Back to Top

[A] Podcast: Exploring CEM

Interview With Karol Jarkovsky

A look inside at how Customer Experience Management (CEM) has changed since Kentico's inception.

Bio

Karol Jarkovsky is the Director of Product at Kentico Software.

Transcript

Cruce Saunders:
This is Cruce Saunders with [A]. I'm here with Karol Jarkovsky with Kentico Software. Karol, thank you so much this morning for meeting me and getting a chance to talk through customer experience management, and content management systems. Can you tell us a little bit about your background?

Karol Jarkovsky:
Yeah, absolutely. Hi, Cruce. Hi, everybody. Thank you for having me here today. As you mentioned, I work for Kentico. I'm actually a Director of Product at Kentico. We are a global software company, and a vendor of Kentico EMS, which is essentially an all-in-one WCMM platform, web content management platform with key online marketing and e-commerce features. As far as my background goes, it's mostly technical. I guess you're going to see from the discussion today that I really love technology, but my past five years involved more in product development and product management than anything else.

I have essentially ten plus years of experience in web development, content management, customer success, technology consulting, and as I said, I'm very passionate about digital. I mean about digital transformation, about digital strategies, how technology influences, and in many cases, reinvents, the customer experience.

Cruce Saunders:
That's a long history and we've had a lot of changes over the last decade, especially in content management and customer experience management and the platforms we use for publishing content out to users. It's evolved a lot since Kentico was founded over twelve years ago. Can you speak a little bit to the difference between content management systems and customer experience management? It's one of the top questions we hear. What is the difference between content management and customer experience management?

Karol Jarkovsky:
Yeah, I agree with you that it's one question that we get a lot as well. I guess there are different answers to the questions, so I'm going to provide the way I look at it. Exactly as you said, I remember when I started at Kentico over ten years ago or something close to ten years ago, content management was mostly about creating the content and then publishing it. Simply, it may be a basic workflow that would just push the content to the website. There wasn't a lot more about it.

Over those ten years, we get to the point where the content management, creating and managing the content itself, is not enough anymore, and instead, what you need to do is to create the content and serve the content that is relevant for the customer, the website visitors. In order to be able to do that, you really need to engage the insights from other applications like analytics, marketing automations, and others that will help you to personalize the content and make it really relevant for the person looking at the content.

I think that when it comes to providing the ultimate customer experience, specifically from perspective of content innovation or technology innovation or the business efficiency, the organization actually needs to invest into technology and data that help them to deduce what should be the content that needs to be displayed and what should be the next best action or best content for a specific customer. That's the difference as I look at it in a nutshell.

Cruce Saunders:
That makes a lot of sense. The customer experience is something that we're all responsible for now. It's almost become the job that we all have, whether we're in marketing or operations, to deliver customer experience across a lot of different channels and it's interesting to me that we're able to do so much more with an understanding of the visitor now, and Kentico is one of the handful of platforms that gives us a deep insight into session data on-platform, because before, to your point, you had all of these different silos of analytics and other software that's outside of the platform that's delivering the content. You can't easily stitch them together.

Now, there's this opportunity to react to the session with personalized content. That's pretty remarkable. Can you talk a little bit about the idea of experience consistency regardless of channel? The customer experience across channels and how a platform can participate in coordinating those experiences.

Karol Jarkovsky:
You are spot-on about the fact that before there were silos, and each silo was represented by a different type of application. As I mentioned, the information from all those different applications needs to come together at some point, so I think that one of the big challenges that are ahead of any platform, not just Kentico but any other platform that is responsible for providing those personalized customer experiences, is how you connect all the data together. I think that's why we are seeing such a boom in popularity of different customer data management tools. [Tools] that are aggregating all the data that then the content management system or digital experience system at Kentico can tap into and can personalize the data, or can personalize the experience accordingly.

Cruce Saunders:
Looking at the experiences that we're managing across various channels, how does customer experience technology and the Kentico platform in particular figure into managing customer experiences across various channels and silos?

