Authentication has a problem: While it's a good thing, it's not an urgent thing for email marketers.
It has a small impact on an individual marketer's deliverability, but in aggregate, it's one of the things that marketers can do to truly help ISPs and the receiver community fight spam and reduce messaging abuse.
As my colleague J.D. Falk, who is a recognized messaging abuse expert, wrote in the Return Path blog this week, "Yet in the background, seemingly far from anything that makes end users excited, email is slowly becoming more secure as authentication -- primarily DKIM -- grows."
J.D. cites a number of cool new features from Gmail in particular that are based on DKIM authentication. Like the key icon feature for PayPal and the new one-click unsubscribe option. Those "shiny new benefits" (as he calls them) may be the incentive that marketers need to stop waiting and start authenticating across all email message types.
What is your status on authentication? Still procrastinating? What do you need to see in terms of benefits in order to participate in this important industry movement?
Please comment below or on J.D.'s post the Return Path blog - and I'll follow up here to share any insights.




Cryptography is always a barrier to adoption of almost anything. This showed with DomainKeys.
DKIM has become extremely complex and convoluted, suffering from a huge case of "design by committee" and "kitchen sink" mentality. I fully expect that to slow adoption.
Derek
Posted by: Derek Harding | August 04, 2009 at 06:02 PM
If ISPs took the stance of "it isn't authenticated so I wont deliver it" marketers would pretty quicky adopt authentication methods. It doesn't initially have to be DKIM, which obviously has overheads, but SPF should be a minimum standard.
Posted by: Jake | August 06, 2009 at 11:26 AM
I agree with Derek. I employ SPF for all my clients, but Domain Keys is convoluted and impossible to figure out. I have tried several times to put together/find a reasonable set of instructions to implement DKIM - but have been completely unable to do so. It requires a PhD.
If this is to become a standard, there should be a simple way for the masses to implement.
Posted by: kelly | August 14, 2009 at 01:30 PM