Marketing AI Institute | Blog

[The Marketing AI Show: Episode 12] AI for CMOs: John Dougherty, Brighton Jones

Written by Cathy McPhillips | Mar 22, 2022 6:36:00 PM

The Marketing AI Show brings you a special episode in our AI for CMOs podcast series, brought to you by Persado. This series kicks off with John Dougherty, CMO of Brighton Jones.

The idea behind the series is to tell the story of AI and digital transformation through the experiences of CMOs from diverse backgrounds who are leading the understanding and adoption of AI in their organizations.

You can listen now on your favorite podcast app, or keep reading for more for a summary and transcript of this episode.

Episode 12: AI for CMOs series with John Dougherty, Brighton Jones

In this week's episode, show host Paul Roetzer sits down with John Dougherty, CMO of Brighton Jones. This special episode kicks off our AI for CMOs series where Paul Roetzer sits down with CMOs leading the way when it comes to piloting and scaling AI within their organizations.

In this episode, Paul and John discuss the role of AI in the world of marketing departments all over the world, and how AI is changing the marketing landscape at Brighton Jones.

As we listened to this episode, we noticed the mission on the Brighton Jones LinkedIn page precisely matched what John said, and how AI and technology were helping their team help their customers achieve wealth in all its forms:

Whether you want to save for the future or celebrate today, give back to the community or explore the globe, your values are every bit as unique as your fingerprints. The challenge is aligning your time and resources to those values, so you can go after the things you truly care about.

For us, protecting and growing your wealth is a foundational part of that journey. But our goal is bigger than your financial peace of mind—it’s your happiness. We want to be the partner that helps you align your wealth, your passions, and your purpose, so you can pursue the life that truly fulfills you—your richer life.

Timestamps

[00:14:30] Personalized orchestrated journeys at Brighton Jones

[00:18:03] Building a CDP

[00:27:13] The data scientist, marketer, and piloting AI at Brighton Jones 

[00:32:03] Data is the new oil!

 

 

Watch the Video

 

 

 

Read the Interview Transcription

Disclaimer: This transcription was written by AI, thanks to Descript.

[00:00:00] Paul Roetzer: Welcome to the Marketing AI Show. I'm joined today by John Dougherty, Chief Marketing Officer with Brighton Jones. Welcome John.

[00:00:11] John Dougherty: Thanks, Paul. Awesome to be here.

[00:00:12] Paul Roetzer: Yeah, it's great to have you. I'm really looking forward to this conversation. John and I have traded some emails and had a little banter before here - there's just so much to get into today.

[00:00:21] So I'm really looking forward to it. This episode is part of a new AI for CMOs series. That's presented by Persado, one of our big corporate partners and we're longtime supporters of what we're trying to do at Marketing AI Institute. So the idea behind the series is to tell the story of AI and digital transformation through the experiences of CMOs from diverse backgrounds, diverse industries, diverse experiences, just knowledge and capabilities around AI.

[00:00:48] But the people who are leading the charge within their organization, they may just be leading the charge to educate people. It might be leading the charge to pilot it and actually get some projects going or in some cases, a bigger enterprise, it might be leading the charge to scale it. Because, what we know is that while AI is uniformly across every research report, forecasted to have trillions of dollars of annual impact, that's an abstract thing to think about.

[00:01:11] You know, it's trillions of dollars a lot, but what we know is most marketers are just starting to explore it. And they're starting to try and understand the impact AI can have on their careers. You know, what knowledge do they need to know to stay relevant or to get ahead in their careers?

[00:01:26] And then how can they apply it to their businesses? How can they actually get started with it? And so the idea behind this series is: let's go talk to the CMOs who are thinking about this stuff, sort of the leaders in the industry who are starting to go through it and see how hard it might be, or you know, find the successes, find the challenges and tell their stories.

[00:01:45] And so that's what we're going to do today with John is take an inside look at what it's been like for him to lead this at Brighton Jones, the lessons he's learned and maybe some of his background at Microsoft prior to that. So John, to get us started, I loved a response you gave me in an email when we first started talking about this, you said "the learning curve has been steep,"

[00:02:04] when I asked you about AI adoption. "Though we have unified data, resolved IDs, integrated personalization at a new site, and achieve some goals, the experience has been humbling and far more difficult than anticipated. I would describe myself as a respectful practitioner, as opposed to an expert."

