Decision Engineering News

Improving the quality of web pages

Tue, 6 Jun 2006 03:52:14 GMT

In 2006 a majority of web sites have poor markup. Browsers are designed to accommodate this, so most users never notice. Site owners who attempt to automate web content management probably do notice. (They find automation harder than expected.) Programmers who create this content, via their code, often have no way to check markup quality. This article describes a simple, low cost way that some programmers can use to improve the situation.

ClickOnce Update

Wed, 01 Mar 2006 21:20:00 GMT

An update on notes about using Click Once technology? The technology promises to offer simple desktop type programs easily installed over the Internet. For my purposes the technology within .NET 2 looks appropriate. A low cost analysis of data till February 2006 suggests that it might be 2011 before this is viable for a general audience. For a specialist audience it could be sooner. (Product and OS releases could change the picture.) This is another review of the evaluation first published in September 2004, with updated data.

Technical Quality of a Web Site

Mon, 14 Nov 2005 03:56:24 GMT

Judging technical quality of a web site is easy, if you know how. These notes give examples of how a user might judge a site and the tools he might use.

The visitor has a real appreciation of how web sites are built and forms his opinion from more than the words and looks. The article illustrates how he looks beneath the surface, inspecting markup, stylesheets and script.

Visualising the rating of a web site

Tue, 18 Oct 2005 03:30:21 GMT

Working with Personas is easier when you can envisage what your visitor is thinking, compare it, see the impact of a change and explain that to others. Graphs help do that. (These examples represent the opinion of the visitor and illustrate his thinking process.)

Through your Visitor's eyes

Tue, 06 Sep 2005 00:48:21 GMT

When designing and refining a web site, a roadmap can save a lot of effort. Part of that roadmap is envisaging how your user actually sees the site. It makes sense for that to guide how you develop and test. An example of a persona that does this, illustrated with graphs of browser and other trends over the past few years.

Setting up a Persona

Wed, 24 Aug 2005 02:37:42 GMT

Describes some of the things you might like to include in a Persona. This persona is used to measure the quality of the site. Like a real visitor the site is not judged just by looks or just by the writing, but a combination of both and more.

Using a Customer Persona

Wed, 24 Aug 2005 02:37:42 GMT

The idea of a customer persona, for web design.

I've found it useful to define a single idealised user when creating a web page or web site. By addressing your words and design to that one person, writing and design are easier. In addition new ideas are easier to accept or reject, when evaluated against the persona.

Are web sites a convenient fiction?

Fri, 3 Jun 2005 01:42:00 GMT

Web site owners often think of their site as a single entity. However many visitor are only interested in one page, they don't look at anything else. If you following through that thought, it could put more work into web design. You might need to have a target user or persona for each page!

If you are a Web Site owner or designer, do you think of a web site as a single entity? One navigation system, one ideal visitor, one look and feel? If you're like me, then you like to stick with one ideal visitor, I call him a persona. One person I hold in my minds eye when thinking through web site decisions. For a site that sells one product, no variants that might well be the right thing to do. If the site has multiple products or many on-page-visitors, it could be wrong.

You can get a handle on this for your own site/s quite easily. Examine the site logs and use your mental image of what the user does. From those you can guess how many people look at one page on a visit and actually read it. As a sanity check think of your own browsing habits and those of others. How often do you use a search engine and look at only one page from many of the sites you visit?

What does that show? Might you have a lot of one-page-visitors? If so it's worth thinking what that means. Here's a few points worth considering:

Your navigation is not used by the one-page-visitor. That's right for some part of your visitors if you had no navigation it wouldn't make any difference! Don't spend too much on the navigation, keep it simple.
Your one persona for the whole site might be wrong. The people who look at different pages might be quite unlike.
If you have a few distinct products (in-a-hurry version, professional-methodical version, test-the-idea version?), you might need a few personae.
Your nice unified design might not be seen by a lot of your visitors. (Which is not to say you shouldn't have that unified design.)

Similar thinking applies to closely related page-clusters. A few pages that a single visitor will look at. How does that impact your design?

Realisation that this is what really happens with your site can be a liberating experience. Some analysis paralysis might be ended, no need to agonise over a design issue if it's irrelevant for most visitors.

Even in this situation I'm inclined to choose a single persona. A well developed persona saves time and is convenient to use. A persona-per-page would be inconvenient and potentially time consuming. I'm currently facing a situation which seems to call for several personae. I'm going to try something different. I'm going to consciously apply what I'm calling a mindset approach. A mindset is a modification to the base persona. It describes what a known visitor is thinking on this visit (and related visits). That involves refactoring the persona. It may be quite an interesting exercise.