power -- no bars means you'll not be getting many calls/pages. Pagers don't have those bars; and those trying to page you don't get a message saying your phone is off the network.

Re:Every Penny Does Count

[SeanJones]

My clerk always calls me the minute I sit down on the loo. It happens with such dependable regularity that I am beginning to wonder whether he has hacked a remote sensing device into my belt buckle.

Re:Every Penny Does Count

[Total_Wimp]

Actually, the medium-sized company I work at paid overtime until a couple of years ago. We were actually all salaried employees in IT, but the company seemed to understand our plight so they voluntarily gave us straight pay, based on our salary, for every hour we worked over forty. It kept a lot of good people from bailing out to startups during the dot.com boom years.

When it was cut, we moved to only forty hour paid work weeks with all overtime comped. People would keep spreadsheets of the overtime they spent patching servers or whatever and they would actually take off a day at a time or a week at a time here or their as compensation. It wasn't' quite as good as getting the extra pay, but it made the department a workable experience for people with families and almost no one left for the next couple of years too.

Recently comp time has been going through a "tightening up" phase. I personally know of several people that are looking to move on. In truth, we have some positive movement in the economy, and we really could use some fresh faces in our department, but the correlation between going to a more traditional salaried environment and people's desire to leave is not lost on me.

I think bosses around the country should keep this in mind if they need to address tough retention issues at their companies. If you treat employee's time with respect, they're much more likely to want to keep working for you. It takes a lot of time to teach a new employee the way things are done at your particular company so keeping them around for a few extra years is definitely to your benefit.

TW

A good idea..

[Anonymous Coward]

We just fired our CEO! =]

Where do you spend it?

[Barondude]

Saving money is directly tied to where you spend it.

Fire mid-level managers who don't do anything...

[Assmasher]

There's bound to be 3 or 4 of them.

Cut Users

[utopia27]

fewer users -> fewer issues -> lower costs.

if it weren't for those pesky users...

turning off computers?

[AmigaAvenger]

generally the electricity isn't going to come out of the IT budget anyway, and you probably won't be recognized for cost savings if you write up a nice proggy to automatically put your computers to sleep at a certain time, so why bother on that one...

Instead, do what businesses themselves do. diversify! If your IT department is only responsible for maintaining a users desktop, then develop an interactive web based help system that goes towards that purpose. Now your it department also has programmers, and your mission is expanded (and hopefully your budget will follow!)

Easy

[Greyfox]

Shut the IT department down completely one week out of every month, and give all the employees of the department unpaid time off during that week. However during that time, the employees of the department will be able to contract their services out to whoever needs them at whatever price they feel is fair. You'll save the company millions!

Oh there might be some outages at the company during that week. I expect that any necessary employees could be convinced to contract their services to whatever department needs them... at about 3 times their average salary...

Are you building instead of buying?

[plierhead]

Bit of an open question really. But are you doing any software development? Sometimes the big drain on the department's budget turns out to be some piece of ambitious development that would be better handled by buying outside.

Dump Microsoft

[RailGunner]

Dump Microsoft Office in favor of OpenOffice.org. Dump IE in favor of Firefox. Time = Money, and time spent cleaning up 0wned Windows boxes is expensive.

You might also consider dumping IIS for Apache if possible.

And yes, you should shut down all your PC's, as it will save money. It adds up.

Re:Dump Microsoft

[u-235-sentinel]

... time spent cleaning up 0wned Windows boxes is expensive.

Many moons ago before I saw the light and abandoned Microsoft products as a way of life, I spent an average of 15-20% of my time recovering 0wned Windows systems. Sometimes it required a complete reinstall because the system was so screwed up. I feel for anyone stuck doing that. Not to mention how boring after the Nth time it happens.

I understand that same company now runs UNIX/Linux systems and has even started testing various Linux Desktop solutions. Companies simply want out (not to mention save a few bucks). Sounds like a chance to propose a migration and build a test case.

Re:Dump Microsoft

[Gorath99]

Making Firefox the default browser is always worth it, but I wouldn't be so eager to throw away MS Office. If all it's used for is typing letters then it could well be a good idea (but then why aren't you using Works instead of Office in the first place?), but if people are using Excel, Access and PowerPoint then you're getting yourself in more trouble than you might think.

I work (part-time) at a human resources department and Excel is used all over the place. If it were replaced with Calc, then we'd have a serious problem. Lots of things that are easy in Excel are pretty hard to do in Calc, at least until you've had time to really familiarize yourself with it. Even worse is that it's a lot of work to convert Excel macros to Calc macros. Those things save a ton of work, but they're sadly not very portable. Maybe in time we'd be able to be just as productive with Calc, but the transition period would not be pretty. Like you said: time is money.

Another problem is with Impress. It's a fine program, but presentations need to look perfect and sadly the conversion from and to PowerPoint files is not yet perfect. Not a problem if all you have to deal with is other people using Impress, but you don't always have that luxury.

And finally, I wouldn't even know what to replace Access with. Granted, no hacker would touch the thing, but office workers generally aren't hackers. Access is about as complex a piece of software as you can put in front of office workers. Anything more complex and they just won't be able to use it at all.

Re:Dump Microsoft

[Theatetus]

And what should they do the first time OpenOffice doesn't open a document from an uber important customer properly?

Same thing I do when I can't open something from an uber important client using MS Office (this happens too): point him to Adobe's "make a pdf" website.

Then again, our uber-important clients don't know jack about technology and don't mind hearing stuff like that (the last Word document that wouldn't open, for instance, was messed up because the VP writing the report tried to embed some weird OLE object in another weird OLE object: we told him to stop doing crap like that, and he was fine with that)

Life Cycle

[Stanistani]

It's funny what has cycled through this category over the years:
First, we had our staff reduced and outsourced to contractors.
Second, we optimized our equipment to take up less space, electricity, and heating cost.
Third, we merged our licenses and maintenance to enterprise contracts.
Fourth, we went open-source and cut off certain high-priced vendors.
Fifth, our help-desk became voicemail, auto-respond, and Indian.
Now we're laying off the contractors and bringing the IT work all in-house again.
Why do I think that someday a pile of transistors will be doing my job?

Linux & OpenSource

[Supp0rtLinux]

I know, I know... call him a Troll. Actually, I saved my company upwards of $15,000 a year by not renewing software and support contracts for our Nokia firewalls and instead replacing them with Smoothwall (a Linux firewall). I was even able to install Smoothwall onto the old Nokia IP350 hardware.

