Accelerating Value
Accelerating Value

Episode · 8 months ago

The Dawning of the Big Ops Era

ABOUT THIS EPISODE

Big Data has finally become a reality for most organizations.

And the increased complexity it brings is ushering in a new era.

We’re beyond Big Data: It’s a Big Ops world now.

Today I’m joined by Scott Brinker, VP Platform Ecosystem at HubSpot, the perfect tour guide to welcome you into this exciting new world.

Join us as we discuss:

- The rise of Big Ops

- How to navigate increasing complexity

- The No-Code Movement

Keep connected with Accelerating Value on Apple Podcasts or Spotify.

Listening on a desktop & can’t see the links? Just search for Accelerating Value in your favorite podcast player.

Today. Every budget approval is an investment deal. If you're a marketer, sales or business leader, you had to promise to deliver value and impact. Writing the wave to get there is hard enough. Finding your way through the storm is even harder. If you're looking for that path forward so that you don't wipe out, you've come to the right place. Let's get into the show. Hey, everybody, this is marks Douce, your host for accelerating value, your weekly podcast about everything that takes to create value, to defend it, to invest in it and to accelerate it, and I am pumped this week. Okay, this is our holiday edition podcast, appropriately dressed in my, you know, kind of slightly Christmas colors, and we are here. We're here with, you know, chief Martek himself, Scott Brinker, to have yet another deep and penetrating conversation, right which which actually we joke about, but that is our goal and we and so we're going to kind of mix it up here a little bit and we're going to be talking about some things that Scott has really been putting a lot of time and focus on, I think, very appropriately, around the roll of operations. In fact, big OPS is is really his core thesis and and it's it has some really interesting ramifications to it and we're, you know, we're also kind of getting into now. You could argue the most important piece right of strategy, OPS and and tactics. I mean they're all really important, but ops is the is the hinge on the door, right. So, Scott, you kind of you kind of come already known here to a lot of people, but tell somebody, tell everybody. You know kind of like where you are right now, like, like what do you really care about the most right now, other than getting the hell out of here and going on holiday? Well, thanks for having me as a repeat guest. I love to yeah, have the the holiday addition here. Yeah, you know. I mean. So for the past fifteen years there more, I've been like thinking about, you know, this intersection between marketing and technology, and you know how it changes what marketers do and you know how business is engaged with their customers. And boy, it hasn't slowed down for a decade. I feel like every year I'm sort of like relearning like what's possible and where the set point is, and I think where I'm focused at the moment is, yeah, you kind of hinted at this big ops concept, which, for me it's kind of just a way of saying, you know, everybody knows big data and for like a decade or more, like we spend all this energy. He's thinking about all, my goodness, all this data. What does this mean? How is this going to change, you know, business as we know it? And in truth, that's taking us a good decade to, you know, get our arms around even just the basics of, you know, capturing this data and storing it and making accessible to you know, being able to do analytics and all sorts of cool stuff on it. Why? I kind of feel there's a parallel where now the challenge moving forward is less about big data just on its own. It's more of this big OPS, which is all of the APPS and Algorithms and analyzes and agents and automations that are all popping up all across the organization, running on top of that big data. They're fed by that big data and they actually feed back into it too. But yeah, this is this is kind of a whole new world. I'll be the first to say I don't have all the answers here. I think it's both dealing with big ops both has an incredible set of opportunities...

...for companies who want to be more affected and efficient in operating as true digital businesses. The boy it's got a whole bunch of challenges, some of which I think we're going to be getting into in our conversation here. is so one way or another, I'd say, yeah, it's. It's a big opps world for me right now. Yeah, I think actually it's. If we're talking about business right, it is probably far more accurate, in fact very accurate, to say that it is a big opps world more than it is a big data world. We have still a situation in most business organizations where there is a lot of data, but big data, right, is a technical term and there's a checklist of either five or seven, depending on which camp you're in. Right and in a particular data set has to check all those boxes in order to be big data. And then you get into the whole conversation about the fact that most of Ai, including machine learning, is a big data analytics solution which has struggled to some degree in certain ways to gain operational traction in in a lot of companies because of the dearth of big data. So if you're studying cancer, you got lots of big data. If you are in business, most of the time you don't. And so how do we, how do we actually talk about big ops? All right, because it's actually, I think, more than it's not something that rides on big data alone. It is it is really, for the most part, what you would what data sciences would call lean data operated right, but it has a lot of moving parts to it. And and how do we make sense of this, not only from a process and technology point of view, but from a people point of you how to, you know, one of the things you hear a lot about, and I know you talked about this a lot, Scott, is we've got to get this to where the the process and technology continues to serve people, not the other way around. Yep, it is. Yeah, I know, you're absolutely right. It's not that you need big data for all the things you're doing in big ops. Like a bunch of these things that are in big ops might very well be those lean data, small data. You know facets, you know what the business is doing. But yeah, the this sort of to Prou to explode like to edge sword of what's been happening with the advancement in technology, in particular around automation and AI. You know, there are now so many products. In fact, I would say, I would dare say, that half of the products on the Martech landscape today have some automation features built into them. There are things that, you know, allow you to, you know, working within that tool, automate certain steps. You know, some of them have a lot more capabilities there than others. Some of them actually, you know, have all these capabilities to not just automate what's in that product, but to like reach out, you know, and over API's automate what some of the other products in the stack are doing too. And then almost all of those products on the Martech landscape today they also have little pieces of AI models in them as well to, you know, s are machine learning that, within the CON text of that product, is trying to figure out, oh well, based on the signals I'm seeing, what does this mean? And so on one level this is fantastic. I mean like the sophistication of what all these products are able to do just in the lens of that particular product. I mean it's just it's, you know, for for marketers, it's like being a kid in a candy store like every day is Christmas, you know, you keep getting all these gifts from the tools. But their challenges. Partly there's the challenge of even just understanding, within the context of tick, their tool, what's possible and how do I use this? Well, but then the real thing, and this is what big...

