Shade Vaughn (00:04):
Hi everyone. Thank you for joining us today for our webinar on Navigating Disruption, analyzing the new truths of enabling AI. My name is Shade Vaughn, I’m the chief marketing officer in North America at Capgemini. I am joined today by Jerry Kurtz, Executive Vice President of our Insights & Data practice, and Dan Simion, Vice President of Capgemini NA Artificial Intelligence, Data Science and Analytics Practice Lead. We’re going to present a brief overview of our perspective on insights, AI-driven programs. And then we will do a bit of a panel discussion between Dan and Jerry tackling a few of the key questions that we’re hearing today from business leaders in our discussions. And then we will open it up for any of you that have questions. Feel free to type them into the chat window. We’ll cover as many of those as we can. I look forward to a lively discussion. So, with that, Jerry, I will turn it over to you.
Jerry Kurtz (01:00):
All right. Thank you. Thank you, Shade. And before we get into the real meat and some of the questions, I just thought that I would start with just some of the fundamentals, because we may have a very varied audience. And for some of you, this is motherhood. This is a mother and an apple-pie slide. But for some of you, I just wanted to provide some context. When we talk about this space of data analytics and AI, we’re really talking about three fundamental components of what we do, those of us who work in the space, what we do. So, first and foremost, everything we do is really to enable or to actually achieve business results, leveraging AI and business intelligence. And so, really starting on the left, we’re ultimately trying to solve complex problems for our clients, leveraging AI technology, leveraging BI and visualization technology.
Jerry Kurtz (01:59):
The big areas really, for the most part, they haven’t changed much in the last several years. You know that some of the big areas include anything to do with the front office and growth, customer, marketing-related analytics. You’ve got the whole world of supply chain and operations. You’ve got the advent of anything to do with the Internet of Things or what we call intelligent industry. And then, particularly with COVID, what we’re seeing more and more is just more attention to anything related to risk analytics. I grew up in manufacturing and the term risk and risk-related analytics, over the last decade, hasn’t come up that much in industries, [outside] of financial services. But more and more, we’ve got to think about all of the use cases about how do we mitigate risk during this new normal.
Jerry Kurtz (02:46):
So, that’s a fourth bucket. That’s not listed here, but it’s coming up more and more in our business for sure. To enable all this, there’s a ton of work that needs to be done around transforming data, whether it’s master data management, transaction data, metadata, you can see the list, a lot of work there. And ultimately, what we’re trying to do there is get the quality of the data and the readiness of the data up. First, we want to get that up, get the cost of this infrastructure down, get the quality up and get the value of all of these solutions as high as humanly possible. But then there’s also, from a workload perspective, perhaps a smaller area, but no less important, which is anything to do with organization and governance. What’s the role of IT? What’s the role of the business, data ownership, data stewardship, your business glossary for things like KPIs? And your organization, how you organize around data, AI and analytics. Lots of partnerships, hundreds of vendors out there from a technology perspective that you need to manage.
Jerry Kurtz (03:53):
And then also, not written here but underlying all that, is just anything to do with change management. These are transformational change programs to AI-enable your business, so you got to be thinking the same way about these programs as you do ERP programs, for example. Lots of challenges around adoption and consumption. So, I just wanted to level set with that. Back over to you, Shade.
Shade Vaughn (04:18):
Great. Thanks Jerry. So, let’s tackle some of the questions that we’re hearing more and more today. How can we be smarter about implementing AI programs and roadmaps and getting more quickly to ROI?
Jerry Kurtz (04:33):
So, that’s a great question. And in a nutshell, I’ve been at this, I’ve done a lot of things in my career, but I’ve spent most of the last 10 years or so in the AI, data, and analytics space. And in a nutshell, this is really the primary learning, what’s on this chart here. Which is, that at the end of the day, most of the companies out there typically are spending between 20 to 30 percent of their available data analytics and AI spend on the business use case side of the house, which at the end of the day is where most of the value lies. But they’re spending 60, 70, sometimes 80% of that available spend for hardware, software, services, people, labor, et cetera, on the data and the governance, more so the data, but the data and governance elements. But data in and of itself is rarely bringing value, it’s more that it enables the value once you apply BI and AI to it.