I also avoided upgrade costs to XP for about 10 of our 50 systems. This last year, we upgraded all from Win2K to XP. However, 10 of the systems were only used by temps, contractors, and consultants and only for web browsing, webmail, etc. So we installed FC3 on them saved the almost $200 each on XP upgrade licenses.

Oh, and I save the whole company countless amounts of money when I installed Firefox and set it as the default browser. Pop-ups went away, re-installs resulting from spyware went away, etc. It saved my time (not having to do re-installs and system restores) and end-user times (not having their system unusable while I fixed them).

Re:Linux & OpenSource

[Anonymous Coward]

The title says it all. We are under the same pressure to save money. We are ditching ORACLE for PostgreSQL (the support contracts are super expensive), not to mention the outsourcing cost of the dba's to maintain. We are also going from Solaris to Debian. In general, all new apps in the works (such as CMS) are open source. We've already done the employee cut backs and, unless they go to outsourcing, can't afford to lose anyone else. The goal then is to reduce costs by eliminating support contracts and license fees. We also just don't take from the open source community. When custom additions are required that would fit the general public (such as LDAP authentication) we send the additions back to the community for them to review and place in the project (if they want).

Re:Linux & OpenSource

[TheCabal]

We saved a crapload of money by not renewing our support with Cyberguard. I always disliked their firewalls, and the one we had was expensive to maintain. All we basically used it for was NAT to the Internet- a previous CTO had lofty dreams that included some hypercomplex firewalling. Obviously those dreams never made it near reality. So we had this $20,000 NAT that we could have gone to Best Buy and picked something up for like $50. I talked the current CTO into letting the support slip on the Cyberguard, and the next time it broke (did so regularly) I'd install Smoothwall or just a vanilla install of RedHat and let IPTables sort everything out. I guess the firewall got wind of that, since it never broke after that. But it saved us a few grand in support costs.

IT is a support organization

[TedTschopp]

And as a support organization, you need to position yourselves as such. There is a certain sense that you are like the electricity or the water. Now granted if you are not in the Business of IT, perhaps the company will look at focusing on their core business and out sourcing IT. Kinda like how most businesses don't generate their own elecricity or purify their own water.

With that said, why don't you look at becoming someone who provides your business complete solutions to their problems as opposed to just keeping Server X up or Program Y debugged. Each of those things can be done by someone else for cheaper. But knowing what your company does, and how to unify business processes and computerize them is not something they can get anybody to do.

So focus on what your company does, and learn their business, and learn how computers will solve their problems. That way you might end up overseeing the group of developers over in India. Learn how to architect a solution, learn how to manage a project. These are the skills that IT needs these days.

Ted Tschopp

bandwidth contracts

[biozal]

We saved a lot of money by renegotiating bandwidth contracts with our ISPs. Moving file servers to Linux and getting rid of the "MS Tax" was also a plus...

Software Licenses

[R33MSpec]

Sorry to state the bleeding obvious but if you can find an open-source software alternative that is as good as or better than the software packages your currently using then why not propose to migrate across?

Re:Outsourced Ourselves

[karnal]

What if my native country is India?

Where do I outsource to then????

Simple, really...

[rongage]

If you want to save the company money, then quit!

Dude, the writing is on the walls as plain as day: ...profitable company...wants to cut costs...

Some bean counter is trying to squeeze as much efficiency out of you folks as possible. If I had to guess, the company is going up for sale soon and they need to make the place look as good as possible for the sale.

2nd job!

[Telastyn]

Use the spare space and bandwidth on the company web server to host porn. Use profits from porn hosting to run IT dept.

*shrug* it's a no worse idea than cutting support costs when support is already overworked. Perhaps such a message [perhaps in more businesslike terms] should be the proper reply.

Who do you want to feel it?

[tverbeek]

Most of the things I've done are either obvious (F/OSS instead of MS, refurb instead of replace, buy used instead of new), shortsighted (cut staff training, support contracts, salaries), or specific to our particular situation. So not much I can really suggest.

Depending on your political clout, it sounds like it might be time to start cutting services (evening/weekend tech support, high-speed internet, etc).

blackmail employees

[rich42]

You could have a semi-official "black operations" program where you monitor employee's web surfing behavior.

If you find out anyone's surfing for donkey porn - tell them it's time to pay up - or their boss will find out.

The money goes back into IT dept. funding so you can still buy that new videocard you need to play Half-Life 2.

Well, it may not be popular but.....

[Anonymous Coward]

Small business with 50 employees:

I fired three of the four IT people, kept the person who knew *NIX and could actually work with my other employees, replaced Wintel machines with Macs running OS X and saved myself almost $350k in the first year and $400k the following year.

Difference went back into the business in terms of reinvestment and profit sharing.

Re:Easy

[Budha_man_99]

Hey it worked for me. And since I implemented this cost cutting measure upper management has taken it one step further. I now get every week off without pay.

Re:Outsourced Ourselves

[mmkkbb]

The Phillipines, the Czech Republic, and Romania.

Read BOFH

[SirLanse]

I read Bastard Operator from Hell for Ideas.
It is available at theregister.com

Automation

[MarkEst1973]

Every day I see people clicking the same things over and over, performing the same work on their computer over and over, typing the same code over and over (in various forms, but the same patterns).

I introduced a new way of thinking in my company. Let's automate more. We don't need grunt programmer's writing easily templatable code. We need smarter senior programmers writing templates.

We don't need to have senior management people writing emails every day reminding us to fill in our timesheets on time. We need one script to send out the alert. And we don't need manually maintained spreadsheet tracking hours and contract rates. It's error-prone, time consuming, and can be better performed by a database.

Anything I see people doing repetitively, I look to automate. After all, isn't a computer nothing but an automaton doing the same thing over and over again?

I've found Python to be perfect for automating a lot of my more mundane tasks. I keep looking for that higher level of abstraction.

The problem is the GUI (*cough cough* Windows *cough cough*) where people can't seem to get around clicking. They can't seem to understand that anything they click on can be written in a script instead.

Hey, but that's just me. If I were a business owner, I'd look to get significantly more from my employees by hiring a really smart guy to automate more work.

If someone automates himself out of a job, you bet your ass I'd find him 10 more jobs to automate himself out of. That guy is worth his weight in gold.

