Starting next week, we’re making our Email Filtering easier to deploy and more intuitive for users. The change is quite simple–instead of delivering mail to a webmail-based Spam Quarantine by default, we’re instead making it possible to tag the spam, add an optional and configurable header, and then deliver the spam through to the users’ regular inbox (filtered into a spam folder, either server side, or in the client).
Making things easier
We looked at how a lot of software anti-spam solutions worked and decided that if we wanted to allow you to get rid of the headache of operating something like Spam Assassin on you own email server, then we shouldn’t create any additional headaches for you to make the switch.
So, we set out to replicate the setup of many of these self-run solutions with the anti-spam software in our data center, maintained by us. That meant allowing you to configure a specific header to add to messages flagged as spam and allowing you to tag the messages as spam in the email subject line.
The net result is an Email Filtering service that is effectively “hot swappable” with your existing setup. To your users, the change can be invisible – they don’t need to change their email client, or be retrained to visit a Spam Quarantine to view spam email. They can continue to receive their email, and messages flagged as spam as they are today, using your own setup.
Fewer hassles
Of course, you get the benefit of not having to maintain those spam filters or worry about things like blacklists, or other abuse issues.
Here’s a graphical representation of the new email flow. You can see on the left, with the Spam Quarantine, users have to deal with two interfaces to email. On the right, with “pass through” enabled, it’s a single point of contact with both wanted email, and spam (tagged and delivered to the spam folder).

Email Filtering starts at just $0.10/mailbox per month. Consider the time spent maintaining your current setup, including downloading and applying rules changes, or modifying settings, and we think you’ll agree that having us filter your users’ mail is the best way to go.



Last week some of the OpenSRS team headed out to San Francisco to attend the 
I’ve thought about it some more over the last couple of days and I want to take a shot at building some business theory and context around this argument. I warn you, I’m going to throw some freshman macroeconomics at you and probably misremember Freakonomics. I might employ the phrase “irrational exuberance” and, just for fun, I think I’ll lob in a venture capitalist conspiracy theory.
So, lots of good reasons to support free Web services. Here’s the problem. These are all way more difficult than they sound and I believe most free Web services today are following a different model entirely – build a huge following now and figure out how to monetize later. Hey, it worked for Facebook. It will probably work for Twitter. But these services have succeeded by getting so many millions of users that they only need to make a few pennies off of each of them to be wildly successful. And maybe one in a thousand companies are going to amass that sort of customer base. It reminds me of the junior drug dealers in Freakonomics who are accepting wages and risks that don’t make any sense because they are fixated on the one guy who got to be the kingpin. Or the college basketball players who never go to class because they’re all convinced they are going to be NBA players.
These companies are subsidizing their users – offering a service that supply and demand would set at $2.95 a month, for example, for free – with resources they are borrowing from an uncertain future. Remember the impact of a subsidy on supply and demand? It shifts the supply curve to the right, lowering the price and increasing quantity. To continue to meet the demand, the service needs to continue to subsidize it themselves, find buyers for their audience or their data or wrestle the revenue from their users.
e on just one of these, it starts to feel more like a lottery ticket. And the key when you consider a pricing strategy really ends up being time. If you start off with a paid service or advertisers or a buyer for your data, you’ll find out a whole lot sooner whether you can satisfy all the parties involved. If you start out offering the subsidy, you could pour years into a service only to find out later that there was never really any way to shove revenue generation into the experience.