What does yours look like? There are a few elements that are must haves:

1. Consistent power: you should have un-interruptable power supplies powering your server and network infrastructure.
2. No water: That sprinkler head right above your servers? – bad idea. If the system goes off for any reason, your servers may not recover.
3. Temperature Control: This can be as crude as a propped-open door and industrial fan. I’m sure you can do better, but as long as the temperature stays in the low 60s or low 70s, you’ll be fine.

Here are the nice-to-haves:

1. Somewhat neat wiring: Fastidious wiring is actually unproductive, but your wiring should be generally organized and labelled.
2. Labels: You’d be amazed at the number of times we can’t figure out which server is which in a server room until we load or eject DVDs.
3. A KVM (keyboard-vide-mouse) switch.

The best server room configuration is one that doesn’t exist. We are big fans of getting your server infrastructure into a data center. It is the only disaster recovery plan that allows us to sleep at night.

How many IT people do I need on staff? We get this question a lot. I’m going to give you a very straightforward answer, with the up-front caveat that each organization is different and individual circumstances may vary.

The general rule of thumb we use is one full time IT person for every 100 end-users. This is true from about 75 to about 125 end-users. But, exactly one full time IT person is dangerous: too much specialized and critical knowledge will accumulate in one person’s head. If that person leaves or is asked to leave, you run the risk of unplanned downtime that is difficult and costly to recover from. Consider two ½ time people instead.

Below 75 end-users, you are probably best off engaging with a managed service provider. From 125 users to 200 users, 1.5 to two IT support staff is the right number. Above 200 end users, plan on one more full time equivalent per 100 users.

What if you want to outsource your IT? (Call us if you do!) With between 1 and 10 end users, you should be able to have a service provider take care of your IT needs on an on-demand basis. From 10 to 30 end-users, you should be in a managed service contract with a service provider. With above 30 end users, you should consider having someone on-site at least weekly. At 70 end-users, the on site presence should be twice per week, at least.

If you’d like to have a conversation about the right number for your business, please contact us.

One of our first clients had an IT Manager who was a mainframe expert. The idea of PCs, Windows, and client-server applications were heretical to him. His idea of guiding his employer into the client-server application world was to bring in consultants to do the work. When we first got involved, we found teams of vendors meeting, at the client site, by themselves, to hash out what was best for their shared client.

Seeing this reminded me of a line that my dad used to use a lot: “A camel is a horse designed by a committee.”

We immediately put a stop to this practice. We here at The IT Management Group are big fans of outsourcing your IT, but you can’t outsource your IT leadership. You need one trusted provider or employee that will act in your best interests and manage all other vendors. If you choose to have a consultant act in this role, make sure you understand what is in the consultant’s self interest. If it doesn’t align closely with your interests, you’ll need to be vigilant about being shepherded into a direction that serves the consultant more than you.

1) Software as a Service (Saas): You pay to use software that is owned, maintained, backed-up (hopefully!), and hosted by someone else. Salesforce.com is an example of this. So is Google Apps. In both cases, your organization is paying a monthly or yearly fee for your employees to use their software.

Usually, all that is needed to use the software is an internet connection. This model places the lowest demand on your internal I.T. resources. It also makes you the most dependent on the
vendor.

SaaS is a mixed blessing when it comes to disaster recovery planning. If your building burns down, all you need is a computer with an internet connection to access your SaaS applications. So, from a DRP standpoint, it couldn’t be easier, right? But, what if the SaaS vendor suffers a disaster? What if they lose your data? Now things get complicated.

2) Managed Services: The managed services model of cloud computing is leasing a combination of hardware, operating systems, and software from a provider, usually a data center. For example, instead of owning your own servers, you could lease a server and operating system from a data center. They would provide to you a physical or virtual server with a server OS of your choice. The provider is responsible for the hardware and operating system, you are responsible for applications.

This model will eliminate some work for your IT staff, but not much. But, it is the model that is most disaster-proof, provided you choose a rock-solid vendor.

3) Private Cloud: In this cloud model, part or all of your server infrastructure has been moved to a data center. You still own your hardware and software, but you are leasing power, bandwidth, and space for your equipment to live. This moves you away from location-dependent computing: you should be able to access your applications from your home or hotel room almost as seamlessly as from your office.

The demands on your I.T. staff won’t decrease under this model, and your disaster recovery planning still needs to take into account server failures and similar localized problems. But, being in a data center, with off-site backup, is one of the few credible disaster recovery plans I know of.