[00:02:20] I loved that paragraph. I hope you don't mind me sharing it, but what did you mean by that? With someone with your background to say "I'm not the expert, but I'm trying," like what, what did you mean when you said that?

[00:02:32] John Dougherty: Yeah, in my career, I've been really fortunate. I've been able to take on some really big challenges and actually have a technical background.

[00:02:40] And and so I just anticipated this would be another big initiative, but it probably is the most complicated I've ever done. And it's a couple of things: first of all, it's the different layers to it. And, you know, first you've got to get data, right, obviously, and then AI. I mean, it's really pretty new, you know, it's like the last 10 or 15 years is really when it's been commercial.

[00:03:00] And so it's just figuring out how to apply it for our business and then last there's, you know, how do you actually activate based on it? And all three pieces are really complicated in and of themselves, but then you also need to get a team of people internally on board and galvanized and I'm really fortunate that I work someplace with a very collaborative culture. That's a really innovative workplace. And then, so all the pieces are there, but even with a great team, that's excited and well-resourced it has still been really difficult every step of the way. I will say though that I think I finally had a peek over the horizon and the potential is, is awesome.

[00:03:38] And it's the direction, at least, our business is going to go, which where personalization and very high client satisfaction is the essence of our business.

[00:03:49] Paul Roetzer: There's a ton of paths to go down there. Why don't you start, give us a little background, Brighton Jones.

[00:03:54] What does the organization do? So as you talk about personalization, understand the kind of personalization you're trying to do there,

[00:04:01] John Dougherty: It's this really innovative company. And it's one of the reasons why I left Microsoft is because I think it's the future of financial services.

[00:04:10] You know some investment companies help with investments. And then some investment companies help with investments and planning. Like Brighton Jones was founded with investments planning and a whole balance sheet management. So that means any asset, any liability we manage.

[00:04:24] Paul Roetzer: Individuals, right like wealth management for individuals?

[00:04:27] John Dougherty: Yeah and so we look at a person's entire balance sheet. And so we factor in, you know, any private investments funding for their kids' education corporate executive benefits just anything that brings complexity to somebody's financial life.

[00:04:40] We manage it. We have these teams with in- house experts, we have in house CPAs, in-house tax lawyers, in-house attorneys so anything somebody needs, we could do through an in-house expert. So we take care of everything. And this new initiative we're launching is called "Beyond the Balance Sheet."

[00:04:58] I'd say a very small part of the industry is offering that now. We're taking it to one more level, which we call beyond the balance sheet and that's helping people align their time and money with their passions and purpose and a lot of people think oh, if I get a lot of money, I'll be happy.

[00:05:15] And what a lot of people find is they get a lot of money and they get complexity and they're less happy. And so there's really there are extras. Like the foundation is great financial management, but you've got to go beyond that to actually figuring out what you can do with that money. To align with your values, what really matters to you.

[00:05:31] Paul Roetzer: So getting more fulfilling basically. You know, I love that.

[00:05:34] John Dougherty: That's what our clients say in the survey but the thing is all of the weight of offering that service has been on our client service teams. And so it's been the lead advisors, the associates, the analysts, the managers they've had to bear that weight just through their personal knowledge of the client. And so what we're starting to do is collect the data and learn the AI and the automation. And so that personalization can be automated based

[00:06:01] Paul Roetzer: On all the data, which again, it's not a niche per se, but with wealth management individuals, what John's talking about here is understanding your customer at a very deep level and letting all the data tell a story that can then be used to personalize their experience.

[00:06:15] And you can do that in any business, in any industry, the same principles hold true.

[00:06:21] John Dougherty: That's right. So our lead advisors just don't have sight lines into the hundreds of attributes that we apply in our data and that our AI looks at. And so we can really better provision our client service teams to serve their clients in countless ways.

[00:06:36] And so it's, we're really taking personalization to the next level through, you know, our. AI initiative, we call it data culture, hashtag data culture, or we refer to it as, you know, data, AI automation, sometimes, you know, just an internal papers. But our, our core offering is about personalization and alignment and by really developing data, AI and automation, where we're better able to do that.

[00:07:02] Paul Roetzer: Well, let's dig into your background a little bit, cause I'm kind of fascinated on a couple of elements here. So the first is your undergrad from Notre Dame was computer applications and economics. How did that get you to a chief marketing officer role? It's not a common thing. I see. When I look at CMO backgrounds.

