The Connected Omnichannel Experience: An Interview with Martin Kihn, SVP Strategy, Marketing Cloud, Salesforce

MK: Well, I tell people, you know, I went to business school so I could just quit writing. And ironically, that's when my writing career actually started to take off because I had something to write about. But I was... It happened naturally. I was sort of at work. I was at the job. As a consultant, you go on-site with clients. So, I went to various cities. I was in Akron. These are not glamorous cities. These were wherever the client might be. And we would go and sit in their offices, and they would give us like a conference room or something, a cubicle. And I would notice that the consultants, they didn't necessarily have a background in the industry because of all these different industries, but they use language in a way that made them sound like they were experts. So, the good partners could sound like they knew everything about everything just because they dazzled people with the way they used language. So, I thought, "I'm gonna write a dictionary. It'll be the Consulting to English dictionary." And that's how it started. And it's actually in the back of the book. There is a dictionary there. That's where it started. And then it just turned into a memoir. I said, "Well, I might as well just tell some stories while I'm at it." (8.27)

SS: Yeah. Well, you're a gifted storyteller and an amazing ear for dialogue was the other part of it. It was quite unique, the book, in the sense that it had this, you know, sardonic perspective on this artificial world of consulting and at the same time, this sort of riveting personal experience that you describe. And the series to that extent sort of brings out the soul of the book. I mean, that much comes through, for sure. I'm not sure all the sexual high jinks that are portrayed in the series, obviously, were true. So, the book...

MK: They weren't. Yeah. This whole personal life wasn't. I mean, I'm just a family man. I've actually been married forever. And I wish that I had done what he did. I would be a much cooler individual. I really didn't. But I understand why Hollywood edited it. If you saw "Pee-Wee's Big Adventure" it's kind of like the end of "Pee-Wee's Big Adventure" and they have Burt Reynolds playing him or whoever.

SS: Yeah. Well, the series went on for five seasons. So, it was...

MK: It did.

SS: ...obviously successful. And you were a consultant to the series the whole time, I understand.

MK: I was. Yeah. Especially in the beginning.

SS: So, you left Booz and then your career kind of followed the trajectory, if you will, of data-driven marketing and it started with Digitas, which, obviously, was one of the sort of pioneers, obviously, in this industry. And you apprenticed there through, I think, the early 2000s. Now the switch over to the agency world, which has its own culture and stories, was that sort of a welcome reprieve from the consulting grind or did it just introduce you to a whole different set of circumstances?

MK: It was much better. I felt so much more at home. And the nice thing about Digitas was it was... The guy who hired me, actually, had been a consultant at BCG. It was David Edelman, who went on to be CMO of Aetna, and is now ex-CMO. A well-known guy. But he hired consultants at Digitas. And our big client was American Express. And they used us...it wasn't all for ad campaigns even though Digitas is sort of an ad agency, but for direct marketing and also business strategy. I mean, the first project, the first big project I was on, was around launching a business card sort of for small businesses. This was called the Plum Card or the... I'm not sure if it's still around. But anyway, it was trying to...it was looking at small businesses' spending habits. So, it was very much like a strict consulting engagement, but the atmosphere in an agency is still my favorite place to work, type of place. It has creatives, you know, the British guys with the little glasses and the facial hair and the wry smirk. And then also very smart people doing SAS models, and then also the kind of smarmy sales guys. And all these groups, very small teams, all working together and there's lots of stress because advertising is really all or nothing. I mean, you're either like nailing it and winning huge clients or you're completely on the way down and firing half your staff. It's this roller coaster. So, emotionally, it was maybe a little bit too fraught for me, but I liked the atmosphere and the people in it because they're very lean and... the teams are very lean. They're very efficient. And there's nothing better than that kind of creative, but also data-informed environment. (11.57)

SS: Yeah. And that's where, really, you set the foundation for the balance of your career because you were thrust into this role where you were doing that kind of data analysis that later on at Gartner, you became their analyst. Is that correct?

MK: Yeah. I joined Gartner because I had...I was at Digitas and then my wife is Minnesotan... this is in New York, New York City, the only place I ever wanted to live. And my wife is Minnesotan and she spouse-napped me and convinced me, I don't know, in a weak moment that we should move to Minnesota. I don't know. To this day I do regret that move. But we were there for four years, and I worked at...I was at a creative agency in Minneapolis called Fallon. A very well-known at one time, like, a creative hot shop. And they did television, so sort of above the line. And I was the analytics team. It was basically me. I did measurement for this TV agency. And it was pretty cool, actually, the job was. But I got recruited to Gartner. I joined Gartner in a group that they were just starting called “Gartner for Marketing Leaders”. And it was the recognition that marketing was becoming more techie and that the types of questions CMOs were asking were much more CIO-type questions than they used to be. They were asking about applications and data, databases, and measurement. And so the whole profession of marketing was becoming more in line with the core Gartner value prop, which is like CIO research. So, I was a part of that team. And it was good for me. It really was a good fit because I really, at heart, really just do research and writing and kind of putting things together. And I don't like having a team, really. I'm sort of, I suppose, a little antisocial. I don't like managing people. I think people should manage themselves. And so, in the corporate world, I'm an uneasy fit, but the Gartner fit was good because I am good at research, I'm good at explaining things and then also public speaking, I enjoy public speaking.