Jerry Kurtz (05:37):
In fact, you’ve probably all heard that phrase, “Data is the new oil.” I push back on that phrase because I feel like it’s the AI and the BI on top of data that is really where the value is and where the quote/unquote value of that oil is. So, at the end of the day, what we’re looking for with our clients is to make sure that you have a roadmap, whether it’s a one-year, two-year, three-year roadmap, where all of these tracks – the use cases, the data, the governance – that these are not just in and of themselves road mapped out well, but they’re aligned with each other. And that if you align, if you sequence, if you ideate and then prioritize and sequence the use cases in the right way, what we have found is we have found that these programs can have very, very high ROI, in excess of 10X the cost of the program. For every dollar you’re spending, you’re saving 10 or more.
Jerry Kurtz (06:38):
And that we’re talking about programs that unlike, I used to put in ERP systems and when we would go after business-case approval, you’d have paybacks of two years, three years, and that’s not what we’re talking about here. The North Star in this space is to be looking for payback in a number of months, not a number of years. So, that’s the North Star here, 10X ROI, payback in one or two quarters. If that’s not what you’re getting out of your program, you should consider re-looking at the roadmap and the approach. We’re not a big fan of doing data transformation for data-transformation’s sake. We’re big fans of making sure that it’s enabling very specific roadmaps on the business side. Back to you, Shade.
Shade Vaughn (07:25):
Thanks Jerry. So, Dan, let’s bring you into the conversation. In obviously a fast-changing business landscape that we’re dealing with now, relying on history to guide our decisions may not work right now. How can we anticipate our customers’ needs better?
Dan Simion (07:45):
That’s a very good question, Shade. Given the actual context, where we all are going through unprecedented times, I think that’s very, very important right now to look at what are the challenges that other companies are facing and how these challenges are getting solved for? So, I’m going to go over one example of one of our clients and the challenge that we are facing with this particular client in the consumer-goods space is that while they are trying to build a sales-recommendation engine that was eventually designed in normal times, I could say, now they are facing the issue that given the new situation that we are in, we need to change the approach. We need to change the methodology. We need to change the way that we are doing the work in order to be able to solve for the business problem.
Dan Simion (08:53):
The idea is that whatever was before was a rigid rule-based recommendation engine. And at the same time, the data that was used was the data for the past three years, which definitely it’s very, very hard given the new condition to be able to actually predict anything into the future. At the same time, by using this type of approach, there are challenges where, relying on the fact that was a high latency to identify the insights, given the fact that it’s taking a lot of time until the data kicked in and was no real time, actually actions that they are taking. Given the new reality and the new normal, obviously they are looking for things that are going to help them out to be much (more) nimble, to get to the pace of real-time insights and using very few data points on the historical data, in order to calibrate the forecast that we’re going to move forward. And at the same time, the products that they are going to be recommended to their particular clients and customers, they change the mix of them and that needs to be reflected in the sales recommendation.
Dan Simion (10:18):
So, this is one example when actually taking action and being nimble and changing the approach for the sales-recommendation engine is generating a huge impact to the value, back to the Jerry’s point on the previous question. If we’re going to move to the next slide, we’re going to give you another example on the fact that, the models generally, whenever we are talking AI and machine learning, we know that we are building these models but these models are not getting better with age. We need to retrain these models all the time and make sure that you account for the model drift whenever the data it’s not anymore representative for the future.
Dan Simion (11:10):
And this is an example, from the work that we have done for a client of ours in the media industry, where we forecast the sales, but what happened is obviously once, let’s say the governor of California has the stay-at-home orders in place, you could see how drift, how that model was not able to predict it here on the blue line. That’s where we choose the pattern to actually change and retrain the model. And you could see the difference between the two of them moving forward. As a result, and if we’re going to make it to the next slide, you’re going to see, actually, the impact that we predicted in terms of the revenue growth for this particular client. And this is helping them out to actually take measures in terms of pricing strategy for this particular movie, giving the fact that the revenue was changed significantly.
Dan Simion (12:25):
So, these are the two examples where we could see how we need to be very careful whenever we are using the AI and machine learning and how we need to retrain or remodel given the new reality and the new normal moving forward. Back to you.
Shade Vaughn (12:47):
Thank you, Dan. Let’s stick with you. How have you seen firms adapting their risk-mitigation efforts and is this as critical as sparking new growth or improving operations?
