Podcast Episode 6
Predictive Analytics as the Ultimate Growth Hack￼
with Pete Gibson
To be successful, you must have a plan. To increase your chances of success, you must know how to use the information you have. In this podcast episode, a proven CIO/CTO for leading brands, Pete Gibson, expounds on how businesses can use data to create forecasts that can fuel the company’s success.
Hello, you’re listening to our podcast, Automating the Enterprise. I’m your host, Dayle Hall. This podcast is designed to give organizations insights and best practices on integrating, automating, and transforming the enterprise.
Today, our guest is an executive technology leader with a solid record of successfully leveraging technology, improving operations and delivering high-performing teams globally. He’s a proven CIO and CTO for leading brands, including Friendly’s, Johnny Rockets, Alamo, National Car Rental and Wyndham Hotel group. We’re incredibly fortunate to have Pete Gibson with us on our podcast today.
Pete, welcome to the show.
Nice to be here.
It’s great to have you.
Well, look, let’s start this. You’ve done multiple technology leadership roles. You’ve been a CIO. You’ve been a CTO. Talk to me a little bit about as you’ve gone into those types of roles into enterprises, large companies, small companies, whatever, what’s your mindset? What do you come in to try and do? What have you been hired to do? What are the things that you really think about entering those kinds of roles?
It’s interesting. Everyone is different. The bit of advice I got to give you is when you go into a new role, leave all your old stuff behind you. Take a toolbox with you, but you leave all the old stuff behind you. I think the thing that you focus on the most is how do you get your IT resources to be more efficient. Because if I get them to be more efficient, then we can do more for the company. You do more for the company, you transform out of whatever the situation was into more of a strategic partner if you focus on the right things. So I normally end up focusing on how do we do business. How can we improve doing business? If we do that, we become better.
Stated another way is it’s not about Pete Gibson working harder, it’s about how can I come up with processes, procedures and doing business to where I can get all of our associates to work 2% or 3% harder. And that becomes a force multiplier for the entire organization.
Yes, because if everyone’s being more productive because of the work that you do, inherently, that will help the business at least bottom line as well as top line. When you just said just there about being efficient, as a CIO or CTO, what does that mean? Is it within the tools? Is it within the people? Specifics around what efficiency looks like for you?
Efficiency really starts looking like teamwork and efficiency. I’m going to say this, everything we do in IT is measurable. I don’t care if it’s software development, networking, whatever. It’s all measurable. You just got to have your organizations set up to go do the measurements so you can figure out how you are performing.
But I often get into situations to where everyone is off doing their own thing within their own team or doing their own thing. So networking guys are doing this and developments are doing this. And then when they get done with it, they pitch it over the fence to the infrastructure guys to implement. The next thing you know, the help desk finds out about it because it just launched and it went live, and customers are calling them about an application they don’t have. And so the tech guys are probably pretty good at doing the coding and things of that nature. You got to work with them. You got to have processes and procedures down there. As an IT guy, the CIO knows a strategy that he wants to go do.
But missing in between there is the operational level. How do you get all the elements to work together? Things like a system development life cycle, how do you do business and developing a technical capability? Not just focusing on the tech, but you got to focus on the tech, you got to build the business processes associated with it and you got to build the technical talent associated with it too. All that needs to take place at this operational level.
A great one that I champion quite a bit when I go into organizations is ITIL. It is great to work across the operational levels. Everyone in the organization working together. This is instead of everyone doing their own little change, here’s our change management process. Everyone is trained on it. They understand it, but it’s all measurable. And at first, they don’t like doing it because it’s free wheeling. But then after a while, they start saying, “Oh, we’re getting really good at doing change. We’re doing a lot of changes.” And guess what, our changes aren’t failing anymore. We’re not having to back out as much because we’re planning. So we’ve become a lot more efficient.
Everyone knows I’m a big CMMI-type guy. And out of that is can you get your development life cycles to be so scoped out, predicted and reliable with software quality assurance, reliable that you’re always making your targets on time and on budget. That means you have very limited rework to do on the other side of it, which means your customers have a lot of faith in you. Not drawn when the days are, well, IT told me it would be here in June. So I’m just going to plan it for August.