SS: Yeah. It's this interesting blend of right and left brain thinking.

MK: Yeah, I know.

SS: You're rare ...

MK: It seems like a modern CMO, they need both the right and the left brain. I always say that.

SS: Yeah. The analytical CMO who also has a creative leaning is...well, that's, in some respects, the T-shaped person. We're gonna talk a bit more about that toward the end of this interview. So, it's quite interesting to learn about that. Now, your time at Gartner was really, I would call it today, the glory years of programmatic advertising. Salesforce as a timely interview just pulled the plug on its own DMP, which must send ...

MK: Not quite. We can talk about that.

SS: Okay. Well, let's do because that was the implication of at least one report, which, you know, would send a smoke signal, I presume, to the rest of the industry about this potential cookieless world, obviously. So, it'd be an opportunity to talk about what Salesforce strategy going forward is gonna be around this, but also, just your perspective on the future around digital advertising. Do advertisers go back to sort of old-school audience targeting, placing, basically, banner ads on certain publishing sites or is Google's path down this concept of a “federated learning of cohorts” really gonna eventually become a practical alternative to the walled gardens? What's your perspective on that?

MK: Well, there's no going back. I mean, I think I understand when people say we're going backwards to pre-RTB because of, you know, cookie deprecation, cookie goes away. This whole idea, the whole kind of, both on the consumer side and the regulator side, and just, in general, the industry is moving away from user-level IDs. But on the other hand, we're not gonna go back because the tech has advanced and techniques have advanced and everything is just much farther along than it was in 2010, you know, when RTB started to become a thing. So, I think that where we're gonna end up is becoming clearer. There will be user-level IDs. There'll be opt-in. So, there will be things like UID 2.0, the trade desk solution will probably exist in some form or other. And all of this assumes that consent has been gathered and stored. And there will be others. Depending on the market, there probably will be some in Asia that are not used in the U.S. and so on. So, there will be user-level IDs. First-party data will be used for media. You can already do it on Facebook, like on Custom Audience, you know, put together a list of emails. So, that's gonna be massively huge. And not only the social networks, we use it but also large publishers. "The New York Times", my hometown paper, is already talking quite a bit around first-party database audiences. And then the other thing is the FLoC. I mean, the FLoC has been paused, the federated learning proposal, and there's a lot of others, FLEDGE and TURTLEDOVE and... well. TURTLEDOVE and FLEDGE, I guess, are the same. But SWANs. And then they ran out of bird acronyms. But there will... The idea of some kind of machine learning-generated cohort is perfectly valid. They just have to figure out the mechanism. But those will exist, I think there's no doubt about it, at least in Chrome.

SS: Yeah. And what about Salesforce's decision? What's the next iteration, if you will? Are you basically gonna fold in DMP functionality into your CDP? Is that the direction?

MK: Well, the long-term vision, we have a whole kind of narrative around this. But I think the DMP as it existed when it was launched won't be required in the cookieless world, but that doesn't mean that elements of DMP don't continue to be useful. What we did announce, now Salesforce has a DMP is called Audience Studio, which was Krux. Krux was acquired in 2016 back in the boom years of DMPs. And we announced that we weren't gonna be selling new versions of Audience Studio to new customers. But on the other hand, we continued to support existing customers, many of whom are relying a lot less on cookies than they used to. They're using other kinds of IDs which you can in the DMP. So, this is not just spin. I mean, we're gonna keep this product going as long as people wanna use it... existing customers wanna use it. But when we talk about media more broadly, like, how you're gonna do ad campaigns in future, I think first-party audiences, prospect customer information that you've gathered with their consent, is gonna be the seed, the kind of core, and then there will be, obviously, be lots of other ways to do media without any kind of data, you know, based on contextual information. And this will be better than it used to be, basically, because machine learning is better than it used to be. So, the machines are smarter. (19.11)

SS: Well, we're gonna get into that too because it's an interesting part of your book. There are many interesting parts of your book that I wanna touch on. Just before we go there, though, you know, you've crossed over in from, shall I say, ad tech to martech to a large degree by joining Salesforce. Is your role at Salesforce to be the chief story storyteller? Obviously, you have a gift for it. You've alluded yourself to your strengths of being research and writing and presenting. Is that your role or is your role more designed to look forward at how the product roadmap can align with where this business is going? What's your mandate in your position?

