Walmart's Analytics Help Them Help You [Shop]

Walmart, the nation’s most prolific retailer, has been collecting customer data for years.  Walmart saves an estimated 2.5 petabytes of data accumulated from more than 1 million transactions per hour; an increasing amount of these happening online (Lutts , 2012).

Most of this customer data is being collected from in-store and online transactions and also from the content of social media conversations.  In addition to Walmart and Sam’s Club, Walmrt has partnerships with banks, vision centers and pharmacies from which they can also glean additional customer information. Customers can cash payroll checks at Walmart, providing yet another layer of personal information including highly private and unique data like driver’s license and social security numbers.  In many cases, Walmart’s customer information database is chock full of information customers willingly provided directly to the retailer. Many retailers offer a loyalty card which tracks information, but Walmart does not. However, they do offer a reloadable prepaid “Money Card” that looks like a branded credit card for which shopping behavior data may be collected from in store and online transactions.

Venky Harinaryan, Walmart’s senior vice president of global ecommerce, said “we see social commerce fueling the next generation of e-commerce where online and retail stores bring a continuous shopping experience to millions of users” (Walmart, 2011). Social commerce allows a retailer to tap into the current customers, but also potential customers who are in the social circles of their friends or followers, who share similar lifestyle typologies and who may be predisposed to purchasing the same items or similar from Walmart.

Walmart’s research division, @WalmartLabs, is dedicated to the convergence of online shopping and social media. The company started employing a platform known as Mupd8 which has the capability of analyzing incoming social media data and issuing an immediate response, such as a product recommendation. According to Accenture’s 2011 global consumer survey, more than two-thirds of consumers search for and read about brands on social sites (Accenture, 2011) so data mining in the social space is a prudent strategy.

Walmart also employs Shopycat, a Facebook app, which analyzes Facebook statuses including Likes,  Interests and post content (such as a friend making a recommendation to another friend) to make a socially-based product recommendation. Social media analytics gives insight into purchase habits and also sentiment analysis and reputation monitoring. It can also identify key influencers who are brand advocates in their social spheres.  

Here’s a snapshot of what came up for me without putting any personal information into the site. I recently did an online search for Ina Garten recipes on Google and on FoodNetwork.com so Shopycat demonstrated to me that it can be eerily accurate.

Wal-Mart is using these applications to analyze more that 300 million status updates a day (Dusto, 2012)! In late 2012, Walmart made Mupd8 open source to developers, a rare move for the dominant retailer.

In addition to exploiting the latest technology to develop customer profiles, Walmart uses Omniture’s SiteCatalyst product to look closely at the performance of their online shopping portal walmart.com. Walmart is interested in knowing how merchandising across this site directly effects conversions by “mining data in real time to quantify and visually reflect the effectiveness of walmart.com and its marketing objectives,” (Khan, 2004). Valuable observations can include things like product placement, campaign duration, price variables and design elements (such as the color of banners) and how these elements contribute to conversion. By reviewing the functionality of their ecommerce portal, Walmart can respond to the online behaviors they observe and adjust elements affecting the clickstream, purchase funnel and conversion rate of the site.

Due to the massive scale of Walmart’s distribution processes, predictive analytics data can also help the company measure inventory and supply chain systems and improve overall operational efficiency. In 2004 when a series of hurricanes hit the southeastern coast of the US, Walmart mined data for shopping patterns that occurred during previous storms and learned that things like strawberry Pop-Tarts and beer have higher sales in advance of a hurricane (Hays, 2004). When a company is able to identify the most popular items in a region or particular store, and manage down to the number of items in stock, the company can deliver better customer service and sell more.

In April 2012, Walmart allowed customers to order online, but pick up and pay for their orders in cash at their neighborhood store after identifying two groups of customers for whom this service would be appealing – those without credit cards, and those who prefer to keep their financial information private (Clifford, 2012). Knowing the buying patterns of these customers is important, since the individual stores actually act as distribution centers, and need to have the product in as soon as the customer is ready to have it, often the same day.

