Institute of Internal Communication

Publish and be damned perhaps, but what about publish and be ignored? Steve Nichols of InfoTech Communications suggests we may be quick to publish content on our corporate intranets, but are often slow to monitor whether it is ever read.

Years ago, we had an easy way to find out if our employee material was being read - we checked the waste paper bins! OK, so it wasn't scientific and didn't give accurate results, but are we any better at it nowadays?

Truth is that we have a host of tools at our fingertips to check the effectiveness of our intranet copy, but we often fail to use them.

The number one tool must be so-called "web metrics". The 'metrics' of web metrics refers to measurement, the science of measuring websites. Specifically, measuring website events, and extracting trends. In this case, those events are human clicks.

These are easily available from most intranets, but often ignored due to the difficulty in making any sense of them.

What should your metrics tell you? They will inform you about numerous aspects of your traffic; the number of visitors to your story (not always the same as the number of "readers"), and how visitors surf through your pages. Your traffic statistics are an indicator of your intranet's performance. They can also be an indicator of your own writing ability.

When looking at your metrics consider the following:

• Does the time of day you publish have any effect on readership? What about the day of the week itself?
• Does the time that the story spent at the top of the story list have a bearing on readership?
• Do images help boost the figures?
• Does the length of a story have an impact?
• Is there a trend in your stats - is readership dropping or rising?

You can probably add your own questions to the list.

If your metrics come in a very user-friendly raw data form, why not see if you can have them in chart form instead? Bar and line charts make it easier to spot trends and the relative performance of one document to another becomes easier to spot.

Web page statistics for intranets can often be very misleading. You can get high page hits on pages that people have found in error as a result of poor information architecture. The hits need to be analysed in context, taking account of the paths through the site (i.e. was the page found by a search, hyperlink, or another navigation option?).

Often a page might be used very infrequently, but the low hit rate is not a measure of the value of that page, which might be an important policy document, for example.

In addition to design (or behaviour) analysis, web metrics analysis can also be used to diagnose server or site problems, and measure the effectiveness of internal marketing and communications campaigns - but only if you look at the metrics in the first place.

Steve Nichols (steve@infotechcomms.co.uk) runs InfoTech Communications and is editor of the IoIC e-zine. InfoTech specialises in online communications and has acted as intranet consultant and trainer for many blue-chip companies including Aviva, AWG, Shell, BT, Standard Life, HBOS, BNFL, Accenture and Australia New Zealand Bank.