How To Succeed with Digital Transformation: A Q&A with PwC’s Tom Puthiyamadam
In a recent edition of the Early Adopter Research Podcast, Dan Woods spoke with Tom Puthiyamadam from PwC’s Digital Practice about the importance of building enterprise platforms for digital transformation. Puthiyamadam leads this work for PwC. Their conversation covered many of the examples of PwC clients that have built various digital platforms using enterprise products to create foundations for transformation. PwC recently finished a 2018 Digital IQ survey that revealed a variety of insights about what companies are thinking about digital transformation, what they’re afraid of, and what people are getting wrong with the process.
Woods: What do you think are the most important mistakes that you’ve seen in the market, based on your Digital IQ survey?
Tom Puthiyamadam: We’ve been in this digital journey for ten or fifteen years. And it originally started with the conversation of digital is about IT, digital is about data only, or digital is about marketing. And right now we’re seeing, for the top financial performing companies in our study, their definition of digital has expanded well beyond that. They’re thinking about new business models, they’re thinking about their workforce and their culture, they’re thinking about how to use analytics across the entire enterprise. So for us, we’re not looking for companies that consider themselves digital as being successful. We’re looking for the companies that are most financially successful and that also have a more expanded definition of what the digital domain is, because if you nail that, that means you probably have a digital corporate strategy.
What’s the wrong way to think about it?
The wrong way to think about it is that your digital agenda is being managed by some special team sitting in Long Island City who is conjuring up a ton of new innovations and it’s a 500-person operation within a 200,000-person enterprise. That would be the wrong way to think of how you actually go conquer a digital challenge.
And so what are the characteristics of the companies that are getting it right?
Number one, they have a real executive, most likely the CEO who is championing the digital journey. Number two, they’re putting employee and customer experience right at the forefront and they have somebody in that C suite who is responsible for both. I think once you have that right to begin with then you have at least the starting point to create a recipe to be successful in this bigger landscape of digital.
It seems that one of the most important things is to have a unified, deep, and wide analytic strategy so you can understand what the customer is doing and so that your decisions about where to go can be informed by real data about what customers are doing. It also seems then that you have to look at the channels that you’re interacting with customers and figuring out how you can create more degrees of freedom and different ways of doing experiments to get this right. I’m assuming that the right path forward is to decide how you’re going to do experiments in order to find what’s going to work for your business?
I think about it in two ways. First, I think consumers want to be able to experiment with you, so you need to be able to provide different options and channels for them to interact with you in a store, online, via mobile, and see what those reactions are. Think about the experimentation model beyond just your four walls, and let the customer play with you within that world, as well as your employees. Second, does the organization have an experimentation culture? Are they willing to dabble and do different things and see what works and what doesn’t? And my only ask is to expand on that a little bit and create models where consumers can play with you before you do a big market launch.
In a lot of big companies, there are institutional disciplines that are really powerful and have led to the success of the company. And one of the institutional disciplines you see often is tight control on expenses. Another institutional discipline you see is a disciplined focus on maintaining operating income levels and growing them. And it seems that in a lot of these areas, you must put stress on those types of disciplines to pull off a digital transformation because unlike most of the investments companies have made in the past, you don’t have a good idea of the ROI. How do you advise clients to get over what made them successful in the previous era?
I find that really interesting for a couple of reasons. One, think about what a lot of outside investment firms would do when they look at how a company is doing, how they’re investing. If you take the average private equity firm out there, they’re not really interested in any of their holding companies making a big digital transformation initiative happen. Why? Because they know the failure rates are pretty high. Why are the failure rates so high? If you’re an outside investor and you have a holding and you want them to jump to the next level, you would think that technology would be a great way to make that next big leap. But the returns and the timeframe for that return is so long, it has most outside investors a bit concerned and a bit weary.
It’s not going to support the IPO in three to five years that would really provide them with a way of cashing in on their investment.
That’s right. And as much as we talk about that model, it’s the same with general corporate models too. A CFO and shareholders are equally worried about this three-year investment, this four-year investment, and when they will get the return on it. And when you dig into why you are not getting the return on these big digital transformations, the number one root cause, from our research, is purely around employee adoption.
You cannot get the employees to adopt the new behaviors and new ways of working that brings in automation. You name it; they just can’t do it. Our focus is making sure companies have someone in charge of what the employee experience is meant to be. Because for the last couple of years we’ve been talking about the customer experience. Then we realized that experience is just a coin. On one side is the customer and on the other side is the employee. They’re interconnected. So our advice to our clients is if you can’t nail the employee experience in any digital transformation you’re doing, your adoption is going to fail.
What you’ve said has implications for a topic I’m interested in, which is how to build high quality, productive, effective enterprise platforms. Most companies are not in the business of building their own technology. Sometimes they’ll build an application or integration, but most of the time they’re going to select, buy and integrate platforms.
Now, that integrated whole is quite a unique construct because the same use of Salesforce, Workday, or a variety of other common enterprise software products can be completely different because you find the processes and ways to automate them that are useful to your business. You had mentioned earlier that there are a variety of ways of thinking about building platforms to support a digital transformation. Could you go through a few examples?
