Some years ago I was the IT director for a relatively large company, something like 1000 employees. I inherited an absolute mess. The chief operating principle had been “spend money as an absolutely last resort” for at least 15 years prior. As a result the entire infrastructure was in shambles.
Email was handled by a tiny DSL provider in the middle of nowhere, POP3 only, no outbound offered, this had to be handled by each of the different branches local ISP provided email account. There were dozens of 10/100 switches throughout the company. As with anyone who doesn’t know what they’re doing sonicwalls were used everywhere but not just any sonicwalls but TZ170s spec’d EOL at least 10 years prior. Of course those were bottlenecking internet bandwidth everywhere. This was a retail chain so most locations had cash registers running on ancient Wyse clients.
Additionally a huge amount of their business came from special orders and in house CAD stuff. Most of the PCs for these people had come from a massive community college auction where hundreds of PCs that were so crappy the college couldn’t use them anymore were sold. These guys bought literal pallets of antique trash computers and sent them to all the stores for employees to then try to run ordering software and CAD, yes COMPUTER-AIDED DESIGN, software on them. Most were running Pentium 3s at best, half a gig of ram or less, onboard graphics, harddrives in the 2 digit storage range, etc. Also almost all of them, if not all, were running pirated windows XP, long past its prime, as well as, office 07 again many years past its use by date.
The Point of This
So what’s the point? What was the impact? How does this relate to saving money?
Well the impact of all these decisions was at any given time there was probably 10 employees who needed a PC that didn’t have one. For each of those there was probably another 10 who had critical software that wouldn’t run (yes that’s 10 times 10 or 100). And literally every employee was forced to operate at a snail’s pace because their PC was so slow, and their internet was permanently bogged down, and their email server wasn’t responding. Who knows how many thousands or millions of emails were lost over the years due to the completely insufficient email provisioning and as a result how many hundreds if not thousands of sales were lost, or orders never delivered to suppliers resulting in missed deadlines and angry clients, perhaps never returning due to this, etc.
Lets do some very basic math with some extremely conservative numbers. Lets say each employee lost 30 minutes a day due to slow workstation (vs something modern but still cheap, like say an i3), slow internet, email outage, or just outright PC failure. 1000 employees times 30 minutes is 30,000 minutes. That’s 500 hours a day. Lets again be super conservative and say all these people only made $10 an hour, the place certainly paid like crap but I do think it averaged a bit better than that, That’s $5,000 PER DAY down the drain in lost man hours.
The Rub
Think of how many brand new very decent PCs you can buy with $5,000 a day. Well I’d estimate about 1,875 PCs per year. Enough to outfit your entire staff twice over. What about those internet issues? Well I replaced all those piece of crap sonicwalls with PCEngine APUs for about 100 a piece. I even set them all up with VMWare so the firewall/router was PFSense and a second VM of Debian acted as a domain controller, man it was slick. All those 10/100 switches were upgraded to 10/100/1000 HPE switches which were only a couple hundred dollars each.
Eventually I put in an on-prem Exchange server and also installed all new servers because their ERP was hosted centrally on more antiqued hardware connected to a crappy cable connection (with no backup) which I also upgraded to redundant fiber connections and a custom built cellular based solution.
So more likely than not I was able to complete rebuild the entire company’s infrastructure for less than they were losing on man hours in a year. Eventually I build an in-house VOIP solution and knocked the average phone bill down by about a grand per location per month as well because they were still using POTS lines with very expensive long distance.
Take-aways
There’s a couple challenges here. First off these numbers are extremely difficult to quantify and/or prove. How do you demonstrate lost man hours due to a slow PC? What’s worse is most CEOs, all but one or two I’ve worked with, are complete boneheads. Most of what I wrote above will just sail over their head while they’re distracted by thoughts of golf and cutting benefits or whatever is going on in there.
So how do you sell these upgrades? Well unfortunately one of the hardest parts of being an IT director is the salesman role. You need to be able to sell your ideas. Obviously a cost benefit analysis is vital, if you can produce one, but sadly the more critical factor is the relationship you build with your boss.
It’s tough considering a good IT director probably is the opposite of a people person but if you want to get anywhere in that role you need to gain your boss’s trust. Once you’ve done that and they can take you at your word you can get just about anything done. Of course if you don’t know what you’re doing you this will blow up in your face pretty fast. The other thing that can occasionally be helpful is real-world ramifications. I was getting regular calls from some auditing branch of Microsoft when I was this place. They had a lot of questions about how an organization that size had practically zero licenses. And yes I did fix all that and they were totally legal when I left. They then outsourced the IT to a crappy local (to the head-quarters in the middle of nowhere) MSP (managed service provider) and last I heard they were doing a great job of getting everything back to running like trash and making really stupid recommendations, like going to Microsoft 365 for only a few hundred thousand dollars a year for something that was already handle in-house with practically zero cost. Did you know most MSPs are Microsoft “partners”? That means they get a cut of the MS365 sales. Not sure how that’s relevant, wink. They also ripped out all the VOIP phones because they couldn’t figure out how to manage them… But you know 3CX is probably just a little too niche and doesn’t work right. It’s more designed for little companies like Pepsi and Holiday Inn, and Wilson (some of their other patrons).
If you’re on the other side of this equation there’s just as valuable of a take-away for you. You need a good IT person. If you have more than probably 5 PCs and they’re critical to your operation you better get one. You can temporarily get away with an MSP but I highly recommend finding a full-time in-house person (and I mean on the payroll not required to sit in a office in the same building as you, this is more meaningless to an IT person than anyone else in your organization). Also if you have a person running your IT that’s doing anything like my examples above they have no idea what they’re doing and are presumably not equipped to think about the long term impact of their decisions, that’s the last thing you want to be true for this role. They should be looking further forward than probably anyone else.
If you found any of this useful subscribe below to be notified when I publish new articles. And pass this along to anyone who might find it useful.