Re:Automation

[TedTschopp]

I've saif this for years, IT should always be looking to put themselves out of a job. (Which interestingly enough is why I'm not concerned with the current fad of outsourcing.) There is always more things to bring under the control of IT automation.

Ted Tschopp

Re:Automation

[archen]

Let's automate more.

It's a good idea, but it's easily snuffed out. Where I work it's really obscene the ammount of redundant, tasks which could be automated... My first and primary block is always management - and they seem to enjoy inventing more work for everyone. Also the users often need to be on board. Right now I'm fighting tooth and nail to get important user feedback, and they just don't care. I think it's important to get user feed back to make sure that you are really automating something and making work easier, not just making the same ammount of work in a different way. When you might as well be talking to a brick wall, progress isn't made.

Last but not least, it takes time and he stated that they were understaffed/overworked. It takes planning and time. I'm getting the feeling that managment there would SAY "automation is good" but then not give you the resources to follow through. Typical I guess. They say they want to cut costs, but will probably just cut jobs no matter what anyway.

Re:Automation

[mveloso]

Yes!

The only way to save money in IT is to fire people. To fire people, you have to automate their jobs.

That's it!

If you want to be really good, go to the next step and try & figure out how to help your company make money. Maybe you can do a quick & dirty data mine on the current databases, so your purchasing can order stuff in bulk instead of having lots of small purchases.

Maybe it's looking at trends, and discovering efficiencies in the business.

Re:Sell Blood

[Cade144]

My company recently became a hosting company as a sideline.

They built this fantastic server room, with climate control, tons of backup power and all the bells and whistles, and found that they could host servers and charge for it.

Think of ways to re-sell your existing infrastructure or other overcapacity.

Where are your costs?

[Daniel Boisvert]

If you're spending lots on new hardware, see if you can wring more life out of the old stuff.
If you're spending lots on software upgrades, see if you can hold off on a cycle.
If you're spending lots on bandwidth, shop for a new provider that may get you a better rate.
If you're spending a bunch on outside consultants, put together some numbers showing how much cheaper it would be to do that work yourself.

Maybe you've got excess server horsepower and could get more use out of what you have by switching to thin clients (and get off the PC upgrade treadmill). Maybe your management will be more receptive to Free solutions now that money's tight. Maybe I'm rambling on without enough information to go on.

Without knowing where the bulk of your costs are, it's damned-near impossible to give you any decent advice on cost-cutting.

consolidate servers

[alen]

I have seen people put a server into production running some app and maybe 10% of the CPU and RAM is used at most. Try to consolidate your servers to run multiple tasks, but don't go as far as breaking the vendor's recomendations.

Another suggestion is run the right tool for the job. Not every database requires Oracle. my company runs on MS SQL server and we are looking at Oracle in the future as we grow. If SQL is too much look at MySQL or some other lower end database.

Don't upgrade unless you need to. We run Windows 2000 and we are looking at 2003 only because we are merging with another company at the moment.

Control resources. I always have people yelling for more mailbox space and file storage space. Tell them to have their department buy it and they STFU.

Do some legwork. Projects and needs come up and it usually means a new server is bought. After a while there is a clusterfcuk of servers running different things. take time out and consolidate servers.

We saved a bunch of money

[SubTexel]

By switching to Geico! Err umm, oops. Sorry Just pops into my head everytime someone asks about saving money. Damn those Ads!!

Re:Turning off the computer is costly

[cdipierr]

Except that your assumption is faulty.

A computer, at night even /w the monitor "on" will probably draw about 100W of power tops. This assumes the monitor is energy star compliant and goes into a typical sleep state and that your PC isn't running some CPU intensive task, so at the very least cut your numbers to 20% and you get $400/month for the 100 machine scenario. This isn't nothing, but you're better off convincing people not to take clients to expensive dinners.

Re:start saying "No"

[houghi]

Unfortunatly if the IT deaprtment starts daying NO, it will also start saying NO to things that are really needed.

What you need to do is to get a GOOD insight into what different departments need and spen your time on that. If then the sales department comes along and asks for something extra, just tell them they can choose what of THEIR other projects should be put on hold.

Also charge the departmenst as if you are your own company. Want a new PC mr. big shot? No problem. Give us the money and you get what you want.

That way the other departments will only ask what they really need as it comes out of their budget.

I got a better idea

[Orion Blastar]

downsize the IT managers who cannot say "No", as they are the ones that force IT departments to overwork themselves.

"Here are the projects I want you to work on."

"But these projects are commercially available for less money than our development costs to make them."

"I don't care, I made promises to other departments that we will do them."

"But it will take a staff of 200 to do these projects in 3 months. We only have 30. We will need more time."

"We don't have the budget for that, so everyone will be forced to work 80 hours a week with no overtime pay."

"In some cases we already have some of these software projects. Like Microsoft Outlook for scheduling and contact management, and Microsoft Project for Project Management."

"I want custom versions of those programs, because I promised them to the other departments."

"Well at least can we have a raise to compensate for all the overtime we will put into these projects?"

"No, in fact, I have to cut everyone's salary in order to help budget more money for marketing and executive pay raises."

Then the IT department has a 90% turnover rate for four years of this, and each IT employee that is fired or leaves ends up costing 150% of the annual salary for that position to replace, which adds more to the IT budget.

Then after being over-stressed, over-worked, and suriving 4 and a half years of this, I get really sick and end up being fired and replaced with someone willing to work for half of my salary.

Re:I got a better idea

[irishdaze]

what the hell? were you working for EA?

Look seriously at Open Source

[sc0nway]

We were in a similar bind three years ago (and two years ago, and last year too now that I think about it....). One of the things we did was replace Bea Weblogic instances on our 40+ production machines with JBoss. It saved the company about $2M in licences. When one the programmers in my group complained of the slowness (We needed to tweak our configuration) I just pointed out that saving the $2M was the same as 30 of us not getting axed. But the morale of the story is that if you are serious about saving money look how much you are paying for software licences yearly and see if there are viable open source alternatives to the big dollar items.

Look at Telecom

[mfarver]

Many smaller companies do not play close attention to their phone bills. They might squeeze every penny when buying new computers or office furniture, but often the phone system and lines are whatever the phone vendor installed 6 years ago. Just like cellphones the phone provider doesn't tell you when a new cheaper plan becomes available.. and the market has heated up.. plenty of competition (at least for business services) means everyone is cutting prices.