OPS is, you know, really wrestling with, is when you have all these tools in a business stack and all of them have different things that they're automating and all them are applying slightly different machine learning models to try and understand the signals that they have visibility to. Now you start to get to something that. Okay, is this a harmonious orchestration, you know, across our organization, or is it a cacophony, you know, of things like, you know, making noise over top of each other? And yes, what a how do we how do we rationalize that? How do we get our arms around it? Yeah, no, absolutely, and I think this is we're being able to maintain the content, or the data and the context right, which is everything. But the data is so key to the alignment right and to to creating that symphony, not the cacophony that you were that you were just talking about. So let's let's delve into this just a little bit, right, and I'm going to start your with the fact that there's a value tube, right, and that people look through different, you know, both ends of the Value Tube, depending on what their job is. And and there's not a right end or a wrong end to the value tube. But if you try and cut the tube up right, you you you've got a problem, right, you're you're creating an unnatural act so that to include strategy, ops and tactics, and these all provide kind of an an evolving, enduring context and reality for the other pieces in the tube. And we're going to just define strategy for right now as a promise. Right, it's a it's a some people would call it a business case, but it's a promise. It says, Hey, if we do this, this is probably going to be good, and it's probably going to be good in the following ways, tactics or the way that we all struggle to, you know, keep that promise across time and space. What do you think ops is today in this to like make it, make it kind of boil it down to brass tax here, and particularly from a big OPS perspective? What is big ops doing that is not tactical and not strategy. Yeah, that's a great question. So if I look at the ends of that toe, like at the very top, you know, for strategy, like we sort of have this feeling like yeah, okay, you know our executives, you know, you know have a clarity around, you know, what they want to set as a strategy and that, yeah, it's just very clear, like you know how that gets initiated and then on the other far into the tube when it comes down to actually executing individual tactics. I mean, this is what the vast majority of us, you know, in an organization like, oh well, these are the things I actually execute and here are the degrees of freedom I have and how I execute it, you know. But between those two is this whole Gulf of I kind of think of it as like the structure and capabilities of the organization from translating that, you know, strategy of a handful of executives, you know, to now something that you know, dozens or hundreds or you know, thousands, you know, of people in an organization are able to, yeah, actually deliver on that strategy, execute their PC. You know, it's not just ideally, hopefully it's not just, you know, the capabilities to execute those pieces, but the structure that you know, pulls them together, you know, helps them have an alignment of all the efforts to stay calibrated. Yeah, you know, and I think what's my my thesis here with big OPS is if you think about the length of that tube, you know, know, for a long time the OPS chunk of that particularly marketing, I just I don't think it was terribly...

...deep. You know, there's a certain amount of I mean we've always had it in the larger the marketing organization, the war it's been, you know. But thinking back like twenty years ago, there's a much more finite set of like oh well, what are the kinds of ops things and structures? You know, marketing needs to support. But the explosion of all things digital, you know right, has just an elongated that tube and like the number of things, the number of moving parts inside the marketing organization today and the speed at which they move because they're almost all digitized, I think has made that ops chunk a much larger and also like more critical to like how do you actually make sure that connection between strategy and tactics is you know, got the proper fidelity. Absolutely so. In preparation for this, right, and I do this for almost all my podcast guests, right, I go out and kind of do some research and talk some people and I'm I'm kind of looking for four points of view beyond my own. So I was looking at synonyms for ops, for operations, meaning specifically what we're talking about, and the first word also turned out to be the word that almost everybody that I talked to who came out with conversationally, which is the word governance. Right, and and we tend to think about that kind of negatively, you know, because you know, I mean, you know, we're even beings. You don't like rules. Right. I think that governance in this context really means a lot of the stuff that you're just talking about, right. It's staying aligned, staying calibrated, staying in the right position, knowing when to move, knowing when to change, all this kind of stuff. You do agree with governance. That's interesting. You know, I would maybe expand it this way and I kind of think of maybe operations as three things. First of all, the actual capabilities. Can I do what I need to do with the capabilities? You know, software process, whatever it is, you know, texecute my tactics. The second is enablement. It's like, oh well, okay, technically that capability exists, but do I actually know how to use it? As anyone like you know taught me that. Where do I go for help, you know, as this range of capabilities expands. Yeah, the importance of having really good enablement, you know, for people take advantage of its critical. And then, I would say the third leg of that stool is the governance, which is make sure we're doing it right. We've got the capabilities, we now know how to do it. Let's make sure we're actually using that right and using it right is everything from staying Aligne to the strategy, you know, making sure that we're, you know, adhering to proper compliance and regulations and all these things that you know are a part of that. So yeah, capabilities, enablement, governance. So I'd like that a lot, would you. And I know this is getting a little tortious here, but but I know that people are going to ask this question. So do you feel like that the decisions as to what tactics to deploy happen at the OPS level? That's a great question. I think I think about in terms of market mix, Martin Mixes, marking mix, happen at the OP level. Well, so this is yeah, and now we start to get into some interesting philosophical questions on this too. I guess my sense is, the further you can push decision making down in an organization without losing control of the car, you know, keeping it on the road going in the right direction we want. You know, the further you can push decision making down in the organization, the faster and more powerful, you...

