Google, yahoo, and hotmail spam filters compared

When a user signs up for slideshare, they receive an confirmation email that they need to respond to in order to activate their account. Frankly, the confirmation rate isn’t as high as we’d like (it’s about 87% since the site launched).
Most of this is due to spam filtering or blocking by the various email providers. For example, my own email provider temporarily rejects emails, causing a 5-10 minute delay in the arrival of the email!
I did some analysis of the conversion rates of users with different kinds of email addresses, and came to some surprising conclusions. This chart sums it all up:

As you can see from the chart, gmail clearly is the best email provider in terms of not blocking email messages. 95% of gmail addresses successfully joined slideshare, more than any other provider. This is a great example of the attention to quality that google brings to the table when it comes to hard algorithmic problems like spam filtering.
Yahoo is in the middle of the pack, slightly below average at 85% (vs. 87% for all providers put together). And the real losers are hotmail (65%: a solid D) and (0%! Brazilians interested in joining slideshare should use another email provider when signing up: I’ve never, ever successfully sent an email to terra brazil).
I’ve done what I can to make sure my email isn’t perceived as spam: for example, I’ve made sure my name servers got a clean bill of health from the awesome utility available at But nothing has had much of an effect: the conversion rates for the different providers are surprisingly consistent over time.
Anyone have any ideas? I’d love for yahoo and hotmail users to have the same success at joining slideshare that gmail users enjoy. What else should I be doing?

One thought on “Google, yahoo, and hotmail spam filters compared

  1. Anonymous January 13, 2007 / 2:41 pm

    Perhaps one simple solution would be, you dont ask the users their email addresses!!.. cuz anyway is not dependent “100%” on email interaction.. no?

Comments are closed.