You're probably no longer in a contract with one vendor anymore, and you often have choice for local service, or even VOIP providers. Ask accounting to cough up the phone bills (hey telecom is an information service, therefore its IT's responsibility)

It is not uncommon to find that a company with 50 employees is paying $2-3000 a month for long distance, internet and local phone service. Often there are a few old "modem" lines no one is using.. too much voice T1 capacity. Whatever.

These days you can get great deals with non-incumbant carriers, epecially in the combined data/telephone market. $400/month for a T1 with shared voice and data is not uncommon. (whatever T1 bandwidth is not used on voice is allocated over to data) A T1 for data or voice only often runs $700-$1000/month. Saving a few hundred bucks per month gets multipled by 12 for thousands per year in savings. There is nothing like saying you just saved the company the cost of your salary.

Application servers

[eyeball]

Investigate replacing some or all of your install base with some form of centralized application services, like Citrix, MS Terminal Services, X11, VNC, etc. It may be a big undertaking at first, but in the long-run, time and money resources would be focused on keeping a small amount of large centralized clusers maintained, rather than a desktop for every employee.

Flexibility would be opened up by allowing people to work from home via the remote clients. If you went with Linux (or a few others with similar capibilities), the desktops could be diskless, further reducing desktop management. Virus/adware/spyware management would almost cease to be an issue.

It's a management mantra...

[Vexler]

It's the management's responsibility - not the IT staff's responsibility - to make sure the company comes out in the black on the balance sheet every year. The average IT staffer doesn't see every penny coming in and going out - that job belongs to the CFO and the accounting department.

Management needs to take a stock of how the cash is flowing and make strategic decisions on how best to save for long-term growth. Buying that shiny and new equipment may not make much sense, until you realize that you are throwing away five times as much money in manhours every year by not biting the bullet and upgrading.

I used to work for a manufacturing facility, and there are a lot of old-timers who think that saving money involves turning off their PCs every night. But they were not looking at how much time they are wasting every day in dealing with old OSs and crash-prone programs. They also did not look at how much time I (the network engineer) had to go over and "fix" their machines by rebooting for them.

Having your corporate culture mumbling to itself "gotta save money, gotta save money" is a good sign that the senior management, together with middle management, has not done its job in formulating and communicating a coherent game plan to the rest of the company.

Re:Outsourced Ourselves

[way2trivial]

you don't know much about it then.
http://tinyurl.com/47c8e/
read up here

Canada: Safe, secure and 'near-shore'
It's about as close as you can get, and its low risk and relatively low prices make Canada a favorite destination for "near-shore" outsourcing.

The Philippines: Low cost, but higher risk
The second most popular outsourcing destination after India, the Philippines has a highly skilled, English-proficient workforce and low cost.

Mexico: It's Close; It's Cheap
Just a short plane ride from the U.S., Mexico boasts a well-educated workforce and lower prices. But the lure of jobs in the U.S. keeps turnover at outsourcers high.

Ireland: Comfort and Convenience at a Higher Cost
Its government is eager to offer tax benefits and grants to companies willing to bring IT work here, making Ireland an increasingly popular destination for software maintenance and development work.

China: Low-level work at lower-than-average cost
Low cost is driving some users to outsource IT work to China, where low-level programming resources can be found at bargain rates.

Singapore: Small but powerful
This small Asian locality has economic stability and a highly trained workforce on its side. But those strengths come at a price.

Vietnam: Nascent capabilities but low cost
A "country in progress," Vietnam offers low labor costs but faces some communications and modernization challenges.

Malaysia
An emerging outsourcing player, Malaysia has invested heavily in a high-tech corridor to lure international business. But a sluggish economy and small workforce have slowed the country's momentum.

Brazil
Brazil is well known for the bossa nova, string bikinis and Amazon forests. Less well known is that, by many measures, it?s one of the world?s major countries. It ranks fifth in both geographic size and population (180 million people) and has the world?s eighth-largest economy.

Russia and Eastern Europe
Its IT workforce is low-cost and highly trained, but Russia's abundant scientific talent remains largely untapped because of government bureacracy and image problems.

Selecting the Right Offshore Vehicle
Opinion: Columnist Bart Perkins says there are different types of offshore outsourcing vendors, and it's wise to pick the type that fits your company culture, requirements and risk profile.

Sometimes it's the opposite.

[TheMCP]

Sometimes buying software from outside is more expensive than just building it in-house, particularly if the outside software is overpriced (as it usually is), and if you can define a reasonable and limited set of features or the software to be developed in-house (this is where most development efforts fail).

As an IT/IS manager, I have in the past been tasked with buying software packages for major company initiatives... and found that all of the decent packages that came anywhere near meeting requirements cost upwards of $75,000, and at that they wouldn't integrate into any existing systems. (This creates another set of logins to maintain.)

For 2/3 that price, I could hire a competent programmer for a year. So, faced with this dilemma and an expensive package I needed to have, I hired a programmer. And the system was developed in two man months, leaving me able to use that programmer's time on other important projects. Oh, and the package integrated perfectly with existing systems, and was expandable.

Okay, so having a full time programmer is $50,000 a year on the books every year, an ongoing expense, while buying the software is $75,000 and then it's over. (Except for the invisible ongoing administrative costs.) But, I saved $25,000 outright by hiring the programmer, and then if I consider that it actually only took two months of time to do it, I actually saved $66,667 by hiring someone. And then I could save more on the next project that I developed in-house.

Re:Sometimes it's the opposite.

[austad]

The scope of your project was one where you could do that. That's not going to work for things like say an Oracle database replacement, or a world-class monitoring solution.

Re:Sometimes it's the opposite.

[Taladar]

This only works if you (as the hiring person) know a good programmer from a bad one.

Wrong assumption

[bluGill]

Your assumptions are incorrect.

For the 100,000 machine office they are not pay per kilowatt-hour like you are. They pay peak-demand, meaning that they pay for however much power they use at the highest moment, as if that was how much they used all day (sometimes even month!). Power at night is free because they pay for it anyway. They agree to this because if you do any management of use at all you can save a lot of money this way.

Places that heat with electric see no change at all because either way energy use just turns into heat. At night things tend to cool down (no humans adding body heat, and the sun isn't adding heat), often a little heat it wanted (but not needed) even in summer months. Some buildings may not even have the ability to heat the building at night if the heat system was designed considering the equipment!

