Donald Taylor’s 2025 Learning & Development Global Sentiment Survey reveals AI’s explosive impact on workplace learning, but the real story is how professionals are moving from buzz to action. They’re using it for everything from quick content fixes to reshaping skills strategies. Yet, the timeless challenge remains: proving L&D’s value. Are we chasing trends or driving results?
If you prefer to watch the podcast on YouTube:
Introduction
Brian Washburn: Hello everyone, and welcome to another episode of Train Like You Listen, a podcast about all things learning and development in bite-sized chunks. I’m Brian Washburn. I’m the co-founder and CEO of Endurance Learning. I’m also your host for this podcast. Endurance Learning is an instructional design shop that can help you put together any kind of training program from classroom to instructor led to e-learning.
But enough about me. Today I’m joined by Donald Taylor, who is the lead researcher of the Global Sentiment Survey, which we’re gonna be talking about in just a second. Hi Donald. How are you?
Donald Taylor: I’m very well. Good to be with you.
Brian Washburn: I’m so excited. It’s been a little while since we’ve talked about one of the global sentiment surveys and, we’re gonna get to that in just a second. But before we do I would love for our audience to get to know you just a little bit better. If you could give us and sum up kind of your essence using a six-word biography.
Six-word Biography
Donald Taylor: I’m not sure it’s my essence, but I, I’ll try to give you a six-word biography. Curious about learning, performance and people. That’s six words. It’s more the professional side than my essence, but certainly I’d say curiosity is at the heart of everything I do.
Brian Washburn: And that’s perfect. Sometimes people try to sneak in an extra word with the ampersand, but–
Donald Taylor: I wouldn’t do that to you, Brian. I would not try to force in an extra word. Come on.
Brian Washburn: I appreciate that. And so thank you so much for carving out some time and talking. You are based in the UK. You’re in Wales and…
Donald Taylor: Wales right now. Yeah.
Brian Washburn: And so it’s great to be able to have this connection, and I know that we have had a chance to talk about this several times over the past few years. But for people who are unfamiliar with this global sentiment survey, could you just set the stage for us a little bit and tell us a little bit about what it is, you know, why you invest so much time and energy in compiling this report year in and year out.
This isn’t just like a one time thing, it’s something that you do. And so and perhaps more importantly maybe just kind of in general, what insights can this survey offer to anyone who works in the field of learning and development?
Donald Taylor: That’s, that’s a lot of questions all rolled into one. Brian, I’m gonna try to answer the background stuff and I, I’ll forget by the end of my answer, the rest of the questions you can prompt me.
Brian Washburn: Sure.
What Is the L&D Global Sentiment Survey?
Donald Taylor: So the background stuff, I’ve been doing this for 12 years now, and it started off as a very simple idea. Ask people what they think is gonna be important next year.
After a bit, it started gathering momentum after three or four years, and there’s a good reason for that. If you ask one person one question, the question we’ve always asked—there are 5 questions altogether—but there’s one we’ve always asked which is “what will be hot in workplace learning and development next year?”
Now, if you ask one person that question, it’s a “So what ?” questions. Its one person’s opinion, but we get something like 3,500 people responding from one hundred countries each year. And when you build up that level of data, year on, year in, year out, you start to see start predictable patterns emerging. Now, there are reasons why you might want to be cautious about the data, and we can look at that in in a minute if you like.
But it is interesting that there are predictable things that are different between countries, predictable things, the way things go up and down. And I think that’s what people find interesting, that what looks like individual choice becomes a predictable matter. And we can look at it and we can use that, and we can use that predictive power to say, well. This is what people are saying now about what they think is important, but what were they saying last year and where will it go in the future?
And the reason why this is important to come back to the data thing is that it’s not a, it’s not like a political poll. If you were in the run up to an election, you’d want to check the opinions of people from every demographic, every age group, every ethnicity. Here we just ask people generally if you want, if you’re interested, respond to the survey.
So the people we get responding are the people who are super interested in what’s happening and what’s new. They’re on the Everett Rogers diffusion of innovation curve. They’re the early adopters and the innovators.
Brian Washburn: Mm-hmm.
Donald Taylor: And what we’re interested in doing is looking at, well, what are those people thinking? Because it’s not true that everything they think of as exciting goes on to be widely adopted. But it is true that everything that gets widely adopted at some point was thought of as being exciting by these people.
So what I’m interested in doing each year is saying, well, here’s what’s happening now. This is what people are thinking. Where does that say we’re gonna be in three and a half years time, four years time, when the world wakes up or doesn’t wake up. So in other words, this stuff we’re talking about hype. Is it reality? What happens next? So that’s why I find it exciting to do each year. You asked me a whole bunch of other questions, which I can’t remember now. You’re gonna have to ask me again.
Brian Washburn: No, no, no. It’s perfect because I wanna jump into this because you’re talking about kind of what some of the, the people who get really excited about some of the, the earliest stuff, maybe get super nerdy about some of this stuff.
Donald Taylor: Yeah.
Brian Washburn: And it’s, it’s stuff that people like me end up adopting two years later.
So I’m really curious about this. And, and for this particular survey, it was the second year in a row that AI has dominated the interest of the respondents, and it’s such a broad topic. Right? So AI is such a broad topic with such a wide range of applications. When you’re taking a look at some of these responses, what does it mean that AI dominated this survey?
AI Dominating the Global Sentiment Survey
Donald Taylor: It’s interesting that it’s only been two years that it dominated. The thing — AI has been on the survey since 2017 and it went through the normal sort of pattern, which is that people were enthusiastic about it and then it started dying away three, four years ago.
So 2021, it was languishing at about 5%. And, you know, it was heading towards people just not being interested in AI at all.
Brian Washburn: Sure.
Donald Taylor: Then Gen A, Gen AI came out, right? We had, we had chat GPT launched and it came out just before the launch of the survey. And then the survey was launched like a week later. Some people picked up on it, some people didn’t. Interestingly enough actually, of the different work groups that we had — consultants, vendors, educators, other, and workplace learning and development. Of those groups, workplace learning and development was the least excited about AI in the first year.
Then the second year, which was so that was the, that was the 2023 survey, 2024 survey. That was when everything just went through the roof with AI and it leaped up to 21.5% of the survey, which, just to explain that very quickly, it’s not that most people, 80% of people, didn’t think it was important. We ask people to vote for one, two, or three things from a list. Almost everybody votes for three things. So if I say 21.5% of the people thought it was going to be hot, that means that three times that many people voted for it. About two thirds of the people voted for it. So it was overwhelmingly popular last year.
And then this year it was even more popular by a small amount. So what does it mean that people were enthusiastic about it being hot this year? By itself? Those numbers just say, “oh, people are talking about AI and they’re excited.” Right?
But if you look at the other question that we asked, which was the other key question we asked, which is, what’s your key challenge this year? Last year? , we get all these answers in and we get like something like 28 to 29 to 30,000 words, which is a small book of, of responses in. We look at all those words and we just categorize them into, into categories.
And last year technology was the top concern and that was led by AI. This year it was strategy and skills and talent, which had been number one before last year. That was one thing. The other thing is in the answer to the question, what’s your biggest challenge this year? And this is a free text box, right? You can write whatever you like in that box, and some people write a great deal, but last year, 60 people just pressed three keys on their keyboard. AI return, bang. That was it. And that was all they wrote for the biggest challenge.
This year only three people did that. And yet we had more words being written this year. So we had people write and say, “well, I like to use AI, but it’s problematic. We’ve got this challenge here.” Whatever. People were describing it in much more detail. And for other reasons, I think there are lots of bits of information that you can look at underneath the surface, but I think that’s a really good example of what I think happened. So yeah, people are excited about AI but in different ways, last year and this.
Last year, it’s what I call the AI shock. People were just, they’d had a year plus of just literally endless news about AI. By the time we did this year’s survey, people were a bit more coherent and cogent about it. It’s more “okay, it’s still happening. But I think I have an idea of what I need to do.” So I think the enthusiasm is there. There’s a more realistic view of what AI can offer this year.
