Going along with the general consensus of those "I'm working on a super-secret-confidential-double-dodeca-awesome project and I can't give anything away" lines, I have a question for you HNers. This isn't a throwaway, this is my account. Mind you, I hardly post, but believe me, I browse. I browse like a badass.
I'm working on a tracking tool and social network for a very (very) specific niche market. It's not particularly small, but it's a market that isn't intimately familiar with the concepts of Web 2.0. I'd wager most of them still use antiquated versions of IE. That being said, most of them are involved in forums of some sort or another. In fact, what I'm trying to do is centralize and standardize what they do in some of these forums. I'm unsure how my notoriously stubborn audience will adapt from phpBB to my proprietary tool.
The purpose of this project? Well, I'm part of this community (not the web-illiterate part). I wanted to build something for myself (started in Excel, years ago, ugh). After speaking with a bunch of people also involved, it seems that virtually everyone I spoke to is enthusiastic about my idea. I've gotten a group of beta testers that I define in a range of browsing skills to see how my program works. It's an interesting challenge, building a web app for people that barely understand the concept, let alone the definition. I've never built a site this large all by my lonesome, so I'm excited to see how I manage, from conception to production without a cofounder or team. This is more of a portfolio builder that I scrape a few hours a week together to work on than a business venture, but it wouldn't be terrible to earn a little advertising revenue.
Have any of you dealt with a large, decentralized, and non web-savvy audience before? If so, any advice?
EDIT: from my comments, revorad broke me down and got me to spill the beans on my project. It didn't take long. I would make a terrible spy. Here's the project info, in a little more detail.
It's a target score tracking tool for gun owners, built around aggregating scores, ammunition makes/types, individual gun scores, etc to let individuals track how well they do (in general, with one of their guns, etc) and track performance metrics across the user base to determine which ammo works well, which guns perform best, etc.
So, a tracking tool mixed with a social network. I guess.
It started as something to help me track my scores and see how I progressed. I made a score calculating algorithm just using Excel that factored in ring size, distance, wind, and a whole load of other factors and it worked really well. I had one of those "this is madness" moments and started porting it into a CakePHP app.
There are a ton of AJAX page interactions, UI/UX elements, etc that are very different from what traditional message boards and forums use.
For the older crowd that uses just email I would suggest not overloading them with information. Too much information on one screen can be a pain to sift through and usually the younger crowd doesn't mind it as much - mostly because they don't read so much as skim, though - but the older crowd will find it to be a chore.
Users understand the back and forward buttons on a browser. Try not to create a situation where the browser buttons would not function as expected. Too much AJAX or wanton use of a lightbox could be considered such situations.
Lastly, visually offsetting the content of your site versus the layout (menus, headers, footers, etc.) can go a long way to guiding usage of your application.