Re:Turning off the computer is costly

[Peyna]

Now, if you don't pay the employees while they wait for their machines to boot...

Your claim is based on the assumption that employees will just sit there and do nothing for the 5 minutes they wait for their computer to start up.

Odds are they'll waste the same amount of time per day doing nothing product whether or not they have to wait for their computer to start up. There are probably many other things they could do while waiting.

The Numbers Fallacy.

[fm6]

Sure every penny counts. But sometimes you can waste more effort counting pennies than you save in unlost pennies.

A lot of companies are on a permanent cost-cutting binge simply because it helps upper management look good with investors. Now, it's often true that these policies get started when a company's wasting money. But they will often continue long after the waste has been dealt with, or even when there was no provable waste to begin with. It's just another example of how corporate policy is set by numbers dweebs, you justify their jobs by the fallacy that every reduction in cost is an increase in profits. It does work because (a) you do have to spend money to make money and (b) as often as not, the apparent cost reduction exists only because of some accounting silliness.

A couple years ago, I had a workstation on my desk that wasn't quite up to what I was asking it to do. A lot of my time (and thus the company's money) was being wasted while I waited for the system to stop thrashing. The standard solution is to request a new workstation. But I thought that was just a little too much to spend. (I'd like to say I wanted to help control costs. But the truth is that I'm fundamentally a tightwad, even when it's not my money being spent.) Instead, I decided to request a RAM and disk drive upgrade which I calculated would make the system much more usable. Here's how it went:

• I put in a request for the purchase. It's only about $300, but as a cost control measure, even $300 purchases have to approved at the VP level. I wait.
• Weeks pass. I threaten to buy the hardware with my own money. For some reason, this threat, though often employed, is usually effective, and I'm told that approval is emmient.
• The purchase is approved. But then my boss tells me that I'm in violation of a new cost control measure, because my workstation has been amortized and now makes the numbers look bad, because of IT costs. I agree to withdraw my previous request and put in a request for a new workstation.
• Weeks pass. A date is set for the replacement of my workstation. But then upper management decides it doesn't like our numbers (we're solidly in the black, but costs are too high. So they impose a spending freeze. No workstation, no RAM upgrade.
• Weeks pass. Freeze continues.
• This goes only for something like six months. Finally, another issue causes me to leave the company.

Is the company making every penny count? No, they're actually wasting money by working inefficiently. They wasted a lot of my time, then tried to buy a workstation I didn't need. But the numbers look good.

Re:The Numbers Fallacy.

[TheCabal]

Is the company making every penny count? No, they're actually wasting money by working inefficiently. They wasted a lot of my time, then tried to buy a workstation I didn't need. But the numbers look good.

At a previous job, the IT budget was on a permanent freeze. In the three years I was with the company, we had only made one major investment in IT, and that was at the beginning of my tenure. Now, we were an Application Service Provider, so our lifeblood was in our servers and how fast we could crunch numbers.

Did I mention that the major investment in servers, all the servers were bought off of eBay and other second-hand vendors?

So I'm dealing with four year old servers, with outdated hardware this is just slowing down more and more, while we are getting more and more customers, and larger customers. I tried explaining to the Powers that there is a fixed number of cpu-hours, and it takes X hours to process customer Y on our current hardware. We were operating at something in excess of 90% capacity. I gave several pleas to free up some money to acquire some faster, more robust servers, thereby reducing X, allowing us to handle more Y.
But as IT is commonly held as a cost center, I got the usual "we don't have the money". I left the company eventually, but heard they hit a hard brick wall as the production environment was saturated almost 24 hours a day- the couldn't bring in any more customers. Sort of ironic that a company can get killed by its own success.

Re:The Numbers Fallacy.

[Fulcrum of Evil]

I gave several pleas to free up some money to acquire some faster, more robust servers, thereby reducing X, allowing us to handle more Y.

But as IT is commonly held as a cost center, I got the usual "we don't have the money".

What the hell were they doing with all the revenue from those customers? Seriously, did you ask them? In those words? IT is not a cost center when your core product is IT.

Re:The Numbers Fallacy.

[TheCabal]

Used the money to hire more salesdroids to "develop more business", and bought an entire floor of a hugeassed building, then crammed everyone into one tiny corner of it.

Meanwhile, the IT processing power remained at a constant while IT staffing actually was cut. Interesting times...

Re:mostly centralization

[Chanc_Gorkon]

I agree with this. While nice for the users (only have to memorize one password) and nice for the net admins (only one auth server to fix when you can't get authenticated), it's horrible for almost everything else. If teh main LDAP goes down and there's no HOT failover(warm standbys suck), then your users will not be able to log in. Also, once a password is compromised, ALL of your servers are at risk...not just the LDAP server. LDAP and other centrally controlled authentication needs to be planned very well and the security needs to be HIGH. Make sure your users call you the SECOND that they think someone is doing something wrong or the SECOND that they forget their password or realize that it's in the wrong hands. Better yet, if you must go LDAP, get some RSA keyfobs and the RSA security package setup. This will inctrase security while reducing the overhead of multiple log ins.

Let the experts do it

[madmaxx]

In our small office, we have cut costs by factoring out un-needed servers. When I started there a year ago, there were 30+ servers running in a cooled server room. They had a OC3, and were hosting their own mail, web, dns, etc.

Now we have 4 servers running internally, and one running offsite. We pay a hosting company to manage our mail and web services, which costs us 1/4 of what we paid our own staff to do. We've dropped our fiber and use business DSL, which is another large savings. We also order all of our equipment from a very capable local shop, who take care of building and configuring hardware for us. As a bonus, or local retailer serves as our expert on hardware choices.

A side-benifit of reducing the number of servers we use, we have a surplus of spare parts. These changes also allow IT staff to be redistributed in the company, doing more important things (like testing, customer support, development). While we still order new parts, we've been able to drop our hardware budget by more than half for the past year.

Resources are better spent on things related to your products and services, so it's important to spend your people on those things as well ... and not on IT. IT, in many ways, is like plumbing or electrical: the business does well to have the services, but should not feel they need to do it all themselves. Stick to the domain of the business.

Reduce Total Cost of Ownership, Document All Work

[TheMCP]

As you no doubt know, your expenses in a system - be it software or hardware - are not only the initial purchase, but also its upkeep, both in terms of ongoing direct costs (upgrade fees, purchase of parts or media), but in time. Every hour you spend working on something is an hour the employer had to pay for you to be maintaining something existing instead of making something better. Employers like to see progress. They're much more willing to pay when they see you're making progress.

