Homepage design and purpose - McGovern vs webstats

In a recent article, Gerry McGovern talks about how the number of visitors arriving at the homepage is decreasing. This prompted me to dig into the University's webstats to see if our site matches his findings.

Gerry asserts that more and more are Googling (or the search engine of your choice) right into the content they want.

His conclusion is one I agree with, and reiterates his usual message with a slightly different slant. Your website is all about the tasks people do on it.

Your customers don't want to get to your homepage. At best, the homepage is merely a series of signposts that will help them head in the right direction. Unfortunately, too many marketers and communicators are destroying whatever credibility their homepages have left with customers by filling them with useless graphics and meaningless words.

But the stats he quoted at the outset of the article are what really caught my eye and prompted me to take a look at our webstats at the University. Gerry says:

Another technology website had roughly 10 percent of page views for the homepage in 2008, and by 2010 it was down to 5 percent. One of the largest websites in the world had 25 percent of visitors come to the homepage in 2005, but in 2010 only has 10 percent.

So I took a look at the University homepage, considering data from people outside our network only - so no staff or current students who might have it set as their browser homepage.

Over the past year, views of the homepage accounted for just over 11% of all page views. Looking at unique page views, the percentage was slightly higher (a little over 12%) but this is to be expected as this figure includes all the visitors who land on the homepage and then leave immediately - what Google Analytics call the bounce rate.

So far so good, the University is in line with the kind of thing Gerry is saying. But then I thought, what if every visit to the website definitely began with the homepage? What would the percentage of homepage views be?

The average number of page views per visit was a little over six, so if every visit started at the homepage we would expect the percentage of homepage views to be about 17%. Not a great deal more than the reality.

I then picked on a few individual months of the past year when traffic patterns would differ, based on the academic cycle. The stats were pretty much the same.

So, yes the percentage of homepage views compared to the whole isn't particularly high, but would we really expect anything else?

Finally I took a look at points of entry into the site. And this surprised me considering the stats so far. The University hompage accounted for 53% of all landing pages. And the variation across the months I considered was greater. From mid July to mid September the figure was higher (56, then 59%), while last month (mid March to mid April) the figure was lower - 48%.

And what about other areas of the University? I repeated the exercise with an important academic website, and a service website with a large external audience.

The academic website's stats were a slightly higher than the University as a whole, while the service unit's figures were a good bit lower.

My guess would be that the business of the service unit is better known to external audiences than the academic unit. Visitors to the service website would be more likely to know precisely what they wanted and therefore could be more specifc with their search behaviour. More people dipping into the academic side of our business don't know what they need to know. And if you don't know precisely what you want, you're less likely to search and more likely to navigate, getting educated along the way, as it were...

But hey - it's all lies, damn lies and statistics. This is all based on what Google Analytics collected for us, and we don't cover every last page the University publishes. Not by a long way. When it comes to what people do, and why they do it, we're just making presumptions when we use stats. You only really learn when you watch people in action.

Which brings me back to Gerry's main point: use the homepage to help people get where they want to go.

Too often marketing and communication behave like needy children. Or like the tailors telling the CEO Emperor about how beautiful his new clothes look. On the Web, content may be king but remember that the customer is dictator.

The decline of the homepage - article by Gerry McGovern