MK: I have elements of both. And my role has shifted. When I first started, it was very focused on CDP because at the time... This was 2018. So, I was at Gartner and I was covering marketing clouds and also a customer data platform as a category, which was, you know, zooming up the hype cycle in 2018...

SS: Yeah. Sure it was.

MK: And then our X... our number one question in that year from our customers, our clients was, "What is a CDP?" because we kept track of all the questions we were asked and the inquiries, but that was my area. And then when I was hired on the Salesforce and there was a real question about, first of all, what is this category that our Salesforce customers are asking about? And secondly, did we have one already? And that was an interesting question. And it was based on data, there were surveys, in particular, Advertiser Perceptions did a survey asking CMOs and marketers what CDP they were using. And Salesforce came in first in that survey. We were...something like 60% of the respondents said they were using Salesforce CDP. At that time we didn't have one.

SS: But you didn't have one.

MK: Yeah. And number two and three were Adobe and Oracle. And only Oracle had talked about launching one. Adobe didn't have one either. So, we asked a very sensible question, I think, which is, "Well, do we already have one?". So, anyway, I was involved. And it was me and a very small group of people doing research, essentially, customer research, product research, trying to answer the question. And where I came... And also build, buy, acquire, do that analysis. Should we just make another acquisition? And the long story short is at the end of that we decided, it is very similar to CRM for B2C. There's enough new to it that we would have to either build or acquire some new capabilities. And the acquiring part will be harder than building it ourself because the integration is, as you may know, very difficult for acquired products. So, we decided to just build it ourselves. (22.14)

SS: We're gonna come back, obviously, to CDPs. But Salesforce now, I think, Gartner has ranked as the leading solution amongst marketing hubs, I think they are called.

MK: Marketing hubs.

SS: Yeah.

MK: Marketing hubs.

SS: That's marketing hubs. Yeah. New terminology now, which probably is a more appropriate term to describe what used to be enterprise suites and cross-channel campaign management and you name it, I've seen it over the years. Going forward, how does Salesforce maintain its pole position now? It's a pretty fierce nuclear arms race. How do you stay ahead of the other guys who have equally large budgets behind them?

MK: Yeah, it's tough. It's competitive. The big competitor in our space is Adobe, but there's lots of others. And in fact, in the CDP space more specifically, they're not only newer vendors, so, what we might call startups, some are quite mature at this stage. But there are a lot of self-developed solutions that have been enabled through a lot of the tools that public clouds make available now. You can build a lot on Google these days. And there are other vendors that are more sort of generic like Snowflake or so on. And there's a lot of open source out there, particularly, for data pipelines. So, we are competing against a whole plethora of other options in the space. I think the way that we were at least thinking is, and I think it's right, is that in the marketing cloud, which is my area, we have been built up over the years through a series of acquisitions that have been integrated more or less over time. And this CDP is different in kind because it was built on the same codebase as Sales and Service Cloud. So, we're closer to the CRM platform itself and the Salesforce "platform." And that's the secret I think, being part of...everyone says they're part of a platform, but, in fact, if you can make something like the app exchange available where literally anyone has developer tools, and they can build an application on top of this set of tools that can then take data in and out from the CDP. So, the CDP...our CDP functions as a kind of just store, repository place to organize customer information. But what you do with it, you know, all that, you could build your own applications. All of that requires more of a platform approach rather than a typical more or less monolithic application, which you call a CDP. So, we're trying to break it down into components. And the whole “platformarization” is in service of a larger narrative that Marc Benioff is pushing, which he calls business transformation, which is another kind of trendy term, I guess. But what he means by that, I think, is that the different parts of the organization work together better because they're all using the same kind of unified profile underneath. So, single view of the customer, if you will. And then this availability of tools, so that you can do your own customization or build your own applications on top or, you know, the 8,000...well, it's only 1,000 apps that are in Scott Brinker's, you know, diagram with all the logos, little, tiny logos, those guys could make their application available in this marketplace. (25.46)

SS: So, the future “marketecture” is the other term, of course, that you hear, does that really become the enterprise architecture that is the foundation for, I think, what you call in the book the “connected omni-channel experience” and the role of the marketing hub becomes more tightly confined to marcomm? How do those platforms intersect at some point?

MK: Well, I think this is related to CDP again. The reason CDP appeared as a category...as I said, it's not new. It's not like people woke up, I don't know, four years ago and said, "Oh, now we need to have a clear view of our customers." I mean, this has been around forever.

SS: Sure.

