Let me tell you a quick story that I’ve shared before on this blog but is still relevant today. For 10 years I walked into the same department every six months, buying business suits and clothing. Each visit, I started from scratch on my own, lost, trying to navigate a maze of racks.
A couple years ago, I started working with a personal shopper. He keeps a database of everything I have bought and looked at. I call him, and he considers my history, my current needs, his store’s inventory, and my real-time feedback to pick out five full outfits that work together and match everything I have previously bought. This shopping experience is much easier and painless for me. Why would I ever shop anywhere else?
Like it or not, my initial experience is pretty similar to most shoppers on any e-commerce site. How can e-commerce organizations duplicate my personal shopper experience on their websites?
Just like in the real world, you need to understand each shopper, respond in a relevant way in real time, and operationalize your strategy within your company. In this post, I’ll use analogies to the real-world experience to bring these three areas to life.
1. Understand Each Shopper
Just like my personal shopper, your website needs to understand each person at the individual level.
Understanding an individual begins with assessing what that person is actually engaged with. If I’m in your store and I picked up a shirt, but immediately put it back on the rack, you would be able to easily tell that I was not interested in that shirt just by observing me. But if I picked up a second shirt and spent several minutes looking at it (viewing the price, looking for the right size, holding it up to observe arm length, etc.), you would be able to tell immediately that I was much more interested in that second shirt.
But in the digital world, both of those interactions would likely be viewed the exact same way by an e-commerce site. They would both represent a “click” on a PDP. That’s because most e-commerce sites don’t consider engagement when assessing what a shopper is interested in — they just judge by clicks. But it’s very clear that I’m more interested in the second shirt than the first.
Question: How are you assessing true engagement in your digital channels?
Assessing true engagement allows you to put shoppers’ behaviors in context to understand their affinities and preferences. If you’re observing me in your store looking at those same shirts, you can learn a lot about the specific types of shirts I like. To start, you’ll notice that I’m spending all of my time in the men’s section, never in women’s or kid’s. Maybe I’m spending more time looking at blue shirts, specifically dress shirts in a high price range. If you get close enough, you may observe that I’m spending most of my time looking at wrinkle-free shirts.
The point is that you can learn more about my behavior when I’m browsing in your store beyond just the fact that I’m interested in a specific product. You can learn that I like the color blue, I gravitate toward a higher price point, and I might be a business traveler. The trick is to consider the attributes of the products I like to determine my preferences.
Question: How do you map context (e.g., categories, brands, price, etc.) to customers to gain a deep understanding of their individual affinities? What might that look like?
Next, you need to understand what I’m looking to accomplish in each shopping trip — my “in the moment” intent. Let’s say that in your store I’m looking at a blue, wrinkle-free dress shirt. I leave the shirt section and head over to a different department for a few minutes, but then I return to that same shirt. I start to walk away, but then I come back to check the price of that shirt. Then I start to walk away again.
If you were observing this behavior in a store, you might reach out to me to see if you could answer any questions about the shirt or if I needed a different size or color. But what about the digital world? You may just let me go, even though it looks like I’m really interested in dress shirts today — specifically that particular shirt.
Question: How are you determining your visitors’ intent in your digital channels? What might that look like? What is needed?
Unified Profile for Each Shopper
All the information you gather about your shoppers won’t do you any good if you can’t bring it together in a central place. Imagine if several different people are observing me in your store, and they each see me take different actions. If they don’t talk to each other about what I’m doing, then they will all have a very different view of my behavior. In the digital world, you can bring together past purchase data, in-store data, call center interactions, your own predictive models, survey data, and more to form a unified customer profile. Without it, you won’t be able to act on any of this data that you collect.
Question: Does your customer data reside in different islands? How are you bringing it together for insight? For action?
2. Respond in a Relevant Way Across Channels in Real-Time
After you have all of this data on engagement, affinity and intent together in one place, you can act on it. This is how you show your shoppers that you understand them.
One of the most important aspects of responding to shoppers, both online and offline, is that your response is in real time. Let’s say I’m still in the store looking at a red shirt this time. I’ve tried on the red shirt, I’ve checked the price, and I’ve compared it to other shirts. Meanwhile, you stand there watching me but you don’t interact. When I come in two weeks later wearing that red shirt, you shouldn’t ask me if I’m still interested in it! The moment for that response has definitely passed.
If you can’t determine my affinity and intent in each shopping trip and then immediately act on it, then you’ll always be responding to your shoppers next time, not this time.
Question: Does real-time matter to you? How are you doing personalization today? In true real time?
With all this information you have about me, it would be a shame to waste it by presenting irrelevant recommendations. Why recommend something to me based on what other people liked, rather than what I like?! In your store, you may know that most people who buy jeans also buy socks. When you see me buying jeans in your store, it’s natural that you may recommend socks to me. But if you suggest white socks, that could be a missed opportunity because I haven’t shown any interest in white socks. I’ve already shown that I’m most interested in business/dress clothes, so I might have responded to more professional socks in a darker color. Or better yet, I may be interested in a higher-priced sport jacket to go with my professional attire. That’s completely unrelated to what other people like, but it could work for me.
Each person on your site is unique. While relying on what other people viewed or bought may be a good place to start, it doesn’t demonstrate that you understand your individual shoppers.
Question: What kind of individualized approaches do you want to take with recommendations based on all the data points you have about your shoppers?
Full Experience Across Channels
If you recall from the intro, my personal shopper knows me very well. He can personalize to me. But the cashier and the other clerks don’t know me and would not be able to speak to me like they do. This is like the experience you get with most digital retailers. On a PDP on the website, I may see a somewhat relevant recommendation. But throughout the rest of the site, the mobile app, the call center, in the store…nothing!
Many retailers lack the ability to personalize consistently to shoppers across all their channels. Personalization doesn’t need to be confined to one spot on your website. All that data you collect can be leveraged across your full site and across email, mobile, search, and more.
Question: How do you want to personalize the full experience? Are you prepared to personalize across channels leveraging common behavioral data?
Machine Learning Plus Human Wisdom
To pull off everything we’ve talked about so far, you need machine learning at every level. It’s the only way to scale your efforts — you shouldn’t need to create hundreds or thousands of rules to create an individualized experience for each person. But many marketers are understandably uneasy about the prospect of putting their shopping experience in the hands of a machine. That’s why you need a solution that gives you control over your machine-learning algorithms.
Imagine that you’ve given your in-store staff specific categories to recommend to shoppers as they are shopping. You may want to promote socks more heavily in a specific week. So if a floor associate sees me buying dress shirts, she may recommend socks to me. But she can use her own understanding of what I’ve been looking at to direct me to the most relevant pair of socks for me.
The right solution will allow you to test and tune your own algorithms, adding specific layers of control where you need, so that you can trust that they are effective.
Question: What are you doing with machine learning today? What do you want to be doing?
3. Operationalize Your Strategy Within Your Company
Finally, after all the work you’ve done to understand and respond to your shoppers, you need to ensure that personalization is operationalized within your company.
You need a solution that is easy to integrate with all of your other tools and data sets. You need to be up and running quickly. You need something that will allow you to continuously test and iterate, so you can learn what campaigns work best and move quickly if something doesn’t work. You need something that enables machine learning-driven experiences and uses predictive analytics to help you identify opportunities and potential problems with your personalization campaigns. This all comes down to selecting the right personalization partner that will provide you with the greatest level of control, while giving you all the support you need to be successful with one-to-one personalization.
To learn more about how Evergage can help you understand each shopper, respond in relevant ways in real time, and operationalize your strategy within your company, request a demo today.