Do niche web applications still strive for legacy browser support?

Question

I'm developing an organization web application geared towards creative thinkers, with productivity tools for freelancers and small group collaboration; designed responsively, the app will work on mobile, tablet and laptop/desktop.

It's completely anecdotal but my target audience (on laptop/desktop) is typically aware of Internet Explorer 6-8's inadequacies and/or at least "hip" enough to use IE9, FireFox, Safari or Chrome.

---

I'm curious about how common it is for niche web application developers to disregard legacy browser support, especially when supporting them will more than likely introduce problems surrounding integral application specifications.

The user-experience and overall quality of my application will be far greater if I don't spend time problem-solving and making concessions for legacy browsers; anything I should be conscious of if I decide not to support IE8?

Edit: I've been looking through http://gs.statcounter.com/ to get an idea of browser version usage in different geozones, and it seems that IE7 is around 3-4% in most of my target geozones. I updated my post and removed specific references to IE7, seeing as I feel comfortable not supporting it.

Answer 1

I would go ahead with your intuition of not supporting legacy browsers if you feel they only constitute a small part of your target population. However, I would recommend clearly highlighting that the browsing/interaction experience would be greatly enhanced if the user uses one of the modern browsers (mention those browsers clearly) and ensure the users see this information prominently.

The way you display this information is dependent on your application but ensure that people are aware of it and don't tend to skim past it.

From an anecdotal point of view, I work on projects for a client which uses only IE and one of our projects required me to design a internal portal for them. The client had certain requirements which would have needed CSS 3 support and though 80 % of the client uses IE 9, there were reservations about whether the remaining 20 % would get affected by this decision. We decided to go with the CSS3 emphasis and we introduced logic in the code so that any detected IE8 users were immediately informed that though they would be able to use the portal, some functionalities would be unavailable to them. Within one month of rolling out, we found that 60 % of the users who had visited the site using IE8 were now using IE9.

Answer 2

Another nice point with StatCounter is that you can easily spot the trends too, so if you know how long your project will take to develop you can get a good idea of the usage of that browser in the future too. ("Currently only 3% of our target audience use IE7, and by the time it is actually release that figure will be less than 2%...")

Not only that; take into account the expected lifespan of the product. If you're spending an additional 20% of development trying to get it to function in a browser that only 2% of people are using right now, if the product lifespan is expected to be several years then you'll have built in redundant code early on that isn't needed soon after launch. The more redundant code the more there is to go wrong (and the harder it is to debug) during the life of the application.