Just talked about this full life cycle, using models like ITIL, the tech, the business processes and the talent. Those are all the three things you have to think about. We hear this a lot, digital transformation or business transformation projects fail, and you talked about having this process. Why do you think we’re still seeing so many digital transformation initiatives fail? What are we missing? Or what are people missing either within IT or from a business level?
One is I don’t think they’re well defined. When I talk to the average CIO and I say, so what is your digital transformation strategy? At the end of the day, you don’t have enough capital to do it all, so you got to have a strategy to do it.
And then as the conversation goes on and I start peeling the onion back and peeling the onion back, I find out that they don’t really have a strategy. I will tell you that I’ve been successful and my teams have been successful by always trying to focus on the revenue stream. You focus on the revenue stream and you improve the revenue stream, IT moves from a commodity play into a strategic play. So now how do you go do it? Why does it fail? We’re back to what we were talking about before. What is your development life cycle? What is your infrastructure policies and procedures? Why are you failing? Why haven’t you solved those basic fundamental items of working together?
Because networking’s going to do networking and development’s going to go do development. So I go in, I’m a big fan of to get the chaos out of organizations. Well, we’re a matrix organization. I’m going to go to a cross-functional project team, including the business customers. They track and they work together all the way through the doggone project, following a process. And it could be a good architecture, a good design and then go into agile to do the various sprints or it could be a waterfall approach. Anyway, but just well-defined as to what you want to go do.
And then at the end of it, you say, is the code ready? Yeah. Is the infrastructure ready for this? Yeah, yeah. Network ready? Business partner, are you ready? Help desk or service desk, are you trained? Are you ready to go? So when we send this thing live tomorrow night, we’re ready to take this thing live as a business service. And it’s much more than just the tech.
I love that. I love thinking about it as being a business service. You mentioned a couple of things there. I’m going to just push on a little bit.
First of all, digital transformation projects or initiatives should be well-defined. I love focusing on the revenue stream, which I think is really important. And we do talk to some customers who say, “Okay, we want to implement Workday. We need an integration tool to help move the data around and see all the data.” What is that initiative for the business? What is it supposed to drive? They’re just like, “Oh, we’re going to implement a technology. It could be a CRM. It could be an ERP.” We start from that.
Yeah, and even on the smaller-type stuff. I’ve seen CIOs try to buy a tool to get themselves out of a bad situation. And so they buy the tool, they implement the tool and it fails because they didn’t think about the business processes that are associated to get that tool to be effective. They didn’t think about the people, i.e. the technical people to do it. Also, the people trained to understand what this is going to go do. And then how are you going to nurture and grow this tool long term? So they just buy a tool, put it in. They listen to the marketing stuff from the folks and it’s going to solve your problems and so forth. And they forget all those other things. And so the tool goes in, and guess what? It didn’t solve the root cause. They haven’t worked all this other stuff, and it failed miserably.
Now I was just talking to a company. They have all types of data issues. Part of it is their invoicing is wrong. And so what are they going to go do? The guy says, well, we need a data lake and we need to do master data management, this, that and the other. And you go, well, that’s interesting and good but probably the calculations back in the system are what is at thought here. And no master data management tool, ETL tool or anything of that nature, is going to go solve if the source data is messed up to begin with.
Yeah. You just mentioned about you’re listening to the marketing hype. And I’m a marketer, so I know all the things that we’re good at. But I think one of the things that within our own organization we try and do is make sure that when people come to us with those kinds of initiatives, we drill into the process of that business outcome. Are they trying to add to their bottom line or top line? And really focus on that because then they’ve got more likely to be successful. Because what usually happens is there’s always someone to blame. Oh, we implemented that and we haven’t solved their issues. Like you just said, are they going to go and they now need an ETL tool? Or are we going to build a data lake? Well, if the process or the technology is broken further down, it actually doesn’t fix it. It just perpetuates this issue. Then you’ve just paid more for technology.
That’s correct. You got to maintain it, but you got to battle obsolescence on that technology. And you’re actually better to consolidate and get rid of the technology.
Yeah, that’s great. Something you said earlier, which I love, which is in IT, everything is measurable, which I think is an excellent point. Now when we first talked in the past, we talked about predictive analytics and the fact that you can use that as a growth hack. I’ve never actually heard that together. So why is predictive analytics a growth hack specifically? What should we be thinking about in terms of data analytics within the modern enterprise?