Which one of these are right for my organization? The answer is most likely a combination of all three. Companies under a certain size should consider using SaaS exclusively. It keeps costs down and eliminates the need for an internal I.T. resource. But, as companies move to an enterprise mindset, it makes sense to begin to own your own applications and perhaps eventually, to own your own hardware and software. At least, that is what makes sense as of this writing. In a future blog post, I’ll discuss where I see enterprise IT computing in the next few years.

But, often, it isn’t a case of being motivated to hide information, it is a case of being unmotivated to document something that is boring. Many people end up in the technical field because they enjoy trouble-shooting and solving problems. But, once a problem is solved, they are bored with it.

Regardless of what the motivation may be, it is unacceptable for a technology professional to hoard information. Not only should your systems and processes be documented, but most everything that most I.T. staffers do should be recorded in an organized and searchable manner. This will greatly decrease the amount of time reinventing the wheel.

No organization should ever feel held hostage by their I.T. Department, any particular staffer, or by any vendor or group of vendors. A senior executive should have complete access to all documentation and should audit it on occasion to make sure it is current.

If you’d like a checklist of documentation you should have, and some suggestions for simple and quick audits, please contact me.

Look very carefully at all of your maintenance agreements. Many companies accept that fact that they need to pay tens of thousands and hundreds of thousands of dollars in maintenance and support. You probably don’t.

If you have a support agreement for your ERP system, call the organization providing support and ask what the alternatives are. Wouldn’t it be better to not renew that $30,000/yr agreement if you knew you can buy support as needed in $5000 blocks?

Look at every monthly recurring cost. How many of these do you really need? How many can you cancel?

The ideal I.T. department will be able to move quickly to lower costs dramatically. If your I.T. department has been doing its job the last several years – staying up with updates, refreshing machines on a regular schedule, mitigating the risks of depending on vendors and staffers – then your company should be able to easily postpone all I.T. initiatives in a financial emergency. You should be able to cut staff, defer purchases, and cancel support agreements.
These steps are not ideal. Eventually you’ll have to restart spending on I.T. But in a financial crisis, your I.T. department should be able to save a significant amount of money.

In the case of corporate email, instead of having an email server or servers on site and paying someone to manage those servers, you would instead connect to an email server managed by third party, paying a monthly fee per user. There are some very compelling reasons to move some application to Saas, but the hype doesn’t always withstand sober scrutiny.

I strongly considered moving a client from their on-site Exchange server to Google Apps Premier Edition for their email. The end users would not notice much of a change – they would still use Outlook for email. I’m still figuring out how collaborative calendaring would work and how we would replace the public folders being used in Exchange, but I’m sure Google’s solution would be sufficient, if not superior.
The three issues that made me decide against moving the client to Google Apps are:

1. Control : Is the client ready to surrender control of a mission critical service to an organization that is largely unaccountable? Google App’s Premier Edition claims to have a 99.9% uptime guarantee, but the Service Level Agreement isn’t satisfactory – if they fall below 99%, then the client must request a service credit for 3 to 15 free days of service. Fifteen free days of service won’t make up for a day of email being down. Right now, if the client’s email goes down, they call us. We are 100% responsible for making sure it works. No excuses, no other vendors to blame. And if we don’t perform, the client is free to find a service provider who will.
2. Cost: Currently, Exchange 2003 is running on a Windows Server 2003 box. Besides patching it, there is no maintenance required outside of an unscheduled reboot every other month or so. So the cost of sticking with what they have is essentially zero. The cost of moving to Google Apps Premier is $50/user/year. For a 50 person company, this would mean writing a check for $2500 just to interrupt the status quo.
3. ITMG’s Business Model: This last concern is weak, and is similar to #1 above, but I can’t simply dismiss it. Is it a good idea for me to steer my clients toward using a service I have no control over? The worst case is this: I make the argument for moving to Google Apps, the client agrees, I move them over, and then there is a two day Google Apps outage. I can’t do anything at all to help the client in the immediate term. I’m uncomfortable with this…it positions ITMG as a sales organization (Buy some Google Apps!!) and not a service organization. As my clients know, I have no problem working ourselves out of a job – but we do this by making client systems more reliable and predictable, not by asking them to trade a reliable system for one that we can’t control.

All that said, I think Google Apps is an excellent choice for many organizations. C! Tech Solutions. just moved the Ferndale Chamber of Commerce over to Google Apps, and they have been thrilled with it. For small organizations, for less corporate organizations, or for organizations that can stomach the idea of not being able to control their own destiny, Google Apps seems like a no-brainer. And, Google is an excellent company with a delivery-of-service record far exceeding the picture I paint when discussing concerns.