...know, your organization is generally going to be. And so, other things being equal, I would like like to see a lot of the capabilities and ops help feed people to make more decisions at the tactical level, you know, and make sure that, yeah, through the governance that's coming through that that there are generally they see the feedback that, yes, this is these are the right kinds of decisions, this is the impact this decision is having, you know. And in cases where you can't do that, where it's got to be at a higher level to coordinate across a larger group, then yeah, I would see those sort of higher level umbrella decisions being obst decisions and say, I think you're right on the money. Right. So, ten years ago when I was at Honeywell as Cmo, right, and we were using really heavy duty analytics, you know, kind of pre software. So we had a large analytics team doing all this for us. The whole goal was to say, okay, these are our strategies and these are the strategic outcomes that we absolutely need to have at the end of the day, right. And we realize, however, that in each given market that how those strategies are going to be successfully expressed and executed or going to be different, they're going to look different, and if we try to kind of do a one size fits all, we're just going to, you know, jam the American way into into every market in Europe. And you know, there's a and we have the fundamental belief that inside of every Swiss or German or, you know, Dutch or French, that there's a little American waiting to resonate with our with our stuff. Right. Clearly, that is that was not that was not going to be a successful approach, right. And so we had to give people the freedom to execute against those strategies in ways that were locally and tactically appropriate for their markets and maintain the alignment all the way through, right, and and that. And so I agree with you that. And we got so MRE the results that we got at a country level were unbelievably better when we did that. That makes a ton sense. So what is the brain going forward for all this? Right? So, I mean we're talking about a situation now where, and you've posted a lot about this, both explicitly and implicitly, that the complexity is just exploding, right, and that's just in the area that is, you know, let's say, the the part that we all control, right, that's exploding, and then all the stuff that we don't control out there in the market place, that's also exploding and becoming far more variable than it ever has before. The speed or the cadence of change is is ramping big time. How do we how do we think about what that? What that North Star or that GPS, or you know, whatever you want to call it? Right, how do we think about that going forward, given the fact that, for all of it wonderfulness, right, the unaided human brain has probably reached its maximum potential in terms of how many variables that can hold it in its head at one time. Yeah, it's a great question. I think part of it is the recognition that it is a complex system, and complex not in like, you know, like the you know, colloquial adjective of there's a lot of stuff going on there. You know, I mean complex almost in like the scientific complexity theory definition of it, and just the nature of complexity is...

I think it's very rare for you to just be able to see like the whole system in an entirely deterministic way. I think, Ay, I just don't think that's going to be possible for a very long time. And second I'm not sure that seeing the entirety of all the complexity in a highly deterministic way is actually what we need to be incredibly successful. It's sort of feels like, you know, you have like these these like local regions, you know, throughout an organization and you need to understand the interaction effects within them, you know, and then you need to make sure that you know, as we're talking earlier, in this whole like governance thing and ops, you know that there are feedback mechanisms to make sure that at a high enough level, things are staying connected and they're staying a line and that there becomes signals when things start to diverge or a nomalies or exceptions, you know, started peering in places we don't want. But I guess, yeah, maybe, maybe phrase it this way. If like on a scale of zero to a hundred, and on the zero side, like we just have no idea what anything is happening, and I've pissess like yeah, like you're no evil, see no evil, like what you know? There's that into the spectrum. And then the other end of the spectrum is is like all knowing, perfectly deterministic. Everything is connected, everyone can get the whole thing orchestrated in their brain. I would say like right now, because of this explosion of big ops that frankly, most companies, most of it right, this is new, this is a new phenomenon. It is a result of technological advancement. Now we're starting figure out, like oh well, what are the implications of this? We're much closer to the zero end of the spectrum right now. I'd like to see us get further, you know, maybe in the top half, you know, the over the fifty mark, and saying like okay, well, listen, yeah, we might not have all knowing, you know, deterministic vision of all of this, but we have the way the systems like work together. There's a much more cohesion than there is chaos, and so that's kind of a way of dodging your question. Of I'm not sure if there's a singular brain that controls all of this, other than at the end of the day, whatever executive group is setting the strategy, what you want is to make sure that strategy is the conscious decision upon which then all the other complexities that you know, merge out of that keep a lining themselves back towards. Yeah, I think that you know. So I guess where I would come out on that is that the real goal here is to separate the signal in the noise, right, and in this case, you know, let's talk about marketing and marketings impact on different business outcomes. The fact that a lot of marketing is noise rather than signal doesn't mean that it's worthless. It doesn't mean you should stop doing it. It just means that it's not a significant factor on a one to one to one basis. Right, it's not a significant factor in terms of meeting whatever goal you're trying to meet, and all that stuff costs money, right. So the ability to say to the business. Hey, this is a stack rank, right, of what we're doing and it's relative importance across this kind of timescale and that kind of timescale and so and so forth, relative to this other stuff, right, which we believe is necessary. But we are going to spend less on some of this stuff and spend more on the stuff that's really driving the business right now in a big way, acknowledging the fact that this may change going forward based on market conditions, of all kinds of things right, and I think that that the one of the less of which there are, of course, many of the last two years is is about this right, that that just because something was understandably good pre covid or understandably effective pre covid,...

