Matt (BlackBelt) Jones asked me and some others some questions about ROI for wireless in the workplace. I sent him a direct reply, but thought I'd make the answer public here.

I'm afraid that I don't have any hard ROI numbers, but maybe some of my own analysis will be worthwhile. I offer it to you for what its worth:

1. I don't see wireless as a replacement for wired networks in most instances, at least in the US. Many buildings are fairly new and either already wired or easily wired after the fact. When I moved into a building a few years ago, I was able to put in wires for about $200/drop which included 2 Cat-5s and a phone. As a one-time build out expense, its too cheap NOT to do. Until I'm running phones wirelessly too, I've got to drop lines anyway.
2. I think wireless becomes compelling when one considers places like campuses or even conference rooms. There, wireless is the only logical solution. My plan at Utah was to build out conference rooms, cafeterias and the like and let user demand drive it into other parts of the building. Maybe that will still happen.
3. Wireless is even more compelling in any environment where most workers move around or where laptops have become the standard option. If it were up to me, I'd issue every employee a laptop (instead of a desktop) just to see what happens. I think its changes how people work. When a company is mostly buying desktops, however, its hard to justify wireless vs wired on a cost basis (at least for fairly new buildings).
4. I think the only reasonable security solution at present is to put wireless hubs on unrouted 10. or 192. nets and force everyone on the wireless net to authenticate using a VPN. This has two big advantages: (1) its pretty secure and (2) we already rely on VPNs to secure us in lots of cases, so its massing our security efforts. The fewer total security solutions an organization has to deploy, the better it can make them. I think including security in the wireless standard was a big mistake. Security ought to be a layer on top of the carrier and it ought to be flexible.
5. A large organization is foolish to not get out ahead of its users on something like wireless: (a) lots of people want it; (b) its cheap---cheap enough to put on a corporate AMEX; and (c) its a security nightmare. This is a recipe for disaster. The only way to fight it is to install wireless networks correctly ahead of user demand. If you don't, they'll all go out to Best Buy and be plugging them into the corporate network without any security turned on at all. Ouch!

I've always believed that wireless is a companion to and not a replacement for wires in a fixed environment. Maybe I'll be proven wrong. That said, I think that I'd be inclined to provide wireless access everywhere I could because I think it changes interactions. It allows me to take my laptop to my colleague's offices and sit with them and collaborate easily. I'm much more likely to do it if I don't have to jack in and out all the time. Its difficult to put an ROI figure on that, but it can't be discounted. Good organizations pay attention to fostering interaction.

Does anyone know of ROI studies on wireless in the workplace?