So, this means you need to do two things:
1) Reduce the amount of time you spend on maintenance.
2) Document everything you do.

So, let's look at these a little more closely.

Reducing time spent on maintenance
Examine your obvious unnecessary expenses and see how you can eliminate them.

Having problems with viruses and spyware, or spending time on antivirus and anti-spyware software? Replace IE with Firefox and replace Outlook with... well, anything you like, really. That'll prevent a lot of viruses right away, and that's an enormous savings. It cost my organization $45,000 in staff time every single time a new Windows virus hit the net, and that's AFTER installation of antivirus software. The antivirus software never seemed able to keep up. Also, start replacing simple desktop stations with Mac Minis. MacOS X doesn't get viruses or spyware. I'm not saying you should take perfectly good stations out of service to replace them, but as you replace older systems with new ones, start putting macs in instead of new Windows boxes.

Macs also tend to stay current several years longer than Windows boxes. So, you could amortize the purchase cost over an extra year, or perhaps even two, and save money on desktop machines that way.

Wasting time setting up software on desktops, or maintaining the software on desktops that were already rolled out? Get a Ghost server so you can just ghost the machines. If someone's software is malfunctioning, don't go muck with their system in person, just ghost their system remotely and move on to the next task.

Lots of your time sucked up by idiot users on repetitive problems? Spend a little time writing a how-to white paper, and when they call to ask that same old question, get the person doing triage on incoming support calls to just give them the white paper so they don't have to bother a tech. The faster you get that person off the phone or out the door, the less dollar value your employer spent on your time dealing with them.

Spending time administrating servers? Reduce the number of servers. A smaller number of larger (expensive) servers, well backed-up and with substantial redundancy, is much cheaper to maintain than a large number of smaller (cheap) servers, because you only have to do each maintenance task once for one large server instead of umpteen times for umpteen little servers. I've actually seen organizations that literally had more servers than employees, and they couldn't figure out why they were spending so much on IT. Yeesh!

Problems with viruses and security on servers? Servers going down from time to time? Replace your Windows servers with Apple XServes. They're fast and easy to configure, can integrate into your existing LDAP login environment, can support both Windows and Macintosh clients (your users never have to know), and can easily be set up for RAID and redundancy. Apple also has superb offerings for on-site maintenance agreements.

Documenting all work
Employers often think they can get away with making you cut a person in IT because they don't understand what IT does, so in their mind IT doesn't actually do anything. You need to show them how much you really do. This means very anal-retentively documenting EVERY action of EVERY staff member, and indexing it to the customer as well.

I mean, if the phone rings, there should be a record in the computer of who called and who they talked to and for how long and about what. Got a stupid user who requires constant hand-holding? S

Frame the problem better

[linuxwrangler]

You need to find the answers to some questions:

Are all divisions being pressured to cut or just IT?

You say that IT is overworked but is the IT operation efficient?

How central is IT to the companies business?

How aware is senior management of the contributions of IT?

How does your company compare to others in your industry? For example you may be profitable but if you make your investors $.01 per dollar invested and your industry average is $.10 then your company probably has a problem. Also, how does it compare in use of and expenditures for IT?

Are there indications that the company is facing problems that will require belt-tightening?

How is IT's performance perceived throughout the company?

Is IT's capability being underutilized by the enterprise?

How resistant is the organization to change?

I could come up with more but you get the idea. With some digging you will soon be able to determine what is _really_ happening.

If someone in power is targeting IT only and setting you up to fail they are probably just setting the stage to outsource. Polish up your resume.

If you find that the company is doing fine then this could be a scare-tactic method to lower or eliminate bonuses and raises. Your call whether or not you want to stay.

If IT has a reputation as a bunch of BOFHs then you have been digging your grave for a long time. If you survive in the short term, this needs to be fixed. Sure, some users can be a pain but users are the reason IT exists at all.

If changes can make IT more efficient, suggest them. Just be careful not to confuse efficient with effective. Doing an unnecessary thing efficiently is not helpful. You may even find that its time to wean remaining users from costly legacy systems.

Think like a businessman. Have you renegotiated with your suppliers? Phone time, bandwith, hosting, loop and similar charges have plumetted over the past few years. Are you paying yesterday's prices or staying with an overpriced vendor?

All the time be sure to remember to judge savings against profit, not revenue. I just dropped our DS3 loop charges by $12,000/year. That's not even a quarter-percent of the revenue of a $5,000,000 business but if the profit of that business is $50,000 then that saving just increased profit by 24%! A lot of businesses are just barely in the black if they are profitable at all. And that loop-charge saving is just 1/10 of what I saved by switching vendors a couple years ago. Costs count.

At the same time don't get boxed into "IT is just an expense". Years ago I read a story about senior managers at an auto company all discussing the painful reality of sagging sales and they started spending a lot of time trying to decide just which factories should be closed when one of the managers said, "I have an idea that will save us _lots_ of money. Let's close _all_ the factories." He got a laugh but most importantly he broke the tension and refocused the meeting.

Consolidating Servers

[Stephen Samuel]

Although you can't get money back on the extra server licenses, it does save you the cost of ongoing licensing and support for the retired machines, as well as the rack costs of the machines (if they're hosted remotely). You can also consolidate them into the two most recent boxes, and thus avoid the possibility of the older machines dying sometime soon.

Re:Outsourced Ourselves

[birukun]

*sarcasm*

You left out California.

In many ways a country by itself. Just a short plane ride from the U.S., 7th largest economy in the world, highly trained (and hungry) IT workforce, Low-level work at average-cost (Lots of Chinese, Mexicans, Vietnamese, and Malaysians), and well-known for string bikinis.

Downsides are high taxes and English deficient workforce.

*/sarcasm*

easy

[batura]

Block slashdot.org in your proxy server!

you want a diagnosis without giving symptoms

[avi33]

Without trying to sound too snarky (there's plenty of that posted already) it's really hard to say without knowing what you're spending your money on. There are huge industries out there that will do this for you (and say, pocket 50% of the savings as a fee) so keep in mind that in asking Slashdot, you get what you pay for. I've been through similar circumstances at past employers, and there are a few easy places to start looking.

