Tackling AI Adoption: Practical Strategies For Upskilling Teams

Tessa Burg: Hello, and welcome to another episode of “Leader Generation,” brought to you by Mod Op. I’m your host, Tessa Burg, and today I am joined by Stefan Tornquist. He’s the SVP of Learning and Research at Econsultancy. Stefan, thanks so much for joining us. We’re excited to have you.

Stefan Tornquist: Hi, Tessa. It’s great to be here.

Tessa Burg: So we were just talking a little bit before the recording started that Econsultancy recently did an internal analysis of AI’s impact on knowledge work skills. And that’s something we are, as marketers we’re constantly thinking about that. How is our jobs gonna be impacted? What will be the impact to our clients who are also in professional services? So tell us what triggered performing that analysis and what were some of the learnings that came out of it?

Stefan Tornquist: Sure. As a skills development company, we’re really interested in how, what is the practical effect of AI on a given area of skill? So, you’ll hear people say, “Oh, well, you know, data and analytics are gonna be completely upended because of AI.” And that may be true in some respects, but what we want to know is to be able to take advantage of AI at a very fine-grained, “I’m doing X job,” What do people need to know? What kind of background skills do they need to even be able to assess, “Here’s the advantage I can get. The AI is correct in its analysis, it’s incorrect in its analysis.” So this was part of a larger effort to build kind of these skills chains and figure out what is absolutely dependent, and then what could you just sort of dip in, use AI for, and then dip out without having that background. The impetus to do the study was in terms of the most fundamental knowledge work skills, where is AI able to play a role today?

Tessa Burg: And this type of analysis and study is definitely not new to you. You’ve been in the research space for a really long time. Tell us a little bit about your background and where you started and where Econsultancy and your role has evolved to date.

Stefan Tornquist: Sure. Well, let’s see. I’ll go, the year is 1997. And a buddy of mine gets back from art school, and I’m finishing up my master’s in political science. I am going to go and try to become a college professor. At least that’s the assumption. And he says, “You know, it’s weird how these banners on the internet take you from the website you want to be on to a different website.” Fast forward to 1999, we’ve got $35 million of other people’s money and one of the first rich media startups. So that’s how I got drawn into this industry, and I’ve never gotten out. When that company got acquired, I became a sort of roving analyst for a while, and eventually became a research director, and a variety of other titles at Econsultancy. So for most of my career there, which is now 15 years, I think, I was research-focused for about 10 of that. And then for the last five I’ve been thinking about skills, curriculums, and building up our product set around skills development, some of which is delivered on demand, some of which is delivered live, and everything in between.

Tessa Burg: So you lived through how the internet, and even specifically media, impacted marketing and advertising at that time. Like you’re in the profession, you naturally have this curiosity for the why. Why am I going here? Why does this work? How do we deliver value to brands? And now you’re sitting in a research seat and recognizing that, “Hey, there has to be training and we’re gonna have to move again.” What similarities do you see between these two times with the internet coming in and really disrupting the way brands communicated, engaged before rich media and the way AI is coming in and asking us to rethink what skills do we need to facilitate experiences and communication today?