...does not mean that it was it continue to be effective, or that it is effective in the same way today. And so the ability. You know, the more I talk to business leaders about this, for them it's ultimately not about or not as much about proving marketings impact and, you know, Roi and the strictest sense and you know, kind of all that historical stuff. It's what they're really trying to say is, I want to understand how it's operating in my larger universe that I care about and when things change, how do we optimize or reoptimize for that? And we're how can we understand that whow? We have an opportunity right now to spend a lot more money and get a lot more impact on the s curve kind of idea because of stuff that's going on out there in the marketplace, and then we need to understand kind of where that tops out and where, if we kept spending more and more and more money, it wouldn't make any sense. I think that right there is really the the goal of big ops in many ways is to say, how can we absolutely crush it, understanding that the definition of crushing it and the opportunity out there in the marketplace is ever changing, and then make sense. I mean it's interesting because when you have an organization of any size, I think you could actually argue you have this problem all throughout the company. Right. It's not that just any given marketing activity has like effectively negligible impact. I mean any given phone call from customer service, any one call, one interaction, negligible impact. The software developer writing any one line of code they write, negligible impact, you know. But all these things like they they aggregate up, you know, and an aggregate, you know, and not even like necessarily like, you know, full coming wide. Agree that just even local, you know, a team aggregation. That's when they start to really have big impact and I think that's, yeah, just one of the challenges in all this data we're seeing in marketing is just trying to figure out what is the right resolution granularity, at which right to pay it at what the the signal is very often the impact of the aggregations, you know, and making those choices of like where we choose to aggregate, you know, and it's the individual pieces that are really yeah, it just noise, noise, noise. Well, the irony is, of course, is that the it? But they almost all become noise the bigger the model, right. So the the history on this has been that data scientists have tried to kind of have a one or two mega models, right, and and there is there is validity to that approach. But the problem is is that you get into a situation where most of it is turned into noise, where if if you had a more targeted or more discrete model, in the words your question was more focused and you were trying to answer that more focused question, you wouldn't see it in quite those terms at all. A lot of stuff that ultimately becomes noise in the big model becomes really important in the small model, right. So when we talk about so I think one of the things kind of going forward here is a version of the UDHLOOP. Right. The udhloop me means observe. So this is the amount of time that it takes for any of us to observe something, decide and get oriented about...

...what it means to US side, what we're going to do about it, and then act in a it was originally coined in the Air Force for combat pilots and obviously the you know, in that kind of situation, speed is life and and you definitely don't want to have slow udloop. But it is become very much of a thing in data science and you see it as a slang term in many areas of a company where they're talking about this. What do you think is going to if you got kind of like look into your your you know, Crystal Ball, you know we're here at the end of two thousand and twenty one knowing what you have, you know what you've seen over the last two years. What is the what is the Ooda loop look like? I'm not asking you how we accomplish it. I'm saying will marketers be okay operating at the same decision speed and action speed that they have historically had, or do they need to be prepared to go faster? Yeah, it's a great question. I feel like this is also one of these things where it comes down to the level, right, because there's a new de loop on, like, okay, I am running a specific campaign, then I am testing particular things with the landing page, you know, and what sort of the speed by which I see what's happening there and how I react to it? You know, when you get higher up, you know in aggregating activities across marketing teams? Yeah, then it's more of like Oh, well, is our collective, you know, paid search marketing, you know, performing and like swear what's the feedback loop there? And then, ultimately, you know, at the next level off it's like, Oh are you know, acquisition, you know, strategies of all kinds. You know where, and so I think all of them are speeding up. I think this is where there's this weird combination where you know, we're talking earlier about, you know, using technology to push more decisions to the edge of the organization. You know obviously then, that you know requires as people on the edge of the organization to have a Localo to loop, to be able to, you know, respond there. And in some ways then it feels like the aggregation above them. I know this isn't maybe the quite ways to use it. It's like a bunch of those things. Actually, I don't even know if you need like the immediacy like it. It's like the trying to decide, like okay, then what become the signals at the next level up that you even want, like a manager or director or Vice President of marketing, you know, to be observing something like where is that resolution of observation helpful? So let's use let's use last week's example with Peloton and Mr Big. Right, so Mr big is in the new you know, in the in the new series of sex in the city, and he dies. Sorry for the spoiler, he dies on a peloton cycle. It Hammers Peloton stock by ten percent and within a very, very rapid turn, Peloton PR comes back with a statement about the fact that it wasn't Peloton, it was his bad lifestyle choices and it started this whole conversation online, which is almost turned into a meme at this stage, right, where if they had, if they had waited even another twelve to twenty four hours to respond, it would not have been anywhere near as effective. Right. That's kind of the the idea here, right. That that there, you know. And this is also where marketers, who have valued orchestration above almost all things for a long time...

