It is nearly two years since I first touched Google Analytics and I’ve been in search engine marketing industry for approximately one year. I never call myself a Google Analytics expert, not only because I am not experienced enough in online marketing industry, but also due to massive data it provides and there are so many ways to analyse.

Today I want to sum up some problem I met and few mistakes I made using Google Analytics once I just started my career as an online marketing consultant. This entry is to share my experience in the past, and also, remind myself there is still a long way to go and I have to keep hard-working.

1. Using one Google Adwords and Google Analytics account for multiple website campaigns.

This is the problem I met with the company who has a Google Adwords account and a Google Analytics account for consulting different websites. Someone may ask, what’s the problem? Isn’t it OK to have multiple campaigns in one account? How is it related to Google Analytics Data?

Let me explain a bit, if you have one Google Adwords account for ONLY one website, that’s definitely OK. You can create no more than 25 campaigns for advertising different section of your website. But, if you have quite a few different websites to consult especially they are targeting different industries, then it will create a mess in both Google Adwords and Google Analytics. Google Analytics has to be linked to each corresponding Adwords account in order to collect the data and differentiate it correctly.

- Whenever link Adwords to Analytics, Google suggests to create a new Analytics account for each corresponding Adwords account.

Why? Why can’t I create one Adwords account and one Analytics account? It’s easy to remember and convenient to switch between each campaigns.

Here above is why, the data is messed by multiple campaigns and the data from corresponding campaign is not correct too. . I checked the code implementation using SiteScanGA by EpikOne and there was no problem. The reason of this is Google Adwords is passing all of its campaigns data to Google Analytics, therefore, you will be receiving multiple campaigns data if you don’t set up a filter.

Here is how to solve this problem, set up a custom filter, select “include”, choose “Campaign Target URL” in filter field and then fill in your target landing page into filter pattern. E.g. your website is www.example.com, then fill in example\.com, if you have many landing pages, use example\.com[^.]*. More information about regular expressions please click here.

*I also would like to address other problems it causes

- What if any client wants to terminate the contract with you and take the control of their Google Analytics account?

MCC would be a solution to give the control away and keep the history, but even if the control is given to the client, their entire campaign is still in your Google Adwords account. And another big problem is, will you really take the risk to give the permission to client? There are still other campaigns for other companies! I don’t believe anyone will do that.

- Have you noticed Google Adwords Help Centre mentioned Quality Score factors including account history performance?

Your account history, which is measured by the CTR of all the ads and keywords in your account

Let’s say you have a new Ad campaign, you set up well-organised Ad groups, created compelling Ad copies, and landing page is absolutely brilliant and relevant to the keywords you target. After few days running, some of your keywords have quality scores of 8 or 9 and conversion rates of them are quite high, but they are just NOT 10 and they are so relevant to your landing page that you cannot even think of any other way to improve them. You may want to email Google Team again! However, think about if you have any poor performed campaigns in the history, it might be the culprit and there is no way you can get rid of it, history is always well-kept in Google database.

2. Had not filtered out clients’ IPs and your IP

Who do you think that will provide most beautiful figures in your analytics data? Your clients and yourself.

- Clients will check their websites so many times a day, and more employees they have, the more serious the data will be messed

- Me, as a consultant, will spend a decent amount of time checking my clients’ websites . I have no idea how much time you spend everyday, but I am sure you will check them, right?

So the clients may ask you this, I have X amount of visits everyday, Y amount of page views, why I haven’t seen our sales increased a lot? You first reaction would probably be, because some visits are from your company employees. Once clients ask for accurate data, you are blind, how will you differentiate it?

My own story was one day, an old client (an existed client of the company before I am employeed) called the office and was willing to ask few question about Google Analytics. The call was put through to me, all he wanted to ask was did his clicks count as visits in Google Analytics. I answered yes, then the tough question came, “how can I know the amount of visits from me?” There is no way I can tell him unless either I filtered his IP or set up advanced filter to record the IPs.

So how can you exclude the traffic from yourself and clients? No matter the IPs are static or dynamic, here is the solution from analytics help.

3. Not set cookie duration

This problem doesn’t cause serious consequence, but definitely worth checking.

There are 5 kinds of utm First-Party cookies that used by Google Analytics

- _utma, visitor identifier, expires after 2 years

- _utmb, session identifier, expires after 30 minutes of inactivity

- _utmc, session identifier, temporary cookie that will be destroyed after you quit the browser

- _utmz, campaign values, expires after 6 months

- _utmv, visitor segmentation, expires after 2 years

Normally there will be _utmz, _utma, _utmb and _utmc cookies set on the websites, _utma is to identify you as a unique visitor, _utmz is to identify how did you go to this website. (search engines or google or other referral sites) You barely see _utmb and _utmc is because _utmb expires after 30 minutes of inactivity and _utmc expires after browser closed.

However, if you do not change _utmb value on your website (30mins), here is the scenario.  A visitor found your website on Google, he came in and started view your product pages, then closed the browser because he didn’t want to buy the product at the moment. He made his decision after one day, found your product page in the browser history and bought you product. You would not be able to track the whole process because _utmb cookie duration was too short and _utmc was expired. In most cases, buying cycle is longer than 30 minutes and customers will occasionally close their browsers .

How can you change the default duration of _utmb cookie? Modify it using  _setSessionTimeout() method in your GA code. More details about this tracking API, please click here.

Or there is a View-Through Conversion Tracking available in Adwords to track conversions within 30 days. However, it is not compatible with Google Analytics

4.  Misuse of advanced segmentation

Technically, it does not really mess the data, but if you don’t use it right, it will consume you loads of time and mess up your brain.

How many combinations of advanced segmentation you can set up? The answer is in the Google Help Docs. With such a large amount of different combinations, you could be easily lost in the data universe.

I used to play with it a lot, spent hours and hours trying different combinations, but it doesn’t help too much to make client’s next move. Why is that? Simply because I was doing that without a clear direction. For example, for most of the websites, you should focus on analysing relevant data to improve conversions. Here is a good example how to use advanced segmentation for better conversions.

5.  E-Commerce Google Analytics duplicated transactions

This is not what I’ve experienced, but I saw it in Google Help Forum and I feel I should put it here.

E-commerce tracking is by far the most complicated case to deal with in Google Analytics, codes have to be generated with different item information and then be implemented to corresponding receipt pages. If the e-commerce tracking data is messed, it will be a disaster.

The data can be messed up by duplicated transactions. How does it happen? As long as the code is in receipt page, if customer refresh the page accidentally or return to receipt page to check purchase detail, the cookie will duplicate the transaction. It may not happen to you yet, but you definitely don’t want it happen right? Here are few solutions provided in Google Help Forum

Of course there are many other things to be taken into consideration in order to not mess up the data, such as Regular Expression using, domains and subdomains tracking, internal site search..etc. Google Analytics is an extremely powerful tracking tool, but you would only deliver maximum effect by making minimum mistakes.

Have you experienced these problems before? Would you like to share your experience of any mistakes you’ve made in the history and how did you sort them out? Or is there anything I can help at all? Please leave the comments to let me know.

Related posts:

1. Advanced Goal Completions Segmentation by Google Analytics
2. Is Google intrusive in the sense of collecting user data?
3. Nedstat Web Analytics Series 1 – Use Direct View To Determine Customer Behavior
4. How To Easy And Fast Create Massive Google Adwords Ad Groups
5. How to find paid search keywords that customers really use