Hi Seth, tell us about yourself, your background?
I have been in the information management space for over 25 years. I founded Earley Information Science in 1994, and we have worked with a diverse roster of Fortune 1000 organizations across industries—life sciences, manufacturing technology, financial services, insurance, healthcare and others The work that we do is really about making information more usable and more findable, and ultimately more valuable.
Online commerce was booming in 2020, and so did consumer reviews. – How can brands better utilize this data to improve their customers’ experience?
Voice of the customer information is found in the comments, feedback, and the complaints that people are posting on social media. Companies can use sentiment analysis to understand what is contained in a post. If someone says “Not bad,” you know that should be interpreted as “good,” or if they say “I wouldn’t say it’s terrible,” that is another positive. If you just do a search term and find the word “terrible,” you might say oh, that’s negative, but in fact, that’s not the true sentiment.
Text analytics and machine learning can be applied to these comments to understand them in the way they were intended. Of course you should always be listening to your customers across multiple channels, including comments that are posted in review forums. Surveys could be used too, but there’s a lot of unstructured information in those posts and in that customer feedback. Building an analytics program around monitoring social media and customer feedback is very important. It can then be correlated with things like new product introductions, new feature introductions, upgrades, or enhancements. So again, it’s important to be listening to this and giving customers a mechanism to provide feedback and, of course, acting on all that input
Seth’s tips for personalization
What tips do you have for companies that want to improve their personalization strategies?
Personalization is a very challenging domain. You can tell this is the case because many algorithms just do some simple analyses—looking at shopping basket analysis, for example ”People who bought this also bought that,” That’s a that’s kind of low fidelity personalization. If you purchase shoes in the past doesn’t mean you’re going to be purchasing those same shoes again. Retargeting ads for the things that a customer has already purchased doesn’t make any sense. Many personalization strategies are fairly poorly conceived from two perspectives.
One is that often, there is not an architecture around the data content, the product information and the customer. Those elements need to be carefully modelled. The second is that the orchestration around personalized messaging is poorly understood. In the context of the journey stage, effective personalization requires very intimate and nuanced knowledge of customer needs. Where is the customer in their process, and what is going to help move them to the next phase? What represents a personalized message?
If the customer is trying to make a selection, it’s essential to identify the piece of content that we’re going to present, so the content, the product information, and the offering all need to have attributes that can be matched with the customer signals. This requires an information architecture applied to customer, product and content data.
The personalized content or message that is going to resonate with one group of users may not resonate another. Many years ago we built an information architecture that supported personalization for a mall operator, and the problem is once we asked the marketing team how they differentiated across groups, they could not tell us what content should be directed to one group versus another. All the customers were getting the same content, because the marketing team didn’t know enough about the needs of their segments. There are a number of algorithms that can help with this, but it’s a complex topic and it requires a lot of synchronization and orchestration across multiple outlets to do it effectively.
In terms of technology, personalization might require a customer data platform; it certainly requires a product information management system, content systems, and E commerce. All those systems that present that information and that that make up that user experience need to have a way of responding to a customer’s digital body language and then presenting something that’s appropriate for them. I would say the first thing to do is understand the customer, map out their journey, and map out the technologies that are supporting each stage of the journey.
The next step is to start to understand those customer use cases and scenarios and determine what the personalized content will be. You need to start somewhere in a way that is meaningful to the customer, but not overly complex. One option is to look at past purchases, or personalizing email messages, changing headers or offers depending upon the audience. Understanding that customer journey, customer lifecycle, and the technologies that support that entire experience is the way to begin.
Do you think personalization and customer-centricity are going to become increasingly more relevant in the coming year? How so?
Yes, it’s always been about personalization in some way, even back when we were trying to improve usability and findability, but it’s becoming even more important. It’s about understanding the needs of the customer and surfacing the content and information that matches their mental model. Your site should organize products and solutions according to how your customer thinks about their problem, how they go about making their selections. When customers come across a website that is intuitive, it’s like “Wow these guys really get me”, and “Oh, this is great stuff.” In those cases, the customer is reacting to the way that information is curated and presented. The content doesn’t have to be personalized exactly to that individual, but it needs to be personalized to the segment, and test on the user segments.