[00:07:20] John Dougherty: I spent about 10 years at agencies and about 10 years at Microsoft. And that quantitative piece always helped because over my career, you know, digital's become more important. And, and so just the more you can dig into the numbers on that the, the better view of reality we have and the better result you get.

[00:07:37] And At Microsoft, I led audience marketing for Windows worldwide. And again, it was just in that organization, it was super important to show ROI. And so that quantitative background was really beneficial.

[00:07:50] Paul Roetzer: So talk to me about Microsoft. This is one of the paths I wanted to go down.

[00:07:53] One. You, you crossed over at Windows with a mutual friend, Christi Olson. Who's been on the show and is one of my favorite people, if just amazing.

[00:08:00] John Dougherty: No, I, I remember working with Christi and I agree. I couldn't agree more.

[00:08:04] Paul Roetzer: Yeah, we can just talk about Christi for 20 minutes. So go check out Christi's episode if you're new to the podcast, check out Christie. She's fantastic.

[00:08:11] John Dougherty: I listened to that episode. I agree. I thought it was wonderful.

[00:08:15] Paul Roetzer: So you were at Microsoft at a very interesting time in the development of artificial intelligence. And so people who aren't familiar with the timeline, AI is not a new thing.

[00:08:25] It's been around for 70, 80 years as a concept, but there was some pivotal moments in the last, 11 to 12 years with deep learning, the ability. Mainly for machines to do language generation, computer vision. And we're seeing that everywhere in our lives today from, you know, Gmail finishing sentences to, you know, opening phones with your face to you know, integration of GPT-3 into language generation for marketers, all of that stuff, voice assistance, all of these things became possible largely in the last nine years.

[00:08:56] So you were at Microsoft from 2007 as a creative marketing manager to 2016, where he left as senior director of marketing. Was AI being talked about within the walls of Microsoft applied to marketing at that time? I know they were working on language and I know they were applying it to Windows as a product, but were they thinking about it in marketing terms at that point

[00:09:19] John Dougherty: No, it was baked into the product and so marketing talked about the benefits of it, as opposed to thinking about the engineering components that made the benefit possible.

[00:09:29] Okay. And, and we were trying to convey clear, simple messages on differentiated benefits and, some of those you know, we're definitely based on AI, but, but answer your question. No holistically, but yes, indirectly.

[00:09:47] Paul Roetzer: Yeah, it's just like, there's a great book called Genius Makers by Cade Metz, who' s also, we have an episode coming up with Cade Metz on the podcast, but he tells the story of the development of AI in many ways through the lens of Microsoft, Google, Amazon, some of the big players, DeepMind....

[00:10:01] And so that they talk about Microsoft working on AI back in the early two thousands and trying to do with Word and within Windows, like doing language technology. And so I know Microsoft has a very deep history in trying to develop AI. And it's always fascinating to know when did it really become a thing to the wider organization versus this engine that kept trying to be built behind the scenes basically.

[00:10:23] John Dougherty: So I, I read that book and I bought a copy for our, our CEO and gave it to him just because he's very forward looking, he's a visionary. And one of the things that book shows is companies that successfully deploy AI, just pull ahead. It's a sustainable strategic advantage. So thought that book was fascinating.

[00:10:39] I mean, in terms of the detail, and also, like you said, AI has been around for a long time, but it hasn't been commercial that long. And it was just the convergence of the hardware, having a hardware that's accessible. That's powerful enough along with m odels that are effective enough.

[00:10:55] Paul Roetzer: And to your point, kind of moving into the Brighton Jones days, the data we needed more data, we needed the compute power to process that data. So you join Brighton Jones, 2016 from LinkedIn in the VP of marketing role. It talks about lead strategy and execution, segmentation, automation, creative strategy, online marketing.

[00:11:13] So content strategy, brand architecture. As any CMO knows, and even at the VP level, you're touching a lot of areas. So in those early days, at 2016, were you thinking about intelligent automation and stuff? Were you thinking about ways to apply AI or were you just trying to take traditional brute force and get the traditional technology and try and do the basics there?

[00:11:37] John Dougherty: Yeah. And back when I first joined, it was definitely the latter. So the brute force, but again, AI was baked into some products, so , Intercom, our chatbot uses, natural language processing and we use course AI now, which also basically records zoom calls and then analyzes the content.