Karol Jarkovsky:
I think to start off answering that question, I think it's important to realize one thing that for me, personally, took quite some time to understand, but it actually was very enlightening when I figured it out at the end. What I'm talking about is the fact that customers are constantly connected, and everyone has a smartphone, everyone has a laptop, tablet, game console. You've got pocket consoles, the ebook readers, smart watches, smart cars, TVs, fridges, you name it.

If myself or you or perhaps the people that are listening to this podcast and other technology savvy persons look at a list of all those devices, we see different devices that are simply used by customers to interact with a brand online. We see the specific technologies that are running behind each device. We see even technical imitation of each device, and subsequently, a technology that is used by the particular device. Those are transparent to us, but for most customers, at the end of the day, those customers we are interacting with, they don't know and they don't frankly need to know that there is such a difference.

They don't really care how difficult it is to optimize experience for a given channel or how tricky it is to keep the consistency among channels. All they really care about is how their experience is, whether their experience is appropriate for the given channel. Not only that, but [they also care about] experience with their favorite brand, whether that experience is consistent across all the channels that they're interacting with. Technology, such as Kentico or any other digital experience technology is really close with the challenge to accessing and consuming the aggregated context or session data, if you will.

If you take a look at the technology that is available, you can see all those context or session bookers popping up all around. Companies like Oracle, Google, Facebook, and others are slowly getting to the point where they're ready to resell or make the data accessible to the masses. For technology, it's important what next step it's going to be taken to be able to leverage the data that is available. But it's not only just about accessing the session that is shared or is available to everyone, but I think that that variable technology is something that can enable technology like Kentico to provide really the next best experience.

If you imagine that there will be an application that could collect the context data about the person that is wearing that variable and figure out if there have been any injuries sustained in a bike crash, for example -- obviously they don't want to get involved in a bike crash, but these things happen and if the person riding a bike gets off the road in the middle of nowhere, it's really hard to rely on passing drivers to help you, right? That's why the information from the variables, the context that it can provide may be very important, and the technology can also leverage information and connect to other technologies, so the organizations can use the experience management platforms to actually type in the information that then can be used to provide an automated insurance claim or damage report. Through the experience platform API, connect with the closest repair store and stuff like that. I think the biggest challenge is how to connect and leverage all the data that is available and how to use it to find that next best experience.

The thing is that the next best experience cannot be something that can be figured out based on some manually codified set of rules. You have to actually have a technology that is using some kind of machine learning that is able to understand that customer context of session that is available, and to then present the right set of content or services or products or actions to provide that next best experience. This is because any single person cannot be expected to build the rules that are necessary to support all variability, granularity, and multi-source nature of the context understanding that is required to present each customer with the best next experience.

Cruce Saunders:
Wow. Okay. There's a lot to unpack in there. You spoke to some really interesting other topics that I think relate to this. Essentially, what I'm hearing about is almost like a services bus across silos that gets unified at the CEM platform, and there's all these interaction points that are happening with operational systems, because the picture you're painting is much more of a complete customer experience beyond just the marketing funnel, and I think sometimes in customer experience management, a lot of the stakeholders are marketers and so, a lot of times, the focus is on automating processes through the buying cycle, but there's all of these post-purchase processes, which are key customer journeys, and you were describing a couple of those and they interact with external systems.

One thing I heard about in your keynote from a conference last fall is this not mobile-first, but API-first concept where we need to design with web services and with APIs in mind, which I think speaks to this idea of the services bus that integrates process across an enterprise. Can you talk a little bit more about your vision for API-first and how that plays for decisions that our listeners might be making in their organizations about building software and applications and web properties that interact with their CEM. How do they think about APIs and how do you think about APIs?

