Web Metrics
A full version of this article is available at http://www.managers.org.uk/ (written, July 2003)
Web metrics measure internet sites as a business tool for communicating with customers. Knowing how to measure a website is as important as knowing what you want to measure.
- Are you attracting new people to your site?
- Is your site sticky? Which regions are not?
- How proficient is your conversion of browser to buyers?
- What site navigation do you wish to encourage?
- What attributes describe your best customers?
- What is your churn rate?
- How do you measure loyalty?
Web logs
If you are hosting your own website you can readily access the web log files. These are long files which contain lots of seemingly random data and are not easy to read. However, looking at your log files this is what you should be able to decipher:- Hits – entire site, average per day, home page. Hits are not a single user. They can be part of page or a graphic. On average 16 hits is equal to 1 visitor
- Page views – average per day, document view. This is more useful as you can see which pages are the most popular
- Visitors – unique, one-time visitors , more than once visitors, length of visit, origin of visit
- What type of browsers your visitors use, who they work for, what time of day they visit their site
- Where they were referred from e.g.
- a search engine and the words they used. Knowing what people are looking for can help you attract more customers and whether it is something generic such as ‘running shoes’ or specifically you by your brand ‘Reebok’
- a portal e.g. Oxmedianet. You can judge whether your website listings are working
- Monitoring – for deciphering what people want and how to get them to stay longer on your site
- Feedback – trying to provide a better user experience by personalising the site according to demographics
- Leverage – increasing customer profitability by cross-selling, customer profiling and providing customer satisfaction
- Strategic – optimising your business model, getting rid of low-margin customers, tracking customer life-time value
Measuring the site: benchmarks
Some benchmarks you can easily calculate using your web log information are as follows:- Stickiness = total no. of time spent on a page/ total no. of visitors
- Freshness = average content area refresh rate/ average section visitor frequency
- Skip factor = no. of visitors who skip intro scripts/ no. who don’t skip * 100
- Migration = average no. of exits from an area/ average no. of visits to an area * 100
- Focus = average no. of pages visited in a section/ total no. of pages in the section * 100
- First purchase = required clicks to purchase/ actual clicks to purchase