This is a common question we get asked by clients who assume they measure the same thing, which of course they do not do. This misconception has arisen because of the claim that Google Analytics can be used to measure the effectiveness of different marketing channels.
This claim is partially true; Google Analytics can be setup to accurately measure marketing channel effectiveness but this requires the cooperation of your website developer. The quick methods of setting this as defined in the Google help sections will not achieve this at all, and are in fact dangerous as they give very misleading information.
This article will explain some of these differences.
Firstly AdWords (or most PPC reporting) will report on clicks (taking account of repeated clicks) and conversions within a month. This reporting is subject to the restriction of the user having javascript and cookies enabled.
By contrast Google Analytics reports on visits to your website and will look at the latest source when assigning a conversion.
The differences between the two are therefore:
- Multiple clicks by one user may only be reported as one click in AdWords. In Analytics this will also be one visit with separate page views.
- Detected fraudulent clicks from obscure IP addresses will be discounted by AdWords but not by Analytics. This is not normally an important issue when dealing with conversion data.
- AdWords will record a conversion if someone clicks on the ad and then makes a purchase on the same computer without deleting cookies. In practise this usually means within one month. There may be a gap of several days and they may re-find the website using natural search, but this will still register as a conversion. By contrast Analytics takes the latest source. This will tend to undervalue the first method of finding the website (often a generic search) and over value the last method of finding the website (product or company name)
- If your website has redirection then the gclid parameter (by which Analytics knows the referrer was AdWords) get removed and the subsequent pages visited will not be counted as from the AdWords referrer. This is a major issue with some shopping carts which ‘lose’ some of the AdWords leads in the shopping cart process and these are incorrectly attributed to direct or organic search.
- The location of the tracking code within the page also has an impact; if one is near the top of the page this will always register, whilst one at the bottom of the page will not register all visitors as partially loaded pages will not be counted. On image heavy or slow websites (Shopping carts) this can have a noticeable impact with one or both type of tracking under recording by 2-10%.
The last three issues can all be sorted by changing the way the website behaves, but is not always an easy task. The issue of first / last contact requires code on the website which reads existing cookie information and feeds this back into the conversion process, especially the final sales or order page.
Webpage redirection can also be sorted, but on many shopping carts with open source modules implementing this can be tricky if other factors such as SEO must be considered. It is however achievable and for best marketing tracking should be done.
The location of tracking code should be the easiest to sort, but there is here a matter of principle: which is more important a working website or perfect tracking? Having 3rd party code near the top of your source code can slow down the time the page takes to load if there are issues on one of the servers used for AdWords or Analytics. This is surprisingly common and the decision to keep all tracking at the bottom of the code may be the best overall.
The conclusion from the above is that most often using Analytics to measure and compare different marketing channels can be dangerously misleading. It is likely to undervalue the PPC and referral channels and overvalue the direct channels.
Analytics can be used to accurately measure the impact of different marketing channels but this is not quite as easy as claimed by the setup guides for Analytics.