MK: Of course, it's not a new desire. And they're also...it's not a new set of tools. I mean, in the late '90s if you talk to, I don't know, like, Unica, you know, some CRM on the B2C side, they would be talking about single view of the customer and cross-channel orchestration, probably email, direct mail, maybe the website early on. That would be not the early '90s. That'd be the late '90s. But it's similar terms and also make it available for analytics. So, it's the same general gist as a CDP. The difference, though, now is that there's so many more channels and there's such a greater requirement for really rapid decision making and orchestration. So, you need different architecture, obviously. But the CDP itself is... I don't see how it could be just for marketing. It really is...because it's focused on customer data, so the C part, or prospect, anyone in the organization who needs to deal with customers would be using the same data source, as I see it. So, that includes the service team, obviously, the sales team, for sure. But even... I don't know. Definitely, you know, anyone in customer experience or research more broadly. But people are doing product development. So, it's got a lot of applications outside of marketing.

SS: Sure. Does the CDP, and again, we're gonna get into much more detail about it momentarily, but does the CDP sit at the center of this galaxy of orbiting hubs that are dedicated to specific parts of the enterprise, or does this reconfigure the enterprise around the customer, which is actually far more interesting speculation to have? Basically, eliminating silos is really the end goal there.

MK: No, you're right. It's the latter. That's the goal. And it is a journey, though. It's a multi-year journey. It's a journey for all of us. And by that, I mean, I don't know any customer who has reconfigured completely or any of our customers around their customer, but many of them are doing really well. And we all know examples like, you know, Amazon or Netflix or... just to take two clichés who, you know, every recommendation they give me is something I wanna buy. So, they certainly know me a little bit better than I wish they did. (29.04)

SS: It would take powerful artificial intelligence, I think, to understand you, Marty, even just based on this fractional conversation we're having today. You're an interesting person.

MK: Well, the secret is, I always say, everyone has something and if you put a Bernese Mountain Dog in any kind of ad, you immediately have my attention.

SS: Well, and let's just pick up on that because you've written...you got your writing career reignited with "House of Lies" but you've written, how should I call it, spoof on, I think, learning to be an asshole in business which I haven't read, but is next on my list. I've met many in my career.

MK: Me too.

SS: You wrote a book on dogs of all things. This book now, though, talk about a pivot, customer data platforms. What led you to write the book, because you could have written about the wider scope of martech, right, the role of different solutions in the enterprise today? But you decided to focus specifically on CDPs. Who is your intended audience? The marketing generalist who's heard the term but doesn't quite get it, or the martech pro who really wants to understand how all of these pieces start to fit together? What was the objective behind the book?

MK: Well, I co-wrote it with Chris O'Hara and, like, he's my colleague at Salesforce. And I knew him before. He's a friend of mine for many years. And also, a writer. He's written previous books, including some on chicken wings and barbecue.

SS: Well, he's like you. He had writing as an early ambition, didn't he?

MK: Yeah, definitely. We have a lot in common, me and Chris. But we were sitting... This was actually after the lockdown was imposed last year. And so we did have a thought that we should do a project. So, we take advantage of this...the time that we don't have to commute, you know, we don't have to travel around the world. And we wanted to explore the space, customer data platforms. The original intention was to do kind of a sequel to Chris's book that he wrote with the founders of Krux, it was called "Data Driven." And it was more around data management platforms. So, we said, "All right. Let's update this for the kind of CRM world and CDPs." And then I thought, "Well, there's gotta be a lot of books already on this topic." And there aren't, actually. It seems to be a green field. So I thought let's do the first. And then I thought, "Well, we can't get it out fast enough, you know, because with publishing..." My experience with publishing is it takes like two years. But we did manage to... we went to Wiley and they said they could get it out before the end of the year, you know, just around the holiday season, which is really fast. So, we have to write it quickly in three or four months, and then get all the permissions and get it proofread, etc. But it was like nine months end to end. And we did. We wanted to be the first. It's not all about Salesforce. So, we tried to be broader than that. And also, as you mentioned in the beginning, it encompasses a lot more than just CDPs. We ended up having to talk about the organization and how marketing can work better with other departments and analytics, things that are more about martech. (32.20)