Karol Jarkovsky:
Yeah, the API-first concept is not really that new. I think it's very similar to what's going on with the cloud. We always had the servers that you could rent from someone that was managing it for you, and you were essentially only utilizing the operation system, the platform that was running there. I think a similar thing is going on with the API-first where the concept of building robust, scalable and open APIs, which is a very brief description of the API-first approach, was here a long time.

What I think is new about this, and then the reason why I essentially talk about it in the keynote, is because we believe that with the raise in popularity of IoTs, Internet of Things, or in general the increase of all those different connected devices that are able to consume the content, I think that's the main reason why as an architect for the enterprise, you have to be aware of the fact that the mobile screen, the smartphone screen, is not really an edge interface anymore.

There are hundreds of different devices using tens of different interfaces that are going to, at the end of the day, consume the content. Then you will have to optimize the experience for those different devices, and you don't want to have replicated content. You will have to share the same content. Essentially, we will go on a very low level. The content repository that contains the content has to provide an interface, universal enough, that -- universal and robust and scalable, of course, enough to be able to serve all those different devices.

What I'm saying is, mobile first was born at the time when the industry realized that everything is headed to mobile, so moving from the desktops to mobile devices, and we got to that point almost two years ago, but in the two years in line with what you mentioned in the beginning, with all that progress, we got to the point where we really need to think beyond mobile, and that's why the mobile-first strategies are being slowly rendered obsolete and the API-first is taking over.

Another point of view is through the latest front-end development technology. If you keep an eye on all the technologies front-end framework that are coming out that makes performance optimization easier, that makes developers more efficient, optimizing the presentation for a specific device that gives developers tools that enable them to create a better experience for their customers, you will notice that there is a new and better technology, front-end technology in particular, available every couple months.

If you think about it, if you would create your customer experience based on a coupled API, meaning that the business logic would be highly connected with the presentation. If you were to improve your customer experience, it would take a lot of effort. It would take a lot of development time, a lot of resources, and so on. That's why APIs, especially the coupled APIs that are at the very core of API-first approach where business logic is completely separate and isolated from presentation logic, are becoming so popular and increasingly important. As a vendor, our focus has to be on the core API that allows us to manipulate the data while digital agencies or organizations can use essentially any front-end framework to build a content presentation and then maintain it and replace it, but with better framework. They will provide a better customer experience down the road.

Cruce Saunders:
Terrific. With all these APIs moving data back and forth, like you said, there's a lot of data being aggregated and there's almost this requirement, long-term, for machine learning and other forms of artificial intelligence to get involved. Can you talk a little bit about your vision for how machine learning and how cognitive computing starts to interact with customer experience management platforms over time?

Karol Jarkovsky:
Sure. Before I actually answer the question, I'm a huge fan of Tim Urban, and if you don't know his blog, I strongly recommend to go to waitbutwhy.com.

It's incredible reading. I love it. Sometimes, it's slightly difficult to grasp all the concepts that Tim is presenting, but it's very, very visionary. Visionary, but also very realistic, because it's happening and cognitive computing and AI is not impacting only customer experience, but it's impacting every aspect of our lives.

To you question, we actually already see real world uses of cognitive computing that are having very concrete impact on customer experience, and there are several examples of its use in advancing personalization, customer service, or customer engagement platforms in general, and you can find cognitive computing, for example, that is comparing personal medical data of a patient as they're leaving hospital and then the data are collected over time from using the variable devices and based on the results of comparison, the doctor is automatically notified if the condition is changing or if there are any symptoms of coming problems.

You can have cognitive personalization engine that recommends you custom spice flavorings. It's actually something that I tried when I lived in the US, where you basically share the information about yourself, and based on the information that you share, the machine learning algorithm that is searching the data, looking for the patterns, it's recommending the custom flavorings for you. I specifically recall a study from Accenture that was talking about customers really getting increasingly frustrated with the quality of the service they were receiving while they were interacting with the organizations and surveys found that over ninety percent of customers have to contact a company multiple times. Ninety percent of those are put on hold for too long, eighty-something percent are forced to repeat their issues to multiple agents as they are transferred. My recent experience with one of the airline carriers, for example.

