No longer a Bloomba user

Last month I downloaded a copy of Bloomba and used it as my primary e-mail client (for non-corporate e-mail). I was very impressed with the functionality overall: the searching mechanism really is how e-mail clients should work. But I decided not to buy it, for a couple reasons:

• Price point. I haven’t paid $60 for personal software in a while. Though I grant that Bloomba is useful, I couldn’t justify throwing $60 at a problem that I’m not feeling enough pain on. Where the problem is significant — at work, where I get hundreds of messages a day and get countless attachments — there’s no way I’m giving up Outlook (unless Bloomba can handle my calendar and contacts, that’s a non-starter). So Bloomba would exist as my personal e-mail client. And right now the options are free (Outlook Express, Eudora) or $60 (Bloomba). And it’s not worth $60 to me to solve this problem.

• Spam filtering. When I switched webhosts last month as well, one of the things I picked up I didn’t even know about was spam filtering. TotalChoiceHosting includes, among other things, SpamAssassin for spam blocking. Having the spam filtered at the server as opposed to at the client is much, much more effective. If you’re always on a broadband connection, this isn’t as big of a deal. But the first time you try to check your e-mail on your Treo and SnapperMail tells you there are 248 messages new since this morning, and 232 of them are spam, you don’t really want to spend the 20 minutes downloading all the junk just to see that Mom wrote to say hi. Bloomba requires SAProxy to run, which is a client-side implementation of SpamAssassin. Nice, but it’s another app running on my PC, and it requires me to download all e-mail. (I suppose I could’ve just continued to use the server-side spam blocking, but then I’ve got Bloomba needlessly running through a proxy — it can’t function without SAProxy.)

• Fit & finish. I’ll freely admit I’m a demanding user, but there were some rough edges on Bloomba that ultimately frustrated me to the point that I decided not to stick with it. Some easy fixes that would dramatically increase the usability of the product:

– Replies automatically set the “From” to the account which received the e-mail. Right now, regardless of which account received the e-mail, the reply is set to the default account. When you manage e-mail from multiple accounts (as I do), this gets confusing for your correspondents. – Ability to delete threads. Bloomba does a nice job of grouping e-mails by thread, but when you delete the top-level item in a thread, I think it should delete the thread (or at least ask if that’s what you want). Instead, you just delete that message, requiring you to either individuall delete messages or expand the thread, shift-select all messages in the thread, then delete. That’s just too many steps, and slightly counter-intuitive. – Bayesian learning. Though SAProxy claims that it gets better at filtering spam as you train it (you can either delete a message or “delete as spam”, which is supposed to give SAProxy additional context for what is likely to be spam), in my experience of 4 weeks of heavy use, it didn’t get any smarter at detecting obvious spam. No matter how many e-mails I got from “Wotch” advertising some worthless product, they all got through. This has to improve, or else users will get frustrated (as I did) that you were deleting the same messages as spam over and over.

Bottom line — I really liked the application. But I think it has some rough edges that need to be polished, and I’m doubtful that they’ll find a lot of casual buyers at $60.

Interestingly, Rafe Needleman at AlwaysOn looked at Bloomba last week and concluded that the business model is questionable. Looks like they have their work cut out for them. I wish them luck, as I’d like to see this model mature to the point where it’s an acceptable alternative to Outlook; the concept of instant searching of all e-mail (and attachments) is awfully seductive.