SS: Yeah. I think the parts of the book that really resonated with me was not so much the technical side of it, which was interesting in of itself, particularly, for more or less neophytes but...I've been around the block a few times, but the speculation about where marketing is going, the role of AI. We're gonna come back to that question because I really did find those parts of the book interesting. But again, just in terms of who you were aiming the book at, yes, there's a gap and that... I had a conversation, a podcast interview, actually, with one of your colleagues, Mathew Sweezey. I asked him that question because he's actually written one of the more successful books on marketing technology that's still selling seven years later. I asked him why he thought that was the case and he said, "Because I didn't write it about technology. I wrote it about marketers using technology." That makes a ton of sense. And it seems to me that that part hasn't evolved a whole lot. Again, a subject we're gonna come back to. I wanna go, just dive into CDPs and you mentioned this, that data has always been the nemesis of marketers. It's a four-letter word for them. It was for me early in my career in database marketing, certainly, and has continued to, obviously, be for organizations. And David Raab said to me that it's because organizations didn't really appreciate the value of customer data. They weren't, you know, that concerned about customers and now they are. And it brings me to my question, which is organizations are appreciating now the value of that data, are appointing data czars and so on to oversee it. Obviously, privacy compliance has become essential. Do you see, and I know the answer because the answer is in your book, but do you see CDPs ultimately eclipsed as IT wades back into the fray and reasserts its hegemony? Do you see them... Do you see the shift away from CDPs toward these central data hubs or data management stacks? Is it an either-or, or is there a hybrid model that will evolve out of that? And you state in your book quite clearly, the answer to that is you think the CDPs will become the single source of truth for customer data. How does that dynamic play out, though? I mean, you got marketers on one side who are behind most CDPs, and then you've got IT on the other side saying, "Wait a minute. This has become too important to the enterprise to give over effectively to one department."

MK: Well, I think the category CDP continues to evolve. So, the way it looks today is not how it will look in five years. So, I think that it's not a static thing. It's growing and organic. And also, there are over 100 vendors called CDPs and no two of them are the same. They're all rather different. And in fact, many of our customers have multiple things called CDPs working side by side doing different things. So, there might be multiple so-called CDPs within a stack. I think the tension between IT and marketing is one that's getting worked out in real-time. And it differs by organization, but in general, I think that there's more appreciation now for the difficulty of the task at hand and the resources and time and expertise that are required to do something like maintain a single view of the customer in-house. I mean, I've been asked many times, like, every week, "Can we just do this ourselves, just use open source and get our good data science team on this?" And the answer is, "Yes, of course, you can." You can do almost anything yourself if you have enough people and you have enough time and you have the right kind of talent and you're willing to invest in it. But the solution, like what Salesforce is offering, so Salesforce CDP or some of the others, is essentially outsourcing some of that headache for you. For instance, you know, Facebook, for instance, changes their API 19 times a year. Do you wanna have somebody on your team dedicated to monitoring Facebook's API so that you can reconfigure it every time it changes so that your process doesn't break, your social advertising goes down? The answer for most companies is gonna be, "No, I don't wanna do that. That's not what I do. That's not what makes me special. So, I'll outsource part of it." And that would require a CDP vendor. That doesn't mean that in-house you don't still have control over things like master data management. CDP doesn't replace MDM. It also doesn't replace enterprise data warehouse, probably. (37.04)

SS: Yeah. I kind of envision the CDP almost as a client system of that data stack. All of the data transformation and clean up and so on that goes on in the plumbing eventually yields a golden record of some kind that CDPs then can append engagement data to. And they're almost then the CDP's role to, actually, to help feed data into other systems that allow the sort of real-time personalization that you describe in your book.

MK: Yeah. It's like plumbing. I mean, it's probably too simple to say it's cleaning up your customer data, but it does do that. And it's also supposed to be useful. So, the idea is not...it's not a data lake. I mean, Gartner always said 80% of, you know, the data you dump in your data lake will never be looked at again, it’s dark data that's completely useless.

SS: Or they're data swamps, I think, is the other term.

MK: It's a swamp. But it's supposed to be useful. And so data should not really be going into the CDP that... I mean, you're saving for a rainy day. This is stuff that you actually have a use for.

SS: You need to action.

MK: That's right. It's more thoughtful.

SS: Yeah. Again, we're gonna come back to a couple of these subjects. But Gartner has predicted that, again, one of those, you know, I think, provocative predictions that are designed to stir conversation, but they're suggesting 80% percent of companies who are pursuing a 360-degree, I'm sure you've read this, will abandon those efforts because it's, as you point out, really hard to do. What's your response to that? Is that the case? Given the reliability of companies going forward on data, will they really give up, wave the white flag?

MK: Yes, I know that research you're referring to. In fact, that prediction was made by a good friend of mine and a fellow Bernese Mountain Dog owner. And if you actually read it, Gardner did say that, that they are abandoning single-view of the customer. They also have another piece of so-called maverick research saying that. "Don't even bother pursuing single-view of the customer." But their point actually is more nuanced than the headline makes it sound. What they're actually saying is that if you define single view of the customer as a complete view of your customer containing every single data point that you could and should have collected about them and that you're trying to create a truly golden record that is absolutely accurate and persistent and exhaustive, you know, MECE, collectively exhaustive, mutually exclusive, that kind of thing, then that is probably a waste of time. And I have to say I completely agree with that. That's not really what a CDP is supposed to be. It's not supposed to be a data dump. It's supposed to be something different. (39.55)

SS: That's the data lake.