A big part of I.T. Management is risk management. In this case, at this point in time, the status quo is more than adequate, and the costs and risks associated with moving to Google Apps don’t justify the benefits.

Plant A was much easier to work with throughout the implementation. The supervisors were not pushy and they did not blow up when they system had problems that might have slowed production. Instead, they looked for ways to keep production flowing, letting us sort out the problems with the implementation at our leisure.

Plant B was not easy to work with at all. They demanded many changes. They required inordinate amounts of documentation and training. And when there were problems that affected production, they refused to find ways to work around the problems. They waited on us to fix whatever the problem was, even if it meant a loss of production. You can imagine that every time the new system caused a drop in production, the executive team heard about it.
Fast forward a year….Plant A was still easy to work with. The supervisors still kept production flowing. But the ERP implementation was a mess. The support burden was higher than it had ever been. The many work-arounds created during the implementation had become cemented processes. As a result, the new system was just as inefficient and fragmented as the one it replaced.

But Plant B, with the unreasonable management, was a resounding success. They rarely needed technical support. All of their employees were well trained. There were no work-arounds to impede the flow of information. The new system was performing exactly as hoped, with some departments seeing a significant decrease in busy-work.

Plant B took the time up front to get it right. At a significant short-term cost, they made sure that the processes and technology introduced by the new system worked, and that all employees knew how to use the system. They refused to accept work-arounds, because they knew how quickly a temporary work around becomes standard operating procedure.

Due to a variety of factors (none having to do with the ERP system) one of the plants as eventually shut down. Can you guess which one?

There are some legitimate reasons to do so. Among them are to make sure scarce bandwidth is not being used for frivolous tasks and to ensure employees are not engaging in illegal, unethical, or harassing behaviors. But when I’m told that “I want to make sure our employees aren’t wasting their time on the internet,” my response is “How would you like them to waste their time?”.

To be sure, the internet can be a time-sink. But so can the water cooler, the cafeteria, navel gazing, and phone calls. Producing reports for an executive that details how employees spend their time on the internet is not going to keep people from wasting time. It may well discourage people from wasting time on the internet, but I’m confident that they will find another way to waste time.

Another strategy is to deploy a firewall that prohibits certain sites and services. For instance, at one client, we block access to streaming media, instant messaging, social networking, gambling, porn, and hate groups. In my opinion, we should be blocking access only to porn, gambling, hate groups, and the like. This would allow employees to waste time updating their Facebook status, but is that such a bad thing?

The ubiquitousness of the internet and modern communications has caused many people to blur the lines between work and home/play/non-work. The younger the worker, the more likely she is to not feel that her job responsibilities are something she takes care of only M-F, 8-5. She may work remotely from home every night from 10 PM until midnight, because she feels she does her best work then and she doesn’t have to deal with interruptions. So, if she still comes into the office for a typical 8-5 shift, is it a problem that she spends an hour during the middle of the day updating her mountain-climbing blog?

Technology provides different tools for supervisors to use to manage their people. But, it doesn’t replace managing people. The best managers will see technological tools not as tools of enforcement, but rather as tools of engagement.

The truth is, there are a lot of terrible I.T. people out there. And, you don’t have to be a techie to find them. Here are a few ways to tell if your I.T. Dept needs an upgrade. If you’d like a more exhaustive guide, please request our white paper.

Your systems and infrastructure aren’t documented. You should be able to receive an immediate response to a request for current systems documentation. You don’t need to understand it, just know it is there.

Your I.T. staff can’t take a week off. Even if you have a one-person I.T. Dept, if someone in I.T. can’t take a week off – with no phone calls or emails – you have a problem.

You have firefighters and heroes in your I.T. Dept. Excellent I.T. Management isn’t glamorous. In fact, it is relatively anonymous. If your I.T. staff is constantly putting out fires or saving the day, they suck.

You’re told something can’t be done. Given enough resources, anything can be done. If you have a request that is difficult or costly to implement, you should be told why in business terms and presented with alternatives. You may decide that, given the cost, it shouldn’t be done, but don’t let your I.T. people you tell it can’t be done.

Your employees look for ways to bypass the I.T. Department. If your I.T. staff is unresponsive or unhelpful, your staff will figure out how to get things done without them. Proper I.T. Management requires some measures that end-users may not be fond of, such as restricting access to certain web sites or personal email, but your I.T. staff should be in constant communication with their clients – your employees – so that they are viewed as helpful partners, not enforcers of rules or the place where support requests go to die.