After reading my previous blog post, you now know that your Google Analytics code is implemented, tracking and reporting data. But how can you be sure that the data being sent to GA is accurate? For smaller sites with less than 250-500K+ hits per month (as of July 31, 2013), this might not be of great concern. However, for sites with a significantly larger amount of website traffic, asking this question could be critical to the well-being of your company as well as impactful to major business decisions your company makes.
What is sampling and how does it work?
If your site receives over 250-500K+ hits per month, then you have probably heard of Sampling in Google Analytics. Basically, when monthly traffic reaches that 250K mark, the volume of data becomes too large and slows down report queries. That’s where Sampling comes into play- GA takes a subset of data from your website’s traffic and reports on the trends found in that sample set. This method is widely used in statistical analysis because the reporting of this subset of data is statistically significant- meaning that the results are very similar to the results of analyzing the raw data as a whole.
Sampling occurs when you use advanced segments, in-line filters, and secondary dimensions that require GA to go back into its database and make a query so you can analyze the data. For large quantities of data, this can take a significant amount of time, therefore slowing down the UI. Sampling allows GA to cut down on query requests as well as speed up the overall data reporting process.
It is important to note that GA does not sample data collection. GA by default will collect all of your site visits, but to speed up processing and the UI, it will only pull sampled data past a certain amount of hits.
One useful feature within Sampling is the configuration of custom reports through your tracking code using the _setSampleRate. This should be used only for sites that have a significant amount of traffic. It allows you to select a certain percentage of visits you want GA to record. Standard GA only allows data collection up to 10M hits a month, and if you exceed this limit, GA might not record your data. _setSampleRate allows you to monitor and customize this in order to avoid violation of Google’s TOS. Google Analytics Premium, however, records up to 1B+ hits/month right out of the gate. Another bonus? GA Premium is a flat fee at $150K a year up to 1B hits per month. There are other tier options if your company gets more than 1B hits a month, contact us for more info here.
Within the interface, you can also utilize the Sampling slider which allows you to avoid unnecessary sampling. When data is sampled, the box in the upper right hand corner will appear. This will allow you to change the sample rate from 1K to 500K (default is 250K).
While “Faster Processing” obviously delivers quicker results, “Higher Precision” typically provides reports with a 90% confidence interval.
Google Analytics Premium Offering
But what if you don’t want sampled reports? What if you want to see all of your raw traffic
data, unsampled, for exact reporting? For enterprise-sized companies, sampled data isn’t ideal. While the sample reports are statistically significant, sometimes the results can be off by 10-20%. When dealing with buyer behavior, budget, and revenue, having exact numbers in terms of monetary decision-making is crucial.
Because standard GA has restrictions and limitations, in the past many companies have implemented 2 different analytics tools on their site to track
data. They liked the clean, easy-to-use interface of GA, but needed the benefits of a premium analytics tools such as Omniture or Coremetrics. With the development of Google Analytics Premium (GAP), companies now have no need for two different tools to track visitor data. GAP provides all the same benefits of standard GA (clean UI, ease-of-use, and fast reporting) as well as the added benefit of seeing their unsampled data.
Added Benefits
In addition to providing unsampled data and an increase in the number of hits processed per month, GAP customers also have access to:
- Faster data processing. Data is refreshed in 4 hours or less 98% of the time (instead of 24+ hours for Standard GA).
- Advanced analysis tools such as 50 custom variables as opposed to 5.
- Increased number of data rows in reports. GAP allows up to 3M rows for unsampled reports, whereas Standard GA only offers 50,000
- Implementation support and training from Google experts and Authorized Resellers (that’s us!). We ensure you are utilizing GAP effectively.
Not completely sold on Premium yet? Amin Shawki wrote two excellent blog posts that further discuss Enterprise Google Analytics and the 5 Google Analytics Tracking Tips for Global Enterprise Sites.
Here at InfoTrust, we are Google Analytics Certified Partners as well as Google Analytics Premium Authorized Resellers. Please feel free to reach out to us if you have any questions or if you are interested in learning about what GAP can do for your company.
Also, be sure to sign up for part one of our Enterprise Analytics Webinar Series, “Is Your Google Analytics Data Accurate?”
