What Marketers Reveal About AI Adoption and ROI Gaps
David Kohl Morgan · April 17, 2026
ai in marketingmeaningful roiai adoption
Introduction to David Kohl
(0:05) Welcome back to another edition of AI Insiders with AI Marketer's Guild. I'm your host David Berkowitz and I'm here with someone who's on my short list of favorite Davids. David Kohl, I got to know very well through his work at TrustX and one of the most
David Kohl's Expertise and Morgan Digital Ventures
(0:31) thoughtful folks looking at areas trust and privacy and areas that are often deprioritized by too many. He's been at the forefront of it for quite a while and now is running Morgan Digital Ventures and came to me with some ideas for
Purpose of the AI Adoption Survey
(0:55) creating some research and trying to see what folks in the marketing and ad industry are thinking about and how they're using AI. You might have seen me share this at different points with the AMG community. Maybe some of you took this survey and wanted to welcome David, hear what you're up to, why you did this, what you found.
Interactive Session Format
(1:25) As always, for those who haven't been here before or been here in a while, keep this super interactive. There are even some we could do some live cuts of the data. Anything anyone wants to go and dive into and explore further, we'll have fun discussion. Welcome David. Thank you, sir. I think the feeling is mutual. You were also on my
David Kohl's Opening Remarks
(1:50) list of favorite Davids. It's a David love fest. Okay, I won't ask where I am on that list, but I'm glad to be anywhere on there. Thanks for having me on. This is going to be fun. We will
Session Agenda and Survey Catalyst
(2:07) definitely do a little talking at the beginning, I'll show you some data, and then I want to reserve plenty of time to play around with the tool in more of a Q&A style. We'll definitely reserve time for that. The catalyst here started after I left TrustX last fall and quickly got back into the market because
AI as a Transformational Catalyst
(2:29) AI seemed like this new transformational catalyst that everyone's embracing in one way or another. I wanted to find out where are people? Where are we on the maturity of playing around to getting value? My consultancy focuses on growth and
Morgan Digital Ventures' Focus
(2:51) differentiation. I help marketers, agencies, publishers create a distinctive value and get value out of that that turns into economic growth, revenue and profit. I saw AI as this catalyst as a tool but wasn't sure where people are in terms of adopting.
Anecdotal Findings on AI Adoption
(3:14) What I heard were three things anecdotally. A very small number of companies, mostly ad tech and martech companies, were talking about AI in terms of a measurable goal. We figured out X is going to help our company and we have initiatives going on. That's the gold standard in getting value from a new
Common Responses to AI Adoption
(3:40) technology. That was a tiny number of folks in my anecdotal conversation sample. Most of the folks I was talking to back in the fall were saying anything from "I'm paralyzed. I literally don't even know where to start. I'm sitting back and watching others." The other group was
Lack of Goal-Oriented AI Strategy
(4:02) "we're playing with AI," or as I like to put it, "we're throwing AI spaghetti against the wall to try to see what sticks." Very few were organizing around a goal, a differentiation and a competitive advantage. The catalyst for doing the survey was to see if the anecdotes aligned with what the rest of the market's saying, or just the
Survey Takeaways Overview
(4:25) David Kohl sample. That's where we started. What I'm going to do, because I'd rather not talk, I'd rather show you, is I'm going to share my screen. Let me share with you a couple of takeaways. I'll start with the stuff that to me wasn't surprising, and I don't
Widespread Experimentation with AI
(4:46) think it'll surprise anyone else. The first thing that wasn't surprising is that whether you're an agency, brand, consultancy, platform, publisher, the predominant responses to the survey, folks are somewhere between experimenting with AI and the very early stages of implementing it as a production
Current AI Adoption Stage
(5:11) day-to-day tool. Not surprising. This survey was conducted between middle of January and the end of February. That makes sense for a month or two ago. Practically speaking, it's probably the same even today. Not surprising. Everyone's between experimenting and implementing.
Content Creation Leads AI Benefits
(5:31) The second thing that I didn't think was surprising is when we asked "where do you rate yourself in terms of maturity by function" with "not using" on this chart on the left in the gray, and "getting measurable ROI" in the orange on the right. Not surprising is that content is king. That's the OG for AI. We all used it for writing. We
Efficiency and Time Savings
(5:58) started playing around for image generation. Not surprising that content was at the top of the list of where folks are getting the most benefit from AI. The third thing that I also felt was not surprising is that to the extent folks are measuring some benefit, if you look at this top section "how AI impact is being measured," time saved and
Measurement Focus on Efficiency
(6:23) efficiency was the top response. About half said that if they're measuring anything, it's that they're getting efficiency and time saved out of AI. Not surprising. I want to come back to the maturity by segment, this is where I got my first head scratch. You'll notice that platform
Agency Self-Assessment Surprise
(6:47) companies – martech and ad tech platforms primarily – rated themselves as perhaps a little further ahead. "We're now implementing." Not surprising. But what did surprise me is that agencies rated themselves so high on the maturity curve. I have two theories here, and they're theories.
Theory 1: Client Pressure
(7:09) One theory is that agencies are under tremendous pressure from their clients to stay ahead. They're investing, they're organizing, and they are themselves driving AI tools in that implementation phase because of their competitive pressure. That's one theory.
Theory 2: Agencies "Drinking the Kool-Aid"
(7:32) Another theory is that agencies are having a little bit of drinking the Kool-Aid. They believe that because they're playing around with AI that they're ahead of the curve relative to other organizations. I thought of that second theory because if you look at the third line here, the 16 companies that responded in
Consultants vs. Agencies Maturity
(7:56) the consulting line, consultants also have pressure from their clients to show that they're advanced, to be ahead of the curve. Consultants seem to be rating themselves further behind, in fact, of the group, the most behind. The question becomes, are consultants too honest relative to agencies
Honesty in Self-Reporting
(8:20) who are a little bit drinking the Kool-Aid? There's no way to know, and we didn't ask people for long form explanations of all their answers. There are some insights that I can show you on what people said, but in the end we have theories. That said, we think it was surprising to see
Open Discussion on Agency AI Maturity
(8:44) agencies at the top. When we get into the dialogue, since I know there are a lot of marketers on this call, I would love to hear how people think about their agencies relative to the maturity of usage of AI. That was one of the surprises. Another one, come back to this chart I showed before. Remember I said time
Lack of AI Impact Measurement
(9:07) saved 45% of respondents say that's the impact AI is making. A quarter of respondents said that we are not measuring the impact of AI. There's a phrase, I'm sure everyone heard, "what gets measured gets done." If you're not measuring meaningfully,
The "Spaghetti Against the Wall" Approach
(9:31) you may not be getting anything done. I think this comes back to that anecdotal evidence I had early on that folks are using AI but they're throwing spaghetti against the wall and seeing what sticks as opposed to focusing on differentiation, focusing on something distinctive that can drive economic value.
Questioning "Time Saved" as a Metric
(9:55) On the measurement front, I've been curious about this for years because the first metric that started to pop with AI usage was this time spent. Even there, I'm not sure how many take that through to real efficiency metrics, especially
Actionable AI Metrics
(10:20) if you're not somewhere like a big tech or somewhere that's laser focused on that. I even wonder with the metric that most people are reporting and saying that this impacts, can they do anything with this? What's actionable based on this as opposed to something where you
Focusing on Business Outcomes
(10:45) see some results come in? David, the last three: customer engagement, revenue growth, and cost reduction are results that could be measured and should come from being more efficient at doing your job. They get insights around customers.
