Advanced Segmentation for Google Analytics – the why without the how

This is purely an informational post about why you, the site administrator, should begin looking into advanced segmentation, a free and powerful tool that is offered to us lucky Google Analytics tinkerers. And how large fortune 100 retail companies who sell things on a day-to-day basis WISH they had the luxury we do with being able to segment our day-to-day customer traffic.Retail stores have entire departments allocated to watching and evaluating metrics that come streaming in from their multitude of stores. The one chain in particular is Best Buy.

A good friend of mine recently got hired as a lead developer at Best Buy’s corporate office, his one full-time year long job is to program new software that outputs all relevant metrics for a particular store. I sat there thinking, where do they get these metrics from?

Did you know, when you walk through the door of a Best Buy store a trigger goes off letting the store know that +1 person has entered the store. So, on any given day they know the total amount of visits to that store. Sound familiar? It should :

What else do they know? Best Buy knows how many items and what items were purchased in a day. Great! What does that have to do with segmentation within Google Analytics? What does this have to do with my website? Well, if we think of each call-to-action and each navigation item on your website as a product category within a retail store, you’ll get a better sense of this analogy. Talking to a sales representative can be considered visiting a FAQ, search, or even a product details page, Checkout can be considered making a purchase, etc. Each page visit is similar to Best Buy being able to know how many people have visited a specific area of the store. However, they have yet to gather that metric. But, what they can gather is the amount of purchases made from that section which gives them somewhat of an idea of how many people visited that area on a given day. (hardly accurate) For instance, they set up a fancy demo of a Nintendo Wii and they noticed a spike of Nintendo Wii purchases. What could you imply from that?

*each page visit is similar to Best Buy being able to know how many people have visited a specific area of the store.

Now that we have made a correlation between a retail store location and a website…

Did you know that Best Buy started segmenting their customer base on May 7, 2004? That means, they created set customer segments that acknowledged the differences of purchasing habits between groups of people. They then applied this knowledge by modifying several lab stores, so that their product line was relevant to the traffic that was coming into their stores. Each lab store conversion cost the company up to $1 million dollars. Spread that across 1,000 stores and that shows how much of an investment that ONE business made towards catering to the most popular segment that visited that one store. That comes close to $1 billion dollars!

The following quote from Wikipedia (I have highlighted the key points, but the entire quote has really great information) :
“Customer Centricity is the name of a business movement centered on catering to specific customer needs and behaviors. Best Buy’s concept of customer centricity means configuring its stores to serve the needs of the particular customer segments that predominate in the area of that store. Some of the ways that the Best Buy company transforms its stores for a customer segmentation, is using different types of store signage, fixtures, lighting and even uniforms. One of the things the company has done for some segments is to create a personal shopping assistant, so that a customer can call and make an appointment for their shopping trip.

The company has created “lab stores” (separate from regular segmented stores) to test the area’s acceptance to the theme and segment products and services. While the renovations to its stores are expensive, sometimes nearing $1 million per store, CEO & Vice Chairman Brad Anderson claims that stores that have already been transformed have doubled their growth rate versus stores that have yet to be transformed. In 2006, Best Buy continued to expand on the customer centricity operating model by opening or converting 233 U.S. Best Buy stores to the customer centricity operating model. During that same year, Best Buy operated 300 segmented stores, or 40% of the U.S. Best Buy stores.”

How did they discern segments and how did they figure out which segments frequented a specific store the most? Unfortunately, I do not know the ins-and-outs of how Best Buy figures out their customer segments, but it is most likely based off of the types of purchases people are making at these respective stores and a whole lot of additional research. I am sure if I did know their “secret formula” it would probably be some form of corporate espionage if I divulged it in this blog. So, my guess could be way off. However, one thing I am sure of : they are dumping a whole lot of money into segmentation.

*phew, imagine a stack of $1 billion dollars?

I want segments AND I want content catered to me.
So, why shouldn’t we take FULL advantage of the FREE tool Google Analytics offers? Our “retail store” not only tracks purchases, but it can also track what pages are not working. I understand if you have a couple thousand visitors a month, the data will be extremely messy. However, with segments you can isolate your customer base and find out what pages are working for your site and which pages are not. You can even isolate your bad pages, to see just the traffic visiting these bad pages and what is happening to these users.

Based off of these customer segments, you can begin to mold your content into something more relevant. Doubling your websites effectiveness should make most people happy, so – why not follow what Best Buy did and find out what type of visitors you have and what pages THEY like. Based off of that information, you can find out which pages are not being utilized and try to apply elements of the pages they like to improve the traffic flow of the “bad pages”.

I could go through many types of segments that I would consider using, but it really depends on what type of business you own and your prioritization of value on your website. Like Best Buy, we want to avoid visits without purchases or in our case visits that result in a bounce.

*We’re lucky. Imagine if only 80% of the people who visited Best Buy did not make a purchase. They would be right there, next to Circuit City.

One example of a segment that I find very effective

*Please note that All Visits : 3.97 is taking into consideration the 9.56 pages/visit for the one segment I have set up. Meaning, there are plenty below 3.97 that is pulling site average down.

I created a segment that allows us to know the actions of just the customers coming from a search engine who are NOT searching for our company name (consider that a referral vs. a true organic result) and time on site > 1 minute. Meaning, this user did not stumble across our site on accident. These people are viewing this site and attempting to digest the content.

My next step? Attempting to find out what worked for these people. What page did they land on? Where did they exit?

Advanced Segmentation is an extremely powerful tool. However, its effectiveness depends on you.

Note : Unlike Filters, Segmentation applies to HISTORICAL data and does not prevent incoming data from being stored. You can switch Segments on and off whenever you need to bring up “all” the results.