Well, let’s look at it a lot of different ways. If you can figure out what’s going to happen in the future, you can correct it before it becomes a major issue, and that’s the whole premise of it. Of course, I get my predictive analytics and heavy statistical stuff back from my government days. But yet when you’re in car rental, you are in hotels, one of the major revenue practices is revenue management and yield management. Of which, you look at setting prices for your different product sets and the changes all the time. But one of the big things that you’re forecasting is demand. So if you can figure out demand, then you know how to set your price. You set your price, you can generate revenue. That’s pricing and so forth.
But let’s try it another way. Let’s look at it and say, you’re a franchisor and you have franchisees. A lot of franchisees leave the system on the order of 6 to 8% a year. If I can figure out which ones are likely going to leave you, they won’t re-up with the brand or they’re just going to flat out give you the keys and walk away. If I can figure that out and give you the list of those and then have the field team try to salvage them, you’re going to generate a lot of revenue.
So case in point. Now you lose 6 to 8% of your franchisees every year. Out of that, let’s just say, I can identify 80% of those that are going to go do it. And then your field team saves 20%. I will tell you, that would generate a tremendous amount of new revenue for you. And if you’re a public company, that can probably impact your stock price.
Absolutely. This is the power of it. If you see that happen, you have the model, you can have your field team at least try and address it.
Try to address it, depending on what you want to go do with that brand. If you look at what a lot of the forecasting professionals do is, yes, they forecast what it is, but then they record the forecast. You’ll see that without the heavy statistical stuff, you’re just looking in the pack and making an estimation. You’re probably right 30% of the time. Where if you use predictive analytics, you can get it up to be 70%, 80%, 90% of the time.
What I like about how you describe that model with the franchisees is, in general, and marketers have been targeted to use predictive analytics in a different way for driving demand, but it’s always the same types of models. But what I’ve seen happen is they put these models in and they have all this data, but they can’t actually turn it into something tangible to go and do. So it ends up being analysis paralysis. We’ve got data everywhere. It’s interesting. We don’t put it into action.
I think the examples that you’ve just given, and I think you kind of did something similar at Friendly’s recently, right before Covid, you’re actually really using that analytics piece and improving the business. You’re putting it into action.
Could you talk to us a little bit about the food delivery system again? Whilst we could see that there was something going on, I don’t think anyone really understood the scale of the pandemic.
Well, it was actually fortuitous. So go to Friendly’s. They had not invested in technology in years. As I was getting there, there were some little conversations about delivery. This is 2018. This is before delivery got hot in the restaurant industry. I said the inspiration isn’t in the restaurant industry. The inspiration is in hotels, car rental, airlines and so forth. Well, this is the way they operate with third-party administrators all the time. And so we want to go do this. We’re going to teach you how to move out of servicing them and moving people into your direct channel, where here’s a new concept for you, where we will yield more.
But then delivery opened up and I’m actually educating them and saying, hey, what’s the cost of a hamburger? They’re going to get this much revenue and so forth. So this is what our profit margin is going to be. Oh, by the way, let’s take that up 10%, 15%, okay? And then as we were doing a couple of delivery services, we had all the analytics in place. And then I would say, hey look, some of these products went up 15%, stop selling. Then other products continued to have strong demand. So that means we have price elasticity in the product set.
So now it’s stuff like revenue management and yield management. And then a light bulb starts going on. So we start yielding on the product set versus just pricing the product set. And then next thing you know, it starts becoming effective.
And then the other thing was you teach him about how every restaurant has its own demographic. Every delivery service to that restaurant has its own demographic. So DoorDash will be an older crowd, but the young college crowd like their local college favorite delivery service. So they all have their own demographic. They all have their own product sets and so forth. You’ve got to understand that to start moving products and setting prices in there. We were teaching them about channel management. We were working pricing up and down.
The other major change in that was e-commerce. We need to have a strong SEO. We need to have an online ordering capability so that you can beat all the delivery services. So how do you go do that? Your delivery services are important to you because they have much higher marketing dollars than you have. So what do you try to go do with them is we would try to leverage them to the max capability. So when you leverage them, what is our placement? And al these things say, you’re in the world of e-commerce now.