Companies today are buying the best cloud platforms out there. They’re stitching them together to build a data lake of some kind to support those platforms to give them the insights that they want.
But ultimately, what are their real assets? Do their assets go beyond just the data that they have on their customers or employees? I think it’s really hard to be a product-based company or a service-based company. But to be a platform-oriented organization—that has a lot of merit for growth.
One of our clients, a big office supply retailer—had the same questions. How do I leverage our assets to compete differently? What do I have? I have a great real estate footprint next to most of my consumers. I have a great customer service staff, I have a great inventory, I have a reasonable digital layer that’s able to interact with consumers. How do I begin to marry that all together to compete with the largest online retailers out there? The thought was why don’t we package that up and understand the true need of the small-to-medium business customers.
So the first step is to have a platform that can harness everything you know about your customer. And if it’s an internal initiative, everything you know about your employee who’s the customer of the internal initiative. In essence, a lot of the systems out there allow you to integrate data from lots of different places, either using graph technology or journey mapping, can be really valuable in creating a wide and deep view of what’s going on.
That’s right. But you need all those views. What is that graph database going to give you? How does that tie to the journey of a consumer? How does that create a new, better outcome for that end consumer or employee? For a small-medium business, what are they concerned about? They love their customers; they love their product. That’s where they start—that’s where they started their business. And how do you give them the ability to focus on that entirely and let the data and the infrastructure be managed by somebody else so it allows you to get back to what you love? It allows you to get back to what grew the business in the first place.
That means building a new sort of platform that will support a Software as a Service interaction between the office retailer and the customer. Now all of a sudden, you’re doing something that looks a lot more like Salesforce in terms of the challenges you have in creating a high-quality SaaS software. When you did that project, did you take something like Salesforce, an existing enterprise platform that was mature, and tailor it and private label it? Or how much did you find that you needed to do through integration and creating a new surface user interface layer?
It was a combination. We actually used enterprise platforms but we had to tailor them a fair bit in this setting to make it work. But as soon as you get into this business, you’re in the business of prediction. You’re going to predict how much that small-to-medium business need in supply x, y or z. If you’re going to be prescient about what your consumer needs, there’s no way to do that unless you’re able to harness all the data that was sitting in four different silos in the organization. Now you’re trying to bring that all together. What happened at the retail store, what happened online, those conversations between you and the service rep—all that needs to be able to come together if you’re going to be able to provide that type of service to their customer.
Given what I know about most companies, it seems like most people are going to have to build a lot more analytics muscle than they have right now. Do you agree?
We are still at the early points of all things AI. One percent of the executives from our study said they know exactly what their AI investment’s going to do for them. So I think we’re still on the ten-yard line of what the analytics capability needs to be at our clients. And once they begin to get that, you’re going to see these leaps forward. Right now we’re stuck in this digital morass, and we’re not going to get there until we see bigger investments and a bigger acknowledgment of what AI data analytics can do to help you leap forward.
Sometimes when I talk to people about this enterprise platform idea, I get pushback. Some people say, “Dan, are you saying that I can just go out and buy off the shelf stuff and somehow get a differentiated advantage from that? If I can buy that stuff, my competitors can buy it too. Where’s the differentiation?” It’s a good point because as data platforms become more powerful, they become available to everybody, especially if you’re talking about a product-based strategy. But I think that there are a variety of aspects that give you differentiation if you pursue an enterprise platform strategy. The first is the ability to master your data supply chain. All these products are going to work better if they have more and better data, so creating an agile data supply chain is really important. The second part is having an automation architecture that’s powered appropriately by AI and ML. And so you can buy all the same AI and ML tools, but the important part is actually using them where they have a good fit and provide a lot of advantage so your people are more powerful and your processes run better. That is not a trivial thing and I don’t think it’s very easy to work out no matter who you have on staff. And then the third thing is user interface/user experience design. I think that that’s a far underrepresented skill in a lot of companies that are trying to do this. If you get all of those things right, I think you have tremendous differentiation. What have you seen in the various implementations that either proves or disproves that theory?
You frame the issue well because I think all the cloud technologies out there are utility platforms. I can almost flip that idea on you a little bit which is I think the only way you can really differentiate is nailing that experiential layer on top. The only way you can actually fuel that experiential layer is with the data and analytics that allow you to make better choices. Right now we’re working with a big hospitality client. We’ve been talking about actually using IoT—sensor data about what your needs and desires at a resort. Most hotels don’t even know where you’re at in the hotel, let alone what you need. What if I actually had facial recognition and sensor technologies across the entire hotel? All of the data is now being adjusted by my data orchestration layer. Because of that facial recognition technology and that data I’m able to know that Dan is actually at the pool and he’s been there for probably two hours. He hasn’t had a drink in his hand the whole time. I can now predict and get the wait staff to send you your favorite pina colada at the poolside. That’s an experiential layer that no hotel can do today but if you had the IoT structure in place, I could tailor an experience for you.