MK: Yeah. They're saying don't build a swamp or don't create a swamp. And I agree. So, if you're trying to build that, but to be fair to our customers, I don't know that people are really trying to build that. Some people maybe.

SS: Yeah. Yeah. Well, I mean, there were other issues in terms of how the data is being used that needs some evolution before that. Let's talk about that. A big part of marketing spend over, say, the last 10 years has been targeted advertising, programmatic advertising, digital advertising, obviously offline media as well, but that's eclipsed by digital ads now. But now customer experience has risen right to the top of the C-suite agenda at least according to a CMO Council report that came out recently. And I think they're expecting the CMO to help lead the way. And their disappointment in CMOs are that they're not showing the requisite leadership to do that. I talked about Delta earlier, in your book you offer it as an example of a seamless experience. Is that kind of gold standard where you're really improving the operational day-to-day experience that the enterprise is offering customers? And if that's the case, what does marketing's job really become? Are they going to be point on helping to design that optimal experience, not necessarily deliver it? Again, I'll use that term you used in the book, connected omni-channel excellence. What becomes marketing's mandate in this future data-driven experience, customer-first world that we're moving toward?

MK: I think marketers have taken it upon...we, I'll say us, marketers, taken it upon ourselves to fill a void. And that's caused the professional marketing to change quite a bit over the last 20 years. I mean, I was in business school almost 20 years ago and... actually, 20 years ago, literally. And at that time, you could tell people are interested in marketing just by looking at them. They were kind of better presented and they were, I would say, a little bit more superficial. I was one of them, you know, interested in meeting celebrities. I’m being reductive here, but not too much. I worried about fonts and big campaigns. And it was not a statistical profession, but now, you know, two decades later, it really is.

SS: Yeah. That's quite a switch from targeted advertising and marcomm. And to your point, most marketers are trained, you know, really are drawn, let's say, to the profession because of the creative aspect of it, not necessarily the analytical aspect of it. So, what becomes marketing's future mandate within the enterprise? Do they yield control to other potentially new groups, customer experience being one of them, that can deliver that desired experience? Do marketers get into the designer's chair and design the experience while somebody else actually delivers on it? (43.05)

MK: Oh, yeah. The point I was getting at was that the key thing here is, it's the customer journey, but it's also who in the organization knows the customer. Who really can take the outside-in view? Most of our...when I was at Digitas, most of our clients...and they even knew this, actually, they're like, "We're sitting in our building here and talking about who's out there, but we don't really know. We're kind of designing, in a sense, artificial journeys for artificial people." The marketing team is supposed to... And I think this is the role that they've assumed, and this is what I mean by we've adopted a role and filled a vacuum. They are in really good organizations, I think, not the voice of the customer, but the kind of soul and heart of the customer. So, they are the ones who know the implications for having a single view of the customer and how that can be useful - use the Delta experience. And I like the Delta experience. I mean, behind the scenes, I think, it's more complicated than it looks the way they've executed it. But it has been a major investment for them. But they look at the whole travel experience for their high-end business travelers, those are the ones they're looking at, frequent business travelers, and all of the different digital touchpoints, and every single, not just the look and feel, but what information is in it? When is it available? How does it come across? How can you listen to the signals that people are giving you? There's a lot of complexity in that journey, but they're trying to just make sure it's all connected and coherent and that it delivers what people are expecting and then just a little bit more. And in order to do that across channels, they just think about it tactically. They need to join the kiosk with the mobile app, with the web, you know, the website. Those three things alone are three different channels and it needs to be in real-time. If I change something on the kiosk, I expect it to be reflected in the app right away. And so everyone needs to be... In retail, in particular, there's some massive problems, you know, connecting online and offline. But I think that marketing is, even though a lot of that is operational, might be supply chain and a lot of it might even be related to service, which is a different department. Marketing is supposed to be the one who understands the customer best. So, I think, ultimately, they'll steer most of that.

SS: Well, it's interesting, because it's certainly something that I talk about as well that the marketer has to be the customer advocate. But my observation is that that's actually marketing's greatest weakness right now, they really haven't figured this out. And largely, I think, because they're driven by performance marketing goals, that their job is to get the next sale, not create new value for customers or, indeed, have an influence on the business model of an organization.