...as a kind of a demonstration of competence, can get into a challenging situation because they want to think too long right before they do something, whereas pr people typically are are more wired for this, right, but they also a lot of times lack some of the contextual awareness as well. So a lot of times what they do can be riskier than what a marketer would do. So this is all about kind of calibrating risk and reward within a very, very, very tight and getting tighter decision loop. Do you feel like that that is a reality, that that we're going to actually have to deal with pretty pretty quickly now, or how do you think about it? I don't want to be I don't want to I don't want to lead you into a to a conclusion. Yeah, I know it's an interesting example. I kind of feel like, yeah, it's like there's the the exceptions and then there's the rule. I mean like, you know the example you gave there, right. I mean like, okay, that doesn't happen every day, but when it does happen, reaction speed is incredibly valuable and we just know this is today's digital environment is such that, you know, if something's going to go really well or really badly, it's impact can spread, you know, like wildfire. So there's having that sort of you know, reaction capability when those issues happen. But I do do still kind of see those as the exceptions, you know, more than like the actual day to day operations of the company as a whole where, yeah, just you know, most of the things that are happening and interactions that are happening around the company don't have that sort of outsized impact and you are really trying to manage to the aggregate across them. And so that's why I guess I'm just trying to I don't know, answer to your question. Are Complicated, complicated. Further right, most of the stuff that we do normally so kind of the not Coloton type thing right, also has a fairly time lagged sense of effects. Well, all right, so this is the thing. Like at the individual level, I would argue they don't. At the individual level they're instantaneous. It's the fact that they only become meaningful when they aggregate to a certain level. And so, depending on what you're talking about, you know, I mean, you know, hold time on your you know, customer support line or open rates on your email campaigns or whatever, at some point that becomes an aggregate thing where like, Oh wait, this signal is turning in a direction, you know, other than what we expected. And so yeah, there's maybe a delay in aggregation simply because you need enough of a sample to have the meaningful aggregate, you know, view on it, and that sample gets collected over time. But the individual data points, you know, that are feeding into that aggregation actually do have an immediacy. And I guess this is one of the questions of do you have the front line of the organization, the edge of the organization, being a tune to and reacting to the individual points that are not meaningful higher up on their own individually, but they are meaningful at the local level and an aggregate, they add up to being incredibly meaningful, but over time, no, and not. So let me give you a specific example, though, of what I'm what I meant. So a phenomenon that many, many marketers talk about right, particularly in B Tob Right, is that they will go out with a campaign and right out of the...

Bat, right off the bat, it just seems like nothing happened. Right, it's sort of you can see people like going to the website and you can see people doing this without in the other but there's not any sense of action for some period of time, right, and it's it's time lag, right. It's the fact that people percolate, it's it's, you know, the fact that particularly is the risk and the cost of something go up, people take more time before they take action. There's a lot of things going on there. But if you looked at it at, you know, thirty days, if you look at the campaign at thirty days, you might erroneously conclude that the campaign was a bust and the you needed to move on, whereas if you had the right analytics right, you could say, actually, you know, we're kind of in what barbed you guys smokers called the stall, kind of in the stall and when it comes to getting to the right temperature in brisket. Right. So that's kind of what I meant by that, trying to kind of make a little less techy. I love it, but I'm very cute. Okay, yeah, I'm so, okay. So, and it's something, though, that's the stall. There's nothing more disconcerting when you're smoking brisket than the stall, because even though you get to point where you can you know this is sort of going to happen right, it never actually happens exactly the same way twice, and so you're kind of like going, wait a minute, last couple times this stall lasted only about eight minutes and we're now at twelve minutes and it's still stalling. What the Hell? Right? And and the temptation is to go and do something right when actually that would really screw it up. And so that's really what I'm talking about, right, is that the ability to say, okay, you know what, no, no value, no impact is created right away. How do we judge this right within the context of a new loop, so that we kind of were saying, okay, now it's clear this is from several different points of view. This is a this is not worked and we need to disconnect it and move on to the next campaign, versus knowing that it just needs a little more time. Giving thoughts on that. I mean, yeah, having been in Betab sales for most of my career, yeah, I I can definitely appreciate that and I think, you know, there's no perfect answer to any of this, but I do think you know, if you say, okay, this is a campaign, that's a very measurable activity and then there's this other end further down in time of actual close sales, you know, and until we get the closed sale, you know, like it's hard to attribute actual dollars to you know, was the campaign successful and owl is he our eye to it. And so if that distance, you know, to closure is days, weeks, months, yeah, you you can't be in a mode of it would seem on the surface to be much harder to know, like hey, is this the right stuff or not? But I mean I kind of feel this is where the digital environment is, you know, so incredibly helpful, is because between you know, that campaign launch and the end goal of a, you know, a large purchase, there is all this other activity. You know, that happens and historically. We can look back over time over all the previous campaigns we were ran that ultimately led, you know, to you know, sales and what those ratios were, and we can compare all the data points of the activity that happened,...