Underutilization of AI in Ad Measurement
(11:07) Let's come back to something for a second. Ad measurement. Many of us are in marketing and media, and ad measurement was rated the second most "not using." The light gray bar on the left. Ad measurement is a great way for AI
AI's Potential in Ad Measurement
(11:32) to show its capabilities. It can take massive data sets and look at insights and use those insights to drive. Where was it? Here it is. To drive things like better customer segmentation or to run media more effectively, which would result in revenue growth. To me, there's a circle here:
Functional Maturity: Ad Measurement & Media Planning
(11:57) we're not focusing on some of the differentiated activities, and it shows up in the functional maturity under things like ad measurement, media planning. If I compare agencies to brands, on the left side you have 14 agencies, on the right side you have 12 companies
Agency vs. Brand AI Usage
(12:24) in the brand or advertiser category. You'll see here that agencies are using it for content creation, so are brands, but to a lesser degree. When you get back into measurement, you would think the agencies would be using it more regularly for ad measurement. You see
Client-Side Measurement Lag
(12:45) it's fourth on the list. We're not seeing client side do that much at all. If I go back to segment maturity, let me pull back. You can see that these were the ones we showed before. If I go to role, you can see that, well, let me pull out the founder role. There's only one respondent there.
Role-Based Self-Assessment
(13:13) Here we go. This is how the roles in the companies thought about themselves. There are some interesting little insights for you here. Bill, do you have a question or are you talking in the background? No, it's the webinar. I'll mute.
Barriers to AI Adoption
(13:37) We're not stuck in a pocket. All right, there we go. Let me cover the last thing on the insights here, and then what I'd love to do is open the floor to ask questions about the data, and we can go through it. The last thing on the insights was the skills, resources, and budgets
Internal Obstacles vs. External Complexity
(14:01) were getting in the way of AI adoption. On the one hand, it makes sense. Everyone's budgets are tight, there aren't enough people. But I would have thought that vendor choices and complexity, which was noted by 40% of respondents, would have been the highest barrier.
Vendor Choice Not the Biggest Barrier
(14:25) Particularly since a lot of folks talked about being paralyzed, stuck. There are, I don't know, hundreds, David? Hundreds of companies that call themselves powered by AI today. I would have thought that one of the major barriers, or much higher, was the
Skills, Resources, and Budgets as Primary Barriers
(14:46) sheer choice and not knowing who to hang your hat with. But in fact, it's internal skills, resources, and budgets. Those are the big ahas that I found. The tool lets us take a look at how one segment compares to another.
Inviting Q&A on Data
(15:08) This is all respondents relative to brands and advertisers. I can do that same look at the functional maturity, all versus brands and advertisers. Rather than me drive, what I'd love to do is pause here and turn over the floor to ask any questions. We can look into the data together.
Sally's Question on Agency Media Planning
(15:33) Sally has a comment in the chat that she would have thought that agencies would be using AI for media planning and buying by definition, as AI builds on the programmatic platforms of a decade ago. I'll show you what we see here. If I compare, let me make sure I got the right
Comparing Agencies to Publishers
(15:54) thing. If I compare agencies, all right, let's not worry about it. I can see what I have all the agencies. It's giving me the roles that I've selected. Agencies are on the left, and on the right side I have publishers,
Ad Measurement and Media Planning Discrepancies
(16:18) to juxtapose the folks that are generally buying media from the folks that are generally selling media. You can see here that ad measurement, media planning are, well, ad measurement I would think both publishers and agencies are doing. We're not doing media so much media planning on the publisher side.
Agencies Lack ROI in Media Planning
(16:39) But on the agency side, you see that ad measurement and media planning are fourth and fifth down the line. In fact, on media planning, nobody on the agency side is calling themselves at a maturity level of delivering actual ROI. This idea of
Client-Agency AI Tool Agreements
(17:00) marketing agencies using AI, they can boast about it in general terms on the one hand, but then they probably have to be smart about it. Get agreements with every client, what AI tools, these are the AI agents to use, and maybe how we do it, something like that.
Client Comfort with AI Tools
(17:23) We think it'll work for you. But then there's going to be reservations from company to company. There's going to be some agreement, and then the companies are going to have their own ideas about which one they're most comfortable with. I'm wondering how that's handled, or is there any
Client-Agency Coordination on AI Tools
(17:46) knowledge about that? Are you asking, Christian, how are the agencies coordinating with their clients on which tools are authorized and which tools should not be used? Not the, I mean tools, methodologies, any of that, because they can boast that, "Oh, we're
Client Dictates AI Tool Preference
(18:11) we're so mature on this," but the client, they're taking their lead from the client in every case. If the client says, "Oh no, we don't like Chat GPT, we think it's the output's terrible. We like Claude or Deep Seek or whatever."
Self-Reporting vs. Client Reality
(18:31) Everybody's got a lot of feelings about AI, and everybody's very nervous. You can boast how mature you are, but that's according to you, right? That's self-reporting. Maybe there's some information about, and it's probably early days, I'm sure it is, but
Tool Agnostic Methods
(18:52) some information about what they do to get a sign-off from the client, or the client's lead in terms of, "Oh, the client trusts Gemini, fine, we'll trust Gemini and run our stuff with Gemini because our methods are tool agnostic."
Q1 Survey: Dominant AI Tools
(19:13) We did ask each respondent for the tools that they're generally using. This was a Q1 survey, and Chat GPT dominated as the default entry point with Claude and Gemini behind them. I have a feeling post everything that happened with the federal government and the change in branding after the Super
Agencies' Wide Tool Diversity
(19:42) Bowl with Claude, I have a feeling that we'd see a different answer today. But within the agencies themselves, agencies reported the widest tool diversity across all kinds of different functions. They reported LLMs, the obvious choices, meeting assistants, synthetic personas, platforms
Client Comfort Drives Tool Diversity
(20:08) for figuring out personas, for figuring out to get to a readiness for a change from traditional search to AI overviews. Then they had a couple of the larger holding companies were developing their own hubs. Christian, to your point, there's probably a recognition that not all clients are going to be comfortable with
Unknown Client Approval Process
(20:31) all tools, which is why they have such a wide diversity of tools in house. I don't know the actual process for getting approval or validation within the client base. In the survey tool itself, which you
Accessing Survey Tool for Details
(20:49) can get access to, you can read through not only the summary by company type, but individual responses by company type to what they're using. I think you'll see that the diversity within the agencies is strong. Thanks. I just, when marketers market themselves, I'm
Skepticism Towards Marketers Marketing Themselves
(21:13) I'm leery. That's all. When marketers market themselves or when agencies market themselves? Agencies market themselves. Marketers marketing myself is terrible. It's terrible.
Brooke's Question on Agency Barriers
(21:34) I don't trust any marketer to market themselves. Every marketer needs a marketer to market the other person. That's it. Thanks. David, it's Brooke Pets, if you can hear me. I wanted to ask a follow-on question or share a thought on this specific to the media planning and ad measurement. Thinking again, these barriers
Agencies and ROI Demonstration
(21:52) specific to agencies, why are they not? Is it barriers on the client side? I would expect agencies to be creating use cases, case studies that demonstrate how using AI the way they're using it, the tools, regardless of the tools they're using, but with their methodologies, how it is driving marketing performance, how it is
Lack of Platform Adoption in Agencies
(22:12) delivering ROI, and use that as a sales tool. I'd be shocked if they aren't doing that. I haven't seen it yet. Nested in there is my question: if the agencies aren't wrapping around certain platforms and tools for media planning and ad measurement, why is that? Maybe they are, but it looks
Explaining Agency Lag in Ad Measurement
(22:31) nebulous. I don't have a good reason why agencies aren't further ahead on ad measurement and then using the ad measurement for planning and buying decisions. Brooke, it does seem intuitive to me that that should be front and center, particularly for media agencies, which I'm sure many of these agencies are.
