Eat Your Pancakes to Fuel AI Big Bet Outcomes, with John Rossman, Managing Partner at Rossman Partners

Transcript

Our transcripts are generated by AI. Please excuse any typos and if you have any specific questions please email info@digitalshelfinstitute.org.

Lauren Livak Gilbert (00:00):

Welcome to Unpacking the Digital Shelf, where industry leaders share insights, strategies, and stories to help brands win in the ever-changing world of commerce.

Peter Crosby (00:22):

Hey everyone. Peter Crosby here from The Digital Shelf Institute. Incremental growth in the future across e-commerce and agent eCommerce will be unlocked by two major AI driven trends, increased scale with increased specificity, higher discovery, driven by specific use cases, higher conversion rates through personalized experiences, and all of that delivered on vastly more endpoints and agentic conversations. To be part of these trends, early means taking some big bets on AI transformation. The question is how John Rossman, author of Big Bet Leadership, your transformation Playbook for winning in the Hyper digital era joins me for a delicious conversation on the stack of pancakes you need to cook up to transform for seizing competitive advantage in the AI era. John, it is so great to have you back on the show. We are so grateful to tap into your brain once more.

John Rossman (01:21):

Is this the fourth or fifth time I've had the opportunity?

Peter Crosby (01:25):

Oh God, now you're testing me.

John Rossman (01:26):

It's one of those too, but

Peter Crosby (01:29):

It's something like that. We keep having your back because you keep making sense.

John Rossman (01:32):

I feel like the fifth Beatle at this time. So

Peter Crosby (01:37):

I think the fifth Beatle is named Claude. That's right. That's right. So I mean, the reason why we reached out again to you is you published a newsletter fairly recently on an AI strategy framework that you call pancakes, which is what prompted this conversation. And before we get into the yummy three layers of this stack of pancakes, please set the context for our listeners. You have a lot of conversations with your clients that are trying to figure all this out. Why is a framework so important and urgent right now based on

John Rossman (02:15):

Your experience? Perfect. Yeah. So the complete setup is yes. I write a newsletter called The Digital Leader Newsletter on substack. And as you know, about a year and a half ago, I released a book called Big Bet Leadership You Transformation Playbook for Winning in the Hyper Digital Era. And what it's targeted to is for executive leaders on how to do their job of actually how do you lead complex, major transformations? Well, there's no bigger topic out there about major transformation than the opportunity of AI within businesses. And so I envisioned kind of a plugin framework into the overall big system to help frame how do we think about the opportunity for AI in our business? And the goal of any good framework, it is a layer of abstraction. It pulls you down from just one big concept. We need to deploy ai, but it also hides all the details of the thousand different things that you could do.

(03:24):

So it's a midpoint framework for how to think about AI in an organization and to segment, prioritize, name and frame what those opportunities would be. So I called it pancakes because A, I try to write something that's halfway interesting and pithy, but also because the concept is you don't want to just have individual pancakes. You want to have a stack. They're separate but cohesive in nature. And creating that cohesiveness of between these separate ways of thinking about AI in your organization, I actually think helps the overall both thinking of what to do and the success that you're going to have doing it. So that's kind of the setup of why I think having a framework on any major situation is really helpful. It helps bring everybody along in your thinking. But in particular, that was the goal I had with this AI strategy framework that I called pancakes.

Peter Crosby (04:31):

Yeah, I mean, when you drop into your clients, do you find there's been some research lately about how little of these projects are achieving ROI and folks like MIT have points of view on what should be done to increase that return on investment. I'm just wondering, are you finding a level of frustration out there that's driving this need for organization or is there something more going on? What's the mindset?

John Rossman (04:58):

Well, there's a lot of variation in that mindset is one thing I would say. There's no simple way of framing what the mindset is out there, but in general, I find a lot of interest and optimistic about the opportunities for the business and in using ai, and everybody still has a little bit of a wow factor, which I think is appropriate. It really is a new superpower and a new way to work. But what I consistently find, and this is kind of what the MIT research points at, is a very fragmented and cursory approach for how to think about using AI in the organization. And so what most companies are doing is they're deploying it for individual use, but even within, they typically are not really well organized in how do you do it well, how do you do it consistently? How do you share best practices among people? How do you put the right management frameworks in place? Because as all the news articles will point out, it's not perfect. Mistakes get made all the time. I notice that in my work, you really do have to have an expert eye when using it, but it's still a superpower situation.