Dan Simion (13:01):
This is a very good question. And, as Jerry mentioned at the very beginning of this session, we saw that the risk now and mitigation risk is not only staying with the financial institutions, but it’s going across different industries, different sectors. And given the pace that we are seeing the changes right now, there is no choice but to act now. And that’s a code from the general director of the World Health Organization. And whenever we are talking about how to mitigate the risk, given the impact of COVID-19, we are thinking in three major buckets, I could say. The first and foremost, I could say it is around the workforce empowerment.
Dan Simion (13:51):
We need to make sure that the workforce, it’s feeling well and they feel that they are going to a secure workplace. And any outbreaks that are going to happen, they are going to be monitored and they are going to take action immediately. And that action needs to be taken by HR. And that’s why we are talking about the HR empowerment, because they need to identify what are the issues, to identify the workforce optimization, and what are actually the messages that need to go to the workforce in order to make sure that everybody’s coming back to work in this new normal. And nothing is more important than taking the buy-in for the leadership. And that’s what we are talking about, leadership empowerment. And that’s where the leadership needs to manage the crisis. And to manage the crisis means to have the right balance between the team, the workforce that they have available and, at the same time, maintaining the operations. And not only maintaining the operations but perform in these new conditions.
Jerry Kurtz (15:14):
Yeah. I wanted to just take a step back. We talked earlier about risk-related AI and analytics becoming more and more prevalent, and it’s keeping us busier on a daily basis. But I thought I would just bring to life a few additional examples. And, for example, in consumer products, you’ve got the whole issue around food safety and quality and tracing and product-related risks. There’s environmental risk mitigation that needs to happen within our client base, a supply-chain risk tied to COVID or other challenges on a day-to-day basis. Companies that where, again, history does not really give us a good indicator of future, now we need to do a different job at our predictions.
Jerry Kurtz (16:09):
And, for example, financial forecasting and related risk. What’s the revenue? Trying to do proper algorithms around revenue forecasting is something that we spend time on, or just re-looking looking at sales-forecasting engines, given that the new normal, the future is very different from the past. R&D-related risk and investment optimization is an example of a space that we’re spending time in. That example earlier when you’re making movies, is it going to be a success, is it going to be a flop? And then just really assessing where you should be putting your marketing dollars. And then, last but not least, during this time of crisis, human-related safety and related risk. All of these areas need analytics, machine learning, more so now than ever.
Shade Vaughn (17:04):
Thanks, Jerry. We have a question in the chat window. I think this is probably a good chance to tackle it. You’ve more or less answered it. The question is: how do you think differently about leveraging data to drive decision making? And I thought it might make sense, you’re in the benefit of being able to talk to business leaders in a lot of different industries. I think that the Quarantine for Business offering that your team has developed is a good example of this, might be worth talking briefly about it. What are you seeing? How are business leaders thinking differently about it? And has it changed our thinking about the abilities that we’ve built up or offers that we’ve developed?
Jerry Kurtz (17:49):
The main thing that comes to mind, first and foremost, is that more and more our clients are wanting to make sure that they are not looking in the rearview mirror anymore and just applying traditional BI, traditional, what we call descriptive analytics to their business to say, “How are we doing today versus last month or last year?” They have more and more need to be able to look forward and to predict what’s going to happen in the future and apply machine learning to their business and AI-enable their business. So that’s one thing. Dan, you want to talk briefly about what we’re doing in the COVID, back to work or some specific examples related to COVID?
Dan Simion (18:33):
Sure. More than happy to do so. So, as Jerry mentioned, the data right now is, looking backwards is not anymore good. But at the same time, what we are doing in COVID and work-from-home situations that we are right now, we are trying to make sure that we gather all the data that it is coming in in real-time, and using things like, let’s say, knowledge graph, for instance, where we are making sure that we are taking advantage of multiple data sources and structure multiple data sources in order to help the companies to take decisions in real-time, based on the breadth of information that it is available right now.
Dan Simion (19:21):
It’s not just having the traditional data sources that Jerry mentioned. And, when you’re looking backwards and trying to compare what was versus what it would be versus what was last year, that’s not good enough. You need to go after and compile, and that’s a lot of work in the data space. To make sure that you bring all this data and you apply technologies and apply things like machine learning and AI, and at the same time, leveraging knowledge graphs in order to be able to drive the insights in real-time.