Stefan Tornquist: Well, on the business front, it feels really familiar because you have this enormous, arguably over-investment early on. Right now feels very much like the late 90s. It doesn’t feel like the second generation where now we’re building companies with value. It feels a lot more like that first generation where, how much has been invested so far, you know, a trillion dollars or more. And when money like that gets invested, people want to see high-end results and they want to see them in the next couple of quarters. And so we’ve kind of set ourselves up for, what does Gartner call it, you know, the trough of disillusionment, in the next, you know, week, 10 days? I don’t know. Pretty soon. None of that has anything to do with the actual practical effects of AI and it’s sort of slow-then-fast impact, I would argue. It’s that we kind of went in… You know, it’s sort of the fault of marketing, right? Marketing from Nvidia, Google, Microsoft or whomever came in and said, “We’re gonna change everything, and we’re gonna change it without you having to substantially affect your tech stack and we’re gonna do it quickly.” None of that is true, right? It still requires… its technology requires backend integration and all the rest of it. So on the business side, it feels super familiar. On the actual skills side, it is fundamentally different. For the first time we’re not dealing with a technology that can only compliment what we do. It really can sometimes replace what we do. And I guess you could say automation in marketing is nothing new, but there is something fundamentally different about a tool that can learn and produce that kind of bulk content, which is so much of what a marketer does. Whether that’s writing emails or ad copy, I think it is fundamentally different that this thing can do that previously core skill. Not necessarily to final draft, but it can, it can, you know… If you think about marketing as 80% or 90% bulk, meaning we’ve got 20,000 different iterations of banner ads. We’ve got 50,000 different triggered emails to our customer base. All of that kind of… We need to do digital circulars with discounts by products and that kind of thing. That kind of bulk marketing, AI can do an awful lot of that. And it really only leaves that kind of 20% bespoke level to us for the time being. That was a very long answer. I guess it’s like automation but broader spectrum. What can’t you automate is the question now.

Tessa Burg: Yeah. And you brought up a really good point that it’s going to replace work where it felt like in the digital age, and maybe I’m wrong because I also lived it, but I didn’t have much experience. I was super-duper young. So I’m going to say something that was coming from the eyes of a 17-year-old who was building their first website. But to me it felt like I was being tacked on to a really strong team. And no one on the team actually also had to learn how to build the website. Like I got hired and contracted to build the website. So when I think about my experience through that transition, it was, I was being added to teams of marketers who were also still continuing to operate in very similar ways, where now, we’re asking people, everyone to operate fundamentally differently than they do today. And if they don’t or if they fall behind, there’s a good chance they could be replaced, or they won’t have the ability to level up to the strategic thinking that AI will afford us. A lot of company hear these stats. They hear it’s gonna replace 85%, 90% of the work that’s being done, but there might be an over-reliance on external research, and you might be taking for granted that you initiated doing an internal analysis. But if other people are taking a step back and saying, “Okay, I hear this, I see external research.” How can they get out of their own way? Like, I feel like it’s the time factor. I don’t have enough time to do an internal analysis. I don’t have enough time to see, to benchmark what things are going right now. And I can barely get people to even take a simple training on what is AI. Like how, what can you tell them, or how do you get that process started for yourself, if you’re sitting right now listening to this, at your own company to do that internal analysis and start taking the necessary steps to evolve roles and upskill and reskill?

Stefan Tornquist: I’m trying to remember, we did a guide for companies a while back and we tried to frame it from what company personality type do you have? Because depending on where you’re coming from, this might be the best approach to integrating AI. And it was pretty clever at the time. I can’t remember what the four personality types are. But everybody knows their own company, and knows like, “Oh, okay, we’re very top-down driven. I need to get the CEO or somebody very highly placed to say, ‘This is our AI initiative and this is how it’s gonna work, and it’s gonna be a part of your manager reviews and so forth.'” If you don’t work in that kind of traditional structure, then the approach ought to be greed and selfishness instead of fear. Meaning people who are, people who aren’t afraid of losing their jobs should be excited to get to do the most interesting part of their jobs more, right? AI is at a stage where if there are… Let us hope that whoever’s listening has interesting parts of their job. Chances are AI is not gonna replace those. So when we talk about AI being able to contribute across 85% of the skills we analyzed, it doesn’t mean it can do them. It just means that it can be helpful to varying degrees. And fortunately the places where it’s most helpful, where you can just about offload your activity, those tend to be at the lower end of the spectrum. But anyway, I want to answer your question about onboarding. I think the first step has to be kind of emotional, because, you know, those of us who are in fast-moving agencies like Mod Op, you probably don’t have a few hundred people sitting in the Midwest who are scared to death that their jobs are going away because AI is gonna eat them. But that’s true at a lot of companies. And the less you know, the more fearful you are, you don’t want to engage in it. And we know going back, the history of change management is the history of emotional blockers. The reason 75% of change management doesn’t work, it’s not because the tech wasn’t right. You can always make the tech work, as you well know. It’s about, you know, do people embrace it? Do they actually change their processes and adopt new ways of working and new technology? This is like that, but even more personal. You know, I saw a bit of research around how the use curve for ChatGPT is pretty sharp. Like, a lot of people who use it a few times don’t use it again, but that’s on the Open AI website. That’s not the logged-in version. And one interpretation was, it just scares people. People don’t like interacting with this because they immediately start thinking about like, “Well, what do I do all day? I write copy.” Or whatever. So you’ve gotta get over the fear first. And this is a great moment to do that. You know, we’re at a moment where a lot of questions are being asked about the ultimate ROI of AI I think this is great air cover to actually get into where it can be really useful in sort of day-to-day productivity across a wide set of skills.