Peter Crosby (06:29):

Accelerator. Yeah,

John Rossman (06:30):

Accelerator. One of the things I notice within organizations that are deploying AI individually is that they aren't sharing best practices and quality reviews like how to do it successfully across the organization. They're just deploying the tool. And what I've noticed on medium to complex type of work tasks is that there's a massive difference between an average prompt versus a well-engineered prompt. And the combination of a framework plus a well-engineered prompt plus the right AI plus an expert to do review, that combination is just a game changer combination in solving more sophisticated problems. If what you're using AI to do is help me write a better email, an average prompt does just fine. But if you were to use a prompt to do something like optimize this delivery schedule, you want to have a engineered prompt and you probably want to have that prompt takes into effect a framework for what does it mean to optimize. And so you want an expert that can review the outputs to validate or tune that output. So that's an example just within what I call the first layer of this pancakes AI strategy, how you need to do it better to actually gain the benefit out of AI in the organization.

Peter Crosby (08:08):

And so you said we've sort of dove into the pancake itself. We're at layer one, and that really is about the individual productivity augmented by ai. And as you said, that's just sort of one of the pieces that you need to get. And the exciting thing about that, and maybe daunting for a lot of people also is that actually, I was talking with a leader in this field who is on eCommerce, is running an eCommerce team, and he said a lot of his job has become engineering prompts and then passing that knowledge onto his team.

John Rossman (08:49):

That's right. And the MIT study, what it points out, it kind of like, Hey, 95% aren't progressing beyond pilot. And it claims there's a big difference between getting individual productivity versus seeing p and l impact. I'm a little dubious of this research, and it's partially because we know that we are seeing hiring impact due to ai that is p and l impact due from ai, but let's just accept the study for what it states. That's another good reason why you need to do it in a coherent coordinated manner is there's a difference between seeing individual productivity versus gaining the benefit of that from a business standpoint. And so you do need to have a cohesive approach even within this entry level layer of AI strategy in order to consistently get great outputs, quality, speed, reliability, consistency across an organization. Those are the things you need to have real lasting effect with any type of tool or approach.

Peter Crosby (10:09):

So layer one,

John Rossman (10:11):

Individual productivity,

Peter Crosby (10:12):

Individual productivity together. So you want the next pancake to add more wonderful flavor to this thing. So what's coming next?

John Rossman (10:20):

Yeah, I like a big stack. And so the next layer is really workflow redesign, leveraging ai. And the secret on this layer of end-to-end functions and processes is you first and foremost have to rethink that process or what that function is. And the way I think about it is you really have to understand what's the transformation, what's the job to be done for this function or for this process? And it's funny how oftentimes people can only talk about the steps that go along the way or the organizations that do it, they can't actually talk about like, well, what's the inputs and the outputs relative to this function or this process? So the first thing is you have to think about it like a box. I don't need to know everything that goes on within the box. I just need to understand the inputs and the outputs.

(11:20):

And if I can do that of actually understanding that, then I can zero based design or re-engineer the best way to get that work done. What we oftentimes find is that there's a dramatic inconsistency in understanding what is actually the inputs and the outputs that we want from every function and from every process. We have an inconsistent approach as to the rules or the policies relative to that. So stripping away all that calcified approaches and inconsistency is the first thing you need to re-engineer the work. Then you actually have to think about, now that I understand the inputs and the outputs, what's the best way to deploy AI to optimize for those and then couple humans around that in order to support the ai. What you want to be careful about not doing is just taking the same policies, the same data, the same approach, and just applying AI at the small little steps that we do as humans to do the work. So in the article I talk about the better truth. And so the better truth relative to this is that AI is the worst it's ever going to be right now, right? The thing we can count on, the thing we can count on

Peter Crosby (12:46):

Unless it gets evil, in which case then we have a different conversation.

John Rossman (12:49):

Some people would say that's better. Peter Evil is relative, right? Eye

Peter Crosby (12:54):

Of the holder

John Rossman (12:57):

Is that this is the worst AI is ever going to be. You want to be able to take advantage of the improvements that we are going to see. And the way to do that is to really think about the work from an inputs and output standpoint and then deploy AI and then couple in humans to help do the things that we can only do so well. And that is the end to end workflow re-engineering that we need to do. So for example, if you were to think about an SNOP sales and operations planning process that a lot of manufacturers and CPG organizations really struggle with in doing well, if all you do if try to mimic all the policies and steps that you do in your current SNOP process and apply AI to those, you're just, as they used to say, paving the cow path, what you want to do is completely rethink what are we trying to optimize for in this SNOP function? Think about it in that way. Apply AI and then put the right steps in place in order to manage and optimize that process