[00:11:56] Paul Roetzer: Real quick there, so people might be familiar with otter.ai is used on zoom. It's baked into zoom. So if you get a transcript of a call, that's actually using otter.ai, you're saying you have an AI that not only takes the transcription, but analyzes what's in that transcript.

[00:12:12] John Dougherty: Yeah. And what it does is it looks at how much we talked, how much the other person talked, how many, what words came up the most, what the longest monologue was.

[00:12:22] And then it assesses, the call, what went well, what didn't go well, what can we do better? And it's all just in the pursuit of adding more value, just understanding both existing and potential clients better. And I mean, to me, that is what AI is all about.

[00:12:37] Ultimately it's just machine s enhancing our ability to understand something and are better understanding the probability that something will help. I mean to me, that's the crux of it. And it's just a better view of reality both now and forward-looking.

[00:12:52] Paul Roetzer: Really well said it; I agree.

[00:12:54] It's just additive technology, because it's such a great use case. Say you had a thousand calls that you had to previously analyze. There would have been a human doing that, someone would have had to go through all these transcripts and tried to identify this stuff.

[00:13:08] So it's almost like a function that couldn't have even existed prior that you can now scale across every phone call and have all this rich information that you as a CMO can act on that data. Like it's insights that you can actually make tomorrow. Start making relationships better based on this information that you just didn't have previously.

[00:13:29] John Dougherty: And it makes the call a better experience because rather than taking copious notes, the person can really focus on who they're talking to. Because the notes are to your point are being transcribed. So it's just better in a lot of different ways.

[00:13:41] Paul Roetzer: There are all these areas you touch, even now, today as the CMO with segmentation and automation and branding and digital media and events and messaging and inbound marketing.

[00:13:50] And again, like every CMO, there's lots of areas you're involved in. So outside of this call transcription and analysis, what are some other areas where you're seeing AI bring benefits to the personalization and experiences of your customers?

[00:14:07] John Dougherty: Yeah, I think there's a couple of big areas and the first is just like personalized orchestrated journeys.

[00:14:14] And so we look at an audience and then we figure out a bunch of experiences that are relevant to that audience. And then we have a target audience. And so we want somebody who's a MQL to become a sales accepted lead. And so we use AI to orchestrate a personalized journey, to get somebody from MQL to sales accepted lead.

[00:14:35] And so we use our CDP for, orchestrated personalized journeys. Obviously we do segmentation, which is, finding groups that are homogeneous, distinct and actionable that we otherwise wouldn't. And when you're looking across hundreds of attributes, you're just going to find stuff that you otherwise couldn't find.

[00:14:53] We do predictive modeling and so we try to figure out who would be willing, who wants to join one of our community impact circles? We would figure out who do we need to call with regard to, our investment philosophy, like who do we feel like we should update on our investment philosophy and then last, storytelling and that's like using natural language processing across all of our articles before we've thought about it in terms of keywords.

[00:15:17] And that's a really human construct that , a human construct to help get the results that you want in Google. Whereas natural language processing looks at affinities. And so before we would say, you know, 5 29, but there would be several affinities in that One would be education funding my kid's education or caring about my kids.

[00:15:35] So affinities are very different than keywords, and then you combine that with predictive modeling. So once you know what affinities are in an article, and once you know what affinities somebody has, you can make sure you're only serving people articles that they want to see.

[00:15:51] And so it's just better storytelling. So those are the big buckets. I know we'll use it more for measurement later. Honestly, we just haven't been able to get into it because we are still very, very early and I just want to emphasize this experience has been humbling. The learning curve is steep.

[00:16:04] The learning curve has a long, long, long, long way to go; every week is a new gotcha. But I think we finally turned a corner and it's like when I realized the investment is going to be well-worth it. It's just going to really take us to a level that honestly, I just don't think our competitors can go.

[00:16:24] So it's going to give us this real advantage.

[00:16:26] Paul Roetzer: And I love how you frame that, you listen to that list of all the ways you're using it. And there might be some CMOs or marketers listening that are just like, Ooh, wow. Like they're so far ahead of where we are. I'm still trying to understand what natural language processing is.

