Push me, pull you

I first wrote the title of this column for another column published in 1996. "Using e-mail, I can push data to your computer," I wrote. "Using the Web, you can pull data from mine." Statements such as that sounded reasonable but turned out to be as paradoxical as Dr. Dolittle's two-headed llama. My push-pull rhetoric had been inspired by PointCast, the application that Wired would feature in an infamous cover story on push technology six months later. When this month's Wired revisited the topic in relation to RSS syndication, the push-me-pull-you beast was no less perplexing than it was eight years ago. Maybe it really is just a mythical creature.

Free IT resource Hear how top CIOs turn change into a competitive advantage. Sponsored by HP Free IT resource Attend the SOA Executive Forum: Breaking SOA Bottlenecks SOAExecForum.com/may2007 Sponsored by InfoWorld

With 20/20 hindsight, I don't think that "Push!" was the idea that leapt to mind when I first used PointCast. As I recall, my immediate reaction was: "a rich Internet application!" -- or rather, since that phrase did not yet exist, "a non-browser-based app that uses Web protocols!"

PointCast was never really a push application and neither is e-mail. There are programs that listen on TCP ports waiting to receive data -- mail, Web, and FTP servers belong in this category -- but they're special cases. Client machines guarded by NATs and firewalls almost never accept unsolicited inbound traffic. What PointCast did was what RSS and e-mail readers do today: subscribe to well-known addresses and poll them for new information.

Don't publishers and e-mail authors push the information to those well-known addresses? Sure. But so do Web publishers by pushing pages to their sites. Suppose I want to be notified of such updates. In all respects but one, there's no meaningful distinction between an e-mail reader that polls hourly for a digest of changes and an RSS reader that polls hourly for the same digest. The lone exception has nothing to do with push vs. pull and everything to do with the locus of control. You can subscribe me to a (poorly managed) e-mail list without my permission, but you cannot subscribe me to an RSS feed without my permission.

Recently I spoke with Dave Lewis, vice president of deliverability management and ISP relations at Digital Impact. His company's motto: "Making e-mail marketing more effective is our single-minded passion." In one of his online essays, entitled "How to Keep B-to-B E-mail From Getting Caught in Filters," his first rule is "Get permission."

I argued that RSS does away with the need for marketers to ask our permission, for us to grant it, for marketers to play by the rules when we revoke it, and for us to trust that marketers will play by the rules. With e-mail marketing, control resides with the sender and permission is a "best practice." With RSS, control resides with the recipient and permission is an inherent property of the medium.

I feel Dave's pain. E-mail direct marketers are stuck between a rock and a hard place. They believe e-mail is necessary because it's an "intrusive" medium, yet they are forced to neuter e-mail's intrusiveness by complying with the opt-in gold standard. Unfortunately, there's no middle ground. With RSS recipients can have, and increasingly will demand, control of the channel.

Dave and I agreed on one point. "You'd be crazy not to communicate with your customers in their medium of choice," he said. My preference is RSS. Trust me with control of the channel, and I'll be more likely to trust you with my business.