That's why I believe that the biggest potential for cognitive computing to impact customer experience is actually customer services. Of course, there are applications of cognitive computing and AI and machine learning and even content production or content altering. Content nurturing processes are also probably the easiest application or the most wide-spread application of the machine learning is in marketing automation and predictive analytics. Those are all areas that are already leveraging the machine learning and I think that -- two examples.

I will get to the customer service example one in a minute. Before that, just an idea how really a single machine learning can help improve the efficiency of users. It's customer experience, but from the marketer perspective, while creating these strategies. I'm pretty sure that you would agree that one of the biggest challenges with digital marketing these days is that a lot of companies lack a comprehensive digital strategy. Most of them are the early stages, putting together the digital strategies. If you have a tool that can provide a smart personal assistant that can help you to start off in your digital strategy and then, as you move through the stages in the strategy, as you execute the strategy and you move forward, it will provide you some hints, some recommendations, some information on what the next activity on the marketing side should be. That would help tremendously, especially organizations that are smaller that are not so mature from a digital perspective. That could be a huge, huge help.

The other example, and the one that is actually, I would say, the first real use of AI or cognitive computing in customer service. It's something that was released just recently. The product is called Satisfaction Prediction by Zendesk, one of my favorite startups, actually, because these guys are very approachable and they are sharing their experience, a lot is very transparent. What they did is, they used machine learning algorithms, searching through the patterns of ticket histories. Zendesk is essentially a CRM and service desk blended in one, and so, as an agent when you are answering a ticket, when you have a ticket open, and you're going through the conversation with a customer via email, there is an indicator, a satisfaction prediction indicator, that is telling you how likely the customer is going to be happy about the outcome of the ticket resolution.

If you think about it, it's all about sentiment analysis. I think the next big application that is going to be probably widespread, is going to be sentiment analysis and predicting what is the mood of the person you're dealing with, because that can tell you a lot about how you should approach that person and stuff like that, right?

I would just wrap up with one more thing. No matter what application of cognitive computing or AI you can think of or you will come across, it always boils down to I would say three main basic elements of cognitive computing that could be applied to a different variety of problems. Those three main elements are: big data, machine learning, and natural language recognition or natural language processing and sentiment analysis. I think that using those three components and looking at, how can I apply these components to whatever my organization is doing, whatever the industry is doing? I can find a very useful and very realistic application of cognitive computing and AI in general.

Cruce Saunders:
Beautiful. Yeah, that is a very clear vision, and it's a good one to hear. We're in the cognitive era already and I think a lot of marketers, especially, but really decision-makers within the digital ecosystem in general are all somewhat confused about how to interact with artificial intelligence. I think as AI and machine learning gets baked into the platforms, we'll make it easier for all of the marketing decision makers, especially, to be able to, for example, deploy sentiment analysis if they can turn on that as a module in the platform. Do you ever see that happening in the Kentico roadmap? Machine learning built on-platform?

Karol Jarkovsky:
Absolutely. I think that there's no other way, honestly, because you think about it -- with increasing amount of data that we have to deal with, right, that marketers, specifically, have to deal with, the more customers that they need to service because everybody gets connected and everybody wants to interact with the brand. There is no other way for the marketers than just get machines to help them to process all those requests, all those inquiries, and help them really provide the best experiences possible. As far as the roadmap goes, we are really looking at including machine learning in one of our products fairly soon. I'm pretty sure it's not going to stay just with the marketing suite, but we would like to see machine learning, as I mentioned at the beginning, be a part of content production process as well. We are currently doing the research on what applications would have the biggest benefit, provide the biggest value to the content editors.