[00:16:40] But I think your point is so well taken that you're still so early and everyone is still so early in this. So you're approaching it from a really smart perspective where you're thinking about the outcomes you're trying to achieve. Like you're thinking about the consumer, the customer and saying how do I personalize that experience? So we create value at every stage of their journey. How do I predict behaviors and outcomes? So we adapt things we're recommending to them to drive an intended action that creates more value for them. And that's your point earlier about like, it's really about prediction. Your predictive modeling sounds kind of like a geeky term, all it's saying is, "we're just trying to predict a behavior or an action or an outcome." And so there's ways AI can help you build these models to try and predict what a specific cohort or specific group may do. And then that helps you rather than traditional persona project of like, let's go write a three page brief on what this person looks like and you sample the LinkedIn things.

[00:17:36] And then we have these three personas and then our entire content strategy is based on these three personas, which may only make up like 5% of your actual audience. Like AI helps you identify almost a bunch of, hundreds, potentially, of micro personas. And that's the key. To do personalization on three personas isn't going to do anything for you. You have to understand these people and what creates value for them, where their pain points are, what their needs are in real time.

[00:18:03] So two things I want to follow on with that is one is CDP. I know you've mentioned how important that is. So talk to me a little bit about the importance of the CDP early on and what it is and what goes into building a valuable one for marketers that maybe aren't that far along yet with their understanding or adoption of a CDP as a foundation?

[00:18:24] John Dougherty: Yeah, it was critical for us. And we looked at different technology tools. We looked at marketing hubs.

[00:18:30] We looked at different packages that did everything. And then we decided to go just with the kind of best of class product. And so we went with a CDP that's designed for specifically for marketing. And the first stage was just a unified data from Salesforce, Marketo, our website, our phone system, a lot of different sources and that's in March. We're still getting data in. We found that we had to do a lot of work to get the data. So for instance, we found that we were sending the CDP a lot of administrative activity from Marketo, a lot of operational activity from Marketo.

[00:19:06] And so like Marketo was updating records and the CDP thought a client had done something, whereas really it was just Marketo reconciling with something else. And so we had to get all of that noise out because we were wildly misleading our CDP. And then we had to get noise out from other sources.

[00:19:23] And then we arrive at this point where like we knew all the data in was, was well-formed and then it was just resolving datas. Like we found out that clients had emails that we didn't know they had, and they were on a bunch of different devices. And so we had to resolve all those things. So. We understood holistically what a client was doing across all of their touchpoints.

[00:19:42] And then the next thing you know, the data is going to be a constant effort. I mean, it's going to be relentless.

[00:19:49] Paul Roetzer: Who owns that on your team, just out of curiosity, is that a specific role or is that someone trained in a specific area?

[00:19:55] John Dougherty: So we have somebody on the team who led in terms of marketing automation, and then they took this on.

[00:20:00] And so somebody internal, who's done a phenomenal job getting the data well-formed, who's partnered with our Salesforce administrator and our VP of IT. And I mean, the team has just moved fast, been tenacious but every week there's a challenge.

[00:20:14] And so all the data that's in now is real data, but we're but we're realizing we need to get more data in to get the level of effectiveness that we want. And then the next level is analysis. So it's using the AI built into the tool to do segmentation and the CDP is really integrated into our website and we're going to integrate it into our email.

[00:20:39] And so then the third piece is activation and It's just crazy. The degree of service we can provide with personalized one-to-one activation, as opposed to one size fits all or to your points, segmented or by persona. And so those are like the three big pieces. I feel like we've got a good data foundation and we need to do a lot more.

[00:21:00] We're working with data scientists to learn about AI. For instance, we have developed a behavioral score for engagement. And the data scientist's point of view is like " you should only look at data that's three months old. " And so, and whereas other behavioral scores, they think, we can look back indefinitely. For some, there's a behavioral score for the totality of somebody's activity. Obviously you need to look back as far as you have data. And so we're learning things like on some behavioral squares, you have some windows, other behavioral supports, you don't have any, so that's an example of like something that we've learned . It's like we don't put any PII or we've never put in an API or we'd never share the content of an email, but we want to know like if an email was sent to a client because would affect the engagement score.

[00:21:47] And so the CDP wouldn't know what the email was about. But it would know that an email went out, so to get effective engagement scores, we need to include email. We need to include zoom. So we had to figure out how to get structured zoom data into the CDP. And that's something we're doing now or we'll do in the next week or two.

[00:22:03] So again, there's no end to either the work that needs to be done or the potential things that we can do. We try to implement three use cases every quarter. And so one use case would be personalized email. And so the articles that somebody would see within the email would be personalized on click.