Tessa Burg:  Yeah, I agree, and I think when leaders get discouraged by the fear. I’ve heard comments like, “Well, just that’s something we gotta do. We just gotta charge forward.” And you don’t actually embrace that there are different perspectives, different emotions. It’s a spectrum of emotions and people are moving up and down from fear to excitement all the time. But you really have to embrace that. And what you did, like doing this internal analysis also, I imagine, helped bring people along and make them a part of the change that’s about to happen, and kinda give them space to express where they’re at, maybe where they want to go. But also–

Stefan Tornquist: Well, I—

Tessa Burg: Yeah.

Stefan Tornquist: —Oh, go ahead. Sorry.

Tessa Burg: I was gonna say also emotion is something that like AI can’t do. Like we can automate a bunch of stuff, but the way we sell, the way we engage is all human and emotional. And so if the process and the change management supporting the process changes, don’t take emotion into account, then you’re really also missing that an opportunity.

Stefan Tornquist: I couldn’t agree more. I think one of the interesting things, one of the findings from the study or observations from the study, so the way we structured it was around what’s called Bloom’s Taxonomy. If you’re not in learning design, forget about what it is. But what it does is it kind of structures activities from the most basic, which are sort of remembering concepts, and then it moves up through understanding, analysis, yeah, applying and analyzing, and then ultimately creativity at the top. And that’s a good model for learning. And we thought it was a good model for, “Well, let’s play this out across 20 skills in each one of these boxes.” And going to your point about the leadership and understanding emotion, AI’s least able to disrupt your job at the top end of that, of this grid. And it’s most able to disrupt your job at the bottom end of the grid. And what that means is people who are in leadership positions kind of spend their, they spend their time in the top of the grid. So AI could be really useful to someone like you or me who is like, “I wanna do this boring thing more quickly. I want a meta analysis of these 10 academic articles.” Whatever it is that would’ve taken us X time, it’s taking the AI enough… It’s adding about 30% productivity to that kind of activity, right? But none of that is dangerous to our job roles. Whereas if you’re in the bottom half of this, this is, you know, understanding, applying, remembering, this is where it’s writing all those targeted versions of an email. It is doing campaign analysis, what have you. Those are activities that can be disrupted. And so the, as is often the case, the people who are making the decisions about it don’t even have the same level of concern. I mean, it may be soon enough, but for the time being, the people that we’re asking to change, and the people who were asking to go do trainings, they can see a very tangible potential negative. But the people making the decisions don’t. So it’s worth coming at it from that perspective. And I think framing it as, we don’t really see, not yet anyway, we don’t really see massive restructurings in knowledge work. We see a lot of companies are kind of not replacing people as they leave, but, you know, it’s very hard to tease out is that just the state of the economy and over hiring during the pandemic kind of working its way through the system. You know, except for there’s the Klarna example and a few others. In the moment we don’t see this huge displacement of knowledge workers, marketers and customer service salespeople. I’m not saying it won’t happen. I’m just saying like right now, the promise is, “How about we save you 30% of your time on the worst stuff that you do?” That’s pretty amazing.

Tessa Burg: I agree. It has been very nice. We’ve experienced a lot of those benefits ourselves, just naturally being excited about tech and what it can do. But one thing that I’m seeing as a challenge is making the time for training, and finding really effective trainings. Like I feel like we’ve run the gamut. Like I’ve had people take data science trainings, so I’m like, “Well, it’s like the TI calculator. You first gotta understand how to do the math before we hand you the really powerful calculator.” Light bulbs and dots. Light bulbs will go off, dots will connect if you understand some basic principles about data science. That didn’t go well. You know, no one–

Stefan Tornquist: Um hm.

Tessa Burg: People weren’t in data science. They did not want to get trained in data science, especially creative folks. And then there were, trying different levels of intro courses, and then it’s like this mixed bag of feedback. “It’s too high level. It’s too technical.” And this is all about like the same trainings and even people within the same departments.

Stefan Tornquist: Sure.

Tessa Burg: Are you seeing these challenges and are you finding that there are more effective ways to train? Or is training itself starting to evolve?

Stefan Tornquist: Yes, to all of the above. So… I mean, one trend we see, and we were talking about a little bit before the call, that there’s a real push sort of, not back to the office, but back to face-to-face training, although it’s usually virtual. So what we see a lot of companies doing that had really, they had over indexed on micro learning. We’re just gonna make everything so short and on video that people won’t have a problem consuming it. That’s not really how we learn, especially in a more advanced concept like data science. Pretty hard to learn in 90-second increments. So, but it sounded great. We’re gonna have this enormous library. Anything you wanna learn, it’s right there. Just go and learn it. Most of those sit dead empty. And then even if you have a more fulsome, but only on-demand training course, you know, I mean, this has been true for a long time, but people don’t want to do it.

Tessa Burg: Um hm.

Stefan Tornquist: You can use a characteristic combination of both. So what we’re both encouraging but also getting demanded from clients is much more blended learning, ’cause people still respond to live interactions and then become much more willing if… You know, if we did a workshop today on like, “Let’s talk about the most interesting trends in data science, and now next week you’re gonna go in and you’re gonna learn some basic data science concepts. And then the week after that we’re gonna be getting together to put them to work.” That kind of thing works much better than, “Great news, over the next eight weeks you’re gonna be consuming 27 hours of video delivered at about this pace.” So that’s one thing. But I mean, I don’t, that has nothing to do with AI in particular. It’s just that we thought that TikTok was the way to deliver training just as an industry. Like, “Oh, it’s just gonna get shorter and shorter and shorter.” If people aren’t motivated, it can’t get short enough. And to your point about people feeling like there’s never enough time, there never is, right? Like you have to really give to somebody a compelling proposition. Like, “This will be good for your career and good for your life if you learn how to do this.” And then you need to put it in, entertaining is too strong a term term, but in a compelling training environment. Now as for AI changing things, which I think was your third note on sort of training, it very much is in the background. Like it’s making it much faster and easier to produce something like what looks like a very high end bit of video is getting cheaper, faster, easier to translate. Doesn’t mean that you can eliminate the people at any one of those stages, but you can really shorten the time. And that’s great. That’s a very practical benefit. Where we’re headed, and I think is really interesting, is everybody should have their own personal kind of AI mentor, tutor, educational evaluator, whatever you want to call it. Something that sort of travels with you, whether that’s just in one corporate environment or throughout your career that really can play a role in saying, “You know, Tessa, you’re already top of the charts in all of these technical skills. Let’s go, you know, it’s time for you to do negotiation training.” Or, you know, “I was reading your email back to the CEO and I think it could be, let’s do some comms training because it could be more forceful.” Any kind of individualized feedback is powerful, ’cause almost everything we have to do in corporate training is by definition, pretty broad strokes, right? Even if it’s a team of eight, you already are kind of teaching to and thinking about a common denominator. And AI really can do both the analysis and the tutoring at an individual level, and that’s gonna get pretty cool. I mean, you look at there are examples already in mathematics, you know the Khan Academy example that everybody’s seen the TED talk around. And they get things wrong sometimes and they’re not perfect. They’re getting things less wrong as time goes on. And speaking for myself, I get things wrong all the time. So, to me that’s a very exciting kind of mode for training and application for AI.