Then another one was, where is the placement? What sells the most? So the CEO, he was a big guy on catering and catering does generate a lot of revenue, but it doesn’t sell in delivery. What sells is dairy. So why give up your prime location on the DoorDash site for catering when you could put dairy there for us to make it easy for the customer to find the product set? And then, what is the organic algorithm or the algorithm for organic placement on DoorDash? They’re not going to tell you specifically, but they’re going to tell you all the different variables that are in there. And then we can start working on those variables.
And the next thing you know is you start doing really well. Now for us at Friendly’s, and here’s the story, we got many delivery services in, just like hotels have. And by the end of 2019, before the pandemic hit, IT, all of their systems doing delivery, was generating 20%, 25%. This was before delivery was hot. That allowed Friendly’s for the first time in years to go profitable because we were private equity-owned. The other side of it is this organization had never gotten a bonus. They got a bonus that year. So they’re all saying, hey, this is pretty good.
And then the pandemic hit. And in New England, the restaurants shut down, so all the dining rooms closed. Everybody went home, and we all lived through it. Everyone went home, but we had delivery in place. Now we went down 65%, but we did not have to close. And this was a brand that wasn’t making money. This brand could have closed. A lot of other brands did close. We were viable because we still had a revenue stream.
And by the summer, we were only down 35%, no dining room. Then later, we were only down 10% in October of that year over year. And that’s just the value of what a good tech organization, IT can go do for a business. Because we were only down 10%, that allowed Sun Capital to sell the two brands.
Yeah. And we’re going to come onto that in a second. But I think what’s really interesting there is, first of all, you were not just talking about predictive analytics to help the business. You talked about IT is helping that bottom line and really helping the business. Then you are even talking about customer journey and customer experience. So there’s a lot of things there.
You used some of these data sets, predictive analytics across the business. You start this new service. The pandemic hits and it sounds like you were able to be less impacted than potentially some of the other businesses.
Outside of some of the specific challenges around the business and growth and getting people to look at the data set, what were some of the other IT challenges you specifically had as you were trying to enable this business to continue to grow? Did you find that things that you thought were automated in the process before when people weren’t around, they were all at home, just didn’t work anymore? Were there specific examples, other challenges within IT?
Yeah. Our biggest one was obsolescence. If you wanted the truth about it, it was just really old equipment that hadn’t been maintained over the years. That was the major one. The average age of the IT team was like 55. They had tremendous, tremendous corporate knowledge. They had a lot of corporate knowledge, but no one had ever guided them to go do the right things. Used to be that the IT team would go into meetings and the business would need something and they would go, “Well, we don’t have any resources. We don’t have any money. We can’t go help you.” I was there for a couple of weeks. At that time, you don’t understand the politics and the stuff going on in the organization. They came to me and said, “Well, the business wants this.” I go, ”Okay. Well, let’s put together three ideas for them.”
Tech solutions come in three sizes. They come in the Cadillac, they come in the mid-grade and they come in the Yugo version. And so let’s put together some ideas as to what we can go do.
I got a call from the chief operating officer and he says, “They went into the meeting and, boy, they are all really helpful and great,” and so forth. The first thing I’d do is I’d pass that back off to the team. Because, hey, it’s just hours and they’re handed out kudos. I understand that the answer was always no out of IT during that time frame. And next thing you know, the IT attitude started changing from, “I don’t have resources, but this is what we can go do.” So then they start solving problems.
And the obsolescence thing, there was no money. There was no capital. But when we went profitable, the CFO got a little bit of money. He threw IT a bone. And that allowed us to buy new servers, and we got rid of all the obsolete servers. They were on the way to modernizing the infrastructure.
That’s good because you’d help the business. I like that.
For organizations, like the ones you’ve worked at, do you find that IT is clearly an integral part of some of the things you said around franchisees and so on and helping them? But for the other functions within the groups, HR, broader marketing teams or sales, are you getting requests from them to get more access to data, to run their own technologies? Is it becoming something that you’ve seen around these organizations that it’s not just around what we want to do for the business, the franchisees and IT, but we’re getting multiple inbound requests from the other lines of business?
Yeah. You get those requests all the time, and you need to service them and you don’t necessarily have the resources to go service them. But then we’ll get it into the change management process and start working our way through it and see what we can go do for those guys. That’s always a challenge and you always got to go solve their needs to the best that you can. At the end of the day, they have needs too.
