Retail And The Rise Of The Machines

Many retailers will consider themselves to be technologically progressive. After all, retailers who weren’t quick to embrace the evolution of e-commerce over the past two decades have failed. The era of brick-and-mortar dependency died as soon as consumers could purchase online. Mobile devices have led to a new era of consumer empowerment, where expectations are high and seamless shopping experiences between all channels are demanded.

This jolt in the evolution of consumer behavior has resulted in many retailers struggling to keep the consumer engaged as they move between sites and stores using a combination of devices. The consumer is one tap, swipe, or click away from abandoning your brand and completing an order with a competitor.

Data has become the most important tool for retailers to keep consumers engaged. Automated messages are triggered when a consumer takes an action, cart reminders are sent when items are left behind, offers are targeted, and products are recommended.

These strategies have worked well for many retailers, but how long can this success last? Do we truly have the time, budget, and brain power to effectively use the mountains of data we have amassed over the past decade to connect with the consumer as they navigate this multichannel, multi-device landscape? The execution platforms in which many retailers use this data require human interaction. Segments must be built. Offers must be created. Cart reminder timings must be set. But does this human interaction help or hinder the progress of a multichannel marketing program?

Retailers must take the step to go beyond the limits of human knowledge, begin to adopt a more progressive perspective on customer intelligence, and trust machine learning, or risk declining shopper engagement rates and lost sales. The ever-increasing, constantly evolving demands of the consumer, the need to sell (or compete) online and in stores, and the emergence of new channels where consumers learn about your products, are challenges that all retailers face today. The complexities and intricacies required to succeed in this ecosystem are only going to become more advanced. Have we met the human limits of e-commerce innovation?

Unfortunately, there currently isn’t a button you can simply press to start a machine that will accomplish all of these tasks and send your site and store sales soaring. This paper, however, will show you three ways you can start to use more advanced customer intelligence data and machine learning to shift your strategies, expand your conceptual thinking, and boost the success of your marketing programs across channels.

Mine The Machines:

Collect the data essential to building a unified customer profile that spans channels and devices.

Predict The Potential:

Harness machine learning and predictive data to better understand your customer and transform your marketing.

Evolve The Execution:

Use unified profiles and data science to promote the products and offers that will lead to higher engagement and increase sales regardless of channel, device or location.

Mine The Machines

Many marketing programs currently use profile, behavior and purchase data to send targeted messages. This common process, illustrated below, shows how marketers capture data, automate their programs, and occasionally include personalized elements before deploying the message.

The messages, whether automated or sent en masse, represent marketing to a specific moment. As the shopper moves along the purchase path, automated messages are triggered based on further action or inaction by the shopper. While these messages can be relevant to a particular point in time on the shopper’s journey, they often take a myopic view of who the customer is and what promotions will lead to a purchase.

A shopper may visit your site, search and browse items, then drive to one of your stores. At the store, they could share photos of the product on Facebook, interact with your app, and use their phone to add the item to their cart on your mobile site.

Should these actions result in a cart reminder email? A retargeted display ad? Product recommendations on your site? A store abandonment email? An in-app push notification? How do you know which one is most likely to lead to a sale?

Determining the optimal messaging option for this single shopper would be a time-consuming exercise. How could this possibly be done for all of your shoppers?

A first step is to break away from this limited view of shopper behavior and start using a unified profile. With all interactions and points of engagement consolidated and linked to the individual, you are providing a foundation for real customer intelligence strategies, insightful trend reporting, and messaging that reaches the customer where they are most likely to convert.

A truly unified profile will collect data from online and instore interactions. Rather than focusing on a traditional, linear path to purchase, a customer’s shopping experience is the sum of the parts. These interactions, regardless of channel, device, or location, work to define the shopper, communicate with them more effectively and move them closer to becoming a customer.

Predict The Potential

After building a unified profile, the next step is to take a step back and let the machines go to work. Don’t worry, there is still an element of human interaction. Even planes using autopilot do not operate completely independent of human involvement.

The purpose of using data science in tandem with a unified profile is to give you, the marketer, the ability to make more informed decisions about your marketing, and streamline, or even fully automate, many of the tasks that are currently required to engage customers and inspire an order

Do you know which products your defecting shoppers are likely to purchase? How much would a 1% increase in engagement impact your monthly revenue? How many of your most active customers are truly chronic product returners?

