In this Digital CxO Leadership Insights video interview, Francis Carden, chief technology evangelist for HFS Research, explains why digital business transformation initiatives are being held back because organizations are too wedded to legacy IT platforms.
Transcript Text
Mike Vizard: Hello and welcome to the latest edition of the Digital CxO Leadership Insights series. I’m your host, Mike Vizard, and today we’re with Francis Carden, who’s the newly appointed CTO for HFS Research. And we’re talking about, well, all those technologies that are driving everybody crazy as they try to figure out how to move these digital initiatives forward. Francis, welcome to the show.
Francis Carden: Hey, thank you. Nice to see you, Mike. Yeah, the word “digital” and “automation,” we’ve not heard those recently, right?
Mike Vizard: Yeah, exactly. And it seems like we continue to struggle and we make advances here and there, certainly automation has come a long way since you and I probably first started out, but my question to you is every time I turn around, there’s somebody out there who just can’t quite harness all these technologies to execute. So what ails us?
Francis Carden: All the legacy stuff. I think that there’s a lot… I mean, everybody says… And what do I hear all the time? It’s like, you’ve heard the same things, like “It’s too difficult for us to change. We’re just going to keep band-aiding what we’ve got.” It’s been going on forever, and I liken this to… And not surprisingly, I think that all the technology that we’ve had for the last 40 to 50 years, we’ve used really, really well, but we’ve been limited all the time. We’ve been limited by, I remember my first program, I was limited to how big it could be, so I had to make the variable names unmeaningful, and then try to figure out how to document it. Because I was limited to the space I had. I’m showing my age now, but that’s been going on for the last 40 years. You can’t index everything. You can’t store everything. You don’t have the bandwidth, you don’t have the computer power.
But if you think about if all that went away and we started the computerization era again, would it be called computerization? But just think of what you could do today. So I think we squeeze just about as much juice as we can out of all these lemons, and we need to start thinking about what you can do now in this new era. I call it the digital revolution era, just because I can’t think of a better name. Does that make sense?
Mike Vizard: It does. Is there still a disconnect between the folks in IT such as yourself who understand what those limitations are and what the business is trying to do? And have we gotten a certain amount of irrational impediments going on because neither side seems to, even 40 years later, know how to talk to the other side?
Francis Carden: You and me have been writing about this for probably as many years, trying to figure out IT and business alignment, and it must sound like Charlie Brown to many people that hear this. But I genuinely, genuinely, genuinely think we’re really close. I’ve seen so much more alignment between business and IT, and I’ve been lucky enough to sit on both sides of the fence. And now at HFS, I sit in the middle a little bit, right? Because I get to be on the analyst side, neither a vendor and neither an end user. So I think that we are getting there and I think that what’s driving that is a little bit of envy from business, seeing their competitors do things that they want to be able to do, go to IT and IT says, “Well, we can’t do that. We don’t have the budget for that, because we’re spending 90% of our budget to keep the lights on.”
So that whole… I just dream, maybe you call me a dreamer, but is this now, with so much envy, with what we can do with so much compute power… I mean, you see the stuff being announced with the GPUs and all that, just the monstrous compute power and growing compute power, things you could do if you unthink what we had to learn to get to where we’re at. I think IT can deliver on what business need, and businesses can be empathetic to what IT has had to deal with for the last 40 years. Maybe. I hope so.
Mike Vizard: And we make those old platforms do the proverbial dance like an elephant, because I hear the same thing you hear. I hear everybody say, “Well, we have legacy platforms. We may not have an API. We’re going to have to write the API,” or, “We need to modernize the front end of it, but the back end we can’t touch.” And there’s just a lot of frustration. And at what point do you just decide to blow them up?
Francis Carden: I go back to this envy point. I’ve been using this in my talk tracks in front of anybody that wants to listen for the last three or four years, surely as an enterprise, as a business leader or even IT, but start with the business. There is an organization you envy. Is that a digital native startup? Is there something that one of your competitors is doing because they took that plunge, they did something? So that envy can often be a driver. It shouldn’t be. But if you are on the innovative side and you can afford it, you could be the one that’s envied, but now turn you being envied, or turn you having the envy to the one being the envied. And I think that the technology’s possible, but we’ve got to start throwing the baby out with the bath water to some degree.
We can do integration, we can do things. There’s so much stuff we can do. But sometimes I look at some of these processes, legacy processes, and why they’re convoluted is because in the computerization era we automated… Good word, still. We automated what humans did prior to computers. So we see a lot of that same analogy in the computerized systems. But if you used to reimagine a process today, I mean they reimagine taxis not by creating an application that allowed you to book a taxi, but by reimagining the whole ecosystem of how drivers communicate, how you get paid, how you pay, you could even shed drives, right? Imagine jumping in a cab with somebody late at night and saying, “Hey, do you want to split the fare?” Imagine if we tried to computerize that. Well, we have, and I think that that’s true of so many processes that are stuck because we keep thinking that’s how they have to be.