And because today’s consumer wants retailers to respond on their schedule, Walmart stores in some cities are offering same-day delivery  for holiday season orders placed online (Evans, 2012) which appeals to a segment of their customer who prefers the convenience of online shopping, but does not want to wait for an item to ship. More than half of Walmart’s online shoppers are now picking up items at their 4,700+ U.S. stores (Clifford.) Real time analytics would play a key role in making sure that stores are ready for the customer demand.

So, we know that Walmart is already developing and utilizing its own cutting edge data management and analytics technology to:

·         Build accurate customer profiles

·         Improve all aspects of the customer experience, both in store and online

·         Increase efficiency in their supply chain and distribution systems

·         Develop predictive online experiences that increase conversions

·         Integrate data across multiple customer channels

So What Else Can The Retailing Behemoth Do?

Walmart can continue to customize the shopper‘s online experience, playing to different segments and subsegments.  Overlaying customer profile data with online site usage data could help Walmart design different online shopping experiences based on factors like the age of the user. For instance, should a Walmart.com designed for person aged 50+ look different than the one offered up to a 21 year old shopper? Variations between sites might not only include merchandise options, but also in the layout and architecture of the site based on visitor flow and on-site engagement data.

I would say another big opportunity for Walmart is in the area of customer acquisition. Walmart predicts this will be a tough holiday season for the majority of its shoppers and therefore the company’s fourth-quarter earnings could fall below estimates. Target, catering to a more affluent customer base, expects to exceed Wall Street projections (D’Innocenzio, 2012).

The average household income for a Walmart shopper is in the range of $30,000-60,000, while the median household income of a Target customer is $64,000 (D’Innocenzio). Walmart already outperforms Target in the area of social media engagement. Continuing to play to this strength would be a good strategy. By mining both the social engagement data and shopping habits of their highest income customers, Walmart could identify their top tier income customers and those in their social circles. Walmart could choose to engage in meaningful ways with these prospects and, quite possibly, adapt their messaging from “deep discount” or “extreme value” to “convenience and selection” or something else that would be appealing to this type of customer.

Also of paramount importance to Walmart, and any online retailer for that matter, are Key Performance Indicators such as site abandonment, cart abandonment, and checkout abandonment rate.  These metrics equate to revenue so they deserve close analysis. These metrics also give a clear indication of some design elements such as how things are merchandised, arranged and communicated. The Walmart site is not for the clutter-averse; there is a LOT going on there. To a new customer, it could be overwhelming. Comparing the figures on these metrics for new versus returning (familiar) shoppers might suggest some site changes with new shoppers in mind such as a Welcome Page, or a quick video tutorial on how to best use the site.

The mounds of personal and behavioral data that is obtained across multiple channels  including online, social communities, call centers, mobile applications,  is transformed into actionable insights that can affect customer behaviors in real time. Multi-channel data integration is key for Walmart in doing what they do best – making it easy for us to consume.

References:

Accenture. (2012). The new realities of dating in the digital age: Are customers really cheating, or are you just not paying enough attention? 2011 Global Consumer Research study. Retrieved from http://www.accenture.com/SiteCollectionDocuments/PDF/Accenture-Global-Consumer-Research-New-Realities.pdf

Clifford, S. (2012, July 4). Luring online shoppers offline. Retrieved on December 9, 2012 from http://www.nytimes.com/2012/07/05/business/retailers-lure-online-shoppers-offline.html

D’Innocenzio, A. (2012, November 18). Walmart and Target: A tale of two discount chains. Retrieved on December 10, 2012 from http://www.cbsnews.com/8301-505145_162-57551520/walmart-and-target-a-tale-of-two-discount-chains/

Dusto, A. (2012, October 31). Walmart’s online marketing technology gets a ‘mup-date.’ Internet Retailer.  Retrieved on December 8, 2012 from http://www.internetretailer.com/2012/10/31/wal-marts-online-marketing-technology-gets-mup-date