[00:22:20] We also can have personalized content in line on the website. So a copy block somebody would see depends on if they're in an audience or if it was based on an affinity. And then there's also modals. And so you can slide something out or roll something down after three seconds based on who somebody is.

[00:22:38] And so those are the sorts of use cases that we implement. And so at Brighton Jones, our structures are that we have people in charge of getting clients and then another category is keeping clients and then getting people and keeping people. And so I am really tightly aligned with the folks in charge of get, keep, and to figure out what use cases to roll out each point.

[00:22:56] One thing I really want to emphasize is I think privacy is going to become a way bigger deal than it is. And so I think that the potential for AI endless, but I think the other shoe is going to be privacy. And so we're getting ahead of that now. Like we're building first party data.

[00:23:11] We're not using third party data and we're really tuned to understanding, like we really listen when people tell us how they want us to use their data. I think data AI automation can be done in a way where it's respectful of privacy and it adds huge value. I think the inverse is true too.

[00:23:31] I think companies can exploit it and those companies who don't add value to the client, but they really explain the data... I think those companies could give everybody a bad name. So as marketers, we need to be ahead of it.

[00:23:42] Paul Roetzer: Couldn't agree more. We're actually... at our Marketing AI Conference in August, we're going to have a workshop on responsible AI, like an optional pre-conference workshop, a three hour workshop where you're going to go in and learn about these exact issues, privacy, first party data, having a framework for responsible AI application within your organization.

[00:24:01] Because I feel it's maybe the most important thing we need to be talking about, but because so few marketers even understand AI and its potential, they don't know to care yet about responsible AI. And I think people like you pushing that conversation is critical because that's been our feeling from day one, in 2019 at our first Marketing AI Conference, I had a panel on the main stage on ethics of AI.

[00:24:24] Because my thing was, I'm going to force feed it to you, like this isn't gonna be a breakout. You're not gonna have a choice to go to this one. This is a main stage thing because you have to expose yourself to these challenges because our whole premise with the conference is more intelligent, more human.

[00:24:40] Yes. We can make marketing better and we can make it smarter and faster. But if we don't make it more human in the process, if you don't use personalization to create better relationships and not take advantage of people, because you have their data, then we're doing this for nothing. Like all we're doing is giving people hacks to do marketing cheaper and get rid of the talent.

[00:25:00] And that's not why I'm doing it. And I could tell you have a similar background and philosophy. It's like, this can go wrong. And once you understand the power of AI and its potential, you also understand how there can be bad actors in its application.

[00:25:14] John Dougherty: Yeah, I couldn't agree more. I I'll go to that session. I sure I have a lot of learn and it's super important.

[00:25:19] Paul Roetzer: So a couple of quick things I want to touch on. You mentioned other areas of the organization that you were working with. Data science and IT-- just kinda touch on that. Like what is the role they're playing in your roadmap and vision for where you need to apply AI and are there any other areas of the organization that you're working with that are enabling adoption and scaling of AI?

[00:25:40] John Dougherty: Yeah. So we're bringing a data scientist on board. Now we've been doing home cooking up until now and And so that's really going to accelerate some of our efforts. Our tech team led the leads implementation of these big tools and it's critical into getting the data into the tools and then making sure it's accurate and well-formed, and it's really, there are no seams. I mean the tech team and the marketing team really are working on the same stuff. It really bleeds over. We're always working with client service to better understand clients and to better understand what use cases we can bring to market.

[00:26:18] We work really closely with finance. And then other parts of the marketing team, obviously, demand gen, automation. It really just kind of permeates everywhere.

[00:26:26] Paul Roetzer: Do you see new roles emerging? Like this is something I've thought about and we wrote about it in our upcoming book a little bit, but it was more theoretical.

[00:26:35] Three years, five years from now, do you see roles existing in a marketing structure that don't even exist today related to AI? Like an AI ops, for example, that's just working with different divisions and departments and analyzing the outcomes, analyzing the needs of the stakeholders and finding ways to drive efficiency through AI --are there, are there roles you're envisioning?

[00:26:56] John Dougherty: Yeah, absolutely. I think you're way ahead of the game. I think that's definitely coming.

[00:27:00] Paul Roetzer: Okay. Yeah. Cause I've even been researching. What are the rules? In what I'm seeing in big enterprises is conversational AI. I've actually seen quite a few people with that in their title, because it does seem like an entry point to a lot of bigger enterprises is conversational.