• Identify where your biggest costs are. Services? Licensing? Personnel?
• Go for the big fish. Saving $200 per workstation for the entire enterprise will create a lot of work, headache, retraining, and frustration. Merging servers to kill off a few licenses will require time and effort, but with a verifiable ROI.
• What services do you use, outsourced, ASP model, or otherwise? You would be surprised how willing your vendors might be to renegotiate terms, even mid-contract. Take the approach of "You're not just a vendor, but also our partner in this business, and you have a vested interest in seeing us succeed. Act like a partner and we'll have a long lasting profitable relationship for both of us. If you refuse to work with us on this, we will waste no time in looking for your replacement when the contract is up." We renegotiated a number of contracts like this, and a couple that wouldn't budge? They were out the door.
• Determine what your core responsibilities are (to the business). Use a minimum of hardware, services, and personnel to reach that goal (in the short term). If you are serving as the "junk drawer" for the entire organization, they will cut their costs and pass the responsibility on to you. Until you quantify exactly what your role is, you won't be able to push back and say "If that goal is important to your department, then you need to find a source to fund that project." Suddenly everyone's pet projects aren't so important when they have to chip in for them.
• You need to manage expectations, as in, you won't noticeably lower the electric bill before the end of the month, but you can say "2006's budget is $50,000 lower due to the licenses that we won't need to renew." Document how you have been able to cut costs, if possible without cutting service. You will probably be faced with a midlevel executive who will say "Okay, now you really have to tighten that belt." You need to be prepared with a statement like "We've tightened our belt 25%, and we can tighten it an additional 25% if we stop supporting projects x, y, and z."
• Personnel decisions - Not to advocate layoffs, but if you're keeping someone around at $90k to do a single job that can be outsourced for $30k, you should outsource it. Ideally, you'd bust your ass to make sure that person was shifted within the company, where their talents could be used to provide more services and capabilities to the company. Again, not to incite a flamewar, because it's a complex and sensitive issue here, but it's probably too late for that.
• On the flip side, for the past couple years, the market is quite ripe to replace agencies with independent contractors. You need to be skilled at finding the good ones, and managing the projects.
• Further beating the dead horse: licenses, services, and personnel.

Re:Terminal Server

[eakerin]

Most linux distos use X11 for displaying GUIs, which has built in support for network transparency.

Calling it a Linux Terminal Server is really just a way to explain the idea to people that only have experience with Microsoft products.

So really, every linux machine can be a "Linux Terminal Server" and every linux machine can be a "Linux Terminal Client". There are a multitude of ways of setting up servers and clients, LTSP is one, Nomachines NX is another, and there are many other methods to manage it (you can easily hack something together yourself if you want).

One other thing that X's native support for networked environments gives you is something like Citrix's seamless windows. You can run an app on your server, and it will display on your client just like applications run locally. In windows terminal services you get a desktop window with your app in it, or a full desktop session.

Establish an expectation of service

[pcguru19]

I've been through this too many times in the past. The simplist way to deal with this is:

1. Establish a budget that meets the paper-pusher's needs.
2. Off of that budget, outline the services, uptime, and response time you can deliver with the money you have. Spend a little time backing up what you say.
3.Publish that to the rest of your company and let them know this is the new performance standard you'll be delivering to the company. Make things that impact executives(like email, bandwidth speed, etc) the things you reduce services on the most and leave the core business needs alone. Sell it as a decision that you weren't willing to compromise the core services you provide to the business.
4. A suggested change if you don't already to it is to charge other departments for an outsourcer when employees create a problem from spyware, malware, or installing random crap on their machine. It's a good incentive for that employees manager to do something about them being an ass and it takes some of the pressure off you. We work on this stuff ourselves and use the outsourcing money to help us out with some tougher stuff elsewhere in our department.
5. Trust me on this one, once you put these rules in place other departments WILL spend their money on you.

Apples and oranges.

[Mistlefoot]

To be frank. Advice on how to save money for a small company is the topic at hand.

You indicate 25,000 computer with 15 techs.

Quite obviously saving $100 in hardware per PC would save you $1/4 million. Cutting back on 15 cell phones....peanuts. You are likely to be inclined to look for savings in regards to hardware - or per PC. A small computer with 30 computers and 3 staff will have far different needs. Saving $100 per computer would be the same as 2 weeks wages..... Peanuts.

I don't know the answer but I do think that you're situation is far from similar.

Re:Apples and oranges.

[Anonymous Coward]

The point is the same, though. I remember starting Word for the first time on a shiny new desktop we had bought while I was in college. It went from double-click to functional almost instantly. To say that again, MS Word ran as fast or faster on a 1996 machine as it runs on a 2004 machine. So for someone typing up documents in Word, an 8 year old machine is just as good as a new one. The same is true of many applications that spend most of their time waiting for the user. No productivity lost, and lots of savings.

Then again, you could get a Wal-Mart machine cheap enough. The software licenses will still cost you, though.

Re:Apples and oranges.

[TWX]

I worked for a small company that had one IT guy (I was in QA) that handled our thirty or so PCs, three or four network servers, the phone/data cabling, programming the PBX, and assisted in building the systems that customers ordered to handle pager traffic. One guy did all this. Admittedly the bulk of the staff was computer literate and we probably could have done it even without the one IT guy, but it didn't require cell phones. He had a cell, but he had a work pager and he'd be reimbursed if he had to use it to call in for something. Workstations ran until they were wholly inadquate for their tasks, and then were reallocated to other tasks (frequently to Quality Assurance for use as test generator machines), while servers were replaced or added frequently, probably more frequently than necessary. Labor was the vast majority of our internal IT expenses, with probably about a third of our internal expenses being equipment. So, probably about $90,000, if that high, and could have been reduced if we'd been more thrifty with our purchases.

Simple Really

[Goose3254]

Keep infrastructure in the house and offshore development. Numbers look better in the short term. Long term deliverables falter. Middle management weenies who would rather gut the company than let go of thier fat bonuses to make an arbitrary number get cut. Competent worker geeks get promoted to middle management where they fail abysmally. Original developers are brought in as contractors to get the project back on track, although at twice the cost. CIO who prompted cycle gets new job with more money at another company. Original company gets bought by CIO's new company. Original company's officers stock options sell for enough to retire to Palm Springs. Second company's employees positions are threatened by management with first company's current staff. Keep infrastructure in the house and offshore development.... Rinse and repeat.

Computers are expensive and unstable, ...