What you talked about is a large set of platforms interacting: the IoT platform, the analytics platform, the operational platform – all interacting in a seamless way. How do you get that started so that you can prove it’s worth making the investment for all of the hundreds of places you might have?
We’ve been talking about that exact scenario. There’s never been an ROI for it because the technology was too expensive. We’ve been talking for five or six years about this use case in hospitality. Meanwhile, no one’s done it because the cost was always too prohibitive. But today the facial recognition technology has become so much more affordable that you could deploy this technology in a hotel and not have the 5–10x cost that you had to spend four or five years ago. The cost of experimentation is going down.
When you think about how you’re going to go forward in a world like this, what’s interesting to me is understanding the fit of a variety of these technologies. Whenever we see an AI product or an ML product advertised or marketed, the claims are exaggerated and when you actually start using it in practice, you find that it’s very good at a certain narrow range of activities if you have the amount of data that you need. But then it’s insensitive and actually quite wrong once you get out of its sweet spot. So how have you advised clients to understand the sweet spot of some of these technologies?
We’re on a tremendous hype cycle for what AI will do. We have yet to hit the trough that comes right after it where executives begin to pull the plug on different investments because they’re not seeing the return. After that, we’re going to see the rise of the application in real business use. And we advise on what’s a moderate investment to provide value to the customer. Because out of 40 possible use cases, 15 to 20 of them will crash and burn.
What you’re saying, essentially, is that in order to make digital transformations work you have to adopt an agile methodology where failure is tolerated in a way that culturally it’s not right now.
Yes. I know we talk about the failure topic often. I think the failure topic is branded wrong because it is sometimes spoken of as these big massive colossal failures. What I’m talking about is small experiments. We want to cultivate a digital culture of innovation. How are you going to do that?
I was at a meeting of one of the insurance clients and we decided we’re going to bring one of our second-year associates who’s on the audit job into this meeting to talk to the global CFOs of this insurance client. And why did we bring this individual in? Because she knew at PwC we want to become a digital firm and there’s all this hoopla around it. So she decided—and if you have a culture of innovation this is what your employees are going to do–she decided to learn how to use Alteryx over the weekend. Monday morning, she came in and said, I think the work that takes you 50-60 hours can be done much faster. She created an Alteryx workflow, experimented with it, tweaked it and in two weeks she nailed it where can do that same task in four hours that used to take me 50. That’s the level that we want to be able to create with our clients. How do you create a new culture and experimentation model where everybody is playing with technology and finding great things that are proven on a small scale, which then you can let blossom and go big?
So it seems like in the hospitality industry but in a lot of other industries, it’s the front line of the workers who have to actually adopt all these new ways of working and this new technology. How do you go about advising people how to make that happen?
I think it depends on the type of employees you’re talking about. Certain employees need to be amazing with their customer experience, and if you could free them up from the day-to-day grind they’re working on, they would actually be better in front of the client. If you empower employees with tools that are designed exceptionally well, I think they’ll be happier. Happier employees are usually better with the end customer.
Many companies underestimate the difficulty of change. But if you look at what happens at a lot of these data lake projects, it’s obvious that we have to have data lakes. There is no choice to go back and use a SQL-based data warehouse to store all our data. We know we have to get to the data lake, we don’t know how many iterations it’s going to take, but we’re just going to keep going until we get there. This is where I see evidence of an impedance mismatch between the top line desire for digital transformation and then the actual tactical operational tolerance for the kind of experimentation and learning from failure that it takes.
To use your example, organizations with a hierarchical culture, based on what we see in our research, are not taking this digital agenda seriously enough. Because if there was truly a CEO sponsoring a digital transformation agenda and he or she knew there were six data lake projects going on, I would assume the CEO would step in and say, “Enough of this. We’re not taking this digital topic seriously enough because clearly, as CEO, I’ve outsourced this to a bunch of other people.” As soon as you outsource something as critical as the data that runs your company to six other organizations and let them do their own thing, and there are people now buried under the lake, it’s pretty clear that the accountability and the governance were nonexistent. So I think that’s what we want to be able to change today. If organizations are taking digital seriously, the CEO is putting their fingerprints all over it versus I think the years past, they gave it to the CMO. Let them play with their little data lake of customer data. If CEOs are paying lip service to it, I think it’s time to take notice and do something different.
You’re arguing that we do have a chief digital officer at all these companies—called the CEO?
It has to be. I think the concept of the CDO at some point needs to fade. Because I don’t think a CEO can at any point outsource the most critical movement in their company. Leveraging AI, automation, all this new technology innovation—I don’t know how you outsource that to somebody else.
Why do you think the CDO no longer makes sense? When did it make sense?
I think it made sense when the digital topic was relegated to IT. It was relegated to the consumer experience. The digital topic is an enterprise-wide topic. Who runs the enterprise? That’s why I say that as soon as you think that your digital definition is an enterprise-wide topic, you don’t need a CDO. CEOs may need leverage from people who will help get the job done because they have a multitude of responsibilities on their plate, of course. But they need to be the ultimate signing authority on what that digital investment’s about.