Theory: Difficulty Accessing Raw Data
(22:53) I have a theory on that coming out of TrustX. For those of you who don't know, I ran an ad tech platform for almost 10 years before I came back to my consulting business in Merkle Digital. My theory on this is that it's harder than we all believe to get a hold of the raw data to then use in the AI models for things like measurement, analysis, and
Raw Data Access Challenge
(23:21) optimization. It's very possible that the agencies would like to be further ahead here, but I will tell you that we made our raw data at TrustX available to all our clients, whether you're a buyer or a seller, but very few clients asked for
LLMs and Data Analysis Potential
(23:41) it because it's a massive amount of data, and at the time, they didn't know what they could do with it because they didn't have the power of LLMs. Now we have the power of an analysis tool where I can chuck the data into Claude or otherwise and say, "Tell me X, Y, and Z." But it's still a massive amount of data,
Data Accessibility as a Barrier
(24:04) and it's still hard to get at the data because most ad tech companies are not like TrustX was, where we freely would give access to the data. Brooke, I'm not sure if that's the answer, but it is a theory as to why folks are further behind on measurement and then using the measurement for optimization and planning. I think
Financial Services Data Barriers
(24:25) that makes sense. I come from a financial services background, highly regulated, so getting the lower funnel data, that's where there are barriers. The agencies are going to get ad response, but often combining that with what happens afterwards is difficult. Thanks. I will say something about the marketers
D2C Marketers' AI Advantage
(24:46) through anecdotal discussions. Marketers that are in a direct-to-consumer or direct-to-customer type of environment. Folks in retail, folks that are OEMs that sell directly on their
Unleveraged Potential for D2C Brands
(25:03) sites, or that frequently have engagement directly within their own environments – those are the organizations that I think could easily put AI to work to get ahead on things like who are their personas, who's coming to my site, how do I use that data to improve the odds or improve the effectiveness and efficiency of my marketing. Marketers have the ability to take some
Marketers Missing AI Opportunity
(25:26) of this on by themselves, even if the agencies don't have the data with which to do it, or aren't doing it for whatever reason. I feel like it's a miss. I feel like the marketing segment, the actual advertisers, aren't embracing some of these capabilities, which are now feasible and easier to do because you always had the data,
Obstacles: Skills and Budget (81%)
(25:53) but now you have a tool set that allows you to make a human query, and that query can give you insightful results. David, this is great. I have a metaphysical question. I wanted to focus in on that 81% which you said on the company side, the biggest obstacle is skills, budget. My question is why? If you imagine
Top-Down AI Push vs. Investment
(26:14) sometimes when you talk to people on the client side, CEOs, want to make AI a focus. I can understand the skills not being there, but are they not investing the dollars? Is it a paper tiger? Curious on your thoughts on that. I don't have a good answer to your question, as much as I would love to
Future Investments and Priorities
(26:35) have had a good answer to your question. We do have a summary next steps as part of this tool, and again, you can look at it either by individual responses or by summary by company type. It talks about what the companies want to do. There's also on this, sorry, it's the priorities tab where folks are making planned investments.
Low Investment in Ad Measurement
(26:58) Here you have how you're measuring, and here you have more of the responses. Planned investments in more automation, in building products. Sorry, strategic priorities there, and any investments in business development strategy. Here again, sorry to keep coming back to this ad measurement, I'm
Strategic Priorities for Ad Measurement
(27:19) so surprised how low on the list investments in ad measurement are when we saw previously that that's an area where it's in my view behind the times. I unfortunately don't have the answer to why, but we can at least see what the strategic priorities and planned investments are between this screen and by reading some of the
Survey Update Frequency
(27:46) next steps. I was curious, how often do you plan on updating this? It's fair, a year from now or even 6 months from now, it may be different once people catch up to their aspirations. I am thinking of doing it in Q3. This was the Q1 survey. Let's let it
Effort Behind the Survey
(28:04) roll for 6 months and do another one. I probably will do it. It's not complicated and expensive, but it does take a lot of time. We definitely had marketing between David and I. For those of you who responded on this call, we're grateful and appreciative. Thank you for that. On the topic of
Separating Media Planning and Buying
(28:25) continuing the survey into Q3, the one thing I was interested, I was surprised not to see more media planning usage for this. I'm wondering if it might be better to separate planning from buying/activation because especially at the holding companies and the large independents, those are two different teams. I think that from a
Challenges in AI for Media Planning
(28:44) buying standpoint, there's probably more accessibility to things that might help you buy one channel or one provider versus another. Whereas on the media planning front, there's a mixing chemicals between strategy and planning, and those lines always blur. It's probably harder for them to nail down AI solutions in
Internal Platforms and Usage
(29:07) that area yet. I am seeing it especially at the holding companies. They do have their own internal platforms, and they're giving the strategists and the planners plenty of tools to use. Whether or not they are using them to full advantage is another question. Walt, I think that's an excellent idea, and I would do it in three
Future Survey: AI Agents
(29:24) ways. I think I'd have media planning separate from media buying and separate from measurement. I think that it's likely by Q3 we'll see further maturity on all that, but what I'm interested in finding out in Q3 is how much folks are using agents, which were
Agentic Buying in Q3
(29:44) a theoretical idea in Q1, but I think by Q3 we'll see agentic buying going on. Walt, I'm going to call you so you can do this with us next time. You'll help us set this up. I would appreciate the input on little subtleties. Looking forward to it. There's another question from Sally
Sally's Question: Human Decision-Making
(30:04) in terms of any stats in the study regarding the human decision-making and problem-solving aspects, and perhaps teams can't agree on how to use AI effectively in media planning. It's a great question. Sally, do you want to add a little color to that? I want to make sure I'm focusing on what you're asking. Sure, and I am not a media planner,
B2B Marketing Context
(30:24) but I've done a lot of B2B marketing for the creative and media space. As I was writing to David the other day, having done a lot of strategic discovery sessions with corporate teams, to then produce brand positioning and content and marketing,
Executive Team Disagreement
(30:46) usually people on an executive team don't always agree. They come with a completely different perspective. The tech guy might have one idea, and the saleswoman has a completely other idea on a priority, for example, any generic priority. I guess that was driving the question, and
AI as a "Life Easiness" Tool
(31:08) somebody else mentioned it, too. Matthew in his comment, "Are people throwing spaghetti at the wall, make my life easier?" In the absence of being able to agree, "Wow, this is what this can do for our company or category," in this case media
Data Slices: Segment Maturity by Role
(31:26) planning. Let me show you two slices of the data that may help us see some answers. I don't have the direct answer, but let's look at some data together. The first one is what I've got on screen now, which is generally segment maturity by role. The roles are
Leadership vs. Executive Perception
(31:47) not radically different. Yes, the marketing and media leadership – these are people that day-to-day are in that function – rate themselves a little lower on the curve relative to let's call the executive management team. I put the two top bars, founders and executive management, together,
Day-to-Day Roles' Realism
(32:07) then the two on the bottom, product and sales. There's a little bit of, maybe the folks that are day-to-day are more realistic about where they are, or because they're in the weeds, they're seeing the maturity better than the other folks on the team. That's a possibility. I'm showing you the
Agency vs. Brand Functional Comparison
(32:27) data, and we're ranking up ideas. Let's go to something that could be a little more interesting here. I'm going to put on the left, because we're heavy here on marketers and we're talking about agencies, this is the comparison of all the functions. On the left are agencies, on the right
Executive vs. Marketing Leadership (Agencies)
(32:47) are brands. Then what I'm going to do is I'm going to select executive management and founders on the left, and I'm going to put in marketing and media leadership on the right. Let's see what the data shows us. The data is showing us that the executive teams feel they're further along on maturity related to
Executive View on AI Impact
(33:13) content, using AI to generate content, to help with strategic decision-making and strategic planning. Their view is that marketing ops is getting returns on investments. Let's look at the same three on the right. The people that do this every day, marketing and media leadership, these are people that said they're in the day-to-day
Executive vs. Day-to-Day Roles in Agencies
(33:36) marketing or media functions. This is inside advertisers. Let's do this. Let's look at it first in the same kind of company. This is folks at the executive level in the agencies versus folks who do it every day in the marketing and media departments at agencies. That's an interesting view.
