Paying attention to your google analytics is one of the most important parts of succeeding online. There is so much to be gained from spending a couple of hours per week looking through the web analytics of a given website.
Something that can be difficult is getting insights on the full consumer buying cycle. Trying to establish all the different touch points a user goes through before purchasing a product from your website can be difficult.
There are many stats packages that provide multi touch point tracking (omniture for one), although it’s also not that straight forward to setup. The most popular stats package is Google Analytics (and is what I use for most sites). The problem with Google Analytics is it works on a last click basis. This means if a user transaction path is as follows:
PPC (Generic) -> PPC (Branded) -> Organic (Branded) -> Direct -> Purchase
The purchase will go to Organic (Branded) as this is the last none direct referral medium. This is a big issue, as keywords being used in a PPC generic campaign may not be getting the recognition they deserve. These are usually measured on a ROI basis, if a keyword is showing no value, it may be pruned incorrectly.
Using Custom Reports for PPC Analysis
There are over 80 standard reports with Google Analytics. I use very few of them. One of the best things about Google Analytics is the custom report feature (and advanced segments). We can start to look at more than conversions for PPC data in order to identify keywords that are being used in the research phase of the buying cycle. For example, I have a custom report with two tabs:
Tab 1 – Data to Measure None Converting Keywords
Tab 2 – Data to Measure Converting Keywords
This allows me to flick between both sets of stats. If a keyword is showing a very poor “Per Visit Value” but is bringing in lots of users who are spending time on the site, then perhaps it’s worth keeping ?
One of the best things I got from reading both Web Analytics books from Avinash Kuashik, is stats that need interpreting are not worth tracking. The best example is time on site. How do you know a high time on site isn’t a sign of poor navigation ?
This needs to be factored in when reading the custom reports above.
New Search Funnels for Google Adwords
Google has gone some way to providing a solution to the problem of interpretation discussed above. They now provide search funnels via the Adwords interface. Below is an explanation from Google:
The new funnels allow you to view a lot more stats about campaigns that may not produce conversions but are assisting other campaigns:
From playing around with the funnels, I haven’t come across any jaw dropping finds:
– Generic keywords are the first point of contact for new users, those who are researching
– Long tail keywords are used by consumers who are in buying mode i.e. done their research
– Branded keywords are usually the last point of contact in the buying cycle
But it does make pruning your list a lot easier. Something I have started to play around with is, how much of a drop in sales will you see, if branded keywords are switched off, assuming you are number one for organic and don’t have a bunch of affiliates bidding on your terms.
The new search funnels are a great addition to Google Analytics (which is getting more impressive each month) … BUT
Multi Vechile Purchases
It doesn’t resolve the issue of establishing how PPC and Organic work together. In the above transaction, organic will still receive the conversion and PPC gets left out in the cold. Two recommendations to combat this are:
– I create first click profiles for Google Analytics. This is done using a couple of hacks, which is a post all on its own
– Always, always, always split your traffic into branded and non branded
What the new Search Funnels can help you with, is establish weather PPC should be used as a branding tool for your company or not (i.e are they assisting other campaigns). Obviously, it only works if the transaction is via PPC (i.e. Organic doesn’t pinch it).
Here are a couple of other great posts on the new Search Funnels from Adwords: