AI-Driven Marketing: Oren Greenberg's Strategic Pivot

Mark Evans: I'm Mark Evans, and you're listening to Marketing Spark. Today I'm speaking with Orin Greenberg, a B2B marketer, growth advisor, and the founder of Curve, a consultancy that helps startups and scale-ups with growth strategy, go-to-market execution, and marketing performance. Oren has spent years working with companies trying to figure out how to grow smarter, move faster, and build marketing systems that make an impact. And that gives him a great perspective on where B2B and SaaS marketing is today, especially at a time when many companies are trying to do less and frankly how to decide how their marketing can break through in increasingly competitive markets. One of the areas that I'm especially interested in exploring with Orin is his deep dive into AI. Like many marketers, he has been paying close attention to how AI is changing how marketers work. Everything from research and planning to execution and measurement. But I also want to understand the story behind that shift. Why did he decide to go deeper into AI? What has he learned along the way? What has surprised him? And where does he think marketers and CEOs may be overestimating or underestimating the impact? Welcome to Marketing Spark One. Pleasure. Thanks for having me. You mentioned off the top that. This shift into AI has been very successful from a business perspective. In fact, you use the word oversubscribed, which I don't hear from many marketing consultants these days. So I tip my hat to you. But taking a step back, traditionally you were a growth marketer, you were an on-demand CMO. And now you've shifted your focus to AI-native go-to-market systems. You're not Talking the talk when it comes to AI, you're walking the walk. Love the backstory into how you made the shift, why you made the shift. Was it something that you studied and thought about and made a pragmatic push into? Was it a spontaneous decision? How did you change gears to do something that is exciting but also a shift in your focus? I think the best place to start is honestly is to say I missed the boat on AI in many ways. I started when OpenAI had APIs called DaVinci and Babbage pre ChatGPT three. This is twenty twenty. I was messing around with this. And I was trying to map the job title and skills landscape for marketers globally. To classify all this free form text was very difficult. So I ended up looking at natural language processing and that was data science. You ⁓ wow. So like you must have really seen it coming and you were really on top of it and you're like pioneering. I had no idea that it was going to become the biggest, most important thing in the next few years. So even though I thought it was a very useful tool, and I remember when I started playing with GPT three, I was impressed by it, but the leap that they had made in the progress from into 2024. I didn't expect such exponential improvement. Which could argue there's maybe the limitation of really the depth of expertise. If I I didn't understand a like a PhD level data scientist who spent his day in linear algebra, and I didn't really understand the potential for growth or improvement from image generation to text generation. Saying all of that, despite my slightly critical view of the potential plus to have been the number one pioneering voice if I'd started early enough. I was still very much enmeshed with AI and dealing with it very deeply, but I wasn't positioned and didn't communicate the expertise publicly until I did. When I made that shift, the reason I made the shift is I had a conversation with my client, 85 mil ARR SaaS, and the CEO was telling me that the valuation of the company has been impacted because of the lack of AI capability. And when I understood that the exit was influenced by it. I realized that the market has now taken to interpreting and understanding the significance of AI. When I heard that, I was like, okay, I'm gonna pivot and focus more and communicate publicly, not just privately, all the stuff I was doing. And that's it. Just referral, word of mouth, and network. And I've had like really brilliant momentum since. This re it's very recent. I only did that pivot publicly, maybe in July or August last year, in long at all. And it's been just nonstop busy since that happened. So I'm working with my smallest is probably like a two mil, three mil ARR, and the largest is like in the billions, like Nasdaq listed. Historically I've worked for very large public companies, maybe about five or six, but never as the AI marketing specialist. So this Nasdaq one like the largest entity I've worked with in this vein since the PIBA. That's sounds amazing. And again, congratulations on the pivot. It's hard to pivot your business when you've done things the same way for such a long time. It would be great to get some insight into the actual work that you're doing with these companies. Are you working directly with the CEO or the CMO or a variety of stakeholders? And then without giving away trade secrets, what exactly are you doing with these companies? Is it Strategic planning? Is it helping them with tactical execution? Is the embrace of the different LLMs? What does that engagement look like? I'd say it splits into two in terms of the target persona. On the smaller companies, it's going to be the CEO. For the larger companies, I'd say probably 50 mil ARR and above. It starts to shift and I end up working with either the CRO or CML. I'd say there's maybe two or three main parts that I help with. The first is the Implementation of AI systems for custom projects. For example, ICP modeling or building a programmatic AEO engine to produce lots of content to get the ranking. The next pillar, I'd say growth audits. And I'd say right now, I've got about five that I'm running in parallel for five different companies. Just lots of different audits of TIN ads, the different kit they use. Is it demand-based or is it Six Sense, Google Search Console, Google Analytics, Google Ads? SEO, AI ranking, AEOGO, so many different terms for it now. Team structure capability. So some of it does overlap with more consulting, professional CMO work. I'd say it's bringing that marketing expertise and then overlaying AI. So it's this kind of this juxtaposition of these two concentric circles, this Venn diagram is like quite niched. I'd say key percent. technology companies and 20% services companies, but only just B2B, right? And then the next pillar is like setting up workflows and automation. So actually using those tools to help increase efficiency. And then to the last pillar is really the AI training. They did they did it, a company raised a billion. There's like a team of 24, eight week program, one hour a week, unlimited support. And I just upskill them. And I'm trying to help that team understand how to work with LLMs. show them lots of different tools that are available and then give them like homework to go do and come back and work as a team to actually upskill and understand, not only conceptually and theoretically, but practically and increase their confidence. But I'd say the Cloud Suite at the moment is by far the most popular. Literally I talked to a company yesterday migrating from ChatGPT Enterprise to the Cloud Suite. So everyone's migrating over into Cloud, which is the right move. They should be doing that. Even though OpenAI was the originator, it's still A really brilliant tool. I use the ChatGPT API and I also use Claude and the Claude API. And I use Gemini. So I use different models for different purposes. So it's not to say they're better as much as they're better for different purposes and different use cases. I'd say there's definitely a very interesting cultural divide in how companies are thinking about their team and AI. I'd categorize it mostly into two. We have the team, we don't want to hire more heads. And we want to maximize their throughput and output and outcome through empowering them through with AI. And that's camp one. Camp two, we're venturing or very near to an exit. We really have to get the IBDA up. We have to cut costs. I'm sorry, 50 to usually I'd say like twenty to sixty percent of the team has to go. We're gonna replace that with either outsourced or AI enabled. So I find the executives their appetite or passion for AI is quite variable. So some of them are like they talk to me and they can't have enough. They just like they just they love the conversation, they want to keep coming back and they're like totally enmeshed in the potential for the technology and they love it. Others see it as a significant strategic focus, but it's not everything. They're not obsessed with it. It's just one of the many things they're juggling and they're aware of the importance of it. There is a third group which is like Extreme. Like everyone needs to know Clot code. They're usually engineers, they're usually people who come from a data background. They're not investing in AI as it is now. They're investing in AI as they believe it's going to be in five to ten years. CMOs are having really hard conversations. ⁓ CMO's telling me yesterday, she's telling her team, you need to you'll be unemployable in two or three years if you don't know PG skills. And I I want for you to be successful. CMOs are like they have to push to elevate their team to use these tools. And I think there's a lot of resistance because of fear and insecurity and lack of comfort. But also to be honest, a lot of the way companies are rolling out the tools is not very skillful. They're just like, Okay, here's your cloud subscription, go learn cloud code. But your task volume hasn't reduced. You just have the access and good luck. Who's gonna go chroll through tens of hours of YouTube videos in the evening instead of spending time with friends or family or chilling? To go and learn something very complicated that they don't really fully understand the value of. So I think there's a the way it's being rolled out isn't always great. That's the smarter companies, they bring an AI trainer to make sure that everyone's elevated and feel supported in that transition. One of the things that I think about when it comes to how companies are rolling out AI is it's often tools and tactics. Everyone's excited about climbing on the bandwagon and they miss the step where what do we want to do with it? How are we going to roll it out? How are we going to trade our people? At the end of the day, what are we trying to achieve here? It could be higher profits so that we can have a better exit. It could be doing more with the same number of people, but there just doesn't seem to be a lot of strategic thinking. Because there's so much excitement. And of course, there's this fear of missing out, which I think is a big factor these days. One thing that I do want to talk to you about was downsizing marketing teams. Cause as you mentioned, I think some of the fear that goes along with AI as marketers look at themselves and go, I'm not gonna have a job, or I'm gonna be replaced by AI. I think a lot of this fear is exacerbated by CEOs. ⁓ Enthusiastically announced that because of their obsession with AI, she had cut her marketing team from 24 people to two. I remember seeing this post. I know who this is. And she was coming at it from the perspective that I have to make sure my company's on the leading edge. I have to make sure I'm not at a competitive disadvantage. And if that means making difficult decisions due to my abrasive AI, so be it. There was conviction in the decisions they had made. The pushback was. Brutal. What was really interesting was that rather than sit back and let the c waters calm, she stepped into it again with something even more extensive that came across as not reading the room. The reality is it is what it is. AI is what it is, and it can do all kinds of amazing things. And that may include smaller marketing teams. Given the fact that you work with a lot of CMOs. Yes. What are they thinking these days in terms of how their teams are structured and How they could evolve in the future. Do some of them look at smaller teams? Do some of them anticipate having bigger teams? What's going through their minds these days? Because they must be under a lot of pressure to generate leads, which is the fundamentals, but also to demonstrate that we understand the importance of AI and we are going to leverage it so that the company can be successful. I'd say most emails I speak to, they have a more holistic view than just lead volume or that as a KPI. I'd say they talk in terms of growth. And driving growth and they have a holistic view, extending beyond acquisition or into retention, App Cell and Crossell. I do think there's very different structures to the teams I speak to. Some are centralized and they're serving either lots of different subcompanies or other teams in other markets. Others, they're obviously just like a dedicated team and one software company. I haven't yet really spoken to anyone, but it's recently that I can think of this adopted an entirely pod based approach and fragmented it. It's usually centralized or One brand and all in on that one. I don't think I've spoken to any CMO recently who's hiring an increasing headcount as they would pre-two years ago, where I went to someone that like, Oran, I'm stressed out, and like, why are you stressed? I gotta ramp up and hire fifty marketers in the next six months. I remember there were conversations and now I can't remember the last time I had a conversation like that. I think it's all I need to either keep it steady or I need to cut it. Or it's not either that, like I don't they don't get into that with me as in there isn't radical change and either, but their focus is like I have to get my team AI first. Like I'm mandated to some of them literally have KPIs provided, like API usage. Not an advocate of that path. I can understand why it's done that way. I don't think it's the right way to do it necessarily. I don't think it's qualitatively effective. I think it's quantitatively effective. And I think Expertise is all about qualitative, not quantitative, right? And what you would do is figure out a program that helps nurture, nourish this person and your team in terms of the career progression and development. And that's how it should be positioned. Even if the drive commercially underneath is fundamentally increase profitability, reduce overhead, drive more output, drive more outcome. But it's how it's communicated. I think it should be more sophisticated. I think fundamentally there's no real change. Between how things have historically been and how AI is. The truth is, two plus years ago, five years ago, if there was underperformers in the marketing team, they'd be cut. If they needed to drive an increase in the battery was an exit, people would be there is no real difference. The primary difference with the A is the technology, which is of a much larger magnitude. Industrial Revolution, Internet equivalent. So the scale and how prolific it is in our combined consciousness and the news and social is so dominant. It feels like a tsunami that's hitting and smashing everything along its way in its wake. And I'd say it's this perception of the unlock of productivity or the way work is changing, which is material. But with both of them, fundamentally, think about what is the mechanism that drives both of them. And at the end of the day, you've still got a PL, you've still got overhead, you've still got cost of sales, you still have profitability, and you're still working towards commercial metrics. That you need to drive. And in a way, that hasn't changed that reality. The truth is when you think about marketing channels, AI