[00:27:14] I've seen some personalization roles that are having AI integrated into the title. Because there's those obvious entry points of personalization, automation, like you said earlier, "this was someone working on automation, so it was a natural extension that they try and find more efficient ways to do automation."

[00:27:30] And so I think you're going to have this evolution of some titles with people in analytics backgrounds, personalization, automation, but then I think there's just there's whole new roles. I've seen creative AI titles, like people are like director of creative AI where they're, I assume, finding ways to integrate language and vision technology into the development of creative. I don't know exactly what they're doing, but I don't know, I'm fascinated by that. What is the org chart of the future look like? And I don't know of an organization that's got the model right now. I think it's just developmental for everybody.

[00:28:01] John Dougherty: Yeah, I agree. I think at least my experience-- and I'm sure people have had experiences to the contrary-- but my experience was 20 years ago. Leader of the marketing organization was the creative and they would go into the tabernacle and come out with a vision. And then, the TV team would turn it into a :30 and maybe a :60 anthem.

[00:28:21] And then the out of home team would take that idea and put it on signs and airports. And then the event marketing team would take it to trade shows. And then kind of lower in the organization was somebody who was trying to analyze the impact of the different tactics.

[00:28:39] And I think there's been a lot of companies that are struggled with marketing mixed models. I mean, what is the impact of each of these different investments? And so that was the structure 20 years ago. But I think it's flipped. I think that you're going to have to be really technically sound and the person before who was doing the spreadsheets, trying to figure out how to allocate money and what's working and what's not, I think that person is the person at the top of the organization that's going to need that skill. And and then I think that AI is going to determine a lot of like the creative direction. And then I think the creative people are going to take the insights from AI to go build a creative. So I think the pyramid has been flipped or it's in the process of being flipped.

[00:29:21] And I think there's new roles you're talking about. You know, the ones that naturally evolved from that version.

[00:29:27] Paul Roetzer: Yep. Yeah. There's a great book I always recommend to leaders, especially CMOs, called The Algorithmic Leader. It's by Mike Walsh and it's phenomenal. And his whole premise is what you just said.

[00:29:38] You don't have to be able to build the predictive models. You just need to know that they can be built and what they can help you achieve. And then you need to know how to find the people to build them. And I see AI in very much the same way. You don't have to understand the eight common machine learning models, and you don't have to understand exactly how neural nets work or what they do or everything like that.

[00:29:58] You just need to know. AI is capable of making predictions about outcomes and behavior. And it's capable of having human-like abilities to understand language, generate language, see images, generate images. Like if you understand that much, then you're better off than like 90% of other marketers right now.

[00:30:16] And now you can start saying, okay, who do I need to know on my team? Like who she is? It's a data scientist team. Is it the IT team? Who should I be pulling in as a CMO to help me create a vision for how we can build greater personalization and performance into our market?

[00:30:33] John Dougherty: Yeah, I wrote that name of that book down. I'll definitely check it out. I, I totally agree.

[00:30:37] Paul Roetzer: Yeah. And like you said, the people who go get those skills now, three years from now, five years from now, you will be so valuable not only within your organization, but within the industry. It's because there aren't a lot of first movers right now.

[00:30:52] And so you don't have to have the background like you have with the undergrad in this area, you can go get this knowledge from a few online courses. Like you can have a fundamental understanding of AI, data science, tech with a few hours invested. And then maybe once or twice a month, you're doing something to enrich that. I love that your idea of three use cases, a quarter, like do the same thing for your own education.

[00:31:12] Like three things, a quarter, you're going to read a book, listen to a podcast, but something to advance you, and you're going to be further ahead than your peers. There just aren't enough people doing this stuff.

[00:31:22] John Dougherty: Yeah, it really changes the way it worked because rather than people on the team talking about, oh, you know, " do we have an offer slide out from the left after three seconds?"

[00:31:31] Or, when do we present the offer? All of a sudden, you know the machines making that decision.

[00:31:38] Paul Roetzer: Well, I could, we could talk all day on this stuff. Okay, so you've been going through this. What is the advice you would give the CMOs who are listening and they're just trying to get started and maybe they don't have the technical background you had, but what are the one or two things that you would advise them to really give them a heads up and you know for a foot forward as they move into this space.