Jerry Kurtz (20:03):
Yeah. The other thing I just wanted to mention, Shade, in answering that question, is that what we’re seeing around leveraging data to drive decision making is that again, I hear the term data a lot, and I hear the term data-driven businesses. And I prefer to say that what our clients are doing or what we hope they are doing is that anything to do with data analytics and AI, including specifically the data side, it should be value based. All the work that’s done, you don’t do data transformation for data transformation’s sake. You do it because you’re trying to bring value to the business. So, center around the use cases, center around the value, and then let the data transformation come along for the ride, rather than trying to become a quote/unquote data-driven enterprise.
Jerry Kurtz (20:55):
And it’s a nuance, but I think it’s a very important one, and it’s more important today given the fact that there are some cash-related constraints out there and that some of our clients are struggling.
Shade Vaughn (21:10):
Good. Good. Let’s move on to our next question we’ve teed up. Is this now the time for cash containment or aggressive innovation? And how do you strike the right balance?
Jerry Kurtz (21:23):
Yeah, so I’ll start and Dan can chime in. But you see this term in the top right there around pursuing the right balance. And I think that it’s key to be very, very pragmatic here. When you talk about, are you going to be innovative and you’re going to invest or do you have cash containment, we recognize that every company is different. So, I’m just going to give you a few examples. So, in some cases, some of our clients we see in maybe high tech or telecommunications, or in some cases, media and entertainment, some of these companies are firing on all cylinders and they’re focusing on continuing to be innovative and ideating, applying more AI to their business. They want a good business case, but they want to invest.
Jerry Kurtz (22:14):
Then you’ve got other situations. For example, some of our clients that are heavily into brick-and-mortar retail, the world’s changed, the foot traffic is less, they need to move more to online. They’re challenged by this crisis that’s been going on the last few months, very challenged. And so, cash containment may be king. But then we have other clients where, in one division, you might have, for example, a CPG company that may be selling some of their products to restaurants, and those restaurants have been closed. They’re struggling and they’re in cash-containment mode, and yet another division of the same company might be selling to grocery stores and might be doing great. And they’re a little bit more on the innovation disruption side of the house. So, at the end of the day, every client’s different and we just need to be pragmatic.
Jerry Kurtz (23:05):
But regardless, the North Star that I talked about earlier, the number one takeaway from this call today, the North Star is to try to establish programs where you’re talking about five to 10X or more value for every dollar you’re spending, with paybacks that are absolutely less than a year and preferably less than six months. And, regardless of whether you’re in cash-containment mode or you’re in disruption and innovation mode, you should still be working towards those kind of targets. Just if you’re in cash-containment mode, obviously you’ve got some constraints that you have to apply to the program that may be a little bit different. Dan, your thoughts?
Dan Simion (23:46):
So, whenever I’m thinking about the right balance between cash containment and innovation, we are thinking about the right balance between, at the end of the day, the two big balances that we are showing there in the slide. That’s around the balance between the investments and innovation in technology, but at the same time, the innovation that needs to be going across these two components: technology, IT, and business. And this needs to be balanced and needs to be in sync.
Dan Simion (24:25):
So, wherever we are talking about the use case, we need to think about what’s the value that it is driven. And at the same time, as Jerry mentioned, it’s not all about trying to do a data transformation for the sake of doing the data transformation. Especially right now, when we have the constraints of the COVID-19 impacts. And we need to think about, what’s the value that it is driven by the organization by this use case to the organization? And we need the alignment with the organization and the processes needs to be in place in order for organizations to adapt these new ways by embedding AI and ML in their processes moving forward.
Dan Simion (25:10):
And I can give you an example of a retailer. We are talking to them for the last couple of days, and this is one of their main concerns, is how I’m going to strike this balance, because I have some stores that are open, some stores that are not open, following local regulations. How am I going to find the right balance? What is my right approach here, in order to be able to move forward with the innovation while I’ll be very aware about the fact that the money are very, very scarce resource right now? Back to you, Shade.
Shade Vaughn (25:52):
All right. So, Jerry, what are some of the biggest challenges that we’re seeing when clients are embarking on data-driven transformation at a time when they’re faced with uncertainty?
