Data Analytics : 3 main problems in e-commerce
How often does your favorite analytics tool drive you crazy when trying to make sense of your online shop statistics? Well, we’ve all been there. As we said in a previous article, « data is the black gold of the 21st century ». Data analysis is at the core of our e-commerce and marketing efforts. Sometimes, however, this crucial step is overlooked because of the many issues that may arise while analyzing data.
That first paragraph wasn’t meant to scare you off or to say that you’re not doing your job well. To the contrary, it comes from all the feedback we gathered at the Paris Retail Week exhibition last week, where we had the chance to meet plenty of e-merchants thrilled to discover the latest innovations in e-commerce.
Many visitors would stop right in front of our booth to tell us “you guys are right, I’m exactly in the situation you describe”. What they were actually referring to was one of our banner stands, stating that e-merchants do not know 98% of their web audience. Indeed, most e-merchants focus on the conversions, considering the rest of the data to be less important. Given that the conversion rate is around 2% on most e-commerce websites, I’ll let you do the math. Only a tiny portion of the data is actually analyzed and leveraged.
The 3 main data analysis issues in e-commerce
Consequences of the company’s size on data analysis
From what we heard, the problems start even before the analysis itself. We met the whole spectrum of e-merchants: from large international companies to single person businesses. While the reasons they were struggling with analysis were different, the consequences were similar.
E-commerce large companies
Large companies are usually composed of many different units (analytics, data mining, marketing, communication, …), each doing a very specific job and making sure not to overstep on the others’ areas of expertise. This often leads to e-mails 10 different people being cc-ed just to be able to gather a single report. While talking to our clients and visitors at the Paris retail week, even acquiring data, let alone the analysis, was described as very challenging. For instance, consider a company with a CRM unit, a marketing unit and a web analytics team. Clearly each of these teams have their own objectives, which they tend to prioritize before replying to other units’ requests. It is more and more frequent for the web analytics team to be swamped with requests from their colleagues and to struggle to serve them or even do their own job properly. At the end of the day, data analysis is a clear factor of stress, tension, and issues between colleagues that eventually hurt the functioning of the company.
On top of this, even if interactions between the teams are going well, they are often way too slow to react quickly enough. For example, adjustments on an e-commerce Web site need to be made very quickly to address specific issues that have just been identified by data analysis. The fact that several colleagues from different units need to be in the loop often makes such quick adjustments impossible to carry out. This may significantly hurt the business.
When it comes to small companies, employees are often required to have a multitude of skills. Yet, data analysis is not an easy job and clearly cannot be improvised. Small companies come with some issues precisely related to their size:
- The culture of data analysis: while we do not want to generalize, data analysis might very well not be a priority when what matters most is managing the suppliers, the sales, the orders, the storage, etc. Yet, many e-merchants came to us with the same question: “How can I improve my conversion rate?”. Somehow, they know that data analysis is crucial to solving their problem but they have not figured out yet which tool they should use to get there.
- Skills: let’s move on and get to Google Analytics. There is a huge gap between knowing this tool and fully mastering it. Its complexity is a real issue here. The data may be quickly biased if the initial configuration is not done properly. A bad configuration can lead to wrong conclusions and further down the path to very bad decisions. The main issue here is the time needed to getting the right training and improving your skills. Not to mention the time to make the most of the huge documentation available for the tool.
- Time: time is money (thank you Benjamin Franklin). Managing it efficiently is crucial here.
- Perspective: you need to step back from your data analysis on a regular basis in order to reach the right conclusions. Avoid jumping from one strategy to another several times and make sure that your data is correct and has been well analyzed before doing so. Sometimes, you may get the feeling that an action you’ve taken to improve customer loyalty or web traffic acquisition is working, while in fact another external factor had interfered. Be aware, Google Analytics does not always give you the full picture.
Financial and human resources: we will get there with time but for now, the resources dedicated to data analysis remain largely lower than, for instance, those dedicated to traffic acquisition. That’s a shame: your new traffic is useless if not optimized and for that, well, you need your data to be properly analyzed.
Data analysis tools
Most existing tools on the market have a content-centric approach. When we talk about content-centric, we mean that all your statistics are related and expressed in terms of products, landing pages, and so on. For instance, creating customer segments based on their behavior is not an easy thing to do in GA. And once such behaviors have been identified, triggering ultra-personalized actions is almost impossible. Despite its recent visual update, GA, as unavoidable as it is, remains “rough around the edges” with respect to user experience.
The purpose of data
Again, we were very proud of our banner stand and to have been able to clearly identify the issues that e-merchants encounter on a daily basis. On a more serious note, data analysis is essential to understanding your customers but it is not within everyone’s reach. As a proof, we have met very few e-merchants who were effectively analyzing the entirety of their data. Most of them focused on the 2% to 3% of visitors who turned into real customers. Conversions are known and analyzed under every possible angle. Yet 97% to 98% of visitors remain unknown and will never generate revenue for the company.
The feedback we gathered during these few days at the Paris Retail Week reinforced our strong belief that a tool – possibly complementary to Google Analytics – that is able to dynamically create behavioral segments and address the entire web audience, would be great to have and we are already on it!
Beyond data analysis and customer segmentation, we will soon be launching a simple and easy to use tool for visualizing your data, your KPIs, watch their evolution, provide recommendations to optimize your site and propose personalization actions to trigger.