MK: I mean, your point is a good one. Certainly, the KPIs might be wrong, but I think that good...good marketing organizations have been trying to move away from the performance-driven approach and too narrow a view of customer. And, in fact, degrading the experience. If you're too focused on performance, the customer experience suffers. So, trying to be more of... And just about every customer that I can think of has thought about that, what you just said, the implications of being too focused on the wrong metrics and trying to improve and to move along. But I think, you know, many have a long way to go. And I can also say, you know, there are good organizations out there that don't really have a well-developed marketing function, or it is just comms. Somebody else owns the voice of the customer or maybe the service team and they do a good job at it. There's not a single configuration that works. (46.49)

SS: Well, even sort of duelling roles between traditional research, market research, and behavioral analysis would be another good example of that insight fragmentation, if you will. Let's then keep going with this subject because it's so important. And I was gonna ask you earlier that, you know, my observations of technology over the... And I have been around long enough that I was there when the early desktop marketing systems were born to see where things are today and tremendous progress for sure. But what it seems to me is that the martech space has sort of tried to identify the immediate needs, obviously, of marketers and not pull marketers forward into new processes or new ways of doing things. Would you say that that's still the case today or... I mean, Salesforce is amongst the leaders, certainly, in providing thought leadership for the industry. Brian Solis, yourself, Mathew Sweezey, that's a pretty impressive set of people. Is Salesforce trying to do that, trying to be that beacon for the entire sales, support and marketing industry?

MK: We definitely are. And we have... There certainly are visionaries on the team and not just the ones you mentioned. They’re even on the engineering side, in particular, some very deep thinkers I run across. The tension is always between what's available today and getting people up and running and adopting tools that they already have and getting value. And so there's always a lot of room for improvement, not always, many times, a lot of room for improvement there. For instance, if you just have Journey Builder, using it better you can get better improvements, and that doesn't require any kind of, you know, visionary insight. But on the other hand, where's marketing going? I think a lot of the... In addition to new channels like voice as an interface, which I'm still, you know, wondering... about the usefulness of marketing...

SS: A lot of people are.

MK: Yeah. But, you know, just in general, modeling AI and machine learning, doing predictive modeling, better segmentation, all that's come along so much and so fast in the last five or six years and just productizing that, making it more widely available, what we call Einstein features, embedding them into the product. I think that approach means that, you know, our tool can be kind of an assistant that's there if you want it. But if you don't want it, that's perfectly fine. You just turn it off. And so that's a way of surfacing new ideas, you know, gently.

SS: Does that suggest maybe, though, that there's a consultative business that should be embedded within Salesforce to accelerate the degree of business transformation that needs to happen over the number of years? I hate to say that given your experience.

MK: Well, we have... I mean, it's a good question. Salesforce has certainly thought over the years about services and what kind of services need to be delivered other than servicing customers. We have a lot of partners, as you may know and, of course, you do know, SIs and also management consulting firms, many of which actually have a dedicated Salesforce practice. So, whenever there's this kind of heavy services-type work that's required we'll point to one of them and, you know, they'll bring in... And it's not just SIs now, you know, the tactical kind of implementation partners. But there's definitely strategic thinking. I mean, I just got off a series of webinars with Deloitte. And Deloitte did the first half and I did the second half about the tools, but the first half was all about the organization and kind of getting your organization in a position where you're ready to start thinking about implementing a CDP. And that's the kind of work that we at Salesforce don't do. That's definitely a partner consulting firm-type. (50.57)

SS: Yeah. Who...

MK: Like my previous life.

SS: Who use a lot of words that people don't understand?

MK: Yeah. I've noticed that the martech consultants tend to be less like that. It's more the kind of high-level visionary, strategy consultants.

SS: Yeah. Yeah. So, I do want, though, to pursue your thoughts on the optimal marketing organizational structure now going forward. So, the other challenge, of course, is, and we were talking about this earlier, about creativity versus analytics. And you've also talked in the book about a COE dedicated to just data, but it strikes me that that should really be a subgroup within another COE which is all about analytics and behavioral insight, which suggests a COE within another COE, which would be marketing ops. What's the hierarchy there? How do you... Like, today, there's lines between those three areas, creating problems in and of itself. How do you eliminate those lines, those silos, and bring that together into a collective pool of multi-disciplinary, but collective pool of people that are organized around the customer?

MK: It depends at what level you need to organize around customer groups. So, how different are our customer groups? I think of it that way. And a classic example would be a lot of times you can do a COE, center of excellence for analytics because numbers are international. You could have someone who doesn't speak English or doesn't speak German and yet they could do data science in a language that everyone can understand. So, you can have a kind of centralized center of excellence around numerical data and analysis. But on the other hand when you're doing communications like social marketing, that's very localized. That needs to be literally in the market because there's too much nuance there. So, I think that's the approach I take to this marketing ops. I think at what level is it important to separate versus to consolidate? And I mean, there's all kinds of structural challenges there, but in general, the COE model is good and I think the agile workflow is good because you have people who have different skills, and you kind of form teams and then un-form them kind of like a movie crew.