Brian Washburn: You know, AI itself, right, those two letters, can mean a world of different things.
Donald Taylor: Sure, sure.
Brian Washburn: And when it comes to AI and learning, did you get a sense for how AI is being used among the different learning and development professionals?
How Are L&D Professionals Using AI?
Donald Taylor: Certainly. Although that’s less to do with this. I mean, we did get some information from this survey, but actually I did a series of three reports last year on the use of AI in L&D, which started on anniversary of ChatGPT being launched. And then we did another one in April and another one in October. So that’s, that was three.
Another one will come out in probably September of this year, 2025. So we can keep that annual cadence of looking at exactly what you’re saying. How is AI being used? And my colleague on that report, Egle Vinauskaite, has a really smart thing that she calls the complexity scale, where we go from simple use to to complex or sophisticated use. And sophisticated uses are when you are tied into the business, you know, using it for things like particularly skills management across the organization. And a simple use is using it for the internal efficiency of the learning and development department.
We know, and it’s not just me saying this, there’s lots of bits of research. Dr. Phillipa Hardman points to this and other people. We know that the vastly predominant use of AI in L&D is for internal efficiency in the L&D department, which means typically generating content faster and handling stuff like meeting notes, using it to brainstorm and so on.
And when we asked people in the report that came out in October of last year, what is the expected benefit that you’ll get from using AI in L&D? And we gave people a list of 12 things. They could choose any number Most people chose three. The thing that got selected way above everything else from that list was create content faster.
Now this is not necessarily what people are using it for, but it’s what they expect use it for which is kind of crazy because there are so many things you can do with AI, but the one they’re focusing on is faster content creation.
Brian Washburn: And so, you said that that’s not necessarily what they’re using it for, but it’s what they expect to use it for.
And I certainly use it to help me with, you know, coming up with a lesson plan or coming up with a case study or whatever. When you come to some of the more complex ways that people are using it, can you give us a little bit of a taste of, of what are some of the other things that people are doing. And this might dovetail into to my next question about showing value as well.
So I’m kind of curious what AI can do beyond, but you know, for anybody who’s listening, like, “yeah, I use chat GPT. I’ll create, you know, lesson plans, or I’ll create, you know, this or that”, what else should they be using it for? What else are other people using it for?
How Should L&D Professionals Use AI?
Donald Taylor: I wouldn’t say what they should be using it for, because everybody’s context is different. It’s up to them to decide what’s right. But I would say if you start with the question being, how should I use AI in L&D. what we naturally do is we think, how can I make my existing processes faster?
Brian Washburn: Yep.
Donald Taylor: And that’s a completely natural response to the introduction of new technology, which goes back generations. If you go back to the very first stone buildings in Egypt in about 3000 years BC the first columns are carved to look like clumps of papyrus stuck together because that’s how you’d hold upa– that’s what a column would look like. By the way, that then got morphed into the existing columns that we have today. So the reason you have flutes in your column is because some guy in Egypt 4,000 years ago thinking it should look like a bunch of papyruses.
So think about it differently. Don’t think “what should I be doing to speed up L&D?”, but “what’s the fuel of AI?” And the fuel of AI is data. Alright, what data have you got in your organization? Well, you’ve got a whole bunch stuff.
The Fuel of AI Is Data
Donald Taylor: One example is in a contact center you have tens of thousands of hours of transcripts, of communication with people, phone calls with people lasting I don’t know, 5 minutes to 30 minutes, and those phone calls have all been written up and by and large, they’re tabulated. What’s a good one? What’s a bad one?
If you run that through a machine and you can use that as the fuel for simulation and training system that will get people who are new to the contact center up to speed on how to do good calls. In fact, you can use it to give them training on new products that have been introduced. You can use it to help people when they’re in the middle of a call, listen to you speaking, listen to the person speaking, and give you prompts as you’re speaking about what you should and shouldn’t do next.
This is not hypothetical. This has been happening for at least seven years and it’s very effective. Back in the UK, HSBC has used this system in order to improve customer relation scores, customer satisfaction scores on calls by 10%. Whether that’s a call taking place in the UK or from a contact center in India.