Peter Crosby (14:11):

Well, because so much of the opportunity for incremental growth for the folks who are listening to us right now across manufacturing and retail is going to come from the digital channels. And there's the digital channels, which is AI optimized e-commerce, so driving towards better outcomes, better conversions, better discovery in e-commerce across more and more of your product catalog. And then there's the opportunities that are going to arise in ag agentic conversations and agent to agent buying. And both of those things need to be scaled up to a level that's not possible without ai. So there's an urgency for getting the growth you deserve and nobody else does deserve in your mind to get that going through ai. And what I love about what you said is that when you think about that use case, the idea that you can have first have AI analyze your existing process for first of all, just explain this process to me based on what you're seeing. And then once you understand and once, like you said, the inputs and the outcomes, then you can have it help say, where can AI now make a difference to which helped me to achieve better scale, better quality, better outcomes, whatever that might be. Does that make sense in this context?

John Rossman (15:38):

Absolutely. And you're going to be able to better take advantage of the advancements in AI that we're seeing almost every week, right? And so you're setting yourself up not just for improvement today, but improvement for tomorrow.

Peter Crosby (15:55):

And I don't want to beat the MIT dead horse, but really what sort of the headline grabbing statistic aside, what I really liked about it was the way that it talked about in order to get ROI more effectively out of these things embedded in the processes that people are already doing, make that where you put, because people learn and get enthusiasm for something that they recognize and yet get the delight of, holy crap, I can do this now. Does that make

John Rossman (16:30):

Sense to you? Absolutely. And part of my response to that MIT strategy or report was no, duh, every type of transformation has a failure rate that is fairly similar relative to that digital transformations, 70% failure rate, 80% have little to no value at all, m and a growth strategies, major technology platforms, all of these big bets, right? High ambition, but high risk strategies that we take all suffer from these similar patterns. And that's why it's not about the technology, it's about how we actually lead organizations in the pursuit of making these types of changes. And what senior leadership loves to do is they love to say, well, if only my people would do something. It's like, no, I believe it's the exact opposite. I believe is a senior leadership failure in most of these circumstances, to actually think clearly about what problem are we trying to solve and what is our outcome?

(17:36):

They fail to create an environment in which you can do what these things need to do, which is they need to be set up as experiments and allowed to do what experiments do, which is they always start from a point of failure. And then maybe you get to iterate your way up to success and you actually need to put the governance in place in order to manage this differently than you do your well understood projects and plans. That is fundamentally the difference between experimentation versus execution of a well understood situation. That's why we wrote Big BET Leadership is for that type of high ambition, high risk change. And I think it is primarily never about the technology. It is always about senior leadership and how they set the conditions necessary for success.

Peter Crosby (18:25):

And having lived through the rise of e-commerce, so much of the transformation was in essence forced by COVID across the e-commerce. Well, the digital shelf was the only shelf and people couldn't go in person to get things. And so all of a sudden these organizations had to move very quickly to innovate in order to keep their revenue coming in at a time when it could have easily gone down to zero if they hadn't figured that out. And I don't want this version, this thing to be driven by something as horrible as that, but I know that our listeners certainly have been at the forefront of driving that change in a very pressurized environment. And that's what excites me about their role and what's coming up. So if we, sorry, let me stop there and see, does that resonate in terms of you kind of have done these things before, it's okay, it just needs a different approach or a different framework?

John Rossman (19:31):

Well, the pandemic was a tragedy, but there were some silver linings that came out of it. One of those silver linings was it showed us when we have to change, we actually can change a lot quicker than we normally give ourselves credit for doing and everything. The e-commerce shift, I helped lead the launch of the marketplace business at Amazon in 2002. Everybody was poo-pooing US Target and all those organizations were just like, ah, this is nothing and everything, but you put 15 to 20% growth on a small category for 20 years. All of a sudden it's something significant. And I think that's the way I would think about the next layer in this AI strategy, which is the disruptive layer is you want to play the game early so that you can learn, establish brand, train your customers, train everybody along the way. You don't want to wait until these concepts are normal because then catching up is extremely expensive and hard to do.