[chris_sawtell]

So stop using an expensive simulation of a typewriter and use a real one instead.

ditched most of our windows servers

[sl4shd0rk]

And you can too. All that TCO talk is garbage. $15,000/year in licensing cost is one thing we saved.

server consolidation

[chunkwhite86]

VMWare ESX server is a very effective tool for server consolidation. Particularly if you have many old legacy servers that you just can't get rid of.

Migrate them all to virtual machines within VMware and say bye-bye to replacement parts, expensive support contracts and pricy KVM-IP devices for outdated servers.

We removed about 65 old (PPro through P3) servers to virtual machines in VMware. In addition to the benefits I mentioned above, we also gained SIX 42U racks of free space, reduced cable clutter, and reduced server room power/cooling consumption.

Disclaimer: I don't work for VMware but I can't say enough good things about their server product.

downtime projects...

[nixkuroi]

One thing we did when I was working at helpdesk was to work on our "own" projects in the company that supported undocument company needs. As a phone support guy, I had times where I wasn't really doing anything (I'll call that time "slashdot time"). During that time, I'd approach other business units (say, the mailroom because they tend to regard you with awe and humility) and see if any of their processes could be automated through some simple programmatic way. As it turned out, they were hand-parsing outgoing mail addresses for capitalization and formatting errors that were fairly uniform in their imported excel documents. By writing a macro for them to insert into their docs on import, I was able to parse and fix the file in 30 seconds where it had taken them 4 days of 3 people working on it all day. Someone calculated that the company would save something like 280,000 a year in overtime and allow those overworked people some time with their families. If your company recognized this as a gain they could capitalize into your department, you could afford to hire a couple more guys and take some of the work off the dudes who are overworked...or give them bonuses or raises (as happened in my case). A lot of the time, people don't even know they can be helped unless you ask them and by helping out other people, you end up helping yourself....PLUS the mailroom reacted with hyperspeed next time I needed a little something mailed out.

re: service contracts

[King_TJ]

I think you brought up an excellent point. Service contracts are *rarely* worthwhile - yet many businesses seem to buy and renew them without even a second thought.

I used to work in I.T. for a place that was constantly complaining about a need to "cut costs" (and in an overall sense, they did - because several of their locations were being shut down as unprofitable, etc.). Unfortunately, we had such things as support contracts with Oracle for our main database that cost upwards of $30,000/yr. to renew - and I don't think we made any use of it except maybe 1 or 2 times in 5 years. (Both of those times, we reported problems which turned out to be small bugs, and "hotfixes" were mailed to us on CD-R disc -- but, these same fixes made it to future point release updates of the Oracle products anyway. I think we could have lived with the issues a little longer, or worked around them, without it costing us over $30K per year worth of problems!)

I also remember a fairly costly maintenance contract we kept up for all of our uninterrupted power supplies. Sure, it covered replacement of worn-out batteries - but at best, it was a "break even" deal over just buying replacement batteries when needed and swapping them out ourselves. If a UPS actually lasted longer than expected, the maintenance agreement instantly became a poor value.

Cost cutting concepts

[hesperant]

Here are a few options that will save money over a long time.

For equipment and hardware issues, Do not turn off systems over night. The only time you age hardware is when you change it's state. That being said, having the machine "Restarted" at the end of a day will give you the benefits of a restart without the actual aging of your equipment. Add to this, keep a constant temperature in the office. A good insulation can go a long way in saving money. You would be surprised at how many hidden costs the liberal use of an air conditioner can have.

For support.
When you can have a support team also engage in other activities such as development and user education it is a winning combination. Give your support staff the ability to help work the common issues out of the system and you reduce the number of times this problem can hurl boulders at you. In essence you solve 3 very difficult to handle issues. The first solution is to provide the IT support staff a method to keep on the ball for the technologies you as a company use. This will reduce call or issue time and increase productivity. Having them institute training will also help you greatly. They have the ability to work with people individually and over a phone in a teacher-student kind of way already. Your just giving them the opportunity to reduce the chances of support need preemptively and allowing all staff to benefit from there abilities. This training will also increase the productivity of other non technical staff tremendously. Most companies use less than half the utilities of the software they spend allot of money on. Allowing your support members the ability to be part of the IT process and not the bumper of the IT process will increase their effectiveness.

Licenses: You can reduce the number of proprietary licenses. In many cases using an in house solution or open source solution for your needs will also have the added benefit of allowing you to tailor the application to your needs. Nothing reduces productivity/profit like being slave to an application that is not well built or has more features than you need.
Keep in mind what applications you use and why. Do not allow constant sweeping changes to your IT department by non IT personnel. This includes you very well meaning and intelligent company owners. Every time you change one aspect of your IT infrastructure that equates to at least 1 hour of extra support time per three people and 1 hour of downtime per 1 people. Every time you get a new sales jockey or developer, Make them use your tools and not you use there tools. Having 6 different applications that do the same thing is a terrible waste of resources. You could also reduce the number of applications used with the addition of training. A groupware client for instance can go a long way in managing your project/messaging/incident/sales/anti-virus/contac t needs without having 6 separate utilities. Reducing the number of utilities a company needs also reduces the system requirements for your staff and allows older machines to perform well.

There are allot of ways to reduce cost without loosing people. Get past the ego's and commercial induced projects and you will find a smooth running machine can keep your bottom line looking quite pretty indeed.

Numbers are your friends

[davecb]

th poster said IT is commonly held as a cost center

Classic oops: if IT serves transactions for sales, it's "part" of sales, a profit center.

As a capacity planner, I usually talk to the business managers and say things like

Last christmas you rejected about 3% of your purchase transactions. You say you're growing by 6% per year, so next year you're going to reject almost 9% of the christmas business.

The hardware to handle 10% more load will cost you $X, and therefor, if the profit on the average transaction is more than $X/(9% of the transaction count) you'll cover your costs immediately.

Npt to mention avoiding pissing off the customers who you'd have rejected!

These are arguements to a profit center: if you can credibly make them, they'll dig up the money and force your boss to take it (:-))

--dave

Cutting IT costs

[erc]

Get rid of Windows wherever possible. Servers can run Linux or FreeBSD, MySQL, sendmail, etc. instead of Windows Server, SQL Server, and Exchange - that one simple move can save you thousands of dollars *per server*. Even desktops can be selectively replaced - for simple office applications, Evolution or Thunderbird work just as good, if not better, than Outlook, and since installing Firefox, I haven't touched IE in months.