Ecommerce Analytics Without Context Leave You Wondering
If you are older than two, you can probably count.
Counting is a good skill. What we can measure, we can change, according to social scientists.
But counting things isn’t the same as understanding them. In 1993, if you had walked into the Gap in Charlotte, North Carolina, I might have greeted you at the door. I may have been the most welcoming teenager at the mall, but I was careless in another aspect of my job.
On inventory day, my manager Karen got really upset at me for miscounting jeans. I remember raising my eyebrows and sharing a look with my coworkers as if to say, “Jeez, lady, what’s the big deal?” To me, I was just counting jeans, khakis and socks, but to her these numbers meant product tracking, theft control and financial management. I didn’t get the big picture and the reasons behind the numbers.
ECommerce analytics are no different. Your numbers are important. You should certainly be tracking metrics such as:
- conversion rate
- average order value
- exit rates for pages
- shopping cart abandonment
- checkout abandonment
- page views per visit
- unique visitors
- time per session
- cost of acquiring customer
These numbers reflect your customers’ behavior but leave essential questions unanswered, such as:
WHY do your visitors act the way they do?
Let’s consider a quick example. Assume you sell various types of snow-cone machines. Looking at your website analytics, you see that your exit rate is high on one of the product pages—a large number of visitors leave your website after viewing the kid’s snow cone machine (similar to Snoopy’s). Potential reasons for this high exit rate could be:
- the price seems too high to visitors
- the price seems too low to visitors
- the product photos aren’t providing enough information
- the spelling error in the product description eroded trust
- the product description doesn’t explain something important to visitors, such as what age group this kid’s snow cone machine is for
- the mobile website pushes the “buy now” button well below the fold
- the shipping cost isn’t clear
Now you can certainly guess at the cause of the high exit rate on this page and take a trial and error approach. Try changing your price or offering free shipping. You could run A/B tests to try out different wording on the buy button, for example. ECommerce retailers commonly do this.
While the trial and error approach may eventually work, it’s inefficient and could be costly. You may spend a lot of time waiting for your numbers to improve and miss significant sales with this approach. The opportunity cost could be tremendous. And the scariest thought is you might NEVER resolve what is going on with this product page by looking at the numbers alone.
In the case of the snow cone machine website, the common reason many visitors leave that one page is simply:
The visitors thought that clicking through to this page would lead them to a standard snow cone machine – not a kid’s toy snow cone machine.
The problem isn’t even on the product page itself. It’s with the naming of the link to the page, and the expectation it set in the minds of the website visitors. This sort of thing could be discovered in as little as one day – maybe even a morning – but NOT by looking more closely at the numbers.
At the Tiny Giant UX Conference in late January, many of the user experience experts spoke to this issue. If you had been in the audience, you would have heard this common theme:
Context is key to understand your data.
MailChimp’s Manager of Customer Research Gregg Bernstein spent his entire talk on this matter.
“Context is what we really need to focus on,” he said. “It lets you get the big picture. You need wisdom, data and context to see the full story.”
Bernstein (pictured left) noted that it’s important to understand the forces – for example, culture and roles – that affect individuals and their decision-making. In addition to surveys, which Bernstein called a “gold mine,” MailChimp also conducts 60-minute user interviews on the phone. He mentioned they also get “fantastic” information from their feedback form. But his explanation of how they do case studies really stood out.
“We will go hang out at someone’s place for 2-3 hours and have a conversation,” he shared. “Our hope is to capture some data that can help us.”
Bernstein’s team visits a user’s place of work not only to listen carefully to what the person is saying but also to observe the environment he or she is in – for example, the noise level, the roles they play, how people interact with each other at this place.
What derives from getting this context is qualitative data that goes into what becomes a “gigantic” spreadsheet of observations. They then seek patterns that can inform their products and improve user experience.
Another speaker, entrepreneur Olga Vidisheva, is on track to have 1000 boutiques on her Shoptiques online platform by the end of 2015. She explained how she’s gotten shop owners to use her product:
“I would walk into the boutiques and ask the owners ‘Why aren’t you online? How can I help you?’ she explained. “I learned the reasons, the sets of problems of users.”
By getting the context of the situation, Vidisheva discovered that many of the shop owners worried they couldn’t take quality photos of their products so they feared going online. Then she set out to solve that issue by providing photography in-house and having them ship products to her – up to 400 photos are shot each day. The problem is now solved for the shop owners, and business growth is now skyrocketing for Shoptiques.
Her clear advice: “Go talk to your users. Otherwise you are going to be irrelevant.”
Aaron Quinn, founder and creative director of eHouse Studio, also supported this idea that data in context is important. After his talk about navigation, I asked him what clients get hung up on.
“There is a tendency to get attached to research, one specific piece of research, and not see whole picture,” he answered. “Mix the research up, get different perspectives and data from different places, talk to people.”
Quinn sounds like another context advocate to me.
Surely, you are already tracking your numbers with some kind of ecommerce analytics software. And unlike teenagers who count jeans, your numbers are likely accurate. Now you can continue to wonder about your numbers and make guesses, or you could spend some time to get the context to understand your user’s needs and obstacles.
Context combined with your analytics data will help you create the user experience that wins customers. Go talk with your users, and listen carefully to them.