Shoppers are overwhelmed with marketing messages, and yet most messages rarely connect with the shopper in a truly meaningful way. Consider how personalized web interactions have become in daily life. Social networks are the primary way of communicating with friends and family. Phones are in our pockets and purses wherever we go. Tablets are in our beds as we fall asleep. Isn’t it time that your marketing messages engage your shoppers in a more individual way?

Imagine this scenario: one of your most loyal customers, let’s call her Katie, lives in New York City but spends much of her downtime (and her money!) in Miami. She researches her purchases while at home in Manhattan, often going into stores to try on items. She asks her friends for input on Facebook and builds wish lists in your app.

Your annual clearance event is launching as Katie takes off for Miami, ready to spend her money. How are you going to reach her? Is it in the inbox or via your app? Does your data define her as a New Yorker or a Miami resident? Do her web interactions classify her as an online shopper, even though she always buys in-store?

Data science and machine learning take the burden away from having to manage these needlein-a-haystack scenarios. Instead, customer intelligence-driven campaigns will reach your customers with the right offer, wherever they are currently located, and on the right device.

This is truly next-level automation, and it’s time for marketers to catch up to the consumer.

Evolve The Execution

You may be wondering if moving toward a unified profile and incorporating machine learning into your marketing means abandoning your current successes and strategies. The simple answer is that no, you do not have to undo everything you have done. In fact, these advances into the world of next-era marketing, where channels and devices are part of the process and not dividing lines, allow you to execute programs that were once unimaginable.

Here are a few ways you can bring more advanced customer intelligence into the equation to expand upon programs you likely have in place today.

Post Store Purchase Messaging

Sending an email after a shopper has completed an order on your site can help set the foundation for long-term loyalty. Targeting customers who have completed a purchase in a store with post-purchase messages can also be very effective, but treating your online and in-store purchasers the same way isn’t the best solution.

Sending the same post-purchase communication to both online and in-store purchasers is a one-size-fits-all option that can lead to a disconnected experience after a purchase has been made. Customers completing an order in a store should also benefit from post-purchase messaging that is relevant to where they purchased.

Purchase data that is part of your unified profile will identify your in-store purchasers and send them messages that are customized to their experience. Valuable customer satisfaction data can be gained to improve the shopping experience, loyalty can be fostered by encouraging shoppers to return to the location, and you can facilitate a smooth transition for shoppers who want to make the switch to your site when researching their next purchase.

Store Abandonment Messaging

Product page and shopping cart abandonment messages are proven revenue boosters. This success should not be limited to your site!

Triggering a message when a shopper leaves a store will require some location-based technologies, which are becoming much more intuitive and cost-effective for every marketer, regardless of budget or resources.

The data science then comes into play when the shopper leaves the store without making a purchase. Data from the unified profile can be used to populate the message with the optimal incentive, relevant product recommendations, and align with the channel most likely to convert the shopper. Based on the customer’s behavioral data, the message can be targeted to the inbox or mobile app, or even be sent as an SMS text message.

Returned Goods Recovery

Expanded return policies have given shoppers the flexibility to purchase several products, knowing they will return some of the items. This lack of accounting for the negative impact on sales caused by returned items results in marketers using inflated revenue metrics to measure the performance of their marketing programs.

You should ask your marketing platform provider how you can account for this data in your reporting to get a more accurate picture of your conversions.

Returns are a fact of life for retailers, and with many offering customers the option to return goods in-store and online, marketers should seize these interactions as opportunities to encourage shoppers to come back and find the right item.

Product recommendations based on the initial purchase that included the return, combined with previous shopping behavioral data, channel preferences, and other site behavior can be used to create an engaging email that encourages the shopper to reengage with your brand.

Conclusion

The days of building marketing messages meant to usher your shoppers along the purchase path are gone. The dynamics of customer loyalty are shifting, and marketers must move beyond marketing to these specific moments in the customer lifecycle.

To truly innovate and meet the ever-increasing consumer expectations for relevant interactions across channels and devices, marketers must rethink how the data they collect is used to build a full picture of the customer. Marketing messages must go beyond the traditional concepts of a segmentation strategy and automating messages along the customer lifecycle to truly speak to the individual.

To succeed in this new era of commerce, marketers will need to rely on more sophisticated data science and predictive modelling that can provide actionable insights beyond the limits of the human brain.