Tessa Burg: I took so many notes during that because it triggered some… Well, I think they’re great ideas.

Stefan Tornquist: You gonna go build a product?

Tessa Burg: I am. Yep, that’s exactly it. But that just, talk about light bulb moment, I was like, that’s exactly what we need. So when we think about what needs to be built into our training platform, and we have one here that we’ve built ourselves, and we always say it’s not just about learning, it’s about facilitating change. And yeah, we have an agent in there called an AI coach that gives you personalized recommendations, but what you were just talking about can really take it to that next step. Like instead of it just being a learning platform, it can also be a door to some of our internal proprietary tools so that as we see you using AI to automate those bottom rung stacks, we can actually see how well it’s going, measure the impact on productivity and efficiency, which we are always gonna do. But what you just mentioned is the reward is that we’ll be able to track your progress, that we’ll be able to show you what it means for you personally to be getting better and better at these skills, and then what’s next. So that sort of… It’s more like we were gonna give people schwag. Like this goes beyond schwag. This gives them the reward of, “Oh my gosh, I’m elevating my career. I can actually track and see how well I’m doing in my job,” and maybe even share back learnings and create these feedback loops with that personal AI coach and tutor.

Stefan Tornquist: Yeah, as long as, you have to worry about privacy, because otherwise people are like, “Oh no, it can see that I’m not working very hard,” or whatever, or that, “I’m not confident in my writing.” But yeah, as long as it’s couched it completely is like, this is here to help, this, you know, it’s like one of those HR evaluations that’s anonymous and so forth.

Tessa Burg: Yeah, no, I like that. I wonder if we make them very personal to people so they’re in their own environments as opposed to something that is shared out publicly across the company. It’s like this–

Stefan Tornquist: Yeah.

Tessa Burg: Something that’s only for them and very private to them, but is still giving them that tracker and that grade.

Stefan Tornquist: Yeah. And by the way, I think this is comfort with AI and sharing data with AI is gonna follow the same path as with marketing, where we would like more privacy, we would like more security, but at the end of the day, I want to use Amazon one click all the time everywhere, right?

Tessa Burg: Yeah.

Stefan Tornquist: And you know, with AI, if it can be truly– To me, I’m curious to see whether the new iPhone 18 in 2025 is the watershed moment where we have a personal assistant that’s really, really useful ’cause if you could just tell your phone what you want to do as opposed to– Anyway. I think that AI, comfort with AI and data is probably gonna follow that same track.

Tessa Burg: Yeah. Yeah, and I think too, it would be okay for people to be messy and make mistakes in other places and then choose what they do kind of share out and then allow them to celebrate what they wanna celebrate openly.

Stefan Tornquist: A hundred percent. And you know, AI’s really good at assessment so it can, instead of giving you, here’s a multiple choice test, or something that evaluates your new R skills, let’s talk about R and developing an app for five minutes. In the same way that you could immediately tell whether somebody knew what they were talking about, AI kind of can, too, because you can see the flags.

Tessa Burg: Yeah. And I love that example. That’s exactly how we do evaluation, is we have people describe to us the process that they would use and then just have them show it a little bit. And you can tell and very quickly it’s like, “All right, you’ve either never done this before. You are kind of messy and unstructured in your approach.” Or, “Man, this is like second nature.” Like, “You just—

Stefan Tornquist: Yeah.

Tessa Burg: —described that process in five seconds and now you’re executing it and this is clearly something you do day in and day out.”

Stefan Tornquist: Yeah.

Tessa Burg: So I love to, yeah anyway, thank you for that amazing idea. I love the personalized approach and just knowing that AI can do that. Everyone can create their own personal road to progress, reward, seeing where I’m at, and they can even decide how much of the messiness, like maybe I need to go off and learn some things on my own, and then I’m gonna practice in this tool and get the assessment. Like turning it on and off can also be very self-directed.

Stefan Tornquist: Um hm.

