OK – we’ve been marketing websites for a while now and I though since we’ve just revamped all the Google Analytics Accounts we manage for clients I thought I’d put this out there. I’ll try and not be lazy and include images.
Google takes extreme precautions to try and eliminate the good traffic from that bad. Have you ever wondered why out of the 100 people that visit your website looking for your products you only get 2 sales? It’s because sadly an awful lot of rubbish continues to get through triggering the Google code and counting as a hit, rightly or wrongly.
Google is still very hit and miss and this Post is about trying to negate some of that.
What follows below is a list for the settings we’ve done to mitigate some of our findings.
Accounts -> Account Settings
We always make sure that Technical Support and Account Specialists are ticked, just in case we need to refer to Google support and make an enquiry. We also agree to Google’s Data Processing Amendment – we don’t always know what these things are – but if Google recommends then I am afraid we cave in and go with it.
Accounts -> User Management
I’m not going to revisit this further down, but usually we will try and give Admin Access to everyone we can. It’s simply really if you’re invited to review someone’s Analytics you cannot do it properly without full access. That doesn’t mean throw caution to the wind, but it does mean you should try and ensure the person you are giving access to is trustworthy in the first place. In other words you should establish trustworthiness some other way as giving partial access to Google Analytics is limiting someone ability to do a good job.
Notes: don’t forget you can see a Change History of your settings here as well
Property -> Property Settings
I give all my Property Names the full URL for easy recognition including the HTTP or HTTPS – so that when browsing many accounts I can see exactly the one I want.
I also select a default view – usually the filtered view (more later) and I also hook Google’s Webmaster Tools, the key is to remember to come back and change these settings if you create more Views.
Property -> Tracking Info
Under Tracking Info where obviously you can find the snippet of code to add to your website, but crucially Google has a ‘Referral Exclusion List’ setting – here you want to make sure that your own domain (without any www’s) is listed. If you have an eCommerce website then you also need to add any Payment Providers like ‘paypal.com’. This setting forces Analytics to more correctly attribute the source of a users, for example it will reconnect up a user who visits PayPal to make their payment when they return to the website for the Thankyou page.
It’s a bit complicated, but worth researching what websites should not be considered referral websites.
Property -> Adwords Linking
If you create more than one View – you must come back and check the Adwords Linking settings and ensure the new View is also connected, a common mistake people forget about and then wonder why there is no data.
Property -> Remarketing
We prefer to create Remarketing Lists in Analytics as a rule, the set up is considerably easier, but we do appreciate this is a question of preference. We do it on a per need bases.
Property -> Social Settings
We list all the social media websites that a website has accounts for that we can. Minus the http:// so twitter.com/atname, www.facebook.com/pagename, www.youtube.com/user/name, instagram.com/user/name and so on. We are keen to know the impact of Social on a website at all times, so this is worth getting right. Data is collected on the Acquisition -> Social in the Reports.
Views – All Web Site Data
The first thing to note, always keep a completely unfiltered version – so that you have an unbiased view of what Google is trying to tell you. By default we arrange the following settings:
– Set up Currency and all other normal settings, leaving most thing default.
– Make sure Bot filtering is unticked
– Link Adwords
– Set up Site Search Settings if needed
– Goals – set up all conversion points accordingly.
– Channel Settings – set up Brand Terms and create Channel Groupings
– Set up e-Commerce settings if desired
– Everything else can be left alone.
Now you can be ready to copy the View to a new View, importing all the settings that you will need for both Views.
Views – www.mydomain.com
Copy the All Web Site Data View to a new view and give it a name you recognise, we always use the URL of the domain. Then make the following changes.
– View Settings – make sure the tick box is ticked for Bot Filtering
– Ensure Adwords is linked and connect Webmaster Tools under the Property Settings
– Finally you can set up your Filters.
Filters is the last and final part and really I should go into more detail, but this is what we do here by default. The real aim though is to have as few filters as we possibly can and it is only recently that we’ve had to use filters at all.
Some Analysts will have no hesitation in recommending all sorts of Filters for all sorts of reason, but the bottom line is that these settings need management and if you have to manage a lot of website it will become a quagmire of settings that need constant attention.
Recently the primary problem we’ve found is that a GA Tracking Code can be used on any website, hosted anywhere on any server. Spammers can literally add any number they like and if it matches a number in the Google Analytics system it will start counting hits. When you’re analysing data you’ll see these hits and question what they are, often inadvertently visiting their just to know who they are. Bam! that’s what the Spammer is being paid for.
The solution is to set up a Filter – that blocks out these unsolicited hits using your GA Tracking Code. OR looking at it the other way around, you only want to include hits from websites you’ve given authority to… such as your own website.
Please visit http://www.analyticsedge.com/2014/12/removing-referral-spam-google-analytics/ for a more definitive guide of this issue.
We call our filter ‘Hacking Defence’ and add all the URLs we want included in our Reporting.
And that’s about it. That’s the set up we under take for each of our clients in Google Analytics. Almost – but I’ll save the next stage for another post, this one is quite long enough.
NB: Blocking unwanted traffic to a website is always better done by the Website itself… Playing with Filters in Google Analytics does not make this traffic go away, it simply excludes it from your analysis, any damage the traffic is doing to your server will continue to be done, unless you can stop it as source.
I hope this has helped.