It's been about a week since we launched DataSphere, and so far, the response has been interesting. There were a few write-ups in some blogs and it was also the centerpiece item on Poynter.org earlier this week.

In addition to the position of a Data Delivery Editor and the purpose behind publishing databases on the Web, the DataSphere was also a chance for our tech team to try some new things that would make the data visible to search engines and accessible to as many users as possible. Whether we succeeded or not in that regard will be borne out through traffic data and other measures that we track on a regular basis.

Given that this was new ground for our tech team and the Web site in general, I wanted to pass on some lessons learned and reasons why we developed DataSphere in a particular way. Before delving in the geeky tech talk (stop reading now if this stuff makes your head spin), I should give some background on how roanoke.com appears on your screen.

We're an internally hosted Web site that runs on Windows servers. We use Microsoft SQL and ASP.NET as our database and programming language. We also use an internally built content management system. We have toyed with the idea of switching platforms and systems, but after making a significant investment in the servers and the software that runs them, it's kinda tough to go back to our corporate parents and beg for more cash to buy different servers and software, especially when we asked for the existing stuff just a year ago.

While we do use .NET for development -- and I know that every programming language and system has its disciples and detractors -- we've made some strides to minimize the extra code that .NET puts on a page. It's not gone completely, but that extra code should not impact performance, readability or the browsers' ability to display a page.

We also took great care to use generic-type .NET controls. This gave us more control over how the page was built and actually eliminated the extra tables that .NET is notorious for adding to a page and the extra tags and attributes that its likes to throw in there for kicks. The only table that appears on the page goes with the results, which is the purpose of a table.

DataSphere is also the first launched project to implement findings by a search-engine-optimization consulting firm that we hired over the summer. Those findings affected the design and placement of elements on the site. In a nutshell, when Data Delivery Editor Matt Chittum gets or creates a new database, he'll work with our IT team to develop a description that includes lots of keywords.

By placing the description first, there's a much better chance that Google and other search engines will pick up the description when it indexes the site. As the firm recommended, the keywords matter, not necessarily the data itself. The next item on the page is related links. By placing links to other related stories (mostly ours) it helps the overall site with SEO placement.

So what are the results? So far, we appear as the third result when searching "DataSphere" on Yahoo! and Google. If you search "Frank Beamer's record," the database detailing the Tech coach's career shows up second in Yahoo! and first in Google. Traffic to the site is just over 2,600 page views since its launch. Not terribly huge numbers, but a start.

• Share this entry
  • Digg this entry
  • Mark in Ma.gnolia
  • Tag with del.icio.us
  • Seed Newsvine
  • Post to Google
  • Add to Yahoo!
  • Share on Facebook
  • Submit to Reddit
  • Related at Technorati
• Posted by john jackson in Technical issues
• :: Comments (0)