The End of Profitable SMB Infrastructure Business
The fact that its harder and harder to make money in SMB infrastructure support is nothing new, you've read about it here over and over. But now some of you are starting to see it and I'm glad I'm not the only one talking about it. Check out this blog post from Brian Williams on one of the leads that fell through:
"Two days later I followed up via phone to see if there were any questions or concerns, at this point no still getting additional quotes. A week later another follow up to see where things stand, we’re almost done gathering quotes but we’ll be honest you’re the highest bid so far so we’re going to exclude you from our final selection.
One of the prospects informed us that the winning bidder informed them that their existing 4 year old PC’s would not need to be replaced (which we suggested replacing), and the winning bidder quoted a white box server with Intel Dual Core processors not Xeon’s as we had proposed."
He goes a little further:
"We believe we’re doing right by our clients in quoting servers with SAS drives, Xeon processors, Dell servers with silver warranty for 4-hour on-site support, keeping equipment on a 3-year replacement cycle and much more. We will stick to our guns because when all is said and done if we start to change our philosophy to win bids then we’ll end up with clients who have equipment that is down, lost revenue and an increased stress level for all."
Now, Brian is an IT consultant so I will defer to him based on the expertise, but I too would have shot down an SMB server with SAS and Xeon drives. Frankly, in all our data centers I have yet to meet a client that has been able to justify a Dual Xeon configuration - from the first time Xeons came out all the way to the current quad core 53xx that people love to spec - do you guys do it just to artificially lift the price of the system so your commission / services look tiny by comparison? Especially if you're rotating systems out every 2-3 years, unless the customer is pounding the SQL end of the system with some heavy duty LOB or has 50+ users (in which case it shouldn't be a single server deployment anyhow) I just can't reason it out. Either way, as I said, I am not an SBSer or an SBS consultant so I am sure Brian had a very good reason for his recommendation.
But what he discovered, and what many of you are discovering, is that its harder and harder to make a buck on infrastructure work in SMB anymore. The entry level dual core servers are so powerful, the SATA2 drives are quite impressive and honestly we are not back in the 90's and early 2000's where there was a night and day difference between a server with Celeron and IDE hard drive, today less than $1,000 a server without OS brings more than most smallbiz shops need packed with max amount of RAM.
You are also competing against the "It's good enough" and "How does Google do it for free?" and the big iron in a small closet no longer flies and any time you are in a sales situation and are trying to defend your solution instead of keeping the client dreaming about the possibilities your solution offers that the others don't... well, that's where you lose the bid.
So its time to sit down, figure out if its time to retrain or refocus. One thing is for sure, if you're looking to grow, you won't be doing it in the infrastructure business. More on the doom and gloom in the new Vladville Newsletter.

Your hardware solutions have to be flexible. What your standards are for a multi-site, 50 user company will not work for the 5 user SMB shop. This all goes back to having standards, yet doing what is right for the client within those standards.
This is going to be a tough year for SBS as it will be a transition year, especially 1Q and 2Q.
1) We await the release of Windows 2k8...and, eventually, Cougar.
2) There is a growing tap on the shoulder that may indicate Vista may end up more like ME and less like XP, especially with the release of SP3...so that is causing more concern and doubt. (Will more business embrace Vista, or cling to XP even longer? Will they try to skip Vista for the next OS completely?)
3) What services will go to the cloud? (As we know some will.)
4) Managed Services or not? Yes for most, but not all...
5) We see further market segmentation thanks to MS...it's no longer 1-75 (SMB), 75+ (mid to enterprise)...now it's 1-5 (or 10), 10-50, 75-200, 200-1,000, 1,000+.. Where do we position ourselves? What products go where?
All of these things are not bad, most are actually quite good. But as the industry settles down from it's boom and the dust settles...who will be left standing...and what will they look like. What will WE look like? What will YOU look like?
SAS drives and Xeon processors, genius :) I'm guessing you beat up your backspace in the wrong direction again :)
You say in your opening that it's hard to make money in infrastructure *support*, but the example you give relates to infrastructure *sales.* Two different things. It's been hard to make money in infrastructure *sales* for at least 10 or 15 years. Now it's ALSO gotten harder to make money in infrastructure *support*. In the example mentioned above, I think Brian must have overestimated the level of trust that had been established with this prospect. Either that, or he misunderstood the prospect's real needs and priorities. Either way, I can't fault him for the server specs he proposed. The fact is that it's virtually impossible to say with any certainty how much hardware a client needs to run an SBS environment. I've asked Microsoft to work with OEMs to provide recommended specs for Cougar for companies of different sizes and usage patterns. I'm not holding my breath. More importantly, what Brian failed to do here was convince the customer that the extra money spent on hardware now would be worth it if for no other reason than that Brian's company was the best long-term partner, and Brian felt more comfortable with the hardware he quoted.
Brian, if you're reading this, make a note in your calendar to call this prospect back in six months and find out how things are going. If you were right, you may get a new customer. If they were right, then maybe you can learn something from how they solved their problems.
I don't think two can be separated - clients tend to shop when they need a solution and if you wont support what they have (Windows 2000, 98) then a sale has to take place. You don't have to sell it, but believe me, your solution ends up looking a lot more expensive even if $0 goes into your pocket.
-Vlad