hasn't changed marketing channels much either. Like Instagram's still strong, TikTok, X, LinkedIn. The only big shift we've seen in terms of channels due to AI is a shift in search, moving away from Google to chat and cloud and perplexity. It's not just the channel. It's like when you think about spend between social and search, those two are like 75% of all spend globally on digital. So half of that is now going through a shift. So it's not like a small shift, it's a tectonic shift, but it's still only one of many channels and activities of how of that have been erupted due to AI. I'm not foreseeing anytime soon new channels evolving. But I do think that there is a big shift now with like agents, and now there's more traffic due to agents than humans in how data is being aggregated and how agents are like the purveyors of content and curation and their filter. And that filter is intensifying because the noise is intensifying generated content and the number of agentic workflows out there. And then I think what's happening is there's going to be like agent wars and this like meta-level of agent versus agent and sophistication agent. And this is like an interesting meta of orchestration and complexity. Now, I don't think it's right to think about like my mental model is not AI replaces people. I think that's bound to happen on the most mundane repetitive tasks. And unfortunately, there people are going to get hit. There's other interesting roles that have been hit, like translation and content creation, and probably soon enough graphic design, even though it is happening already, it hasn't entirely removed that functional role. And those roles are obviously like almost obsolete, if not in dynamite. So I think there is going to be the death of roles. But on the whole, I think a much healthier paradigm is the empowerment and how AI helps people be more efficient and more productive. Does that necessitate smaller teams? I'd say for many tasks, yes. Many areas of marketing, especially digital marketing, yes. But experiential events, yeah. They're probably not going to be as impacted by it. So it's very like you can't say everything, but like very big chunks of a lot of things that we rely on and where a lot of the traffic is flowing from or where a lot of the revenue is being driven, right? When I look at where the marketing budget is deployed and I look through that lens, there is gonna be Quite a bit of disruption in those areas. Big question to ask you. So many CEOs are telling their teams that, in quotation works, they have to use AI. Yes. So a two-prong question here. What does a structured AI adoption plan look like? Going back to my argument about the need for a strategic approach to AI, where do you see the biggest opportunities for B2B and SaaS companies when it comes to leveraging AI, whether it's efficiency, better targeting, factor execution, or something else? It's a lot in one short question. I have colliding thoughts and I'm just waiting for one of them to win. See see which one of the competing thoughts is the winner. Let's just wear the second one, 'cause that's more recent. The problem here is it's very variable by team and how they deploy and where the growth is coming from. So it's very hard to say like all of them should be doing X. What is very clear is that everything to do with research, content creation, planning in terms of a thought partner is critical to have AI integrated into those systems and methodologies. And it's amazing to see how people are still doing things that are very manual. Even using Cloud Cowork. To organize files in your folder and work locally and edit and ⁓ modify lots of bulk actions is critical now. I think it's becoming table stakes because of Clot Cowork. Do you think that where we're moving to is this open claw Hermes always on agent that is a delegator or creating of subagents to execute on tasks? because of the limitations of the context window and making sure memory is persistent. And more that's gonna grow for progressive adoption. But I haven't seen any company yet deploy that practically and efficiently. I'm seeing more freelancers doing this, more agencies, AI first teams, but no company that's like north of 50 mil. I haven't seen like an open claw deployment across the team or anything, right? So a lot of like the hype of what people are very excited about and the reality there's still quite a large gap because of AI slot. Because of hallucinations, sort of security concerns. Because of a competence gap, a skills gap. There's a lot of gaps there. Yeah, different teams, they have different workflow automations they set up, either deterministic using NA ten or probabilistic using clot codes. You've got routines now. I personally set up a lot of stuff with clod codes, like cron jobs. Most of my marketing activity is through clod code personally. So I don't use clod AI and clot cowork for ongoing activities and tasks. I have agents that are dedicated that I've set up that run for me. I do encourage clients to build that too. Who have very limited time and energy in our day. The more you can offload through automation, the more it frees you up to do more strategic work or develop the experiences and learning to elevate the