If you're new here, please subscribe to my RSS feed. And follow me on twitter. Thanks for visiting!
For years, Search Engines have been dominated by the SEO pros who can get the most links from authoritative websites, with good link text, and who know the ins and outs of on-site markup and linking strategies.
While those tried and true “Hard SEO” methods have their firm places at the foundation of the search engine algorithms, the world is changing.
I think that SEO is getting softer.
Surely the search engines want their users to demonstrate happy behavior (sticking around a site) rather than dissatisfied behavior (being repulsed by a site and leaving as quickly as possible).
Whether the user user coming from natural search results or PPC, the search engines want happy users. That’s what I mean by “soft”. The human side.
Since search engines cannot actually know the feelings of their users, they must watch behavior and draw conclusions. They’re likely doing this through toolbars, analytics, and who knows what, in that they can see if a user clicks over to a site and comes right back for a different result or immediately leaves the site (a bounce) and they can see to some degree how much time the users are spending on the website (page views and visit times).
Recognizing this, we set out to redesign one of our already successful sites to improve the user experience, and work on this “Soft SEO” strategy. We spent November through January on a redesign to provide a better user experience, and as a side effect, improve our “Soft SEO”.
Our bounce rate was previously at a steady 70% or so. 70% of people who came by ended up leaving the site after that initial page view. On our site, a lot of traffic goes directly to pages where we feature products offered by our partners (affiliate relationships) and many users would head off to look at those partner sites.
For a “thin” affiliate site, you’d expect very high bounce rate, closer to 100%. The goal of a thin affiliate site is to get traffic in and out as quickly as possible, through an affiliate liink. That is good for converting visitors to sales, but if the user didn’t spend any time on your site, the search engine probably isn’t going to assign much value to that.
Our goal is to develop a community, and therefore we want people to stick around. So we improved the look and feel, improved usability, and we added new features and highlighted the existing features, so that more users stick around, sign up, and participate.
On the day we launched in late January, the bounce rate dropped from 70% to about 50%.
And here’s what happened to our page views. It went from an average of about 1.6 pages per visit to 3.0.
I see this as a successful redesign. Granted, these are only a few statistics that are easy to demonstrate and I’m not suggesting that you look only at these. In fact all of our analytics point to happier users. And the great thing about Soft SEO is that we don’t really worry about how changes are affecting the search engines. These are long-term strategies that just improve things slowly over time. I’m confident that the results there will be positive.
And we’re not satisfied. In fact, we’ve been busy on phase 2 of this redesign for the past 6 weeks. Hopefully we’ll see additional steps in the right directions.
Take a look at your websites and think a bit about the softer side of SEO. How your visitors interact with your website. Do they do things that demonstrate that they’re interested in what you have to offer? Are you providing a valuable service?
Are your visitors sticking around? Or are you quickly sending your visitors elsewhere? If so, you may feel like you’re doing the visitor a service, but this is a service that the search engines feel they should be able to do and would probably rather bypass the extra step.
Added: got a few questions about how conversion rates are impacted by this. My expectation is not much, which is fine with me, as that’s not the primary goal. Though an important goal is of course to not hurt revenue. Here’s a chart for the same time period.
By virtue of this type of data, the graphs are very erratic, so I’ve included some trendlines that show that things are pretty flat, maybe a little bit up.