Filter Correction and Small Sample Size
(33:59) Yeah. Now let's look. Sorry. I apologize. I had the wrong filter on. This is the agency. That's because n equals 1. There was only one person in the agency that called themselves marketing and media leadership, probably because they figured that means a marketer side.
Agency Leadership vs. Brand Marketing Roles
(34:17) Maybe the better is here. Let's do this. You all have to excuse as I move my cursor around. I'm trying to find a good comparison. Here on the left you have the agencies, and on the right you have brands, but you have people in marketing at brands. At the agency side I'm
Divergent Organizational Opinions
(34:38) looking at people on the call themselves the leadership team, and since it's an agency, they're all in media. I'm not saying I know an answer, but I'm showing you that the data shows interesting opinions that are different between different people in the organization. There are still a lot of people not
Low ROI in Strategy Use
(34:56) using it for strategy. On the strategy, n equals 1, there's only one person getting delivery on ROI. We had 10 respondents in this category, so 10%. You guys see my cursor, right? So you had the n equals 10, and only one respondent said most say that
Exploring the Data Tool
(35:18) we're in the implementation category, which is the blue. I encourage everyone to get the tool, to play with this. I got to tell you, there are so many slices here that I've played with it now for a month, I discover stuff all the time by looking at the data.
Accessing the Tool Link
(35:39) I think that's fascinating, we could juxtapose these two. If you do marketing and agencies together, you get a more interesting view. Where can we find this? Dave, will you resend the link at the end? I would put that in here again. David's Brooke again, a
Suggestion for Future Data Segmentation
(36:03) reaction to this. It's interesting, and to delve further, in the future, if there's a way to make sure on the brand side that you have brand advertisers who follow a traditional model, relying on their agency partners, versus a
In-House vs. Agency Model Data
(36:22) sunset who run their media planning and buying and measurement in-house. That may be more B2B, but it struck me that might be interesting to see that lens that may be further down the road, but I would find that interesting. That would be interesting. We had 15 brands. Let me see, what's the n here?
Data Limitations on Brand Models
(36:45) 11. 11 out, wait, hold on. I had some filters. 12 out of the 55 were, I'll call this a brand advertiser. Brooke, we don't know how many of those were in-house versus more traditional agency model. We didn't ask the question. It's a great question to ask.
Question on Publicis Group Participation
(37:04) I was wondering, and you don't have to say this if this is proprietary. I'm a Publicis Group veteran. If any Publicis media folks, Publicis Group is so laser-focused on using technology and I've talked to them recently. I'm surprised, that's all. Sally, we
Anonymity and Agency Self-Assessment
(37:25) collected the data anonymously on purpose to ensure that folks would be honest. Which is why I was a little surprised about the agencies putting themselves at the top. I figured if they're being honest, they wouldn't drink the Kool-Aid in the survey. I don't know who responded. With a very
No Publicis Respondents Identified
(37:46) small number of exceptions, we got a handful of folks who put their name on the list to get a preview. But none of the folks on the preview list had a Publicis email address. Wow, this is neat. Very cool.
Claude's Role in Tool Creation
(38:04) Not for nothing, but I am not a software engineer. I know about that. I said, "All right, I'm going to use the tools." I fired up Claude and I said, "Here are the results. I'm trying to create a tool." What you see on screen was probably two to three hours of work total.
Claude's User-Friendly Capabilities
(38:28) The initial tool took about 15 minutes. Then there was some tweaking, then I learned how to get it on my website, then I learned how to change the colors, then I learned how to create the comparison view. All of it was Claude teaching me and doing it for me.
Claude's Accessibility for Non-Developers
(38:50) For those of you who haven't experimented, a total non-techy non-developer was able to do this. I'm impressed with Claude for that. Was this Claude code, or the 20 bucks a month version? Yeah, 20 bucks a
Data Source: Typeform CSV
(39:12) month. Where did you upload the spreadsheet? What was the question, Khan? What was the source? Was it a spreadsheet? We did the survey in Typeform, and Typeform gives you the ability to download all the results in CSV. Spreadsheet, basically a spreadsheet. I was having trouble in the
Claude for Multi-Dimensional Analysis
(39:33) initial analysis trying to slice and dice it, because it's multi-dimensional. You have functions. These are the different kinds of respondents here. I turned to Claude and I said, "Here's what I'm trying to do." I said, "Well, think of this." Mike, I love you, man.
Recruiting for Future Surveys
(39:53) I did a similar project. It's powerful. My BFF. David, I want to share one other thought as you plan for your next round of this. Sounds like it's a quarterly basis. I don't know if you looked into this, but for recruiting some of the groups within the ANA Association of National
ANA as a Recruitment Source
(40:14) Advertisers, they have an AI Marketers Group, and they have digital media groups. They might have an interest in not only results, but getting their brand partners to participate. I think it's great what you built, and I'd love to see this evolve.
Mixed Trade Association Responses
(40:30) Thank you. I'm not going to knock any particular trade association, but I reached out to them all. Some were, "Absolutely. We'll be happy to send a note out and get people in. We'd love to see the results." Some were, "We don't do that."
Final Insight Before Break
(40:48) It doesn't make sense to me, because I think this is insightful and I appreciate everyone on this call agreeing, but we had mixed reviews from my friends in the trade associations. I want to show one more thing before we break. Let me put one more screen up online.
Importance of Measurement & Consultation Background
(41:11) I think this is important for folks on this call to think about. I said earlier that what gets measured gets done. Before I created TrustX in 2016, I spent 25 years in management consulting at Ernst & Young and Price Waterhouse. Our pedigree was what I'm
The Differentiation Model
(41:34) doing now at Morgan Digital, which is to help companies differentiate and grow. We used to use a model like this. I'll explain it. It's tuned to AI, but the titles on the bottom could be any transformational technology. It's looking at if there's a catalyst,
Catalyst: AI and Gen Tech
(41:57) if there's something happening in the market – the something now is AI and gen tech – you look at what's available now that we know works, what's coming soon, and what's in the future. That's the green circle on the right. It could have been
Value Drivers and Differentiation
(42:14) AI, it could have been programmatic, which is what we were calling it 10 years ago. It could have been digital 20 years ago. On the left is value drivers for your business. It's how you differentiate. It's how you get ahead. It's how you drive growth. There are high value
Identifying High-Value AI Applications
(42:34) differentiators, and medium, and low. What we do, what consultancy does at Morgan Digital, and what I used to do in my days at PwC and EY, is you look for that red star at the top. You say, "If I know this technology is capable, and if I can apply it right here in this high value area,
Lack of Strategic AI Implementation
(42:59) I can get ahead." My most frustrating takeaway from doing this survey is that very few companies were looking for where their red star is. They're playing. They're looking at gen tech. They're looking at AI. They're throwing spaghetti against the wall, but they're not thinking about it as
AI as a Strategic Imperative
(43:20) a strategic imperative to get ahead. Thankfully, I have some clients who are paying me to do that, but the market was not doing it on their own. I leave you today with this picture and say, "Think about this in your businesses. How do I merge the capabilities with what's going to drive growth
Final Call to Action
(43:46) in my company?" That's ultimately what this is all about, and is why we have folks like you here to help us figure that out. I encourage everyone to sign up, get the tool. If you see something in the data that we uncover, or that you
Seeking Community Feedback
(44:03) "Wow, this is interesting. Have you thought about it?" Send David and me a note and say, "Have you thought about this?" Everybody looks at this stuff differently, and we're always looking for insights. We appreciate your feedback to help us learn more.
Thank You and Contact Information
(44:21) Sure. We'll appreciate all this, David. I appreciate all the thoughtful questions as always. You have the survey link, and I shared David's LinkedIn as well. Hopefully, all can get in touch for thoughts and questions,
Upcoming Sessions and Events
(44:38) and ways to evolve this further. We've got things scheduled now through the end of June. Some terrific sessions coming up. Also, let me know if anyone's going to be at Possible. I'll be there with Mark Teschler this year. Look forward to
Closing Remarks
(44:59) seeing you next week and beyond. Thanks, everyone.
