Response to 9 sins of Google Analytics

There is an article on Search Engine Journal today called “9 sins of Google analytics“. As a fan of analytics, i would like to offer some comments on this particular subject.

1. Data ownership

Back in 2006, i wrote a simple piece about “Google conversion tracking pros and cons“. In the article, i did talk about antitrust issue and the fact that Google could use data from Google analytics/conversion tracking to help their Adwords program. More precisely, Analytics data could be used to help Google determine average Cost per Click advertisers are willing to pay for certain type of conversions. Also, Analytics data could be used for tools like DoubleClick Ad planner, which is Free, yet promoting Google Display Network in the process.

Google has become more and more visible in everyday Online life, which shows how powerful the company has become. Yet the privacy issue will continue to be a very important issue.

For a better view of what type of data Google could have, refer to this article “The Evil Side of Google? Exploring Google’s User Data Collection” by Danny Dover back in 2008. Basically every click you do, every form you fill, every search you perform across google properties, your picture, your friends connections, your videos, music, surfing habits, your phone number, mobile number, your pet name etc… ALL could be collected via a vast number of Google services.

That explains why various attempts have been made to infiltrate Google database.

Before this post turns into a privacy discussion, let’s move on to related point. One of the big advantages of not owning the data is that you can’t customize the analytics tool to churn out different types of report as you wish or Integrate the analytics tool with Other Tools that your organization is using effectively.

With a Data warehouse, one could do wonder with the stored data collected from analytics. However, only a very small percentage of companies/users would require this level of sophistication.

2. Limited support i.e. client service for Google Analytics

Yup, this is totally true. Since Google analytics is a free product, if you have any questions about it, you better “Google” them yourself. It’s kinda fair though because the product is Free.

If you don’t know yet, Google sets up a dedicated site for Google Analytics and you can find answers to most of the common questions there, Google analytics support

There is also the Education center which provides a more complete overview/understanding of the tool. This is free as well.

3. Mediocre reporting

The author mentioned that Google Analytics reporting capability is not as good as some other tools. Well i think this point would deserve some strategic thoughts.

Personally Avinash Kaushik is one of my “IDOL” when it comes to Analytics. He is a web analytics evangelist. In one of his book “Web Analytics an hour a day” Avinash shared with us his thought on Analytics reporting in a crystal clear manner. A report would be meaningless if you can NOT take some kinda actions after analyzing the report i.e. Actionable Insights!

So i would suggest that before concluding Google Analytics or any tools have mediocre reporting, we would have to state clearly which reports those tools can’t generate and the meaning/motivation behind the reports. The next step is to evaluate whether the added benefit would justify the investment required (if any)

4. Limits to defined variables

Well my thought about this point is the same with point number 3. It’s true that one can’t define/change the definition of variables with Google analytics easily. The keyword is doing it easily. Google Analytics does allow for custom variables.

I would tend to think that the motivation behind custom variables is the capability to do segmentation ; Right now that capability has improved considerably with Google Analytics. Before this is one of the main selling points for more advanced/costly solutions.

Right now you can easily do segmentation for:

  • New visitors
  • Returning visitors
  • Paid Search Traffic
  • Non-paid Search Traffic
  • Search Traffic
  • Direct Traffic
  • Referral Traffic
  • Visits with Conversions
  • Visits with Transactions
  • Mobile Traffic
  • Non-bounce Visits

Or you can do advanced segmentation using different dimensions like Visitors, Traffic Sources, Content, E-Commerce, Systems and metrics like site usage, e-commerce, content, goals.

advanced segmentation with google analytics

If you want to export data from Google Analytics, refer to this link

5. Slow data delivery

Ok i haven’t tested the exact lag time for Google analytics and i agree that if the need to know data in Real time is very important for your business, Google analytics may not be the best solution out there. I personally feel that lag time is not an issue to me. It’s 6.34 am Thursday 24 Feb 2011 now and i can see data for Thursday 24 Feb 2011 on my analytics report already.

Before jumping in and getting a Real time Analytics tool, i would suggest we can follow the advice of Avinash Kaushik again. Apply the “So What” test three times.

6. Lack of log files

It’s not that common to admit on your blog that you don’t know something. However for this point i don’t really understand what Annie is trying to say. I personally have never used log files for added analytical capability before.

I would guest that the author mentioned about the ability to export your data somehow. I found this link and while never tested this before, it feels to me that Google allows you to export the data via API.


7. No spiders

Personally i don’t see much point of even tracking spiders (search engine bots) via Google analytics. Google already has an excellent tool for Web Master and SEO practicer to use i.e. Google webmaster tool. If you are serious about see the movement of spider, again for the purpose of doing Search Engine Optimization i hope, you have to use Webmaster tool and not relying on Analytics alone.

In my opinion any JavaScript traching technologies would not be able to track spiders because when spider visits a web site it does not trigger the javascript funtion to execute. So not only Google analytics but Omniture and a host of other tools can’t do this either. But the main point is this feature is not much desirable of Google Analytics from my humble opinion.

8. Paid search traffic being shown as natural search traffic

Technically majority of the tracking technologies out there face this problem. And it’s not a major issue anyway because with Google Analytics, you can either turn on auto tagging from your Google Adwords account (auto turn on by default) or you can tag the URL using URL tool builder then you can see clearly traffic from paid search and natural search.

The way Analytics tracking technologies work is that they identify the source (source URL) of the incoming traffic and classify them into different buckets. If they traffic comes from Google/Yahoo, they fall into search engine section. Without telling the tool which part of the traffic stream is from paid search and natural search from campaign manager, any analytics tools would be blind and won’t be able to identify.

9. Revenue from email unknown

I am a bit concerned with this statement because from my limited understanding, majority of other Web Analytics solution can’t report revenue from a particular source automatically.

And all of them, including Google Analytics, offer a very simple solution to track revenue, conversions or any ecommerce data for a particular traffic source using simple tagging. With Google Analytics, you can use again URL tool builder.

After enabling E-commerce tracking function in Google analytics, you can sit back and see data like item sold, quantity, total revenue, the source of the conversion, date of the sales etc… easily. Whether it’s email, or banner ads campaign, Facebook campaign or affiliate campaign, Google analytics can track, provided that you identify these sources and put in proper tracking tag. This is the norm for other tracking technologies as well.

Last but not least, i notice that Annie Wallace, the author of the post on Search Engine Journal is a self-taught viral marketer. So while many of the points of her post are open to discuss further, i would say she succeeds in creating a viral effect for her post with 16 comments thus far, 204 tweets, and 119 shares. The numbers could be more in the coming days!






Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.