[00:32:02] John Dougherty: I think that there's a few things that come to mind and I heard the cliche that data's the new oil and I didn't really internalize it. You know, it sounded right. I didn't really think about it too much, but I don't know... recently I began to understand why that is and that it's true. And so the sooner that you start to unify your data and resolve IDs and just really be able to use it as a tool, the better, and so start with a data strategy and you don't need to figure everything out, but just know that at some point in the future, you're going to want access to this data.

[00:32:41] So start now. And so I think that's that's number one. I think number two is it takes time, like what I thought we would do in a month, we're doing in a quarter. It's harder than I thought. And so really coming out with a robust team and a realistic expectation of how long it will take... we did not have that, but luckily we have this culture of continuous improvement and we have a culture of collaboration and so it worked, we were able to move slower and still, we were viewing it as a big success, but just really figure out the time.

[00:33:16] And then last, I've had a lot of big projects where I've written a half a page or one page brief and just got everybody on board and then was able to figure out stuff along the way. Don't do that with this. You'll find every stone in the stream with your shin and it really hurts. Like you really want to have your use cases mapped out beforehand, and you want to know what data you need for those use cases.

[00:33:43] And you want to know what models you want to use, and you've got a really pay attention to the details. This is not like... you will be figuring out things as you go. However you want to understand in detail exactly what you want to accomplish, what your organization needs, and then what data and analysis and activation you need to make that a reality.

[00:34:05] You've got to have it all mapped out beforehand, do not go in with this conceptual idea and hope to solidify it as you go. If you want to better serve clients, then you've got to figure out like what matters to them. And so like, what's your passions, interests, hobbies, data?

[00:34:23] How do you collect that? And how can you work with client service so they understand that they can better serve their clients if we collect this data, then how do we collect it? And where do we store it, structured or unstructured? I mean, you've got to get all that stuff sorted.

[00:34:36] We did as much as we thought we needed to upfront reasonably, but in hindsight, we would have done a lot more and we would have like figured out, we would have really looked at, "okay. This CDP has a native integration. What does that mean?" You know, does that mean that it can write household data to the contact level?

[00:34:54] That stuff's important. I didn't know to think about it beforehand but were we to do it again, I certainly would.

[00:35:00] Paul Roetzer: That's a great three pieces, but I think the final step is that try and think through the knowns. If you've listened to this, I think you understand the importance of the data and if that's not your world and you hear that and you're like," oh, it's like building pivot tables in a spreadsheet,"

[00:35:15] Like I want nothing to do with it, find the person on your team who does. You're going to need that right hand to really help put the data in the right structure and get the right technology partners on board. It might be the technology partners you already have, but I always say, AI is just smarter technology.

[00:35:31] You're still trying to solve people's problems. You're still trying to create value. You're still trying to personalize experiences, but there's smarter ways to do it. And it requires maybe an evolution of your talent internally. Maybe an evolution of the technology you're using, but you may have both of those building blocks already, and you might be able to just train your team with some new capabilities and you might be able to leverage your existing tech with features you didn't know that tech offered. And so it's not like blow up your tech stack and blow up your team. That's not at all what John's saying. It's not what I ever preached. Be strategic about this, realize the opportunity. Don't give it the half page brief, give this the way that it deserves, because it can have an outsized impact on your growth and your success.

[00:36:16] John Dougherty: I couldn't agree more with all those things. Definitely find the person on your team who has a passion for it, and then try to figure out how you can use the technology you have. Absolutely. The reality is you've got all this great data. You just need to capture it and make some sense out of it. That's like the first big step for sure.

[00:36:35] Paul Roetzer: Well, John, this has been amazing. It's kind of everything I hoped for more with this series is. You know, I can research this stuff all day long, but I'm not a CMO. Like I'm not living what you're going through every day. And I think just to hear it firsthand from people like you that can tell the real stories and are willing to be open about the challenges is so valuable to our listeners and our viewers.

[00:36:54] And so thank you on everyone's behalf for sharing what you've shared with us today.

[00:36:59] John Dougherty: I enjoyed chatting with you. I think you're way ahead. I really do. I think this area's going to blow up. It's just a matter of when.

[00:37:06] Paul Roetzer: Awesome. Well, this has been the Marketing AI Show. Be sure to follow or subscribe to the podcast.

[00:37:10] If you're curious about AI and want to continue exploring the ways it can transform your business and career. So thank you again to John and thanks everybody for joining us today. We'll see you soon, John. Appreciate it.