When you get IT running well and doing a good job, in this day and age, everyone knows that technology is imperative. So if the IT team is doing well and there’s a new initiative in the company, probably the first or second call is to IT. It’s just not the tech side of it. It’s the services side of it also, the second M&A.
Even at Friendly’s here, it was ghost kitchens, right? Who’d they call? They call IT. Because we understand supply chain. We understand technology. We understand a lot of the business. But more importantly, we know how to get people to work together to go achieve a common goal, i.e we’re going to go implement ghost kitchens and not just leave it to the ops guys. They’ll go do the ops. But we get ops, accounting, everyone in the supply chain all working together on the initiative through a project methodology, project plan, so forth, that we’re going to go implement this thing. And if you do it right, you’re successful. And of course, it does include tech. Modify the systems to go support this new initiative.
What I appreciate about what you said is we hear this a lot around lines of business and IT and shadow IT and all that kind of thing. What you just said was if your IT team is on it, adding value to the business, whether it’s through things like this predictive analytics, there is no need for there to be a conflict between lines of business and IT. If IT are on it, there is a process they’d follow if they work with the lines of business. I think the outcome can be significantly better and you don’t have to have this conversation around lines of business putting out their own technologies or implementing stuff. Eventually IT will end up owning. That’s generally what happened.
You’ve used this phrase “peeling back an onion. Using analytics is like peeling back the onion.” We talk about digital transformation initiatives, how you need to plan, measuring everything. Tell me your principles around why peeling back the onion using things like analytics, why is that important? What’s your philosophy on that?
Yeah, good question. Let’s use customer intimacy as a real prime example, right? Average static set of reports come out. This is what happened in the business yesterday or over a period of time. And they’ll look at it at an aggregate level. When we started doing the delivery stuff, the business was happy. Oh great, we’re in food delivery. I started looking at it and I go, “no. No, you’re not.” Then they go, what do you mean we’re not? I said, you’re in dairy delivery. You’re in the dessert delivery business. Because that’s what your analytics are telling. All interesting.
And in the restaurant industry, they do day parts. They measure performance on basically four-hour periods, the breakfast period, the lunch period, the afternoon and maybe the evening period, whatever the day parts are. I said, that’s not good anymore. So we got to get it down to the hour and a half hour. I just want to understand that. And so what we found out was by going through the hour and getting the higher fidelity in the analytics was that, yeah, we do sell food, but our business really starts taking off at 6:30 in the evening. We started doing a lot of deliveries at that time. Oh, by the way, we do some food, but we do more ice cream, and it gets stronger from the 6 to 7, 7 to 8. 8 to 9, it gets stronger. The 9 to 9:30 half hour is just as strong as the 8 to 9 complete hour.
So why don’t we shut off delivery at 9:30 and close our restaurants at 10. If you would have done the day parts, you never would have picked that up. So the more I can peel that onion back to understand, how does Dayle buy? What does he buy? He comes back this many times. Then we can do things that are old, but might be new to some industries. Let’s just do database marketing. I’ve got such a tight set of data and databases coming in there that I can do better marketing. And oh, by the way, we were talking about revenue management. I’m going to move you out of those delivery services and move you into my mobile application. And then on that mobile application, I can do a push notification. Then I find out Dayle likes a certain ice cream, whatever, and that’s what he always seems to buy due to an association analysis. And we got it on sale or something like that. I can push it to you versus the other customer. And I had more meaning to you.
You’ve already shown brand loyalty by giving up real estate on your phone to put our app on there. Now let’s give you stuff that’s going to bring you back and invite you back into the business. A lot of these concepts are really solid in other places. But in some industries, they’re just new concepts they never heard of. But they’re just happy to have a name on a list. They’re going to send everyone on that list the monthly promotions. What I call, they basically spam you.
Yeah. But that’s a great example of looking at some of these analytics and actually moving it forward into personalized experiences, into oculars that are going to appeal to me. Personally, I can’t eat ice cream at nine o’clock at night. But in general, clearly, if that’s what the analytics are saying, you can help the business.
