Here are five technologies that can unleash the power of all the data an online retailer collects.

Do you know your customer? This is a tough question for any retailer to answer, yet a majority will assure you that they do, and they’re partly right. Most, if not all, retailers who sell online have access to a great wealth of customer data. And while retailers often “know” their customers in the aggregate, do they really know their customers as individuals?

Customer information such as age, gender, marital status, geographic location and more are likely scattered throughout all of your information sets, but you might not have it in one place. How do you approach customer analysis if you do not store all of your customer data in one place?

First, we must understand the difference between aggregate and raw data.

The usual, aggregate data

Popular tools such as Google Analytics are great for getting the meat — the aggregate information about your products, customers and how they’re interacting with your site. But it still represents a boiled-down version of your data, which, by nature, means it’s a summarized and sometimes overly generalized understanding of your customers. This limits your ability to drill down on specific customer segments or even individual customers.

advertisement

In aggregate, you’re able to see how a product sold day by day, how many orders came in over a specified period of time and even which group of customers bought a specific product. However, you can’t analyze individuals. For instance, who is Joe Smith, how many times has he shopped here, what has he purchased and what does his wife shop for?

Raw, untapped data

Raw data, on the other hand, can include every interaction and data point imaginable. And if you have it stored somewhere in a standardized format, you can import it to almost any data analytics tool that you can get your hands on. The best part about this data is that you own it and you can migrate it from place to place.

It’s also important to note that raw data was once very expensive –– even a decade ago –– largely due to hardware requirements. Today, cloud services like AWS allow any retailer to build enormous databases for pennies on the gigabyte.

advertisement

The benefits of going raw

Having the database and tools to pinpoint a single customer is important because without it, marketers and developers can be at a disconnect about who exactly their customers are, and what exactly their customers want. Once these teams are on the same page and have the power to analyze a specific customer, they can find patterns in individuals and then market specifically to them without the customer even noticing.

All of this information –– age, gender, clickstreams, etc. –– comes from raw data. It’s the information that hasn’t been tampered with by software like Google Analytics. And with all of that untapped raw data comes endless possibilities.

The caveats

advertisement

Tracking down all this information has some people paranoid. Most people don’t care but, still, they have become more aware of the information stored and used on the internet. And the reality is, the only way to not be tracked or traced is to not use any modern tools and live in a cave. How much data you should collect depends greatly on the comfort level of your customers and the business itself. Using the sophisticated art of persuasion –– offering coupons –– may do just the trick to get a customer to disclose their age while keeping the exchange mutually beneficial.

Another obstacle around raw data are the regulations imposed to keep information and customers safe. Depending on the type of site, different regulations apply. For obvious reasons, there are heavy regulations when collecting credit card data or the information coming in from certain types of websites.

Finally…

Once all raw data is collected, data analysts can, unlike with aggregated data, analyze it with a plethora of tools. In other words, raw data unleashes the power to use customer data in multiple ways, wringing out every drop of information into productive software. Here are five tools to use to help you become well versed on your customers and their needs:

advertisement

Redshift

Column-based data warehouse that runs on AWS, is inexpensive and extremely powerful. It requires planning ahead for storing your data in a structure best suited for your specific analysis.

Kinesis

Amazon Kinesis allows you to create and consume data streams that can be analyzed in real time to give users immediate feedback — a recommendation engine based on clickstream data, for example. The raw data can be stored for long-term consumption and analysis.

advertisement

Pentaho

Pentaho is a “Comprehensive Data Integration and Business Analytics Platform”. It has numerous parts that allow you to create dashboards from a number of sources. There are community-supported and paid versions available, each with their own set of pros and cons.

Various Programming Languages

When it comes down to it, sometimes you need a proof-of-concept before going live. Data stores can be complex to set up, and take time and resources to create queries and reports. It is important to be able to stage these queries and reports on a subset of your data. Languages such as R and Python are excellent for a number of kinds of analysis and are easy to learn. Data analysis can get expensive, and analyzing giga- or terabytes of data can be done without jumping into highly scaled applications.

advertisement

Tableau

An easy-to-use application that can connect to your data stores, Tableau is great for entry-level dashboarding and can grow with your needs. It comes with connectors for a wide variety of data sources, and has a huge knowledge base.

Sumo Heavy is a digital commerce consulting firm based in New York and Philadelphia.

 

advertisement
Favorite