Recapturing personalized customer interaction in the age of ecommerce

Welcome back

This is a guest post from David Stephenson, Chief Data Officer at DSI Analytics.

In days long-past, when a store manager understood and interacted individually with each customer, the manager had a fantastic capability to understand and to help each customer.  In our age of ecommerce, with millions of digital customers, recent technology developments have brought us back to the point where we can recreate that personal experience.   This is critical to understand, as shopkeepers who only interact with customers while ringing up purchases are as destined for failure now as they were in the old days.

Online Customer Journey

If I’m running an online store, I can approach my digital in four ways:

  • Worst:  I simply record the number of sales made (like the shopkeeper who sits behind his register all day).  This is really, really negligent.  I deserve to go bankrupt.
  • Slightly Better: I use basic web analytics.  This is minimal effort, but at least now I can see basic stats:  how many digital visitors, where they came from, time on my site, etc.
  • Standard Practice:  I’ve got an analyst who set up conversion funnels, UTM codes, ecommerce tracking, integration with A/B testing, etc.  I get some high-level insights but am very limited in what detailed insights I can get and in how I can help customers in a more personalized way.
  • Big Data Method:  The way to regain the personalized customer experience of old days is to store and analyze the entire customer journey, for every customer.   Now I have the ability to once again understand the intents and preferences of my customers:  what they originally thought they wanted, how they modified their product search, when they reconsidered products, when they read reviews or sorted by price or size or rating or whatever.  I can see what product features were selected and then unselected, and which of those selections correlated most heavily with purchases.    I can then start to help my online customers in ways I never could before.

It’s the Big Data method that allows me to truly understand and eventually influence customer journey.

The Realm of the possible

These days there are also relatively painless ways to start using Big Data to recapturing personalized customer interaction (See my recent article Useful Big Data: The Fastest Way  for more practical details).

With relatively little effort, we can

  1. Get answers to extremely detailed questions about what customers did while browsing our sites, accounting for the customer journey and general personas of every single digital customer.
  2. Deep dive into multivariate tests to understand in highly nuanced ways how potential product changes affect conversion rates.

How does this look?  Let’s walk through the process that a traditional shopkeeper would follow, and realize that we can again do this with our millions of digital customers.

Customer Intent

Right from the start, the successful shopkeeper finds out all he can about the person visiting his store, looking to understand things such as

  1. How did the customer find out about the store?  A friends’ recommendation, marketing efforts, internet search?
  2. What is top of mind for the customer looking?   A particular product, the solution to a problem, a gift, a bargain?

By getting answers to these questions for digital customers, we can optimize our marketing efforts, our landing pages, and our product mix.

Customer Preferences and Priorities

As the customer continues the shopping process, the shopkeeper will begin to understand the preferences and priorities of the customer.

  • Desired price range and features.
  • Persistence of search criterion, including price elasticity
  • Prioritization of search criterion, as indicated by sort order and item views
  • Engagement with product reviews

The customer may express these preferences outright, or the observant shopkeeper may imply them by watching the customer’s actions…which products are ignored, quickly put back, or re-considered multiple times.  Getting that level of behavioral insight for millions of digital customers is amazing, and it is possible.

Proactive Shopkeeping

The shopkeeper will help the customer.  He will recommend products most likely to sell (or perhaps with the highest expected profit margin!)

The shopkeeper will keep two things in mind when making these recommendations:

  1. What have I already learned about the customer’s goals, intentions and preferences?
  2. What has worked with other customers?  If I know what similar customers purchased in the past, I can recommend those products, thus increasing the chance that I make a sale and that the customer walks away with something that makes them happy.

The shopkeeper will be very clever in how he nudges the customer into making a transaction.  For example, he’ll know when to offer spot sales.  He’ll learn when to make certain inventory prominent and which features of each product to highlight for each customer.

It’s not enough to say ‘most customers bought this product’.  It needs to be ‘most customers just like you bought this product’.    It can’t be ‘sales increase when I run a 10% discount’.   It needs to be ‘sales increase on this product to this customer base when I offer a 10% discount at a particular point in the shopping journey’.

Observant

How does the shopkeeper know how to personally interact with each customer in a way that produces the most sales… this process that we call ‘conversion rate optimization’?  It’s by remembering what has worked with customers in the past.  The shopkeeper observes and remembers the entire shopping experiences of all past customers.  He can then make observations such as

Most people who came looking for phone X and then moved on to consider phone Y and Z ended up purchasing phone Y;  if they also sorted based on price, then conversion rate increased by 7% when a 5% discount was offered on phone Y.

Wow.  Now I know, for a particular segment of customers, how to both make more money and to have a happier customer.

But not intrusive

Two important things to watch here:

  1. We don’t store or use personal data without specific consent.   No problem.   I don’t need to remember your name or anything that identifies you personally.  I just need to remember your general persona and behavioral signature.
  2. We don’t creep people out with what we know.   Don’t do what Target corporation did and send a maternity advertisement to the pregnant girl’s home just because you figured it out based on the vitamins and skin lotion that she bought over the past few months.

And always learning from the customer

Eventually the customer makes the purchase, or perhaps doesn’t make the purchase, but either way the shopkeeper knows much, much more than he would if he’d just focused on the point of sale.  Not only has he increased customer satisfaction and increased the chance of sales, he has developed his understanding of what works or doesn’t work.  He’s learned what is important to customers and can direct his purchases, marketing and product placement accordingly.

Perhaps the customer was very well prepared, and the shopkeeper has even learned something new about his own products, such as cross-sell or substitutionary potential.  He can later use this new information to make better recommendations to future customers.

To sum it up

Using Big Data technologies, today’s shopkeeper can again understand the preferences, intentions and habits of customers at extremely high level of detail, allowing him to understand and help them in much more insightful ways, thus both increasing his various conversion rates and also providing the customer with a much better shopping experience.

All in all, it’s a lot of potential insight, and it can make a massive difference in revenue.

 

 

 

This is a guest post from David Stephenson, Chief Data Officer at DSI Analytics.

David StephensonBio: David Stephenson works in the fields of Data Science and Big Data Analytics.  He has led global analytics programs for companies such as eBay, Axel Springer and Coolblue (a skyrocketing ecommerce player in the Netherlands).  David has also worked extensively in insurance, capital markets, and financial risk management and has served as an expert advisor to top-tier investment, private equity and management consulting firms.

David completed his Ph.D. at Cornell University and was subsequently Professor at the University of Pennsylvania, teaching applied analytics to graduate students in the engineering and Wharton business schools.

David is currently based in Amsterdam.

Check out DSI Analytics and connect with David on Twitter @Stephenson_Data and at LinkedIn here.
Photo Credit: Alaa Ali / ĻΩooĻΩoo ‏εïз‎ © Flickr via Compfight cc

Leave a reply:

Your email address will not be published.

Site Footer