A teaser at MarketingProfs (no link since it's a two sentence teaser) mentions a new book titled Marketing Metrics: 50+ Metrics Every Executive Should Master. The teaser goes on to suggest that 50 metrics is quite a lot to keep track of and that web marketers are likely going to want to zero in on a much smaller set of data. This got me thinking.
I consider web analytics to be one of the most important feedback mechanisms available to web site owners. I also think that most people misuse the data that web analytics software provides. Most of the misuse comes from looking at too much data. There is simply no way that anyone whose job isn't 100% web analytics to be able to get anything of importance by looking at 50 page reports every day or even every week. Such large reports should be looked at quarterly. Anything more frequent will not reveal any actionable trends. Not to mention the amount of time and effort required to frequently pull together such a large report package.
On a daily or weekly basis it can make sense to look at some of the big picture items like pageviews and most popular pages. Not so much to detect trends, but rather to respond to out of the ordinary events such as an unexpected citation on another site or to detect web server errors. These types of reports are more valuable to operational staff rather than to management.
On a monthly basis it's reasonable for management to look at total pageviews and unique visitors. These numbers are easy to wrap your head around and are generally easy to retrieve. Big changes as compared to previous months has the potential to trigger some action which is a good thing.
Every quarter it makes sense to dig in to many more stats. Things like pageviews per visit, top areas of a site, keywords used at search engines, traffic from search engine results, internal search terms, and top non-search engine referrers. These are the standard reports that I think upper management is most likely to react to. They'll probably want more, but I suspect in most cases any more data simply won't result in any action.
The job of the web analytics professional is watch the web traffic data and look for patterns and points of interest that might emerge. Sure, you can, over time, come up with a list of things to look for, but the real skill comes from noticing things that you never thought of. Obviously I can't list examples here since my point is they're not things one would expect. That's a bit of a wishy-washy concept, but hopefully I explained it well enough.
Now I'm even more curious about what MarketingProfs has to say about the subject. I wonder if I can expense the subscription fee 
- Web Analytics - Statistics to Track
- SEO Strategy
- Google Webmaster Features Wishlist
- How Many Keywords Should You Have?
- The End of Click Fraud?
Entries (RSS)