Jerry Kurtz (26:06):
My reaction to this question is that what we’re seeing, and I was also a part of some of the discussions with this retail client the last couple of days and they gave us great evidence of this, is that the challenges that they are seeing and the challenges that we are seeing with our clients is that the technology part is actually the easy part. And I would never say that data science and AI are easy. These are brilliant people working on these projects. I’m just saying that we trust, ninety-nine times out of a hundred, that the technologists know what to do and know how to do it and can solve these problems. But the issues and challenges that we’re seeing are the non-tech pieces of these programs. For example, when we build these AI solutions and data-science solutions, what are the users going to do with it?
Jerry Kurtz (27:01):
How are they going to action it? How are they going to leverage this to actually do their job better and do it differently? So, we think it’s very important to work backwards from that question and make sure that, from the beginning, you’re making these solutions consumable and actionable. So, it’s not just about the data science or AI track, it’s about the visualization and the dashboard and the consumability track.
Jerry Kurtz (27:30):
The other thing that comes to mind is, if you’re going to have a program that has 10X value or more, I believe, in many cases, we are limited by our own creativity. For example, in order to get that kind of value, you don’t want to just ideate and come up with three or four use cases and pick one or two and go, you really need to cast the net wide and then go and try to figure out what are 30, 40, 50 potential use cases of AI that we can apply to the business?
Jerry Kurtz (28:04):
Because then if you cast that net wide, then you’ll definitely have the opportunity to find five or six or seven of those 50 that are the real low-hanging fruit, the stuff that is low cost, it’s high value, the data is readily available, and executive sponsors are ready to support it. Those are the projects you want to start with, and that’s how you drive early payback and high value.
Jerry Kurtz (28:30):
Again, if I take a step back, it’s the change-management aspects. This may sound a lot like an ERP program. It’s totally different in many ways, but you got to be worried about adoption, education, the technologies are changing constantly, communication, change management, sponsorship, that these should be business-sponsored initiatives, again, not just doing data transformation for data transformation’s sake. And then a big one is that different clients are approaching things differently from the standpoint of what’s the role of IT, what’s the role of the business? The answer is it needs to be a seamless partnership between the two to drive these value-based programs. Hopefully, that’s helpful.
Shade Vaughn (29:17):
Dan, anything you’d add to that?
Dan Simion (29:21):
I think that’s covered very well, and there are examples from our clients where we are moving through these data-driven transformation programs and actually where we were in the middle of the data-driven transformation programs, but they need to change the course given the new reality. And they needed to focus much more on the bottom line. What’s the value and how I’m going to bring the value while the data-driven transformation and data-transformation programs are in place? And that’s a couple of examples where we’re just bringing to life these types of issues that we are seeing with our clients. They are facing the reality. We have clients in the hospitality industry, we have clients in retail, they face a lot of cash constraints, and what’s happening right now they are rethinking the programs and they are thinking about how I’m going to bring to life the use cases at a very fast pace in order to be able to get the value that we’re going to fund, actually, some of the transformation programs that eventually they are putting on hold, or they are scaled down given the uncertainty that we are facing right now.
Shade Vaughn (30:44):
So, I’m going to remind you, if anyone has questions, feel free to enter them into the chat window and we’ll tackle those. Jerry, I’m going to throw a wild card at you, we didn’t discuss this prior, but in addition to your daytime job, leading our Insights & Data business line in North America, you’re also our global lead for IoT. And I don’t want to ruin the perspective that you’ll share and what I would expect we’ll do in a future webinar or series of webinars, but we recently made a very significant acquisition and play in the intelligence industry base with Altron, something we’re very excited about and being able to combine operational technology with information technology. And we really see that as the future. Do you mind just kind of teasing some early thoughts that you have on how this is going to impact business leaders over the next couple of years?
Jerry Kurtz (31:48):
Yeah, I think that this whole…I’ve been involved in the IOT space since 2012, and I remember GE coining the phrase, I think it was the industrial internet and the Internet of Things, and that type of thing. And so, it’s been around for a while, but I still feel like many of the companies out there are still just dipping their toes in the water. I’ve worked with clients in the past, BMW and some others, that have been doing this stuff big time for years. But there’s also a lot of companies that are still just dipping their toe in the water. And then when we talk about IoT, I’m not just talking about sensors and the OT and the plants or SCADA, and some of the tech side of the edge of getting things from the sensors, I’m talking about value.