So that’s a different way of looking at it. If my job is performance, is it the best way to help people in a contact center? Is the best way to train them or is the best way to give them some sort of simulation? Turns out it’s actually very effective giving them a simulation. So that’s one example.
And there are lots of other things. Where else is there data in the organization? Well, there’s, there’s data lurking in people’s heads that would be really good if you could surface it, put it into a Slack channel. There are startups at the moment doing exactly that, getting information from conversations that take place in Slack, then make a coherent, referenceable source of it and putting it back in so that next time that question comes up in Slack, a little bot can come in and say, “Hey, if you’re asking this question, here’s something that can help you with the answer.”
The Data in Conversations Reveals Your Organization’s Skills Heat Map
Donald Taylor: Again, that’s the question of where’s our data? The data is in the conversations taking place between people. Actually it’s all over. This one happens to be in Slack, but there are lots of such conversations and so on. The classic case though, where AI is really being used is to say well, we’re gonna use it to understand our skills gap. So we’re gonna understand what skills have we got, what skills do we think we need, and how do we bridge the gap?
That is something that can be done, and I’ve done that in the past — 20, 25 years ago. But it’s extremely time consuming process to do it manually. Get AI to help you and there are some great organizations doing this already. Get AI to help you with that, and you’re suddenly gonna find yourself with the ability to understand the skills heat map of your organization and focus your energies where the skills needs are greatest. Again, that’s not looking at from what do I need to improve in my current process, but if I could do anything with the data in my organization to improve performance, where would I start?
Brian Washburn: I love that reframing of the question, and it seems like we need to do another podcast that is dedicated strictly to AI and AI applications.
But I do wanna get back to the survey because in this survey, you know, beyond AI, you made the point that showing value returned in a big way in this year’s survey. You know, showing value is essential to what we need to do as L&D professionals, because that’s how we can make the case for our budget really, for resources.
And yeah. In this survey, you know, showing value, it does come in a distant seventh place behind things like reskilling, upskilling, personalization, adaptive delivery, consulting more deeply with the business. You know, it’s good that it’s on there. Why do you think that it fell off for a little bit and then even in its emergence it, it’s kind of a distanct seventh.
How Important Is Showing Value in L&D in 2025?
Donald Taylor: It’s quite a long answer, so stay with me. First off, typically in the course of an option being on the survey—and we give people 16 options to choose from—it goes down over time. So mobile delivery was our very first number one back in 2014, and it fell down over time. And I took it off a couple of years ago because it went from being something super exciting to being something that was business-as-usual.
Other things sort of start off exciting because we’re excited about it and then we realize it’s a bit harder than we thought. Like curation. Curation was, was super hot, but over time people realize, actually that’s, that’s pretty hard to do. “I’m gonna turn my attention elsewhere and, and I’m not gonna bother about that so much.”
But either way, things start off with a high score and then drift down. Now there are three options, which I call the “Value Trio” of these 16, which behave very differently to that over time. The three options are:
- Consulting more deeply with the business
- Showing value, and
- Performance support
And they have been predictably different performers for a period of five years between 2019 and 2023. They didn’t go down. They stayed resolutely in the middle of the table. In fact they edged slightly upwards.
So last year, when we had that AI shock that I was talking about, they all dropped down and it, it was a really startling drop. Almost everything dropped down because AI sucked up the votes and they dropped down as well. This year, the only things that rose on the table were personalization very slightly, and those three options, which are all linked to people’s minds around value and AI itself.
And when I say they’re linked together in people’s minds, I think what happens is this. People think, well, okay, we’re gonna consult with the business and see what the problem is. We’re gonna do something that supports performance, and then we’ll show the value afterwards. So it was notable that these three things are the only things that globally jumped back up again.
I think what happened was there was this shock of AI. People were like, deer in the headlights. Didn’t, couldn’t move in 2024. 2025 people said, “okay, look, let’s, let’s get back to what’s important. We know value’s important.” The reason I say that is that in different geographies, one other option, depending on the geography, bounced back as well, and it was always something which had been traditionally well supported in that geography.
