The second to last session of the day at A4U Expo in London focused on setting goals devising action points and fully utilising the Google Analytics platform from the publisher’s viewpoint. It was delivered by Doug Hall and Russell Sutton from ConversionWorks.
Analytics for publishers is a fairly confusing area. Most don’t sell anything, so assigning monetary value for goals becomes difficult, whereas it’s somewhat easier to measure in e-commerce. What usually occurs is the publisher does not properly define value in their goals, and instead focuses on reporting rather than forecasting. The problem with this approach is that you should be using analytics to inform what you should be doing, rather than simply reporting on what you have done. Everyone with a website should be looking at data to size their opportunities, rather than looking at it from a rear view mirror.
Even worse than merely reporting, a lot of publishers are also reporting poor vanity metrics – those that Google Analytics gives out of the box – such as visits, time on site and pageviews. While these can be useful, they rarely explain value. Fortunately, there is an alternative solution in finding value, even if you don’t sell anything.
There were two approaches explained in the talk. The first is rather more simple, but it is that publishers need to act as ‘pseudo e-commerce’ – if there are particular buttons or actions onsite that drive clear revenue from a third party, then these should be set as goals much in same way as shopping actions provide value for an ecommerce site. An example of this might be price comparison, whereby various links derive different values set by the merchants. Track this links in Google Analytics, and you’re acting as a pseudo e-commerce site.
The second approach is for when publishers cannot easily derive monetary value from their goals. The following might all have some sort of value, but it is not easy to derive it in monetary terms:
It’s important to deem which is the most valuable out of these outcomes, and then place an ‘engagement score’ against it. You then place relative scores against the other possible outcomes, like so:
Once this values are assigned, you can build reports that deem where valuable traffic is coming from, and what content is also driving value. The crucial point is to find high value users, and you can then start asking questions like the following:
You can then begin to optimise accordingly.