...you know, from start to finish and all those others, and sort of see where we are in our current one and get some sort of reasonable, you know, sense of like hmm, is this behaving like our more successful campaigns have in the past and the stage of the kinds of activities people are doing? or or no, you know, and in some cases it's going to be no that you know again, that's just not that much of a science right now because there's so many variables that go into a given campaign and how it impacts people, on how people wat there. They're not cookie cutter, you know, sidebyside comparisons. But the more you have that sort of overlap and you sort of are able to use, you know, some interpolation to see like Oh, yeah, okay, now this seems like it's on track. I you know. I mean this is I don't know why I'm explaining this. This is your area of expertise, you tell me, like, but now I mean, you're right. I mean look, I mean, at the end of the day, right, the analytics to do all this have been around for a long time. They're widely used in the physical and social sciences. I mean climate change and everything that we have understood over the last a twenty years or so about climate change on this earth rides on multi variable linear and nonlinear regression. Same thing with pandemics, same thing with almost everything. I mean the the the joke and data science, of course, is that, you know, eighty percent of the world's questions are answered by regression and and that, and it's probably true. The the challenge historically has not been the math or even the data. I mean, obviously it's always challenging to collect data. Has Been Forever. Aristotle wrote about the fact that it was such a pain in the ASS to collect data. So it's not new. But data availability is certainly today, you know, not really the problem. I mean, we can certainly nuance that statement, but it's not the problem. The problem has been historically that operationalizing data science to make make it possible for people to make a better decision than they otherwise would have without data science has been really, really tough. Right that's mainly because it's been human powered. It's so it's not. It's not that they don't know what they're doing and all that kind of stuff. It's about the fact that it's human powered and therefore it is slow. It's slow as held. It's very difficult scale. So hence the the preference for the giant unimodel approach. Very difficult to scale, it's very difficult for non data scientists understand the outputs and it is consequently extremely expensive. So at Honeywell I was spending five, six seven million dollars a year just on analytics. Now it wasn't worth it. Yes, it was absolutely worth it. Did it help us do phenomenal stuff in a environment that is so heavily regulated that nobody really thought that some of that stuff was possible? Yeah, absolutely, it sure did, and I personally in my team, personally made out like bandits as as a result of that. But right, it had to be automated in order to get to that next step. And most people don't, you know, when they we talked about Ai, Scott, I mean most people immediately kind of go to the flavor of the month, which is machine learning. But automation is actually part of AI. It's an AI technology. So there's a and and actually far more accessible and far more easily implemented. Pick your subject right then, then anything else that we can think of. It's probably easiest to start with automation. So that's all where it's going right now, because it...

...has gotten to be so complicated right that that you got to have something to help you out. And MTA, you know, of course, you know. The more data you have about your customer journey, that's as long as it's good data, it's just fantastic. But third party, that demise, a third party data has pretty much put, you know, a period at the end of the MTA sentence. Okay, so I want to just kind of switch you're actually to. You've talked a lot about no code and lowcode software. This is kind of understood today to be code for I don't need to be an expert anymore in order to create something that does something that I want to do. Is that possible, right, or do you still have to fully understtand the process that you that you're effectively wanting to automate, or can you not understand it today and yet still get there? Yeah, no, it's a it's well, I am fascinated by the way this whole no code movement is sort of changing, changing the boundaries, you know, of WHO's capable of doing what. And I think of it very broadly, like it's not just no code like APP building. I think like any sort of thing I like if it's an APP and automation or, you know, a web page, an analysis, you know, a graphic generation. It's like basically all these things that, without being an expert in a particular discipline, I am increasingly able to do some of the low end and over time, maybe even some of the midend cases of that myself. I think where my perspective of how it works is. I know what I want, I just don't have the skills to build it. And if what I want for a bunch of these things is so localized to what I'm doing that there's no way it actually makes sense for me to take a ticket and wait for, you know, weeks or months or someone in it to like look at it and then they look and they're like wait, that's just you and you're doing this and I have to spend time like now. I'm sorry, it's just not worth it. You know, and it's just been a ton of things up business where we're like, oh, yeah, I would really like it if this process was faster easier. Could I have this thing or could I have that? You know. But yeah, the answer is just been well, doesn't really matter because you know I can't make it and you know it's not significant of enough thing on its own to make it worth while for, you know, other experts to come in and do it. No Code in my world like changes that. You know that it's someone who's like wow, I really wish I could like automate this process. And you know, when I get people who are filling out a form here for this event, I wanted to trigger this registration, you know, over in this other system. I want to make sure I'm getting the follow up email. I want to take the tshirts eye make sure that that's getting over to the thing for the production of the swag that's going to go for it. And the people who are running these events like creating this stuff. They know exactly what they want to have happened in that process. They just haven't had the ability to create an automation for that until now. And so yeah, I think you do need to know what it is you want to be able to use no code, you know, effectively. To me, the the the bigger challenge is not that people aren't doing a good job of solving their localized needs through no code tools. I think in generally they are. It's that their worldview, their context is generally very localized and now sometimes, as they're leveraging more and more automation and more and more like Ai, we you know, capabilities within that...