Evans, B. (2012, October 12).Walmart, Amazon, eBay: Who’s the speed king of retail? Retrieved on December 9, 2012 from http://www.nytimes.com/2012/10/10/business/wal-mart-tests-same-day-delivery.html?_r=0

Hays, C. (2004, November 14, 2004). What Walmart knows about customers’ habits. Retrieved on December 9, 2012 from http://www.nytimes.com/2004/11/14/business/yourmoney/14wal.html?_r=1&pagewanted=print&position=&

Analytics Yellow Belt

Today was a day of highs and lows. I enjoyed checking out the Google Analytics reports and watching the traffic on my blog increase as everything synched up perfectly. In my mind, I was transforming into an Analytics Ninja!

 

We were asked to set up some goals and use Google Analytics to see how well, or not, our blog sites were meeting them. There are four types of goals and they are all pretty straightforward:

 

• A URL Destination Goal
• A Time on Site Goal
• A Pages Per Visit Goal
• An Event Goal

The goals I set for my blog page were pretty simple. I wanted to set some benchmarks for Time on Site and Pages per Visit, and I wanted to track an "event," which in this case was the number of people who played the “Hanukkah song” video in my previous post. Google Analytics is set up to track pages, but not things that occur within pages, like embedded video. Video is generally a good way to increase Time on Site, especially if it is compelling or, such as in the case of a tutorial, useful and illustrative of a concept.

 

Goals are pretty important to people who are trying to accomplish something with their web site. Leading visitors to a purchase or getting them to fill out a form are two easy goals you can set up, or maybe you just want them to spend a designated amount of time on the site, which is really all I had hoped to accomplish. Not all goals necessarily lead to a purchase, as it really depends on the nature and purpose of the Web site. I wanted people who found Her Online Space to linger a while.

Things were going so well for me.

 

And then...I got cocky. I couldn't leave well enough alone. I wanted to challenge myself, and post a video to track an "event," simply, the number of people who hit play. Anyway, it crossed my threshold of knowledge and it quickly went south. Being a newbie to analytics, I have no business messing with HTML code. Really, I don't. There is a little code manipulation required to track the "event" and somehow I made error after error and, well, it just wouldn't work. Be my guest and watch the video, but I can't tell if you were there or not (at least at the time this blog post was published). Though a little defeated, I tried to keep it in perspective.

 

I set a Time on Site goal at four minutes since the video clip was about that long, and determined that a logical Pages Per Visit goal was three pages, hopeful that people would not only check out the video, but also some previous posts. When I looked at the goal completions on December 3, I had 20 total goal completions,  10 each for Time on Site and Pages Per Visit and a 0 for Video Plays for disappointing but obvious reasons.

 

 

As a marketing professional, it is important that I at least know the language of analytics. As more aspects of business move online, it only makes sense to know how to interpret the success or failure of various campaigns. Analytics is how you do that. Analytics expert Avinash Kaushik would like us all to be Analytics Ninjas. Personally, I am hoping for maybe yellow or orange belt status.

Rest assured, I will probably not be employed in a role where I need to input tracking codes myself. However, I do see myself in a role where I might need to weigh in on the ROI of a particular promotional effort or public relations campaign. So knowing how to engage in meaningful dialogue with skillful Web site architects and designers will be a good skill to have. Also, I admit, a working knowledge of analytics is a pretty key skill for the modern-day marketer.

 

I was inspired by some additional information on this very topic. Vaughan (2012), writing for Hubspot's inbound marketing blog, suggests that its prudent for marketers to remember that web analytics is not the complete picture and that "marketers really need much richer data to understand the performance of their marketing campaigns, something that web analytics alone can't provide." Naturally, the measurable metrics available on Google Analytics or other platforms are useful in telling us how a web site performs in a technical way. Knowing things like the clickstreams, the funnels, and the duration on site is useful. But it is the job of the marketing professional to interpret this data in a meaningful way. The marketer looks at all the data, across multiple channels (including some that are not online) and makes adjustments in strategy to positively impact ROI.