One of my data scientists, a guy by the name of Wei Jen Ye, extremely bright individual back when he was working at Orbitz, basically found out that you’re buying a hotel room and you’re using a Mac, you’re willing to pay more for a hotel room than a guy on an Intel system. Coming in through a router, we know what your source is. Now I’m not saying gouge a customer, but maybe you upsell this guy a little bit more. If you were the guy, because you have more options, this guy is willing to pay more as a set.
And that’s just the power of it. Most people are just saying, well, let’s get into place and offer them something. No, we want to understand this customer deeper, deeper and deeper so that you can get the right product to the right customer at the right time at the right price point. And that’s what we’re trying to go do. If you do that, as you guys know, you get a sale. If you don’t, you probably won’t.
That is the definition of predictive analytics as the growth hack, which is what we’ve been talking about.
Let’s close the podcast on that specifically. If you are a CIO out there or a CTO, because you’ve done this kind of thing, and they’re looking at potentially, okay, we’ve got all this datal. We’ve got some initiatives for digital transformation. Let’s say, we’ve got to the point where we’ve defined outcomes and metrics. But what are the three or four things that you would say, look, these are the things to remember as you start these initiatives, as you start to pull this data together, as you start to serve it up to the business? What are those two or three things every CIO, CTO should remember?
The first one that I always try to do is solve a business problem. I often hear people in the data world saying, I can’t get tools. I can’t do this. I can’t do that because I don’t have the resources and whatever. And what they’re trying to go do is they’re trying to sell, before I can go do anything, I need a master data management tool. And then you go, okay, that’s good. But you’re on a cost basis when you go do that. If you had the master data management tool, what would you go do first? The first thing you try to do is go do something that’s got to really impact the company. So if we don’t have the master data management, what can we go do to start prototyping and show the business using analytics that I can go do this?
So I may not have all the tools together, but why don’t we just go show them price elasticity? Why don’t we go show them that every day we take a look at what happened in the delivery services, what is selling and what’s not selling? What’s going up, what’s going down? We’ve got all that stuff. And as you can see here, we went up to a certain price and this stuff continued to sell. Why don’t we take it up a little bit more and make a little bit more money? And they go, oh yeah, good idea. Then the next thing is, they said, I can do a better job if I have more tools and more resources.
And here are the two or three things that I would challenge that would have a high return on investment for us as a good company. You don’t sell it immediately, but you just put the thought in their mind. It’s a give and take, but you got to plant the seed and you’ve got to always be focusing on the business value. If not, you’re in the commodity play and that’s a tough sell. Business value, I stand a better chance of winning.
I love that. I think that’s a perfect way to close. Just a couple of reminders of the things again that I heard through this that I think is really important for people to remember is in IT, everything is measurable. I really love that. And I think that is a way of then showing that IT, call them digital transformation initiatives, if you want. But if you’re solving business value, you are more likely to have a successful outcome and you use some great examples of how you can actually help the business. And IT can really have an impact on that revenue stream, which, I think if most organizations operated that way, there’d be so much more harmony between lines of businesses and IT. So I really appreciate those examples that you’ve shared with us today, Pete. Any last pearl of wisdom? Because I’ve taken so much from this already.
The one thing that you and I talked about prior to this is what advice do I have for a young CIO getting into this thing is one is never stop learning. I’m always reading. And as you heard me say earlier, I’m a Harvard Business School, read their business books. And every book I read, I get a little nugget out of it. If I get the opportunity, I try to implement that nugget. You’ll find the average person out there doesn’t continue to learn. So they’re just going along with the flow here.
The other thing I always recommend is focus on improving the IT organization. It is huge. Anything you’re going to do to get efficiency in there is really good. And it’s just not the tech, but it’s also the services that you provide are extremely important.
The third thing is just try. Got to find so many people and get behind, we don’t have this, we don’t have that. Well, go in there and go try the free version, play with it, have fun, try it. And then come back to me and what do you think we can go do with this? Even though it’s a pretty popular BI tool or something like that, you know it’s going to work, but you’re getting them to buy-in. But go try it.
That’s brilliant. Well, we’ve come to the end of our podcast. So I’d like to thank our guest, Pete Gibson. Great discussion around predictive analytics as the ultimate growth hack and a whole lot more besides. So thank you to everyone for listening to our podcast today. We will see you in the next episode.