...automation. The thing they are doing locally might have other ripple effects or are there impact, you know, elsewhere in the organization? I might be affected by other people's, you know, like automations and, you know, processes elsewhere, and that's sort of to me, where the big opps challenge really starts to come in. Is it's not being successful in the localized use of a notecode tool, but it's how do we make sure the complexity of lots of people all throughout the organization, you know, doing a bunch of these localized hats with these tools aren't aren't having an interaction effect, you know, that said, overall negative instead of positive. You know, one of the things about that, this really interesting to me is you know, it wasn't too long ago that people in tech would use the term point product. Right, very very specific, little product. It did certain things and if it was ever acquired, it was acquired as a took in and went on down the road. What you seem to be talking about that idea, but in eat at an even more granular level. So like I don't even need what might be called a full product, I just need a piece of point functionality to do this one thing. And your example about, you know, the swag and all that kind of stuff is a really good one. Right. Is that where you see this going? And then the flip side would be is, is there a way that a CMO or a very senior person can get involved with no code without hurting themselves? Wow, okay, yeah, two great questions are I think. I mean, like let's start with that. That that last one is yes, I absolutely think executives can leverage no code tools. But the question me comes because they are typically further removed from operational realities. Yeah, you know, I think it depends on what it is like. For instance, for me is, you know, an executive like I have a data question. Like you know, Hay, listen, you know, I'm curious. So if something I've I wake up with a hypothesis, I'm like, you know, I want to see like the relationship between X Y and Z and ABNC, and previously, you know, I would have had to been like, okay, well, let me ask, you know, the analyst. You know, I'll the analyst is sick today. All right, hands comes in. I they've got ten other things on their list before this. I'm just curious about this. I don't wanted to rail, you know, somebody else's work and you know, and so halfs of time, these ideas that come to mind, you know, they just sort of, you know, slip away, you know, versus, I think a lot of these sort of citizen data scientist tools. Maybe that's a better way of calling these examples. Instead just no code, but same idea. You know. Yeah, it's not going to make me, as an executive and overnight data scientist, but there's a bunch of these really simple, low ind case is of things that actually, yeah, I should be able to self serve. You know, a number of these answers and I think that's an incredibly bad I mean it's not just powerful for executive. That's powerful again, as we were talking earlier, precise probability to the edge. But I think it's something that an executive can directly be able to say, Oh yeah, that's a no code tool that, yeah, I would use because I've got a million questions every day, you know, and the more of them I can instantly self service, yeah, actually, the faster and better the organization moves that will actually would you. So thank you for that, because you just really summed up a big part of what proof is all about. There, right there. You know that that pretty much nails it. So when we think about more and more and more of this kind of product and more doing more things, more automation, more, more man this this...

...word is so freighted with so many different meanings for so many different people, but I'm going to use it anyway. More autonomous decisionmaking. Right. What do you think that does to a marketing culture that is been all about intuition and my ability to orchestrate something great, because I've seen the elephants so many times. Marketers love personal control, right. I mean, and it's and and if they're not unique in that by any means, I'm not trying to say so, but like one of the things that we see a lot with proof is they want they want all of the answers, but the next logical step, which is to take those answers and automate the implementation and updates of plans, they're not ready for that yet. They want to do that themselves. Yeah, I mean I guess it depends on the level at which you're talking about, right, like, I mean if you're doing like a national brand campaign, that's, you know, going to have, you know, commercials on the Super Bowl. Yeah, right, that's the sort of thing. Like I'd be surprised if the CMO didn't want to be like very closely involved in that. But my God, you like go down to the level of like how we responding to these social media things here? What's our you know advertising strategy with like you know Google paid media? I mean the cook that becomes a lot because of this whole thing about the fact that digital just has such an enormous, you know footprint of you know, all these individual, you know touch points. I yeah, I actually don't see as much. I kind of feel like it's for themos away from you know, this feeling like, Oh, I have to have my hand on the wheels of everything, because it is absolutely physically impossible at this point to have, you know, your hands on all of it where it's actually operate. It gets aggregated up to a higher level. And so, yeah, we're probably somewhere between the two and probably depends a lot on the particular business and the particular Cemo as to you know, how much they wanted intuitive control versus trusting more of like this distributed capability that is being orchestrated, you know, through an analytics, you know architecture. I think I feel like we're making a lot more progress towards that ladder, just because, you know, necessity is a mother. I mean like, yes, I need this department a run and I can physically not do it all myself. Yeah, so let me let me set up my last question for you, because it has been a great conversation in this way. So, right up until the end of world, I write, you had gas piston driven fighter aircraft. Right for the most part. There were early, early jets, but for the most part it was it was still ultra high performance piston and we're talking about speeds in the four hundred to four hundred fifty miles an hour kind of range. And the the instrument meation cluster inside the cockpit was pretty basic. It had substantially complicated, you know, since the beginning of the war, but it was still pretty basic and you still you still had aces and particularly really high performing fighter pilots who were so called natural pilots. Right, they flew by the seat of their pants and they just had a feel for it. As jet as the Jet Age took on right, you had a situation where all of a sudden we were exceeding mock the speed of sound and then we went to mock to and the it had profound effects on the design of aircraft. Were maneuverability...

...and instability were synonymous in the airframe and the speed with which things were happening meant that all of a sudden, fly by the seat of your pants kind of approaches were no longer valid. In fact, they would kill you if you try to do them. And so you started to see the automation of a lot more subsystems within an airplane. The instrumentation in the cockpit just exploded and today is even more so. And as now, at the point where you were wearing pilots wearing helmets that transmit brain command and straight into the helmet to try and improve the speed with which decisions are made. Right now, this is an extreme example, right, and it's a life and death example, and for most marketers it's not life and death at all. I mean there are a couple of exceptions, but not many. Where do you see this in practical terms going like? I mean, because the speed has very significantly improved and I think that that is where the ops, the big ops thing is really coming from is. How do you how do you stay in alignment and in governance with things moving that fast and all of a sudden, just just like a dog fight? You know it, you know at Mack one, right, there's a lot of swirling and curling going on right, and how do you keep track with all that? So do you? Do you have us a sense for this? I mean, is there a point at which, for example, that neither marketers nor their audiences can keep it straight? Is there kind of like a wall, or is it just like the sound barrier that no one thought you could penetrate and you go through it and you're kind of in a different set of physics? Yeah, I think the Metaphor Seems Act, you know, with again as you pointed out, that hopefully, you know, less dire stakes, you know, but yeah, it's just all these things are moving incredibly fast to so many of them, and so, yeah, that's changing, you know, the use of these systems, you know, not only from a monitoring perspective, the the observing, you know, but then also the loops in which action, you know, happens and how that ex how the execution of like decision is made and then like the time it takes actually execute on that decision. Does it does it change? In your view? The pilots? Yes, to a certain extent. Again, this is where, like, I feel like the analogy starts to fall apart, because they always do. You know, like the cockpit, as complicated as it is, is still actually a relatively local you know. It's almost like, oh, how does this change warfare when you've got all that localization inside the cockpit, but then you have you know, you know, fleets of, you know, hundreds of those aircraft, and then they're being coordinated with all these other activities and where does that fit in with the strategy? And, you know, I think in many ways that larger arena, you know, is probably more of the the metaphor, you know, for a cml at any you know, large organization. And so, yeah, you need a bunch of those like automated things, you know, and highly intelligent systems that are operating on the local level, you know, different marketing touch points and activities. But you also need to then be a will to leverage the way that's aggregated up and how quickly that impacts like your decision of Hey, listen, we're going to change this tactic here or, you know, hey, this you know, this tactic here is like telling us this facet of our strategy is, you know, missed some...