Tessa Burg: Now I’ve completely lost track of the interview. That was such a bright… I guess this is why you’ve spent your career in research, training and development for moments like–

Stefan Tornquist: Well it’s been fun, you know, with… And digital marketing was just starting to get boring. Like—

Tessa Burg: Yeah.

Stefan Tornquist: —Were like, “Well, will they turn off cookies?” And then AI came along at just the right time to sort of upset the apple cart.

Tessa Burg: Oh, I agree. I was getting so tired of only talking about marketing automation and like what–

Stefan Tornquist: Yeah.

Tessa Burg: segments should get which emails like gosh. I just wanted to like bang my head into a wall.

Stefan Tornquist: Yeah, we made some hay from customer experience for a while.

Tessa Burg: Yeah. I think AI is a very exciting turning point for marketing for tech. Like we are standing up an innovation group and I was just telling our CEO today, “There is no lack of excitement.” Now what’s needed is priorities and alignment to what our clients value and care about. Because with AI we can legitimately create or develop anything. And for businesses like ours in service, that has not always been true. And that’s true in really any professional services organization, not just agencies. That can be true for accounting companies, for consulting companies. You can create your own automated solutions. So now it’s a matter of priority, management, scalability, and that’s where training and change manager on adoption becomes so important because literally create anything but is what you’re creating–

Stefan Tornquist: Yeah. And I don’t think a lot of people really Grok the degree to which this has changed technology, like technology production so profoundly. Like there’s a reason that 90-plus percent of programmers are using Gen AI, because it saves them just enormous swabs of time. Yeah, I mean, you can do things in minutes that would take days.

Tessa Burg: I agree. And what excites me most is it’s starting to show or shine a spotlight on engineers’ creativity. Before, I feel like you had to be the type of engineer who’s also very good at communicating to get recognized. And now I think this has democratized that creativity and talent, and I love seeing like even our team evolve and really be more in the spotlight. So it’s–

Stefan Tornquist: That’s really interesting.

Tessa Burg: Yeah, it’s been– It’s funny ’cause a lot of people have said like, “Oh, this will replace developers. This will replace coders.” And it’s like, well maybe the bad ones, which I don’t know very many of, but like the good ones? This is really giving them opportunities that they’ve never had before and giving them a space and more time for the creativeness that they’ve always had inside of them, but they’ve been doing so much production-level code and fixes and–

Stefan Tornquist: And it’s the same for marketers, right? I mean, you know, tell me a marketer who’s like, “Yeah, you know, I spend 80-90% of my time really on strategy, brand value, that kind of thing.” Like nobody, not even the CMO. Like, I spend 80% to 90% of my time on, you know, such minutia of like, you know, “I wrote this email, but now I’ve gotta get it trafficked,” and, “Here are the nine versions.” And, “Oh, you know it’s hitting a spam filter on whatever.” That’s marketing day-to-day is, you know, fine grain detail.

Tessa Burg: Yep.

Stefan Tornquist: This is the kind of stuff that AI can help with.

Tessa Burg: Yep. Well we’re over time ’cause I got really caught up and excited. This conversation has been inspirational and highly engaging. If the audience wants to be inspired by you and reach out with any questions, what is the best way to reach you?

Stefan Tornquist: Well, I’m pretty much off Twitter these days. I still have an account, but feel free to– I’m on the Econsultancy website or LinkedIn. Sktornquist, I think. But very happy to hear from anybody and talk about the study at more length or about skills development.

Tessa Burg: Well thank you so much, Stefan, for being our guest and taking the time today. It’s been a great conversation and–

Stefan Tornquist: Tessa, thank you.

Tessa Burg: Yeah, I hope that we get to talk again soon.

Stefan Tornquist: Me too. Bye everybody. Thank you.

Tessa Burg: And if you want to hear more episodes of “Leader Generation,” you can find them at modop.com. That’s M-O-D-O-P dot com. And we will keep exploring and keep tracking what’s next in AI skill development and change management, and continuing to help you increase AI adoption in a way that is valuable to your company and your customers.