specialization and value that you can create in the organization. So you confirm, do you feel I've answered a second question in the way you intended? The one thing I've been thinking a lot about recently is The value that marketers, whether they're in-house or consultants, bring to the table these days. Because I think many people believe that AI can do a pretty good job of creating strategic plans and content and design and competitive research. All the things that marketers would traditionally do as a consultant. Where do you think the value that we're delivering these days? What do CEOs or CMOs want from marketing consultants? What kind of value are they looking for? What kind of insight are they looking for? As important, what are they willing to pay for these days? I think it's tricky. With traditional consulting and the value that traditional consulting delivers, I think the biggest problem is that the person who's using Claude cannot discern if the quality of the output is of a high enough standard. And I've personally seen that a lot of stuff has been outputted and it's erroneous. But the person who outputted it didn't know it was because they didn't have that expertise or specialization to discern it. There's two parts to this, the reason for this problem. The problem part one is the executive who's using AI doesn't know that they don't know how to use AI properly. They just think if I type it in, then it's smart enough and it's gonna give me response. And they don't really understand that they haven't inputted enough specific context or broken down they request into with in the correct order to get the response that they want. And they also discount That AI is very psychopanthic and it's very much lucky to applease you or appease you from what you already think. And people are they're looking for they think they're looking for impartial view, but they're really looking for validation. They are given to them, they're happy that they believe their strategy is correct. That's the first part. The second part is the context that's inputted into AI isn't enough in terms of just general context. The specificity required to get the strategy accurate enough hasn't actually been inputted. Like all of your competitors and all of their differentiators, all of their services and all of their products and your entire client base. And if you think about what a a consultant would do to get that level of data and depth to get you an accurate assessment, the people forget that the AI needs that. They just assume, ⁓ it's ingested the internet, which is actually wrong, because it actually has an MOE model. So it has a model of a mix of experts. And what happens is you don't have an AI that answers your request. You actually have a request that's gets sent to A specialized model as part of the general model that answers your request. There isn't an AI that's interested in internet. And also of all the parameters in the AI, it never really uses all the parameters at any given point. And getting technical or specific here isn't to show off. It's to explain that if you're using AI and you don't really understand how you're using AI necessarily. That was the point. And the point is a lot of executives, they have these very broad ideas that are very rough and inaccurate. And therefore, the output and the quality is not the standard it needs to be because they don't understand how the model works. So what happens is you're getting the average of the average, you're getting the mean average of the AI response as slightly personalized. So your very limited context you provide it. What happens is you're getting a response you can't even evaluate if it's accurate or not. And that's the two-part problem with how executives are thinking about it, and they're thinking it can replace the experts. Unfortunately. They're probably right 50% of the time. So half the experts probably are useless and could be replaced with an eye and it's an equivalent. And the other half are really good and competent. But for the decision maker who can't evaluate who which 50% is good or bad, then they end up going, okay, I'm so expensive. I might as well just do it all the hundred percent. And sometimes it makes a fatal mistake and that's very deadly. If that category of the traditional consultant is in decline because of everything I just described, the consultant who is empowered through AI or leveraging AI is at an advantage because now the buyer is understanding that person is leveraging those tools to either be more efficient. So they're delivering more value for the same cost. And the person is and is more aligned culturally to the shift in the market and the competitive forces that company is under and that executive is feeling, then they need to be cutting edge and need to be delivering better performance. And that person is communicating, they get it rather than being resistant to it. Someone the other day said the marketing person is very capable, very smart, really lovely, but a complete laggard. You can hear this friction there between this person's OAI is useless. And the CEO is we should all become AI first and AI native. Imagine what's happening in that relationship dynamic where there's so much friction because