...key assumptions and we now want to factor those assumptions back in and update the strategy and then, you know, communicate that out. I'm well, actually had that lost in my metaphor there somewhere. I'm like when I was at honey well, right, we all had the opportunity, if we wanted to, to take some very basic flight training and then, of course we wanted to go further. Week certainly do that, but one of the things that struck me so powerfully in my few lessons was the the difference between visual flight rules and instrumented flight rules or fly by wire right, and the way that the instructor pilot made the point was that he had me fly into a very dense cloud bank and then surrender the controls to him and as far as I could feel, all my five senses were saying to me that nothing was any different than when I had flown in right except that when he had me fly back out again, we were upside down, and he said, you realize that if you had just assumed that you were still right side up and you pulled up on the stick to climb, you actually would have gone into a dive and crashed right made a big, big impression on me at the time, and that was also kind of very much what the transition on the pilots side in combat flying. That was what was going on between say, nine hundred and fifty nineteen eighty, right as they were taught to not use their eyes right, to trust their instruments and not their feelings, for lack of a better way putting it. Do you think that we are kind of in that same place today, not only with markers but with business people, or do you feel like I mean, we're are. Where are we on that? Yeah, I mean I think it's important to recognize that all that instrumentation didn't just miraculously happen one day, right, like Oh yeah, you didn't have this yesterday and now we all magically figured it out and we guarantee this is going to work perfectly. Right. There were some whole process by which, like you know, they were even figuring out, like okay, how do we do this? How does this automation work? Is this goop? Well, that didn't work, you know. I mean, and I kind of feel like that's again more of where we are. You know, an automation and big OPS and business right now is yes, we know we have to get to a place where we rely on more of these systems to have just the ability to operate, you know, in this highly automated digital world, but we're still figuring out exactly how those automations need to work. Do we for the instrumentation? Do we actually have the right data? I mean like, for instance, you could imagine if you had that instrumentation in the fighter jet, but you know, one of the signals they had in quite wired up yet was the altimeter. So like yeah, we got a whole bunch of data on which direction you're fitting and we can make all these it. We don't know exactly what height your add but how important could that be in like making an automated decision? You know, it's I think we're just still in a state where we've got a lot to learn and and that's again why I kind of phrase this big ops thing kind of like an era, and I think we're at the very beginnings of the era. Is just there's a lot to learn and figure out. But yeah, and when you get to the other end of this, I think, yeah, the the metaphor you have. Therefore, this is kind of exactly what I think we expect to see. What a great conversation. One of the best things about having a podcast is is having this kind of conversation and because, you know what, you never know at all. Ever, you always have a great opportunity...

...to learn and you it. I think it really shows once again that you also don't have to agree on everything in order to have a really great conversation. So, Scott, I just really appreciate it and I wish you all the best for the holiday. Guys, this is this is going to go live during the holiday. So we've gotten a lot of really interesting feedback. Scott this is sort of really interesting. The number of people who have said they're going to really catch up, not just on our podcast but on podcasts in general while they're on their holiday break is pretty amazing, and we're up to about seventyzero listeners on this podcast today and almost all of them. Their behavior fits in kind of with with that last comment, because they're not streaming it, they're downloading it to listen to it. So it's kind of a it's kind of an interesting little thing there. But thank you so much again for for your time, for your way you talk with me about these things right, because so many people are really trying to wrestle with it and they're trying to figure out like what's the most important thing, like how I even begin right. And so maybe in the New Year sometime, third podcast with you, Scott, would be and this would be more in line with I think you know, your core thing about Martek is to a day knowing what we have learned. Right. We're do you start and and do you is the starting place just one place, or is it several places that then converge together? So we'll go there. We'll explore that if you're up for it, and that could be a really cool third show. So again, everybody, have a wonderful holiday, Scott. Thank you so much, sir. Thank you, mark. It was a pleasure. Have a great holiday as well. All right, we'll be back after the first year and hope to see again on the flop. Thanks. The sooner you can optimize your marketing spend, the quicker you can start delivering clear, measurable value to Your Business. That's exactly where business GPS from. Proof analytics can help. Learn more at proof analytics DOT AI. You've been listening to accelerating value, where raw conversations about the journey to business impact help you weather the storm ahead. To make sure you never miss an episode, subscribe to the show in your favorite podcast player. Until next time, you.

In-Stream Audio Search

NEW

Search across all episodes within this podcast

Episodes (35)