On the one hand, earlier this year we only migrated 70 of 2700 pages and files forward into a new site, and only 3 people have asked for anything back.

On the other hand, even when they’re not the culprits, I often think it is a management failure letting staff write irrelevant content, leaving the web team in a difficult position when they know they should refuse to publish it. Going back to the above example, even though we have the stat’s to prove no-one but him read the page (pure vanity publishing), one of those guys will get their stuff back because he’s on the board. Same-same in my last government job when senior managers wanted to see themselves online. Very frustrating.

I needed to look up an old house deed on our county’s website. It’s nice that this is available online (they’re public records, after all) rather than me having to schlep down to the courthouse. What’s the first thing you see when you go to this website? http://web.co.wake.nc.us/rdeeds/

If you click on ‘view a message from the registrar’ (lower right), you get a reinforcement in case you didn’t figure out who the registrar was. Different suit but similar pose. Now, the latter link is fine for a picture but I didn’t expect it on the home page.

Regarding the university client with 26,000 pages — it’s a rough start. However, looking for pages that are not highly visited may be because of poor design, weak content, and a confusing structure, among other variables. How do we really know what our web visitors are thinking?

However, the “task completed” metric and the examples of how easy it is for site visitors to sign up for a course, etc. are just the tip of the usability iceberg. While it’s imperative that the site be usable enough for people to be *able* to complete a task, from the perspective of an e-commerce site with the *task* of selling a product or service, the completion percentage is not likely to mean the site/page is un-usable. That percentage will vary by offer, competition, perceived value, brand image, page layout, amount of information. Optimization testing is going to be the area that gives the big bang for the buck in e-commerce.

Thanks for all the good info.

I’m wondering, for instance, if analytics can’t be of help in noting an absence of task completion. For instance, peole coming to an online inquiry or order from and then leaving without completing it. Might that not signal an intent to complete the form that was killed by initial reaction to the form?

On another note, I have a university client who just asked my thoughts about the best way to move forward with reviewing and rewriting of 26,000 website pages.

My first response was simple: review your analytics for the past year and see how many pages you can find that relatively few people are visiting. Make those pages candidates for elimination. With any luck, the result will reduce the scope of the overall project that the client is now hoping to complete in three months.

A useful element in Google Analytics is the geographic distribution of visitors. That might, for instance, give some insight in turn into how to best target paid search ads for students in an online degree program.

While I agree it’s a nice idea that press releases are published as more natural, flowing articles, not templated press releases, this throws up all kinds of issues with who should read what. You potentially cause confusion for everyday non-journalist users. Which ‘news’ should they read? 1) The news feed in the main site or 2) the news feed in the Press & Media section? If there are two news feeds coming from inside an organisation, then should I be reading the press room’s news items because the information will be more credible because it’s less likely to spin to hyper cynical journalists?

No - there needs to be a template / distinction between general user news and communications aimed at journalists. Press release in general can be improved and are not ideal, but it’s evident who they’re for. Turn press releases into fluffy features or articles and journalists will ignore it completely and call the PR team direct on the phone.

One idea would be simply list bullets of facts and events, then supply the phone number of the best PR contact to follow it up. This is pretty much what journalists do - skim a release for what’s being announced, then calling relevant contacts to get the real story or more detail.

Sadly, fluffy or wordy press releases will continue to exist while journalists copy, paste and tweak press releases when ‘writing’ news.