But if we re-imagine some of these processes, some of them without paper, some of them without all the interconnections, and the velocity at which we can build and play with these things now, isn’t anything like it used to be. Because listen, you remember this, we used to do proof of concepts, that it would build something. And you know what? Then it would be like, “Right, we’re going to live next week.” It’s like, “Mo, no, no, we didn’t build it.” They go, “It’s too late. It’s live.” But now we can dabble, we can play, we can build, and if what we build quickly, we can move into production but with all the risks removed, with it being solid, with it being robust, with it being agile, which I know we can, then maybe we’ll start to get people to re-imagine and be the ones that are ultimately envied.
Mike Vizard: Are we too wedded to our existing processes, and should we accept that they’re fundamentally flawed? Sometimes I talk to people and they tell me they have this great process and then you get into it with them, and there’s more exceptions than there are rules, and there isn’t a process.
Francis Carden: The old exception conundrum. The only way, and I’ve been dreaming about this day, probably every five years for the last 40 years, that we would get to this point, and we are at an inflection point now, where… You remember this. So I have this slide and I’ll try to replay it in my mind, but basically we used to have to choose a hardware vendor. Then we had to choose an operating system, then we had to choose a database, then we had to choose a computing language. Then we had to choose what UI we want to do, we want to be fat client, thin client, or a thin client in a fat client. We had to choose a mobile device. We even had to choose the browser, and the list goes on. You don’t have to choose most of those things anymore. You could start building an app tomorrow, me and you. We pick something, we want to run a nice vendor and it runs in the cloud and it’s all computer, it’s all secure. It’s like bam. And then we could go live.
That should mean that business should be allowed to go into these playgrounds to build and reimagine their processes. And you know what? It might not be that leap to go live. We might need some integration. We can bring back enterprise to tie this over. But you know what? Sometimes it’s just better to forget the old stuff. How does a digital native get going? How does an Uber start? I mean, they don’t go and say, “Well, I’m going to go and buy the 7,000 legacy systems,” Mike. There was a digital bank, I used that as an example, they started up and then another bank said, “Well, they don’t have the 6,000 applications that we have.” I’m like, yeah, they don’t, but they’re still a bank. I can still get mortgage, a loan, a credit card.
And it was like, “Yeah, I suppose you have a point, but…” Okay, it’s not as simple as that. I know that. But it can be, because you’re going to miss out. Just look at the compute power you have now. It’s not 30,000 bytes to fit in one program. No one cares how big your program is. No one cares how big your data sets are. In fact, the bigger the better because you can have AI then crunch that data and not lose anything, not forget anything.
Mike Vizard: We got to 12 minutes without mentioning AI. But here we are. Will AI force a lot of these conversations? Because it seems like, to me at least, there’s the training of the models, and then there’s the deployment of the models, and the models need to be closer to the point where they are consumed. So are we looking at some new era where many of these issues will finally be addressed because AI models are going to force us to make some decisions we didn’t want to make?
Francis Carden: Yeah, just look at how fast it’s accelerated in… I don’t know when whole GenAI, chatGPT came into your life, but it’s measured in months, maybe a year at most. It’s not that long ago. And the greatest minds in the world are now accelerating on one path to even that’s being blown out of the water with what the possibilities are. And in the business and IT world, we’re all going, “Yep, okay, so what we are going to do with it now? What can we do with it?” So you could argue, you can throw AI at all your legacy stuff. I mean, you can, but should you? Well, yes, you probably will have to, but imagine what you could reimagine if you put all this together, reimagine your processes, embed AI from the very beginning, because it’s just going to be inherited.
I think AI is going to be inherited, everything you do, you won’t have to add AI. And think of this term, we keep saying we’re going to add AI to that. We’re going to use AI to make this process better. It’s just going to be inherited. For every idea that comes up, “Oh, we could use AI on, pick an example, on my process data, determine the best process path.” Yep, you absolutely can. But maybe we should build that into the very things we build, so that the process path can be predicted. It can predict when a process is going to fail. You don’t have to run after faults on it. So I think it’s those two camps. Yeah, there are going to continue to be great minds thinking about how we could use AI on all this legacy stuff. I worry that it’s going to keep legacy around for even longer, the Antichrist of what we’ve been talking about. But I do think it’s going to benefit us for what we reimagine the future in this digital revolution era, because it’s just mind changing, mind-blowing.
Mike Vizard: I agree, but I also worry that we’re just going to perpetuate these islands of automation that have existed, and now we don’t have these things connected with each other because somebody isolated a slice of something without considering the implications of the whole.