So, in the Netherlands, for example, it was performance support, which bounced back this year. In Ireland., it was coaching and mentoring, which bounced back this year. And it’s, it’s almost like people last year were just bewildered, and then this year thought, no, we’re gonna get back to what’s important. Globally, that’s showing value, and for each country there’s a specific thing they bounced back on.
Now the other question you asked was, “but Don, look, we know value is important, but why is it never higher than some mid-table position?” And the answer to that to me is that we have more than one group of people voting in this survey. We have some people who vote early who are super enthusiastic and who are absolute flag wavers for what’s they consider to be super important in L&D.
They are typically the people who vote for showing value. And I follow the votes each day as the survey comes in. Showing value always starts off high and then it trails off a bit as the rest of the votes come in. When the majority– see it’s almost like the innovators want showing value to be high, but then the majority of the rest of the votes come in and it starts to sink down.
And I think that is the answer to your question, is that showing value has a really strong core of support. But it’s not widely accepted amongst my voters as being super important.
Brian Washburn: The way that you frame that makes so much sense in terms of, you know, it depends on who’s answering and, really kind of the, the motivations of the answer as well. You know, the last thing I’m gonna ask you about the report here is in your forward you say that one of the primary aims of this report is to provoke discussion. What do you think is an essential discussion that people in L&D should be having based on some of the data that you found here?
The L&D Global Sentiment Survey Should Provoke Discussion
Donald Taylor: I actually put questions in the report in a few pages to try to stimulate this thinking. And that was because I was kind of asking myself, why are people still answering this? Why are there still people interested in the results of the report? And they told me, “well, it’s ’cause we use this very often in our management group., in our L&D team. We just look at the report, we say, well, you know what’s happening and why.”
And I say, I think the first question I suggest people ask is why, right? And I try to present answers to why, but I’d also ask them to think about why do they think it’s true because I might be wrong. And then I’d ask them to think about– second question would be, does this apply here? So it might be that, for example, in the UK showing value for some reason, is seen as being a very highly rated option. So you might ask yourself, “well, does that apply here? Does that apply in my country? Does that apply in my organization? Is it something which is really important here or is there something else which is more important?”
And I think just using it as a stimulus to that sort of very simple conversation is worth having. But I’d also ask you, you can go beyond that because I think AI is a very good example of something where something has, and actually reskilling, upskilling, which is number two on the list, is another example of something which has been incredibly well supported, but means different things to different people.
And I think it’s worth asking yourself, “how much is this hype? And how much is it reality?” And the answer is always, it’s a bit of both. So ask yourself, “well, okay, where is there value in this thing? What do we have to be cautious about? Who can we listen to that we trust?” And then you start opening up that conversation with your team.
You’re starting to have surfacing this, I hope, useful information sourced from within your team, which will enable you to help you set your strategy for the year.
Brian Washburn: Thank you so much for offering a little bit more insight because it’s one thing to look at data and just kind of be fascinated by the data and get lost in the data.
Donald Taylor: Yeah.
Brian Washburn: It’s another thing to be able to use it as a jumping in point to, to have some of these conversations. I would love to talk with you more about all sorts of things, but we’re gonna run outta time here. So just real quick, if you could let people know where they can find out more information about you and or about the, the global sentiment survey.
Donald Taylor: If anybody wants to find out more information about me, good luck to them. I mean, I’m all over the place. Yeah, LinkedIn is the best place for that. And my website is Donald H. Taylor. And I’m Donald H. Taylor everywhere. DonaldHTaylor.co.uk. DonaldHTaylor on LinkedIn.
Brian Washburn: And, well, I look forward to hearing from you. Donald, thank you so much for, giving me some time and, having this conversation because it is always really interesting to, actually dive into data, especially data around how are people feeling about what is important these days. Thank you everyone else for, for listening.
If you found this helpful and do you wanna. grab a link to this recording and put it on LinkedIn because this is a really, really helpful conversation that I think lots of people should be having. If you ever wanna nerd out about anything that is training related let’s grab virtual coffee. Send me a LinkedIn note. And until next time, happy training everyone.
