Google’s roll-out of product enhancements is both exciting and frustrating in equal measure. Exciting because they offer something new to play with (aka in-depth testing) and frustrating because their product teams obviously come from the stealth-school of marketing.

A case in point: back in May ’09, they blogged about the upcoming addition of pivot tables in Google Analytics. Great stuff even if you’re not a spreadsheet geek, sorry, I mean ‘power-user’. So when can you get your hands on them? Well, they’ve been staring at you for some time now but you probably didn’t see them. No one I speak to seems to have noticed them, so if you’ve missed them, you’re not alone.

Pivot tables? Yawn….

Hang in there because they are really, really useful  (if you’re not sure what a pivot table is, check out the Wikipedia entry). Here’s an example of how bounce rate analysis in Google Analytics comes alive with pivot tables.

Bounce rate is a measure of single page views followed immediately by an exit from the website. Basically, visitors are arriving and leaving from the landing pages without engaging with the rest of your content. The higher the bounce rate, the more visitors you’re losing before you even get chance to convert them into prospects and/or customers.

For bounce-rate analysis, I usually start of with the Content >> Top Landing Page report in Google Analytics set to a comparison view:

Some nasty red bars in there indicating that those pages are performing worse than the site average in terms of bounce rate – a good way to prioritise your analysis and optimisation efforts. Notice also the pivot table selection icon – very easy to miss.

Before pivot tables, the next step would be to look at a ‘problem’ page (is it OK to be high bounce?) before drilling down to analyse entrance sources and entrance keywords. In particular, I’d look for signs of low relevance or incorrect visitor expectations with respect to each traffic source and, for search traffic, each of the search queries/keywords.

Now, with the new pivot table functionality, it’s much easier to cross reference and compare dimensions and metrics. I can create tables based on any of the dimensions available in Google Analytics while simultaneously viewing up to two metrics. For example, let’s see the landing pages again with bounce rate and pageview metrics but this time pivoting around the dimension of traffic source:

Interesting stuff: the top landing page (4) that I’ve highlighted was looking OK in the comparison table view with an average bounce rate of 24.58% – about 10% better than the site average – but now I notice that Yahoo visits landing on this page are only bouncing at 22.82% while visits from Google and Bing are producing the highest bounce rates – 30.77% and 32.46% respectively. Why the difference? At this point, I’d be rolling my sleeves up and digging deeper…

The point here is that pivoting the data provided me with additional insight. Without it I might not have investigated this page further because it’s average bounce rate looked OK and certainly better than the site average. Now I know there might be issues with the traffic from Google and Bing and probably lessons to be learnt from Yahoo traffic. Effectively the pivot table is helping me employ best-practice analysis, in other words, analyse segments not site-wide averages.

No support for advanced segments! Use profiles instead.

If you’re now thinking ahead about how you can get fancy by applying pivot tables to any of your advanced segments, think again: they’re disabled as soon as you fire up a pivot table. Not a disaster if you follow best-practice and create separate profiles for each major segment you’re interested in analysing such as organic traffic, PPC traffic, new/returning visitors etc. Just head over to the appropriate profile and off you go. Here’s an example of that approach.

Suppose I want to see which Google AdWords keywords are driving sales of each product on an ecomm site. One way to do this in Google Analytics is to open up a profile I created that captures ONLY Google AdWords traffic then create a ‘Product Overview’ pivot table, pivoting around the keyword dimension and looking at metric, product revenue:

What can I get from this view? Keyword 4 in this instance is interesting as it delivered £3k5 in revenue for product 4 and I wouldn’t have expected it to deliver any. Maybe I need to tweak the AdWords campaign to exploit this opportunity further.

Try experimenting with pivot tables in Google Analytics and let me know what successes you’ve had using them.