(20:43):

So the third layer on the pancakes AI strategy is the secret, I call it secrets, right? Every type of business tends to have what I refer to as a secret, which is if we could solve for this problem, you fundamentally change the complete dynamics. It's an unlock relative to the business. I work with a steel manufacturing organization. Their secret is about being able to hedge and forecast demand relative to commodity prices, because if they can hedge commodity prices better today, that's really where their profit comes from. I work with a construction management design construction management organization. The big unlock in that business is actually having access to skilled labor. How do you train, retain and hire the right skilled labor? And if we can use AI to change their ability to do that better, you completely change the dynamics of their business. Now in CPG for manufacturers, maybe that secret is ag agentic ai, right?

(21:58):

I would fully encourage companies to be doing thought experiments and tracking very closely what is going on in that realm. Even if you don't do anything today, you should be studying this extremely, extremely closely and figuring out how do we participate? How do we create the basis for being able to do this extremely well and get in early? But there's probably other major unlocks in the business that are wicked problems. These are really hard problems. And what you want to do on really hard problems is you want to have a theory of change. This is how we think something like this might play out. And then B, is you want to play it early and small, right? You want to think big, but bet small. So you want to start experimenting along the way so that you can actually have an experimental path relative to solving a really hard problem in the business.

(22:57):

But when everything in the business has to be an 18 month or shorter payback, and all we do is one or two year planning cycles, and we suffer from short-termism of optimizing quarterly profits, that's really hard to do. That's fundamentally why I still think Amazon is one of the most interesting companies out there is because they understand how to both pursue short-term results, but also to lay some bets along the way that aren't going to pay back quickly that are the seeds for what the future is, and they are shaping that future. That's what more companies need to do, is figure out how to take calculated risks in very affordable manners, very manageable manners in order to create and participate in what we know is going to be the future trends within industries.

Peter Crosby (23:53):

So John, the layer three being about finding the secrets or the secret, what is that process? And that's probably another podcast, but I'm just thinking, I'm putting myself in the listener's shoes, which is, okay, that sounds great. How do you make that pancake? How can I help my organization make that pancake? How do you think about that?

John Rossman (24:24):

Well, my typical response to something like that when I'm doing a keynote or a workshop is kind of like, well, that's where the keynote ends and the consulting assignment begins, right? Because it's very particular to the organization relative to where that is. But in general, from an industry standpoint, manufacturers, it really is understanding if we could solve either this major risk or constraint in the business or some other dynamic in the business, we completely unleash a competitive advantage. That could be growth, it could be access, it could be reducing a risk. Well, for most manufacturers, it's the fact that their customer relationship is managed through a third party, through a retailer. The thing, and I could be wrong on this completely, but the thing I would be thinking about is how do I create a set of moves where I'm creating more opportunity for direct relationships with end customers?

(25:30):

That would be one thing I would be thinking through. The other thing that I would be thinking through is how do I actually improve my product or service relative to my customer? Well, things like either localization or personalization is a killer feature that most large companies have a tough time executing against, right? That's what small companies do, is they create a localized personalized product or service when we get to scale. But imagine if you could solve that for a large conglomerate. If you could solve the localization and the ability to personalize product or service, or maybe it's to create localized product, well then you've created a completely different basis for how I compete. You can't imagine how you do that today with all the constraints and costs and issues that go on, but that's why it's a secret, right? Because if we could solve it, it would change the dynamics of our business. That's how you start thinking through that layer of the pancake stack. And as I write about it's this is not the place for most organizations to stop. You want to earn some momentum. You want to earn some insights along the way through the first couple of layers of this AI strategy before you start really developing a perspective or pursuing a perspective relative to solving the secrets of a business.

Peter Crosby (27:04):

Yeah, a good friend of mine calls it competing at machine speed. That's going to be the speed of this next year. And just to close out this pancake layer, I just wanted to ask you talk so much about needing to move early, especially with this area. And I'm wondering of your clients, how many of them have traditionally been early movers and how many are finding we need to change the mindset of being sort of the fast or medium followers this time?

John Rossman (27:42):

So I think one of the really positive things, and I think this comes from this first 30 years of digital transformation, is that people saw that the late movers have really paid the penalty for not moving quickly. And so I'm thinking of primarily e-commerce capabilities that the late movers really pay a price relative to that. And so I think you are finding executives much more open to saying, well, we got to figure this AI thing out. They don't know exactly how to do it, but they are sold on the fact this is new, this is different, and we cannot suffer from being the late responder. We don't necessarily have to be the first responder, but we want to be a close follower to the first responders relative to this. That's what I'm finding much more is that executive teams and boards are much more like, okay, we got to move now. We need to understand how to lead this differently. We need to understand how to develop a cohesive approach relative to this. But we do understand I'm not having to sell the concept that this is new, different, or that you need to act. I'm not having to sell on how to go about it. That's where the real conversation starts.