Francis Carden: They have to, the risks are too high. And I’m seeing a bit of a clawback. I don’t know if you’re seeing the same thing, the people you talk to. So I’m seeing a bit of reticence going, “Yeah, we never really considered this. Okay, so let’s take a step back and think about it properly.” Now, is that going to stifle innovation? I hope not. It shouldn’t do. But you can’t half-do AI. If there’s any risk, speaking of it, any risk that something that could go into a data store that shouldn’t go in. We’ve had that risk for many, many years. And here it seems a little bit worse because with most of these large language models that we are thinking and talking about might exist somewhere else, we don’t even necessarily know where it’s going to exist. In fact, I saw a presentation from a big vendor the other day where they’re using AI to predict the best AI to use at the right time all the time.
So I’m like, hold on. Yeah, that makes a lot of sense, because we can’t figure it out. So let’s use AI to figure it out. At least advise and recommend. So I’ve got family, friends, and some of them are scared… That’s said the polite way. They’re scared out of their minds about what they’re reading about AI. And I think you can’t say, “Don’t be scared,” right? We all have got concerns. Some are more scared and some are concerns, but the innovation that this is going to create for us all, I think might just pull everybody out of that legacy quagmire and say, “No, we ain’t going to do this on the legacy stuff. We don’t even know what’s in there. God forbid we throw what we don’t know is in there over to somewhere else where we don’t know where it’s going.”
Can you imagine that? So hopefully that’s not going to stifle the innovation. What it will do is it will get people to go, “Let’s reimagine this process out of this computerization here. Let’s imagine what the process should be today. And then we inherit AI at the right time in the right place securely and safely.” I’m seeing that. I’m seeing that with software development, AI and software development. AI can build the software for you. What was the latest one? AI can generate code. I’m sure you’ve got a comment on that.
Mike Vizard: All right, well let me ask you this then. Is this an opportunity to, quote unquote, “Bring out our dead” in a way that we wouldn’t have done before? Because we now have the freedom to think about, we can go back and say, “Hey, it’s a whole brand new ball game. It’s a whole new era and we’re going to make some different decisions,” and maybe we are going to go back and reevaluate some long-held cherished beliefs. And it’s okay now, because we have, AI is the great get-out-of-jail-free card.
Francis Carden: It has to be proven by accelerating this… We were happy in the days of proof of concepts. Now, proof of concepts are not just little tactical tasks, they can be entire business processes. Let’s just build this from scratch. What should it look like? In fact, we used to have this mantra where we would tell people to describe their process, and it’d start off with one wall. It’d end up running out of walls, because your process with the variability would be all around the wall and in the street next door. I mean, it went on. But actually, if you reimagine what you want the process to be today as if you didn’t know any of that other stuff, it would look nothing like that. Because the baggage from all the connections, because we used to have people walk into banks, we used to have people write checks.
We still will, but they should become the exceptions. In fact, I think that’s the game changer here. So we can reimagine these processes, go play in your sandbox, go build it, dream of anything you want to do, go build it, working with IT and business for rapid evolution of these new generations of technologies and go, “You know what? Oh, we can go live now. All right, let’s do it.” Even run that as a side business. Keep your older customers or your customers that don’t want to give up their favorite color checkbook with their logo on it, right? Keep that. But over here you take the new generation into this new era, but don’t let the old generation of legacy systems stifle your innovation from what the next generation is going to expect from you. Otherwise, you’ll get left behind. And I think getting left behind is going to happen faster than it’s ever happened before.
Mike Vizard: So what’s your best advice to your fellow CTOs out there? Because I think they come in two flavors. One is overly aggressive about all things new technology and the other one is holding on for dear life, “Don’t mess with it, it’s working and I don’t know why.”
Francis Carden: Go show the business what they can do. Go hand in hand with them. Have empathy for what they’ve had to deal with. And show… They will have empathy for you because I think there’s empathy both ways, between IT and business. We’ve talked about that for a long time. Have that respect that they’re just frustrated, because your competitors are going to come and bite your backs. Basically. They’re coming up thick and fast. And you don’t even know who your next competitor is going to be necessarily. And that’s accelerated. We’ve been saying this for years, but it really has accelerated. So if IT can just recognize that business has got good ideas amongst the bad ones, let them go play in some sandboxes. Let them go build, reimagine it. They’re not going to do any harm.
And if they come up with something that actually then is sold back up to the board or to go into production, then you can make sure that governance and all those rules are in place to make sure all that happens and then turn it on. Turn it on, and then that means that you can carry on keeping the lights on, doing what you do, but slowly start turning some of these lights off for God’s sake. Some of these lights should have been turned off long ago, and I don’t think CTOs really like managing 6,000 applications just because they have to, right.
Mike Vizard: All right, folks, you heard it here. To make a reference back to Ben Franklin, “If we don’t all hang together, we’ll surely hang separately.” All right, Francis, thanks for being on the show.
Francis Carden: Hey, my pleasure.
Mike Vizard: All right, and thank you all for watching the latest episode of the Digital CxO Leadership series. You can find this episode and others on our website. We invite you to check them all out. Until then, we’ll see you next time.