(0:05) Welcome back to another edition of AI Insiders with AI Marketer's Guild. I'm your host David Berkowitz and I'm here with someone who's on my short list of favorite Davids. David Kohl, I got to know very well through his work at TrustX and one of the most
David Kohl's Expertise and Morgan Digital Ventures
(0:31) thoughtful folks looking at areas trust and privacy and areas that are often deprioritized by too many. He's been at the forefront of it for quite a while and now is running Morgan Digital Ventures and came to me with some ideas for
Purpose of the AI Adoption Survey
(0:55) creating some research and trying to see what folks in the marketing and ad industry are thinking about and how they're using AI. You might have seen me share this at different points with the AMG community. Maybe some of you took this survey and wanted to welcome David, hear what you're up to, why you did this, what you found.
Interactive Session Format
(1:25) As always, for those who haven't been here before or been here in a while, keep this super interactive. There are even some we could do some live cuts of the data. Anything anyone wants to go and dive into and explore further, we'll have fun discussion. Welcome David. Thank you, sir. I think the feeling is mutual. You were also on my
David Kohl's Opening Remarks
(1:50) list of favorite Davids. It's a David love fest. Okay, I won't ask where I am on that list, but I'm glad to be anywhere on there. Thanks for having me on. This is going to be fun. We will
Session Agenda and Survey Catalyst
(2:07) definitely do a little talking at the beginning, I'll show you some data, and then I want to reserve plenty of time to play around with the tool in more of a Q&A style. We'll definitely reserve time for that. The catalyst here started after I left TrustX last fall and quickly got back into the market because
AI as a Transformational Catalyst
(2:29) AI seemed like this new transformational catalyst that everyone's embracing in one way or another. I wanted to find out where are people? Where are we on the maturity of playing around to getting value? My consultancy focuses on growth and
Morgan Digital Ventures' Focus
(2:51) differentiation. I help marketers, agencies, publishers create a distinctive value and get value out of that that turns into economic growth, revenue and profit. I saw AI as this catalyst as a tool but wasn't sure where people are in terms of adopting.
Anecdotal Findings on AI Adoption
(3:14) What I heard were three things anecdotally. A very small number of companies, mostly ad tech and martech companies, were talking about AI in terms of a measurable goal. We figured out X is going to help our company and we have initiatives going on. That's the gold standard in getting value from a new
Common Responses to AI Adoption
(3:40) technology. That was a tiny number of folks in my anecdotal conversation sample. Most of the folks I was talking to back in the fall were saying anything from "I'm paralyzed. I literally don't even know where to start. I'm sitting back and watching others." The other group was
Lack of Goal-Oriented AI Strategy
(4:02) "we're playing with AI," or as I like to put it, "we're throwing AI spaghetti against the wall to try to see what sticks." Very few were organizing around a goal, a differentiation and a competitive advantage. The catalyst for doing the survey was to see if the anecdotes aligned with what the rest of the market's saying, or just the
Survey Takeaways Overview
(4:25) David Kohl sample. That's where we started. What I'm going to do, because I'd rather not talk, I'd rather show you, is I'm going to share my screen. Let me share with you a couple of takeaways. I'll start with the stuff that to me wasn't surprising, and I don't
Widespread Experimentation with AI
(4:46) think it'll surprise anyone else. The first thing that wasn't surprising is that whether you're an agency, brand, consultancy, platform, publisher, the predominant responses to the survey, folks are somewhere between experimenting with AI and the very early stages of implementing it as a production
Current AI Adoption Stage
(5:11) day-to-day tool. Not surprising. This survey was conducted between middle of January and the end of February. That makes sense for a month or two ago. Practically speaking, it's probably the same even today. Not surprising. Everyone's between experimenting and implementing.
Content Creation Leads AI Benefits
(5:31) The second thing that I didn't think was surprising is when we asked "where do you rate yourself in terms of maturity by function" with "not using" on this chart on the left in the gray, and "getting measurable ROI" in the orange on the right. Not surprising is that content is king. That's the OG for AI. We all used it for writing. We
Efficiency and Time Savings
(5:58) started playing around for image generation. Not surprising that content was at the top of the list of where folks are getting the most benefit from AI. The third thing that I also felt was not surprising is that to the extent folks are measuring some benefit, if you look at this top section "how AI impact is being measured," time saved and
Measurement Focus on Efficiency
(6:23) efficiency was the top response. About half said that if they're measuring anything, it's that they're getting efficiency and time saved out of AI. Not surprising. I want to come back to the maturity by segment, this is where I got my first head scratch. You'll notice that platform
Agency Self-Assessment Surprise
(6:47) companies – martech and ad tech platforms primarily – rated themselves as perhaps a little further ahead. "We're now implementing." Not surprising. But what did surprise me is that agencies rated themselves so high on the maturity curve. I have two theories here, and they're theories.
Theory 1: Client Pressure
(7:09) One theory is that agencies are under tremendous pressure from their clients to stay ahead. They're investing, they're organizing, and they are themselves driving AI tools in that implementation phase because of their competitive pressure. That's one theory.
Theory 2: Agencies "Drinking the Kool-Aid"
(7:32) Another theory is that agencies are having a little bit of drinking the Kool-Aid. They believe that because they're playing around with AI that they're ahead of the curve relative to other organizations. I thought of that second theory because if you look at the third line here, the 16 companies that responded in
Consultants vs. Agencies Maturity
(7:56) the consulting line, consultants also have pressure from their clients to show that they're advanced, to be ahead of the curve. Consultants seem to be rating themselves further behind, in fact, of the group, the most behind. The question becomes, are consultants too honest relative to agencies
Honesty in Self-Reporting
(8:20) who are a little bit drinking the Kool-Aid? There's no way to know, and we didn't ask people for long form explanations of all their answers. There are some insights that I can show you on what people said, but in the end we have theories. That said, we think it was surprising to see
Open Discussion on Agency AI Maturity
(8:44) agencies at the top. When we get into the dialogue, since I know there are a lot of marketers on this call, I would love to hear how people think about their agencies relative to the maturity of usage of AI. That was one of the surprises. Another one, come back to this chart I showed before. Remember I said time
Lack of AI Impact Measurement
(9:07) saved 45% of respondents say that's the impact AI is making. A quarter of respondents said that we are not measuring the impact of AI. There's a phrase, I'm sure everyone heard, "what gets measured gets done." If you're not measuring meaningfully,
The "Spaghetti Against the Wall" Approach
(9:31) you may not be getting anything done. I think this comes back to that anecdotal evidence I had early on that folks are using AI but they're throwing spaghetti against the wall and seeing what sticks as opposed to focusing on differentiation, focusing on something distinctive that can drive economic value.
Questioning "Time Saved" as a Metric
(9:55) On the measurement front, I've been curious about this for years because the first metric that started to pop with AI usage was this time spent. Even there, I'm not sure how many take that through to real efficiency metrics, especially
Actionable AI Metrics
(10:20) if you're not somewhere like a big tech or somewhere that's laser focused on that. I even wonder with the metric that most people are reporting and saying that this impacts, can they do anything with this? What's actionable based on this as opposed to something where you
Focusing on Business Outcomes
(10:45) see some results come in? David, the last three: customer engagement, revenue growth, and cost reduction are results that could be measured and should come from being more efficient at doing your job. They get insights around customers.