Cognitive applications or cognitive AI or cognitive computing it’s a field of artificial intelligence that is meant to reduce the cognitive load on the user. Reducing the cognitive load on the user means that when they website they don’t have to work really hard to find what they need; it should be I want it to be surfaced for the user. If it is not easy, people leave your site because they can’t find what they want. No one wants to go through the hard work of sifting through a bunch of irrelevant content—they want to get the content they need, so it is absolutely increasingly important. So again it’s an evolution, an ongoing effort to improve the customer experience. Part of that is understanding exactly what they need when they need it.
Social media pages have become crucial for companies in most industries, especially in eCommerce. What’s the most common mistake you see in a company’s social media strategy?
One big mistake is not providing meaningful or valuable content to the user. Sometimes social media is just shouting at the user, ”Pay attention to me,” or they try to get people to retweet if they like a product. They develop programs that are not necessarily geared towards solving the problems of the user. This can happen in either a B2B or a B2C context. Content needs aligned with the needs of the users and has to be content of value or interest.
I think the biggest mistake companies make is that they think of social media as just a communication, a one to many communication out to their marketplace. But it should be a way to listen, and engage, and respond. It’s about listening and creating a dialogue. We need to think of social media as a conversation, and we need to focus on the value of that social media content for the user.
What’s the most insightful book you read in 2020?
Hmm, I’d have to say that the most insightful book I’ve read in 2020 is my own book, The AI Powered Enterprise. I am joking but I reread it a year after I wrote the book, and I did see a lot of interesting insights from the big picture view. That aside, I am reading a book Goliath’s Revenge, it’s about how large organizations need to reinvent themselves in order to compete with small, agile organizations. It really says that large companies can be agile as large enterprise, they can experiment in different ways, and use their market presence, volume, customer relationships, and resources to really compete with the disruptors. They can disrupt the disruptors. I’m partway through and it’s excellent so far.
It looks like working from home is going to stay with us for the foreseeable future. How should Executives gear up to the changing times?
Many organizations have had remote workers in the past or have been able to accommodate remote work to a greater extent during Covid. Yes, it’s a trend that’s going to continue and even increase. It’s really about making sure the virtual, extended workplace has the tools and the technology that users need to overcome what you lose the in-person collaboration and serendipitous encounters.
When employees are working from home, they lose the ability to just ask their co-worker or go up to the fifth floor ask their boss. There’s certainly an ability to do that with some collaboration tools. But there is something that’s lost, and I think putting together the repositories and the knowledge bases and streamlining that digital workplace, because you can’t have acts of heroics upstream and have a seamless customer experience externally. It’s about working on those efficiency and engagement tools.
Your employees are your customers as well, and you need treat them like customers by meeting their needs. That means mapping out the employee lifecycle and understanding their scenarios and use cases- understanding how they need to look at their information and investing in tools to streamline work and make it make it fun and engaging and productive.
Younger generations entering the workforce have been digital natives – they’ve grown up with digital technologies their entire lives. They have an expectation that you’re going to have certain capabilities. If you have a rickety old intranet and a crappy knowledge base and 10 different applications to enter information into to do the job – that’s not a very satisfying work environment. Companies want to come across as innovative, and it’s not very innovative.
Last but not least, what is your favorite CX metric?
Rather than picking out a single CX metric, I would say it’s better to pick out a single quality, and that is for the metric to be actionable. Likelihood to recommend, customer satisfaction, click-throughs and conversions are all good metrics, but what you really want to see is some indication that people are finding the information, products, or solutions they need. Some of these metrics may be a little deep in the weeds, but they are indicators of customer engagement. You don’t want to see “pogo-sticking,” where customers go into your website and bounce right out.
The way to foster engagement is by understanding the customer journey and picking out metrics that show whether the journey is going well. If it’s not, the metric that you chose should be actionable, meaning it should give you clues about what to do to fix the problem. The best metric is not always going to be the same for every situation. Maybe you need a metric that indicates whether your search capability is adequate, or a “Was this helpful?” question to guide personalization. This approach will help your company pick the right metrics and continually track them for course corrections and ongoing improvement.