Peter Crosby (29:07):

And you mentioned a bit ago just the importance of leadership, and you're talking about again now. I'd love to in our last few minutes, really zoom in on the leadership principles. What are leaders called upon to do in this moment to be able to shift to that kind of agility, speed? What do you recommend and how do you inspire that?

John Rossman (29:34):

Well, the challenge is these things fail for many reasons, right? So there's not one simple thing do, but here's a top five list of the things that they need to do. First is you need to focus on problems that matter. It's too easy to get distracted by little low hanging little things that you could do with ai. You need to focus where the juice is worth the squeeze, because this is going to take resources to do so. That's the first thing. The second thing is make sure that you really understand the problem and that you define the outcome. What's the future? Don't worry about the journey of how we're going to get here and everything, but let's make sure we understand the problem and the future state that we envision in big be leadership. We call that the big bet vector. That's the fundamental vector that we need to have that if you have that, you actually have the basis for managing it successfully.

(30:29):

But so many of these journeys start off with we don't really understand the problem we're trying to solve, and we certainly don't have a consistent, correct level definition of what do we envision the future being? And when you don't have those things, then you can't do the next things that you need to do. You can't set up experiments appropriately. You have to just go big. You just have to believe it. And because you actually don't understand it yet, you can't communicate appropriately against it. You have to talk in vague terms, and you have to talk about these burning platforms of we have to change, we have to do better. That type of communication is actually damaging relative to these types of journeys, because we all do what people do, which is we start making meaning out of it for ourselves. So you need to communicate.

(31:19):

We talk about in big BET leadership, being a chief repeating officer, you need to be the chief repeating officer, but you need to repeat things that actually have caloric density to 'em, right? They're healthy. It's healthy communication, not just bad power intake, right? Yeah. Be better. Yeah, exactly. And that's classically what you get. We have to do better and everything. We have to care more about our customers. That actually is horrible. And then the last thing you need to do is you need to staff and put the right incentives in place relative to these types of things, because again, these things are going to start with failure and maybe earn their way up to success. So you need to have a completely different type of environment in which to have a team that is managed in the appropriate way, because these are different than your typical types of projects or programs. So that's the short list of the top leadership things that need to be done differently to have success in these high ambition, but high risk situations that I call big bets, but everybody's undertaking them. Your AI strategy, your digital transformation, your re-platforming, your acquisition, all of those things have high ambition, but are loaded with failure if you don't manage them correctly.

Peter Crosby (32:44):

Well, John, the pancakes sound yummy. I really appreciate you walking us through those layers and setting up a mindset for our folks who often are those change agents certainly were in the e-commerce era where they had to turn around and really educate an entire organization about the opportunity and then what they felt it would take, and then that spurred innovation across their organization, which is a super exciting place to be, but it's also very daunting. So for those of our listeners who kind of want to figure out their next steps on this, what would you offer that might be helpful resources for

John Rossman (33:29):

That? Well subscribe to my substack. So search John Rossman Substack or the Digital Leader, and you'll find the newsletter. I'm actually going to write a specific newsletter and a report for AI strategy and CPG and manufacturing companies. We will coincide the release of that with this podcast. Oh, that's awesome. Thank you. So that's one that is like I wrote a book with my client, Kevin McCaffery, called Big Bet Leadership. It's a simple book, but delivers a powerful framework of how do you actually lead these very different types of initiatives, big bets, high ambition, high risk situations, and connect with me on LinkedIn, John Rossman.

Peter Crosby (34:12):

That's super simple and incredibly helpful. John, for the fourth or fifth time, I'm sorry I don't remember the number. Thank you so much for bringing your really thoughtful and strategic analysis to the podcast. Super appreciate it.

John Rossman (34:25):

Well, I can't wait to get invited for the next one. I don't know what it is we're going to be talking about, but I hope this isn't the last one.

Peter Crosby (34:32):

For sure not. Thank you so much, John.

John Rossman (34:34):

Thanks, Peter.

Peter Crosby (34:35):

Thanks again to John for sharing his deliciously transformative meal with us. For more transformational inspiration, meet up with all the smart experienced brains that will be at the digital shelf summit in Atlanta in May. All the info at digitalshelfsummit.com. Thanks for being part of our community.