Underutilization of AI in Ad Measurement
(11:07) Let's come back to something for a second. Ad measurement. Many of us are in marketing and media, and ad measurement was rated the second most "not using." The light gray bar on the left. Ad measurement is a great way for AI
AI's Potential in Ad Measurement
(11:32) to show its capabilities. It can take massive data sets and look at insights and use those insights to drive. Where was it? Here it is. To drive things like better customer segmentation or to run media more effectively, which would result in revenue growth. To me, there's a circle here:
Functional Maturity: Ad Measurement & Media Planning
(11:57) we're not focusing on some of the differentiated activities, and it shows up in the functional maturity under things like ad measurement, media planning. If I compare agencies to brands, on the left side you have 14 agencies, on the right side you have 12 companies
Agency vs. Brand AI Usage
(12:24) in the brand or advertiser category. You'll see here that agencies are using it for content creation, so are brands, but to a lesser degree. When you get back into measurement, you would think the agencies would be using it more regularly for ad measurement. You see
Client-Side Measurement Lag
(12:45) it's fourth on the list. We're not seeing client side do that much at all. If I go back to segment maturity, let me pull back. You can see that these were the ones we showed before. If I go to role, you can see that, well, let me pull out the founder role. There's only one respondent there.
Role-Based Self-Assessment
(13:13) Here we go. This is how the roles in the companies thought about themselves. There are some interesting little insights for you here. Bill, do you have a question or are you talking in the background? No, it's the webinar. I'll mute.
Barriers to AI Adoption
(13:37) We're not stuck in a pocket. All right, there we go. Let me cover the last thing on the insights here, and then what I'd love to do is open the floor to ask questions about the data, and we can go through it. The last thing on the insights was the skills, resources, and budgets
Internal Obstacles vs. External Complexity
(14:01) were getting in the way of AI adoption. On the one hand, it makes sense. Everyone's budgets are tight, there aren't enough people. But I would have thought that vendor choices and complexity, which was noted by 40% of respondents, would have been the highest barrier.
Vendor Choice Not the Biggest Barrier
(14:25) Particularly since a lot of folks talked about being paralyzed, stuck. There are, I don't know, hundreds, David? Hundreds of companies that call themselves powered by AI today. I would have thought that one of the major barriers, or much higher, was the
Skills, Resources, and Budgets as Primary Barriers
(14:46) sheer choice and not knowing who to hang your hat with. But in fact, it's internal skills, resources, and budgets. Those are the big ahas that I found. The tool lets us take a look at how one segment compares to another.
Inviting Q&A on Data
(15:08) This is all respondents relative to brands and advertisers. I can do that same look at the functional maturity, all versus brands and advertisers. Rather than me drive, what I'd love to do is pause here and turn over the floor to ask any questions. We can look into the data together.
Sally's Question on Agency Media Planning
(15:33) Sally has a comment in the chat that she would have thought that agencies would be using AI for media planning and buying by definition, as AI builds on the programmatic platforms of a decade ago. I'll show you what we see here. If I compare, let me make sure I got the right
Comparing Agencies to Publishers
(15:54) thing. If I compare agencies, all right, let's not worry about it. I can see what I have all the agencies. It's giving me the roles that I've selected. Agencies are on the left, and on the right side I have publishers,
Ad Measurement and Media Planning Discrepancies
(16:18) to juxtapose the folks that are generally buying media from the folks that are generally selling media. You can see here that ad measurement, media planning are, well, ad measurement I would think both publishers and agencies are doing. We're not doing media so much media planning on the publisher side.
Agencies Lack ROI in Media Planning
(16:39) But on the agency side, you see that ad measurement and media planning are fourth and fifth down the line. In fact, on media planning, nobody on the agency side is calling themselves at a maturity level of delivering actual ROI. This idea of
Client-Agency AI Tool Agreements
(17:00) marketing agencies using AI, they can boast about it in general terms on the one hand, but then they probably have to be smart about it. Get agreements with every client, what AI tools, these are the AI agents to use, and maybe how we do it, something like that.
Client Comfort with AI Tools
(17:23) We think it'll work for you. But then there's going to be reservations from company to company. There's going to be some agreement, and then the companies are going to have their own ideas about which one they're most comfortable with. I'm wondering how that's handled, or is there any
Client-Agency Coordination on AI Tools
(17:46) knowledge about that? Are you asking, Christian, how are the agencies coordinating with their clients on which tools are authorized and which tools should not be used? Not the, I mean tools, methodologies, any of that, because they can boast that, "Oh, we're
Client Dictates AI Tool Preference
(18:11) we're so mature on this," but the client, they're taking their lead from the client in every case. If the client says, "Oh no, we don't like Chat GPT, we think it's the output's terrible. We like Claude or Deep Seek or whatever."
Self-Reporting vs. Client Reality
(18:31) Everybody's got a lot of feelings about AI, and everybody's very nervous. You can boast how mature you are, but that's according to you, right? That's self-reporting. Maybe there's some information about, and it's probably early days, I'm sure it is, but
Tool Agnostic Methods
(18:52) some information about what they do to get a sign-off from the client, or the client's lead in terms of, "Oh, the client trusts Gemini, fine, we'll trust Gemini and run our stuff with Gemini because our methods are tool agnostic."
Q1 Survey: Dominant AI Tools
(19:13) We did ask each respondent for the tools that they're generally using. This was a Q1 survey, and Chat GPT dominated as the default entry point with Claude and Gemini behind them. I have a feeling post everything that happened with the federal government and the change in branding after the Super
Agencies' Wide Tool Diversity
(19:42) Bowl with Claude, I have a feeling that we'd see a different answer today. But within the agencies themselves, agencies reported the widest tool diversity across all kinds of different functions. They reported LLMs, the obvious choices, meeting assistants, synthetic personas, platforms
Client Comfort Drives Tool Diversity
(20:08) for figuring out personas, for figuring out to get to a readiness for a change from traditional search to AI overviews. Then they had a couple of the larger holding companies were developing their own hubs. Christian, to your point, there's probably a recognition that not all clients are going to be comfortable with
Unknown Client Approval Process
(20:31) all tools, which is why they have such a wide diversity of tools in house. I don't know the actual process for getting approval or validation within the client base. In the survey tool itself, which you
Accessing Survey Tool for Details
(20:49) can get access to, you can read through not only the summary by company type, but individual responses by company type to what they're using. I think you'll see that the diversity within the agencies is strong. Thanks. I just, when marketers market themselves, I'm
Skepticism Towards Marketers Marketing Themselves
(21:13) I'm leery. That's all. When marketers market themselves or when agencies market themselves? Agencies market themselves. Marketers marketing myself is terrible. It's terrible.
Brooke's Question on Agency Barriers
(21:34) I don't trust any marketer to market themselves. Every marketer needs a marketer to market the other person. That's it. Thanks. David, it's Brooke Pets, if you can hear me. I wanted to ask a follow-on question or share a thought on this specific to the media planning and ad measurement. Thinking again, these barriers
Agencies and ROI Demonstration
(21:52) specific to agencies, why are they not? Is it barriers on the client side? I would expect agencies to be creating use cases, case studies that demonstrate how using AI the way they're using it, the tools, regardless of the tools they're using, but with their methodologies, how it is driving marketing performance, how it is
Lack of Platform Adoption in Agencies
(22:12) delivering ROI, and use that as a sales tool. I'd be shocked if they aren't doing that. I haven't seen it yet. Nested in there is my question: if the agencies aren't wrapping around certain platforms and tools for media planning and ad measurement, why is that? Maybe they are, but it looks
Explaining Agency Lag in Ad Measurement
(22:31) nebulous. I don't have a good reason why agencies aren't further ahead on ad measurement and then using the ad measurement for planning and buying decisions. Brooke, it does seem intuitive to me that that should be front and center, particularly for media agencies, which I'm sure many of these agencies are.