SS: The Hollywood model.

MK: The Hollywood model. So, I think that seems to work. And if it's all around an actual project, so, I think the software developers know a lot about how to get things done quickly and well. (53.40)

SS: Yeah. Well, I talk about this. The agile marketing model, obviously, which is short-term and sprints and so on, but there's the longitudinal model that's required here, which is the transformation part about how do we become a better customer-oriented organization? How do we create value for the future rather than for the present? How do we survive the disruption going forward? Somebody's gotta own that strategy and then report into the C suite about, you know, what direction the company has to go in. And to me today that's a big miss for marketing. They're not really stepping up to do that. Sorry for the editorial commentary. I wanna talk about complexity because I think amongst the many challenges that marketers have today, data, obviously, being one of them, complexity is the other one. We have a lot of solutions that aren't fully utilized. Certainly, commentary about that. Is the answer simplification of the technology, you know, no code development platforms is probably a good example of that, or is the answer really lies in, and this is back to the organizational model, training a new generation of T-shaped marketers who are completely at ease with data and, maybe you're the model, completely at ease with creativity as well?

MK: There can still be room for specialists without a doubt, but I do think that we underestimate the power of the creative and the messaging. And I think that it's actually gonna become more important over time rather than less. It's the area where a company can differentiate in the way that it talks to its customers in a more human way. And it's also something that people will probably always do better than machines other than coming up with the right sale price. Machines will probably be better at that. But coming up with a way to communicate with other people that can move them is the realm of creative. And so I like your latter approach, which is the T-shape. And it's not now that we need to, you know, teach people to respect data. I think that computer science is now the number one major where it used to be English. So, it's not that. People know data is important. And I think we need to actually tell them that creative is important.

SS: Yeah. Well, I use this term, creating the relatable brand, and that's to your point about humanistic marketing, not mechanistic marketing. But AI suggest, you know, particularly the strength of your book is getting across this idea, that ultimately AI will be the salvation. It will allow these very complex communication programs real-time to happen working off a well-structured, that's your term, not mine, CDP data set. So, that kind of then, hopefully, the complexity fades away a little bit. Is that fair to say?

MK: Yeah. I mean, things like, you know, if we sat down with identity management 10 years ago, you had to have an exact match between ID, identity A and identity B. It had to be the same email, same name. Then they could do fuzzy matching. And now through AI, you could do some very complex matching scenarios without really a lot of code and without a lot of initial setup. So, the software itself is getting smarter. It's supposedly making our jobs easier. The issue is that we need to understand how it works so that we can, you know, guard against things like bias or inaccuracy. (57.20)

SS: So, in the short time remaining, Marty, just put your futurist hat on, five years from now, not just the martech space, which obviously continues to evolve, but marketing in general, where are we? We've had, you know, just in the past five, six years, tremendous progress, advancements. Where do you see this business? What's your... Part of your job is, in fact, contributing to the future roadmap. What's that future look like to you at this point?

MK: Well, I think just in general terms that consumers will take more agency in the way they're both targeted and measured. And I don't mean necessarily in the weeds of who's collecting what data about them, but I think that we will...in the same way that for, I don't know, movie recommendation engines, we can give it instructions, I think we as consumers will be more open than we are now and less paranoid about marketing, targeted advertising, marketing messaging. We'll see more of the value in it because they'll be better, and they'll be more relevant. But we will give more explicit instructions, I think. So, we will... There'll be ways for us, for instance, to tell Facebook the kinds of things we like and don't like that aren't as clumsy as, you know, what we do now. The steps we have to go through now are just too ridiculous. But the machines will be much more nuanced and sensitive to that kind of thing.

SS: Does marketing keep its name marketing, or does it become the customer experience group within the company?

MK: Well, it'll stay marketing, but it will morph into...become the customer experience crew.

SS: So, is there a new memoir in the works then based on your experience of the past few years here, Marty?

MK: Of course, yeah. Also, I am very interested in cats. I wrote a book about my dog. So, if I can combine cats and marketing and I think there's something there because cats really do rule the internet. So, I'm working on that.

SS: Cats and dogs. It's like IT and marketers in the past, cats and dogs.

MK: They can co-exist.

That concludes my interview with Martin Kihn. As we learned, marketing is in the midst of a transformation from managing campaigns and touchpoints to a “lifecycle” model where the goal is to deliver a more unified experience across channels. But to pull that off, they need a unified customer profile, connected to every system of engagement, and a planning model that takes into account all stages of the customer journey. Marketing should elevate its role from strictly a marcom focus to become the customer champion, and learn to master the art and science of data-driven marketing.