 

The bottom line is this. If you have a web site, or manage one, you need to know how it all works together. The most basic metrics of time on site, page views, traffic sources and bounce rate are helpful but only if you know how to apply them to make good business decisions. Leavy (2011), writing for Entrepreneur, suggests five simple things you should be able to know from your web site analytics.

 

1. Do people already know you, and how are they finding you, and why?

2. Are you bringing in potential, qualified customers?

3. Is your social media presence bringing people to your web site?

4. Are visitors "bouncing" from your home page, and if so, why?

5. Are they visiting the right pages and getting the information they need to "convert?"

 

Any marketing professional worth her salt is interested in improving sales, impressions, awareness...whatever the case may be. It is critical to be aware of all of the useful elements in the tool box, and the possibilities of analytics-based decision making.

 

Marketing used to be accomplished by the old “spray and pray” method, where you spend a lot of money putting out a lot of information to the masses, never really being sure it was working.  Heffernan (2010) describes old-style advertising as unaccountable and that “companies are slowly but determinedly moving from unaccountable media advertising to traceable, analyzable results.” And that is exactly what is happening in the online ad environment where you can view instant analytics and make changes to improve the message, the channels and the audience.

 

I can't say I'll be applying for any web analytics jobs soon and I should probably avoid messing with any code. But, I do forge ahead with the confidence that I have a functional understanding of the foundation of this discipline and how I can apply it to my marketing campaigns.

 

References:

 

Google Analytics. (n.d.). About goals. Retrieved December 2, 2012 from http://support.google.com/analytics/bin/answer.py?hl=en&answer=1012040

 

Heffernan,  M. (2010, June 10). Spray and pray: Why does anyone still buy advertising? CBC Moneywatch. Retrieved on December 2, 2012 from http://www.cbsnews.com/8301-505125_162-44340243/spray-and-pray-why-does-anyone-still-buy-advertising/

 

 Leavy, J. (2011, July 11). Five things you should know about web analytics. Entrepreneur. Retrieved on December 12, 2012 from http://www.entrepreneur.com/article/219955

 

 Vaughan, P. (2012, March 8). Why you need marketing analytics, not web analytics. Retrieved on December 2, 2012 from http://blog.hubspot.com/blog/tabid/6307/bid/31705/Why-You-Need-Marketing-Analytics-Not-Web-Analytics.aspx



Video Makes Blogs More Interesting

As we approach the final home stretch of IMC 642, Web Analytics & SEO, we are being asked to generate more traffic to our fledgling blogs in order to gain a more comprehensive understanding of all of the tools that exist to draw traffic, increase conversions and analyze the success or failure of an online site.

I don't think that I'll be winning any awards for my thought leadership on this topic, but that doesn't mean I did not grasp many of the concepts outlined in this class. I consider myself fortunate to have had this opportunity to enroll in the Integrated Marketing Communications (IMC) Masters Program at the esteemed West Virginia University. As a marketing professional, I know that knowledge of this particular field of study will be increasingly useful in the next decade.

We've worked hard these past few weeks, cramming in a semester of content into a nine week class. The holiday season is upon us, so soon we will all be decking the halls, baking up a storm, and possibly shopping 'til we drop. Actually, scratch that shop 'til we drop thing. IMC 642 definitely taught us that we can sit at our computers in our PJs and wait for all of those highly targeted behavioral ads to tell us what we need to buy this season. Privacy? Who needs it?

Whether you will celebrate Christmas, Hanukkah, or Kwanza, one thing is true. Many of us will be working very hard up to that final assignment  due on December 24 (Christmas Eve)! And, we will all be ringing in the New Year as graduate students, eager to complete our WVU Master's Degree in 2013!

I thought I'd offer a little break in the study action with one of my favorite seasonal videos of all time. I bet you like it too. Don't worry, it's rated PG.   HAPPY VIEWING!