Archive for August, 2005

Is your site a boy or a girl?

Research from the University of Glamorgan’s Business School has identified the gender of 32 other university websites.

Gloria Moss and Rod Gunn’s study found that 94% displayed a masculine orientation with just 2% dislaying a typically female bias.

Although details of their methodology are visibly lacking (Dr Gunn will only say, “the statistics are complicated” – just as well we’ve been shielded from them then), the logic goes as follows:

“Where visuals are concerned, males favour the use of straight lines (as opposed to rounded forms), few colours in the typeface and background, and formal typography. As for language, they favour the use of formal or expert language with few abbreviations and are more likely to promote themselves and their abilities heavily.”

Strangely, the university’s own website was not included in the study, but the researchers assure us that it “was found to be one of the most equally balanced in terms of visual design.” To me, it appears to fit the profile of a masculine site quite well.

Do have a read if your only objective is to make your site aesthetically pleasing to one gender or the other.

KISS and make up

When it comes to a developing medium like the web, KISS (Keep It Simple, Stupid) is a good principle to follow.

This blogger is an advocate of the search marketing success story that is sponsored links. For marketers, it allows you to attract a targeted audience from a keyword or phrase of your chosing and best of all, you only pay when someone visits your site. It brings to mind John Wanamaker’s famous quote ‘I know that half of my advertising is wasted – I just don’t know which half’. While some of the visits through to the site may not result in sales or a specific action or may even be random, the cup is more than half full (to continue my abuse of clichés).

One of the challenges of selling in online media alongside traditional media is explaining the complexities of online in KISS terms. When it comes to planning search marketing campaigns, I will typically put together a spreadsheet showing predicted spend on chosen keywords to cover a monthly period. The two main platforms I use are Google Adwords and Overture.

Here’s where the fun starts. Google’s predictive tool provides estimated traffic and ranking per day, Overture’s per month. This is an area where the major players need to agree a common metric and stick to it. Taking their figures, copying them into a spreadsheet and setting up the formulas to provide comparable stats a pain. Even the tables of figures use different sets of columns. Indeed why not have a simple export function when you are doing the research yourself? (Overture will do this for you, if you use their Fast Track service. As for Google I’ve always done the calculations myself, so cannot comment on their agency service).

Then it comes to the business of delivering on the campaign. Over the years I have run many search campaigns through Overture and Google. One universal is that the figures rarely match the predictions, leaving to red faces and frantic tweakings as you try to deliver on your promises. I appreciate prediction is an inexact science, particularly with Google’s blend of popularity and bid, but this is an area that really needs to be improved.

I will of course keep using their search marketing services for the simple reason that they work for my clients. I just wish they’d make it easier for me to sell in and KISS and make up.

Google experimenting with predictive search?

A Google search today brought up an unexpected prompt. I am used to seeing the alternative spelling prompt ‘Did you mean’ at the top of the page and equally the ‘Dissatisfied? Help us Improve’ link at the bottom of the page.

However, on a search for Hillsborough after the fifth result was a thin blue dividing line with “Dissatisfied? See results for: hillsborough disaster” below. This was followed by further results below this and is the first time I have seen results displayed in this way.

Google dissatisfied

This appears to be a way of associating related terms, so if you chose one term ‘Gin’ it is prompting you to search for the associated ‘Gin and Tonic’ as a way of narrowing your search and improving your search results.

Google often experiment with their search engine results page below the radar before making announcements about changes. Could this form of predictive search be another example of one such experiment in the offing?

As ever, a quick search reveals that a small section of the Blogosphere has picked up on this as well, although many people are unable to replicate the results suggesting this may be a localised experiment for now.

Active Monitoring and Passive Intelligence

One of the major challenges facing those of us that would track online media has been establishing (honestly) the balance between the active and the passive.

While Monitoring is deemed often tiresome, repetitive, junior, low value, loss-leading and ‘agency’, trend analysis or Intelligence is considered insightful, senior, high-value, high-margin and ‘consultancy’.

Most of the available tools and methodologies pander to one of the two extremes, in part a reflection of the ambitions of their buyers (and sellers). The trouble is that not even extremely clever people spend their entire day doing clever stuff. And a lot of the ‘intelligence’ is automated enough to streamline and enhance the more passive aspects of digital media monitoring.

Some of the same considerations apply to blog aggregation tools such as Newsgator and Bloglines.

I look forward to seeing a new generation of tools that reflect a greater understanding of what people actually do, as opposed to what they think they do – or worse still, what the programmers think they do!