Theory: Difficulty Accessing Raw Data
(22:53) I have a theory on that coming out of TrustX. For those of you who don't know, I ran an ad tech platform for almost 10 years before I came back to my consulting business in Merkle Digital. My theory on this is that it's harder than we all believe to get a hold of the raw data to then use in the AI models for things like measurement, analysis, and
Raw Data Access Challenge
(23:21) optimization. It's very possible that the agencies would like to be further ahead here, but I will tell you that we made our raw data at TrustX available to all our clients, whether you're a buyer or a seller, but very few clients asked for
LLMs and Data Analysis Potential
(23:41) it because it's a massive amount of data, and at the time, they didn't know what they could do with it because they didn't have the power of LLMs. Now we have the power of an analysis tool where I can chuck the data into Claude or otherwise and say, "Tell me X, Y, and Z." But it's still a massive amount of data,
Data Accessibility as a Barrier
(24:04) and it's still hard to get at the data because most ad tech companies are not like TrustX was, where we freely would give access to the data. Brooke, I'm not sure if that's the answer, but it is a theory as to why folks are further behind on measurement and then using the measurement for optimization and planning. I think
Financial Services Data Barriers
(24:25) that makes sense. I come from a financial services background, highly regulated, so getting the lower funnel data, that's where there are barriers. The agencies are going to get ad response, but often combining that with what happens afterwards is difficult. Thanks. I will say something about the marketers
D2C Marketers' AI Advantage
(24:46) through anecdotal discussions. Marketers that are in a direct-to-consumer or direct-to-customer type of environment. Folks in retail, folks that are OEMs that sell directly on their
Unleveraged Potential for D2C Brands
(25:03) sites, or that frequently have engagement directly within their own environments – those are the organizations that I think could easily put AI to work to get ahead on things like who are their personas, who's coming to my site, how do I use that data to improve the odds or improve the effectiveness and efficiency of my marketing. Marketers have the ability to take some
Marketers Missing AI Opportunity
(25:26) of this on by themselves, even if the agencies don't have the data with which to do it, or aren't doing it for whatever reason. I feel like it's a miss. I feel like the marketing segment, the actual advertisers, aren't embracing some of these capabilities, which are now feasible and easier to do because you always had the data,
Obstacles: Skills and Budget (81%)
(25:53) but now you have a tool set that allows you to make a human query, and that query can give you insightful results. David, this is great. I have a metaphysical question. I wanted to focus in on that 81% which you said on the company side, the biggest obstacle is skills, budget. My question is why? If you imagine
Top-Down AI Push vs. Investment
(26:14) sometimes when you talk to people on the client side, CEOs, want to make AI a focus. I can understand the skills not being there, but are they not investing the dollars? Is it a paper tiger? Curious on your thoughts on that. I don't have a good answer to your question, as much as I would love to
Future Investments and Priorities
(26:35) have had a good answer to your question. We do have a summary next steps as part of this tool, and again, you can look at it either by individual responses or by summary by company type. It talks about what the companies want to do. There's also on this, sorry, it's the priorities tab where folks are making planned investments.
Low Investment in Ad Measurement
(26:58) Here you have how you're measuring, and here you have more of the responses. Planned investments in more automation, in building products. Sorry, strategic priorities there, and any investments in business development strategy. Here again, sorry to keep coming back to this ad measurement, I'm
Strategic Priorities for Ad Measurement
(27:19) so surprised how low on the list investments in ad measurement are when we saw previously that that's an area where it's in my view behind the times. I unfortunately don't have the answer to why, but we can at least see what the strategic priorities and planned investments are between this screen and by reading some of the
Survey Update Frequency
(27:46) next steps. I was curious, how often do you plan on updating this? It's fair, a year from now or even 6 months from now, it may be different once people catch up to their aspirations. I am thinking of doing it in Q3. This was the Q1 survey. Let's let it
Effort Behind the Survey
(28:04) roll for 6 months and do another one. I probably will do it. It's not complicated and expensive, but it does take a lot of time. We definitely had marketing between David and I. For those of you who responded on this call, we're grateful and appreciative. Thank you for that. On the topic of
Separating Media Planning and Buying
(28:25) continuing the survey into Q3, the one thing I was interested, I was surprised not to see more media planning usage for this. I'm wondering if it might be better to separate planning from buying/activation because especially at the holding companies and the large independents, those are two different teams. I think that from a
Challenges in AI for Media Planning
(28:44) buying standpoint, there's probably more accessibility to things that might help you buy one channel or one provider versus another. Whereas on the media planning front, there's a mixing chemicals between strategy and planning, and those lines always blur. It's probably harder for them to nail down AI solutions in
Internal Platforms and Usage
(29:07) that area yet. I am seeing it especially at the holding companies. They do have their own internal platforms, and they're giving the strategists and the planners plenty of tools to use. Whether or not they are using them to full advantage is another question. Walt, I think that's an excellent idea, and I would do it in three
Future Survey: AI Agents
(29:24) ways. I think I'd have media planning separate from media buying and separate from measurement. I think that it's likely by Q3 we'll see further maturity on all that, but what I'm interested in finding out in Q3 is how much folks are using agents, which were
Agentic Buying in Q3
(29:44) a theoretical idea in Q1, but I think by Q3 we'll see agentic buying going on. Walt, I'm going to call you so you can do this with us next time. You'll help us set this up. I would appreciate the input on little subtleties. Looking forward to it. There's another question from Sally
Sally's Question: Human Decision-Making
(30:04) in terms of any stats in the study regarding the human decision-making and problem-solving aspects, and perhaps teams can't agree on how to use AI effectively in media planning. It's a great question. Sally, do you want to add a little color to that? I want to make sure I'm focusing on what you're asking. Sure, and I am not a media planner,
B2B Marketing Context
(30:24) but I've done a lot of B2B marketing for the creative and media space. As I was writing to David the other day, having done a lot of strategic discovery sessions with corporate teams, to then produce brand positioning and content and marketing,
Executive Team Disagreement
(30:46) usually people on an executive team don't always agree. They come with a completely different perspective. The tech guy might have one idea, and the saleswoman has a completely other idea on a priority, for example, any generic priority. I guess that was driving the question, and
AI as a "Life Easiness" Tool
(31:08) somebody else mentioned it, too. Matthew in his comment, "Are people throwing spaghetti at the wall, make my life easier?" In the absence of being able to agree, "Wow, this is what this can do for our company or category," in this case media
Data Slices: Segment Maturity by Role
(31:26) planning. Let me show you two slices of the data that may help us see some answers. I don't have the direct answer, but let's look at some data together. The first one is what I've got on screen now, which is generally segment maturity by role. The roles are
Leadership vs. Executive Perception
(31:47) not radically different. Yes, the marketing and media leadership – these are people that day-to-day are in that function – rate themselves a little lower on the curve relative to let's call the executive management team. I put the two top bars, founders and executive management, together,
Day-to-Day Roles' Realism
(32:07) then the two on the bottom, product and sales. There's a little bit of, maybe the folks that are day-to-day are more realistic about where they are, or because they're in the weeds, they're seeing the maturity better than the other folks on the team. That's a possibility. I'm showing you the
Agency vs. Brand Functional Comparison
(32:27) data, and we're ranking up ideas. Let's go to something that could be a little more interesting here. I'm going to put on the left, because we're heavy here on marketers and we're talking about agencies, this is the comparison of all the functions. On the left are agencies, on the right
Executive vs. Marketing Leadership (Agencies)
(32:47) are brands. Then what I'm going to do is I'm going to select executive management and founders on the left, and I'm going to put in marketing and media leadership on the right. Let's see what the data shows us. The data is showing us that the executive teams feel they're further along on maturity related to
Executive View on AI Impact
(33:13) content, using AI to generate content, to help with strategic decision-making and strategic planning. Their view is that marketing ops is getting returns on investments. Let's look at the same three on the right. The people that do this every day, marketing and media leadership, these are people that said they're in the day-to-day
Executive vs. Day-to-Day Roles in Agencies
(33:36) marketing or media functions. This is inside advertisers. Let's do this. Let's look at it first in the same kind of company. This is folks at the executive level in the agencies versus folks who do it every day in the marketing and media departments at agencies. That's an interesting view.
Filter Correction and Small Sample Size
(33:59) Yeah. Now let's look. Sorry. I apologize. I had the wrong filter on. This is the agency. That's because n equals 1. There was only one person in the agency that called themselves marketing and media leadership, probably because they figured that means a marketer side.