Cruce Saunders:
Wonderful. I have at least another ten minutes worth of questions, but I know we need to wrap up. There's the Internet of Things and the semantic web, and really, your approach towards strategy. All of those things are very interesting, but will have to wait for another chance for us to talk, but in order to wrap up our conversation today, I wanted to ask you a little bit about your general vision for where Kentico is headed next as one of the very few platforms in the entire world that manages user session data along with content and is able to deliver customer experiences across lots and lots of channels, and react to them in real time with personalization without a lot of custom hacking, but using the features already available on the platform.


That's a very powerful position for Kentico to be in. Love to hear your thoughts on the next year, two, three for Kentico. Where it's headed, broadly, and what you'd like to see become part of the experience for marketers on the Kentico platform. As much as you can share with our audience about that direction, because I know you drive it. Let's end with that.

Karol Jarkovsky:
I think that one thing is very exciting but at the same time, very challenging for any product management person, I guess, no matter what industry or organization that person is part of, is really dynamically changing conditions in the market. You can envision yourself doing certain things two, three, four years ahead, but the conditions in the market are changing so drastically, so fast, especially with all those technologies popping up that it's really impossible to predict on the higher level, feature level, what should or what would be involved in a product.

But exactly as you said, there is a high level vision that every product should have, and really for us, it is about providing developers and marketers with experiences that are natural for them, that are in line with their way of work, and they are experiences that allow them to achieve their goals faster, and whether that means employing more artificial intelligence or cognitive computing for content production and marketing automation or predictive analytics, all those technologies, all those things we've been talking about or whether it means anything else, we will see over time.

What I think is absolutely crucial, not only for Kentico, but for any organization that wants to stay in the digital destruction revolution that we're going through, is innovation of products, services, and even business models, because that's the only way to succeed. That's exactly the reason why, at the beginning of the last year, we had started the concept of internal startups. It's an innovation program that is actually pretty amazing, and I believe still fairly unique compared to other organizations, other vendors and what they do, because we essentially form a group of people, mostly from existing Kentico employees, and we create this so-called internal startup.

These people then, using and applying principles of leading a startup, explore problems and needs of the customers in the market. They're looking for opportunities that we could turn into new products. That's how we actually created our newest software as a service product called Kentico Draft that helps with the content production stage before the organization makes a pick about the content management solution or content experience management platform they want to use.

We have currently two more groups that are looking for problems and needs that are related to content delivery and publishing and digital experience management. We are expecting to find those major opportunities and hopefully we'll cause a disruption in the industry with those. All together basically what we talked about today, all those topics, plus whatever we'll be able to find using lean principles and investigating the market more, is what's going to form the future Kentico product line.

Cruce Saunders:
Wonderful. Agility and responsiveness and the ability to, in real time, engage with the market seems to be a competitive requirement nowadays and it's great to hear the way that that's expressing itself at Kentico with these internal teams. There's so much innovation happening at such a fast pace for everybody. That model makes a lot of sense and hopefully our listeners will reflect on their own organizational approaches and how innovation is managed and look for ways to increase that pace of change inside of their own organizations as we move towards a more cognitive and intelligent world.

Thank you so much, Karol. Is there anything you'd like to leave our listeners with in terms of contact information or follow-up? How to follow you and your work or anything like that that would be a good coda at the end here. Anything you'd like to leave with our listeners to follow up on?

Karol Jarkovsky:
Yeah, absolutely. I would actually love to hear from anyone that has their own experience, their opinions and suggestions and stuff like that on what we discussed and anything else that Kentico is doing. If you can follow me on Twitter @Kentico_KarolJ, that will be appreciated, or you can follow the work that Kentico is doing at the @Kentico tweet handle.

Cruce Saunders:
Terrific. Karol, thank you so much for your time. I know you've got an incredibly busy schedule managing a giant team of really talented developers there at Kentico. We appreciate your vision and sharing that with our audience. Have a wonderful rest of the week and thank you to everybody else who's come to listen to my conversation with Karol Jarkovsky at Kentico.

Karol Jarkovsky:
Thank you, Cruce. Thank you, everybody.


Download Podcast Download Transcript