Vista Not Driving Sales

Let me begin by saying this is not another rant about Windows Vista- I happen to like it. I would never recommend upgrading an existing system to Vista, but if you're ordering a new machine with the right hardware (dual-core cpu, 2gb+ ram) it's a great, reliable experience. Instead this entry will focus on the uptake of Vista amongst SB's from my role as their IT consultant.

In versions past there has usually been user interest generated from TV and print media before I even walk in the door. We'll talk about what's new, check compatibility with their existing applications, and when they're ready, order a new system with the current version pre-installed. Because it's a new system, often the SB owner or decision maker get's first dibs.

At this point it's worth talking about the "viral" spread of Windows within an SB. Once the new system is in place the user (and often users) find themselves more productive and enjoying the improved interface. Not all of this is due to Windows of course, the fact that the new system has a bigger screen and faster hardware helps too, but often users judge by what they see, and what they see is Windows. It's not long before the other users start griping about their own systems until the owner/decision maker relents and eventually replaces all of their workstations over the next 1-2 years. While I'm sure to take flack for the "virus" analogy (and I really don't mean to infer the connotation), with Vista, the virus never takes hold.

Let's talk about Vista at two of my largest clients:

At the first business the decision maker is in this case a consultant's dream user- not in IT themselves, but something above a power user who can handle the day to day IT questions of the staff they work with. A new system was ordered with Vista Business and their company applications installed. While some older, lesser used applications didn't work, generally the experience was satisfactory, and the user has reported enjoying many of Vista's new features. The problem however is learning curve- while this user figured things out fairly quickly, their time is already heavily taxed between their work responsibilities and handling staff IT issues. So at this point half a dozen new systems have been ordered and deployed, all with Windows XP. After running it for 5 years, it's what their staff knows.

At my second example company, the sell was a little harder. Like before, the decision maker (and in this case also the owner), is somewhere above the power user level. They were personally interested in Vista, but also had "heard it had problems". He wanted a new system for himself and was inclined to go with XP. After some discussion of new features, I finally assured him that if he disliked Vista, I would roll the system back to XP at no charge. That was 6 months ago, and a few weeks ago I got the call that said "roll it back". Again, reliability was not an issue (though some applications did not work), but the user found they were not as productive because they were used to XP.

In either example, the new features were not able to overcome the obstacle of retraining a staff 5 years entrenched into XP. These are just two examples, but I could go on. In fact, of all my clients, I support just one Vista system at this point- over a year after RTM. Long-term these companies will upgrade to Vista or its successor. As XP ages and support fades, the cost of maintaining it will reach a tipping point, and of course, at some point it will no longer be available for purchase.

The issue is not will SB's transition to Vista, but will Windows continue to drive sales. For this release cycle, the answer is no- and possibly those days won't return with future iterations either. As a consultant, the initial investment to get Vista into a business as a catalyst to drive sales and training has not been a profitable one, and the losses serve as a real disincentive to push it further.

The Argument for T1 in a DSL World

For the last 8 years or so DSL has been the obvious choice for the majority of small businesses (SB's) I've worked with. The cost/benefit ratio of having a relatively high bandwidth dedicated circuit (7Mbps/896K) for something south of $100 is a fairly easy sell, even to business owners who still don't equate internet access to water in terms of necessity. For general web browsing and e-mail applications it's a wonderful solution- as long as those continue to be your needs.

The issue of course, is feature creep. Let's talk about a particular scenario I recently worked with:

Company X has been a progressive DSL customer since moving off of 56K Frame Relay around 2000. First to 256Kbps, 640Kbps, 1.5Mbps, and most recently 7Mbps. As bandwidth grew, so did the number of internet applications- Basic POP3 mail clients gave way to an on-site Exchange Server, 3-4 small websites were hosted internally instead of at an ISP, remote users began accessing the office via VPN and FTP, and locally run database applications were shifted to an offsite Citrix ASP.

There was no specific day or project that tipped their configuration into a state of degraded performance, as much as the random "quirks" one might expect with any internet application began occurring more frequently, eventually to the point of causing user frustration- especially in the form of dropped Citrix connections.

Bandwidth usage is a difficult thing to assess as it's a constantly moving target- but aside from explainable outages (ie, the user who CC'd a 10MB PDF to 200+ receipients) not a single measurement indicated network congestion as a source of the Citrix disconnects.

My first move was to contact the line provider (in this case Qwest) to verify performance of the circuit. Next I replaced the DSL modem and Linksys router (cheap equipment, well known to fail)- but also to no avail. While I wasn't convinced the router was truly swamped, upgrading the client to a Cisco 871 where I could better monitor and regulate network traffic seemed a solid move.

The core issue however, was not the amount of bandwidth available, but the latency of it. Pinging Citrix servers resulted in 55-110ms trip times. With the T1 circuit, ping times dropped to 5ms! Number of user reported disconnects since the switch: zero- even though bandwidth dropped from 7Mbps/896Kbps to 1.5Mbps.

The fact that a T1 line offers higher reliability and lower latency isn't news. The issue in convincing a SB to switch is often financial, and this is where it's critical to step back out of the IT consultant role and put on the SB equivalant of the CFO hat. In this case that meant making a broader analysis of the customer's telco, long distance, and voicemail needs. Convincing a SB to spend $400-$600 MORE per month on a data line is an uphill battle to say the least, but start subtracting out the costs of POTS lines, user downtime, and the soft costs of IT staff responding to those users and you quickly begin to build a budget neutral solution.

Should all SB's switch to T1? Of course not- for most, DSL will remain the standard-barer. But it's helpful in a world where sales are ruled by Megapixels and Megabits to step back and look at how this alternative solution can serve the broader needs of the business in an affordable manner.