Agency Leadership vs. Brand Marketing Roles
(34:17) Maybe the better is here. Let's do this. You all have to excuse as I move my cursor around. I'm trying to find a good comparison. Here on the left you have the agencies, and on the right you have brands, but you have people in marketing at brands. At the agency side I'm
Divergent Organizational Opinions
(34:38) looking at people on the call themselves the leadership team, and since it's an agency, they're all in media. I'm not saying I know an answer, but I'm showing you that the data shows interesting opinions that are different between different people in the organization. There are still a lot of people not
Low ROI in Strategy Use
(34:56) using it for strategy. On the strategy, n equals 1, there's only one person getting delivery on ROI. We had 10 respondents in this category, so 10%. You guys see my cursor, right? So you had the n equals 10, and only one respondent said most say that
Exploring the Data Tool
(35:18) we're in the implementation category, which is the blue. I encourage everyone to get the tool, to play with this. I got to tell you, there are so many slices here that I've played with it now for a month, I discover stuff all the time by looking at the data.
Accessing the Tool Link
(35:39) I think that's fascinating, we could juxtapose these two. If you do marketing and agencies together, you get a more interesting view. Where can we find this? Dave, will you resend the link at the end? I would put that in here again. David's Brooke again, a
Suggestion for Future Data Segmentation
(36:03) reaction to this. It's interesting, and to delve further, in the future, if there's a way to make sure on the brand side that you have brand advertisers who follow a traditional model, relying on their agency partners, versus a
In-House vs. Agency Model Data
(36:22) sunset who run their media planning and buying and measurement in-house. That may be more B2B, but it struck me that might be interesting to see that lens that may be further down the road, but I would find that interesting. That would be interesting. We had 15 brands. Let me see, what's the n here?
Data Limitations on Brand Models
(36:45) 11. 11 out, wait, hold on. I had some filters. 12 out of the 55 were, I'll call this a brand advertiser. Brooke, we don't know how many of those were in-house versus more traditional agency model. We didn't ask the question. It's a great question to ask.
Question on Publicis Group Participation
(37:04) I was wondering, and you don't have to say this if this is proprietary. I'm a Publicis Group veteran. If any Publicis media folks, Publicis Group is so laser-focused on using technology and I've talked to them recently. I'm surprised, that's all. Sally, we
Anonymity and Agency Self-Assessment
(37:25) collected the data anonymously on purpose to ensure that folks would be honest. Which is why I was a little surprised about the agencies putting themselves at the top. I figured if they're being honest, they wouldn't drink the Kool-Aid in the survey. I don't know who responded. With a very
No Publicis Respondents Identified
(37:46) small number of exceptions, we got a handful of folks who put their name on the list to get a preview. But none of the folks on the preview list had a Publicis email address. Wow, this is neat. Very cool.
Claude's Role in Tool Creation
(38:04) Not for nothing, but I am not a software engineer. I know about that. I said, "All right, I'm going to use the tools." I fired up Claude and I said, "Here are the results. I'm trying to create a tool." What you see on screen was probably two to three hours of work total.
Claude's User-Friendly Capabilities
(38:28) The initial tool took about 15 minutes. Then there was some tweaking, then I learned how to get it on my website, then I learned how to change the colors, then I learned how to create the comparison view. All of it was Claude teaching me and doing it for me.
Claude's Accessibility for Non-Developers
(38:50) For those of you who haven't experimented, a total non-techy non-developer was able to do this. I'm impressed with Claude for that. Was this Claude code, or the 20 bucks a month version? Yeah, 20 bucks a
Data Source: Typeform CSV
(39:12) month. Where did you upload the spreadsheet? What was the question, Khan? What was the source? Was it a spreadsheet? We did the survey in Typeform, and Typeform gives you the ability to download all the results in CSV. Spreadsheet, basically a spreadsheet. I was having trouble in the
Claude for Multi-Dimensional Analysis
(39:33) initial analysis trying to slice and dice it, because it's multi-dimensional. You have functions. These are the different kinds of respondents here. I turned to Claude and I said, "Here's what I'm trying to do." I said, "Well, think of this." Mike, I love you, man.
Recruiting for Future Surveys
(39:53) I did a similar project. It's powerful. My BFF. David, I want to share one other thought as you plan for your next round of this. Sounds like it's a quarterly basis. I don't know if you looked into this, but for recruiting some of the groups within the ANA Association of National
ANA as a Recruitment Source
(40:14) Advertisers, they have an AI Marketers Group, and they have digital media groups. They might have an interest in not only results, but getting their brand partners to participate. I think it's great what you built, and I'd love to see this evolve.
Mixed Trade Association Responses
(40:30) Thank you. I'm not going to knock any particular trade association, but I reached out to them all. Some were, "Absolutely. We'll be happy to send a note out and get people in. We'd love to see the results." Some were, "We don't do that."
Final Insight Before Break
(40:48) It doesn't make sense to me, because I think this is insightful and I appreciate everyone on this call agreeing, but we had mixed reviews from my friends in the trade associations. I want to show one more thing before we break. Let me put one more screen up online.
Importance of Measurement & Consultation Background
(41:11) I think this is important for folks on this call to think about. I said earlier that what gets measured gets done. Before I created TrustX in 2016, I spent 25 years in management consulting at Ernst & Young and Price Waterhouse. Our pedigree was what I'm
The Differentiation Model
(41:34) doing now at Morgan Digital, which is to help companies differentiate and grow. We used to use a model like this. I'll explain it. It's tuned to AI, but the titles on the bottom could be any transformational technology. It's looking at if there's a catalyst,
Catalyst: AI and Gen Tech
(41:57) if there's something happening in the market – the something now is AI and gen tech – you look at what's available now that we know works, what's coming soon, and what's in the future. That's the green circle on the right. It could have been
Value Drivers and Differentiation
(42:14) AI, it could have been programmatic, which is what we were calling it 10 years ago. It could have been digital 20 years ago. On the left is value drivers for your business. It's how you differentiate. It's how you get ahead. It's how you drive growth. There are high value
Identifying High-Value AI Applications
(42:34) differentiators, and medium, and low. What we do, what consultancy does at Morgan Digital, and what I used to do in my days at PwC and EY, is you look for that red star at the top. You say, "If I know this technology is capable, and if I can apply it right here in this high value area,
Lack of Strategic AI Implementation
(42:59) I can get ahead." My most frustrating takeaway from doing this survey is that very few companies were looking for where their red star is. They're playing. They're looking at gen tech. They're looking at AI. They're throwing spaghetti against the wall, but they're not thinking about it as
AI as a Strategic Imperative
(43:20) a strategic imperative to get ahead. Thankfully, I have some clients who are paying me to do that, but the market was not doing it on their own. I leave you today with this picture and say, "Think about this in your businesses. How do I merge the capabilities with what's going to drive growth
Final Call to Action
(43:46) in my company?" That's ultimately what this is all about, and is why we have folks like you here to help us figure that out. I encourage everyone to sign up, get the tool. If you see something in the data that we uncover, or that you
Seeking Community Feedback
(44:03) "Wow, this is interesting. Have you thought about it?" Send David and me a note and say, "Have you thought about this?" Everybody looks at this stuff differently, and we're always looking for insights. We appreciate your feedback to help us learn more.
Thank You and Contact Information
(44:21) Sure. We'll appreciate all this, David. I appreciate all the thoughtful questions as always. You have the survey link, and I shared David's LinkedIn as well. Hopefully, all can get in touch for thoughts and questions,
Upcoming Sessions and Events
(44:38) and ways to evolve this further. We've got things scheduled now through the end of June. Some terrific sessions coming up. Also, let me know if anyone's going to be at Possible. I'll be there with Mark Teschler this year. Look forward to
Closing Remarks
(44:59) seeing you next week and beyond. Thanks, everyone.