Jerry Kurtz (32:42):
I’m talking about how do you AI-enable the business as it relates to things like connected products, medical devices, connected vehicles, connected or what we call digital manufacturing and all the assets you have in your plants that you want to be able to do predictive maintenance, predictive quality…computer-vision type of technologies, imaging. And then last but not least, and these are huge, huge categories, is other asset classes. You’ve got products and manufacturing assets, but what about utilities and oil and gas in terms of all the assets they have out in the field? Real-estate assets, smarter buildings, this space has so much value and so much potential for application of AI. And we couldn’t be any more excited with our digital engineering group, our transformational invent group, our acquisition of Altron. We think we’re well positioned in this space and we look forward to working with our clients more and more to solve some big problems. So hopefully that helps, Shade.
Shade Vaughn (33:48):
Good. Good. Well, I will ask each of you to share any final thoughts you have. Before we get into that, I want to let everyone know that if you would like to schedule some time, we’re offering 30-minute virtual sessions with our experts, that would be happy to do a deeper dive into your questions and challenges you’re facing. We have a number of offers specifically around assessments, business case, roadmaps, data-estate modernization, gaining ROI from your AI investments, managed services. We would be more than happy to schedule that. If you’re interested, please just email Reuben Garcia. It’s firstname.lastname@example.org. We’d be happy to set up a session at the time. Jerry and Dan, big thank you for all of your time and all that you’ve had to share today. Any closing thoughts from either or both of you?
Jerry Kurtz (34:44):
Yeah, I apologize if I’m redundant in any way, but I just want to make sure that if you take nothing else away from this conversation, that if you are running some form of data analytics and/or AI initiative, that if you are not getting somewhere in the neighborhood of $5 to $10 or more for every dollar that you’re spending, or you’re not getting payback within a year or less in some form of something that at least closely resembles self-funding, we encourage you to take a step back, call us, if it makes sense to re-look at your roadmap and re-look at how the use-case track is defined and how it relates to the data-transformation track, and the governance track. If you can get this right, those are the kinds of results that you should be getting out of these programs. Dan?
Dan Simion (35:44):
On my side as final thoughts, I think that you guys need to think about that AI, it’s nothing magic about it. AI, it’s a reality and you need to take advantage of it. And as one of the codes that we had on the slides, you need to take action now and making sure that you are the forefront of using AI, ML type of solutions to move forward and bring value to your companies.
Jerry Kurtz (36:18):
Shade, if you don’t mind, just one other quickie. Some of you may also be trying to figure out, you hear about robotic process automation or intelligent automation. And how might that relate to AI or are these different programs? I wanted to just share one example where I worked with a client over the past few years, where they wanted to automate some of their supply-chain-planning functions. Now there are hundreds of people around the world that do a lot of manual work around, let’s say, distribution-requirements planning or demand planning, and they wanted to automate some of these roles so that they could have people spend time on more valuable activity. And I bring it up because some of these initiatives are leveraging automation technology and data science and machine learning and AI together.
Jerry Kurtz (37:11):
In many cases, these aren’t separate things. These are joint things. So, that’s an example of a capability that we have. And again, just another use case out there, so to speak, or use-case family, I should say. So anyway, thanks everybody. Thanks Shade for hosting this. And we look forward to engaging further, and there are going to be more of these, where we deep dive into some other areas around, whether it’s supply chain analytics, IOT analytics, customer analytics. There’ll be more of these over the coming weeks. So be on the lookout, and we’ll be in touch.
Shade Vaughn (37:45):
Thank you both. And thank all of you for joining us. And as Jerry mentioned, we’ll have a series of webinars you can find that we’ve done already on Capgemini.com around navigating disruption, especially in the four pillars of people, performance, operations, supply chain and customers. And then we have some deeper dives into cloud, cyber security, and obviously today with AI and data. And if you enjoyed this, please feel free to share the content with your colleagues. And we look forward to bringing you fresh content in the weeks ahead. So, thank you.
Jerry Kurtz (38:23):
Thanks everybody. Take care.
Dan Simion (38:25):