data driven decisions

Deciding on Web Content: Data or Gut Feelings?

Omniture’s survey for 2009 found out a lot of interesting things, but one simple fact really leapt out at me.

80% of respondents made decisions about web content without using any measurable data at all.

Omniture suggests that data-driven decisions may actually form less than 20% of the total, because many people will define “It worked before” as a data set. But even if we suppose that 20% of the people who took the survey actually base decisions about web content on some information, that leaves 80% who use “gut feelings” — and maybe creative sparks or something. They didn’t mention that, but I see a lot of random inspiration among clients, so I figure that’s probably part of it.

So what measurable data might you use?

  • Data about search. I like to look at what people are actually searching for. Google keeps track of this, and they should know. While popular searches aren’t the whole story — the greatest number of search queries may be completely unique, just as the greatest number of sentences are — they can certainly give you useful information. For example, I’m currently writing a site for a natural foods store. Local search results show that more people search for their initials than for their full name. Their current website never uses the initials, and they’re not on the front page of the SERPs for those initials. You know I’m going to change that.
  • Data from your customers. Looking at what people use to find your website now can be useful. That natural foods store gets very few people in via those initials — no wonder, since they’re several pages in. As we know, that doesn’t mean their customers don’t search for them in that way, just that they don’t find what they’re looking for. So it’s good not only to look at the keywords that bring people to your site, but also to ask people what they look for. I find that customers often search for brand names and product names, but that retailers often think their customers will search for the name or type of their store. It’s easy to check on this. Testing your site is also a good plan. And asking people why they visit your website — combined with your analytics data, this information can guide your content in important ways.
  • Data from your business. Tyler Katzen of Onix Web Development yesterday suggested asking clients for the percentage of their revenue that came from each area of their business. I was impressed with his cleverness. That one bit of information covers a whole bunch of questions I usually ask people. Gather this data for your business, and compare it with your website: are you putting a lot of your space into something that doesn’t actually represent the profit center of your business?

I often explain that SEO copywriting is as much math as literature, and that’s the truth — if you do it right. With 80% of businesses doing it wrong, your website can stand out for search and conversion if you make sure you’re part of that 20% using metrics to make content decisions.






Leave a Reply