of this resistance to the technology and the belief around it. So I love your comment about 50%. You what's the 50% are smart and 50% reminds you of John Warmaker, that quote, I know 50% of my marketing is working. I just don't know which part of it it is. And the other thing that struck me about what you said is there The validation bias when it comes to AI is that you can put the inputs for a strategic plan and the outputs look really good because it validates your thinking or your biases, what you lack and what a consultant or somebody with experience can give you is perspective and challenge your views because they've been there, they've done that. The AI may be very smart, but it just hasn't had the the experience. Not a lot of people see. So there's it's not AI is an amazing tool, but it's a it's not the be all and end all. And I think that's something that consultants can bring to the table. Great conversation. One last question I wanted to ask you is when you look at the B D B marketing landscape, and I guess it's really hard given we don't know where it's going to be in three months, let alone three weeks from now. What are the most vulnerable areas for AI disruption? We've talked about how content's being turned upside down, increasingly design is gonna be more AI dependent research search. When you look into your AI crystal ball, what do you envision when it comes to disruption? I'm gonna give the abstract intangible answer, which is directional, and then maybe throw in a couple of examples. What's AI's biggest superpower at the moment? It's coding, like unequivocally. So every task marketing related that has a coding component or influences a coding component is going to be more susceptible than anything that's further away from that central point. So everything to do with website design, changing website, forms, workflows, any of that stuff is much more at risk. Anything to do with research and data synthesis, because AI is very good with summarization, right? It's very good taking that content that it has and synthesizing it and interpreting it and explaining it, right? What's definitely happening now in terms of the more pioneering tools? Is that they're very focused on removing the UI and UX and enabling that flow of data directly through the AI, through cloud code, right? But even that's a niche. Right now I'm talking about go-to-market engineering and revenue operations and marketing ops, fundamentals of the data pipeline. I think all of that's going through a disruption now. I think that everything to do with like strategy and orchestration. is obviously very sound and safe. And I you know, so what I find is I'm now able to do marketing work that I would have had juniors do, but without having the juniors because I can leverage AI. So I can actually take on a lot more activity because I'm an expert and I know how to evaluate what the AI is doing and it's so fast and capable that a lot of that lower end work is really at the highest risk compared to that middle and definitely not the top echelon. Right. It was about thirty four functions in marketing. core functions. So it's very hard to go through all of them. Right. Everything to like SEO is going to be relatively standardized, like AEO answers, all of this stuff like on-page optimization, content assisted creation, all of that is much higher risk. Everything to do with constraint parameters that has a finite number of variables is a high risk. Google Ads, Meta Ads, X, all of these platforms where the number of Parameters variables is very limited. So it's relatively easy to identify the right pattern. And even though like creatives can be produced with AI and even on scale, the ideation process of producing exceptional and differentiated creative is obviously still got to be in the domain of human minds. So anything that the AI needs to be able to do that it doesn't have access in its data set to, so it's outside the parameters of what it knows, is obviously still in the remote of humans. Everything to do with human to human is obviously like account management, communication is obviously all still outside of the domain of AI. I think what's really interesting is like how quickly it's advancing and how capable it's becoming and how it's progress like the speed at which it's eating up from the bottom up. So yeah. At the current rate of improvement, I'd say that lower end role and that mid range role is most at risk. Great conversation, covered a lot of ground. Where can people learn about you and what you do? I think the best bet is probably sending out to the newsletter on my website. LinkedIn's been a bit funny recently, so I would have usually said LinkedIn, but it's a bit unpredictable. So I'd say probably the newsletter's the best bet. Thanks, Orin, and thanks to everyone for listening to another episode of Market ESpark. If you found this conversation valuable, subscribe on Apple Podcasts, Spotify, or favorite podcast app. Drop a quick rating and share it on social media. And if you're a B2B or SaaS CEO struggling with stall growths and frustrated with marketing that isn't working, we should talk. You can reach me by email, mark at markdeforts.ca, connect with on LinkedIn, or visit marketingspark.co.