Hosted by Jeff Walter, Founder and CEO of LatitudeLearning
As artificial intelligence reshapes nearly every aspect of learning and development, organizations face an important question. Is faster content creation enough, or should the industry be focusing on creating learning experiences that genuinely improve performance?
That question sits at the center of this Training Impact Podcast conversation between Jeff Walter and Scott Hewitt, CEO of Real Projects. Over more than two decades, Real Projects has evolved from a small digital consultancy into an award-winning eLearning company that specializes in helping organizations deliver engaging learning experiences at global scale. Rather than chasing every new technology trend, the company has built its reputation on combining practical production processes with thoughtful instructional design, localization expertise, and an unwavering focus on quality.
Throughout the discussion, Hewitt explains why successful digital learning is not simply about creating more content. It is about creating better content, delivering it in the right language, organizing it effectively, and ensuring learners receive information that is meaningful within their own context.
Like many successful entrepreneurs, Scott Hewitt did not begin with a perfectly defined business model. After experiencing multiple redundancies during his career, including roles in multimedia production and the oil and gas industry, he decided to create his own opportunities rather than continue searching for employment. What began as a general digital services company gradually evolved into a highly focused learning organization.
Initially, Real Projects accepted virtually any type of digital work. The company developed websites, managed projects, built intranets, and created custom multimedia applications. While that broad approach generated valuable experience, it also made growth more difficult because the organization lacked a singular focus.
Eventually, the company recognized where it created the greatest value. By concentrating exclusively on eLearning development and off-the-shelf learning content, the entire organization became aligned around one mission. That clarity benefited both employees and customers while allowing the business to continuously improve its production methods.
Hewitt reflects that entrepreneurs often discover their true niche only after experimenting with many different opportunities. While those early years may seem unfocused in hindsight, they frequently provide the experience needed to identify long-term strengths.
One of the more interesting themes from the discussion is how experience outside the learning industry shaped Real Projects’ approach to instructional design.
During his time working within oil and gas operations, Hewitt was exposed to highly disciplined production environments where version control, project costing, quality assurance, and repeatable processes were essential for success. Those operational principles eventually became central to how Real Projects develops learning content today.
Rather than viewing eLearning as a purely creative exercise, the company approaches development much like a manufacturing process. Every stage is documented, repeatable, measurable, and continuously refined.
That production mindset becomes particularly valuable when organizations need to scale hundreds of courses across multiple languages while maintaining consistency. Instead of sacrificing quality for speed, Real Projects focuses on building reliable systems that make both possible.
This philosophy reflects a broader trend across learning and development. As organizations expand globally, repeatable production processes become just as important as instructional creativity. Creativity captures attention, but consistency builds trust across an entire learning ecosystem.
One of the strongest messages throughout the conversation centers on a distinction many organizations overlook.
Translation and localization are not the same thing.
Many companies assume that translating English training into another language is sufficient for global learners. According to Hewitt, that approach often misses the very goal organizations hope to achieve.
Real Projects currently offers hundreds of learning courses across multiple languages, including German, French, Italian, Brazilian Portuguese, Latin American Spanish, Indonesian, Thai, and Vietnamese. Expanding into those markets required much more than simply converting words from one language into another.
Localization involves adapting content so it feels natural to native speakers. That includes adjusting terminology, cultural references, graphics, audio, typography, abbreviations, and even subtle differences in tone.
For example, an artificial intelligence course translated into German cannot simply preserve the English abbreviation “AI.” Native German speakers commonly use “KI,” requiring updates throughout graphics, illustrations, navigation, and supporting materials. Similar challenges appear in cybersecurity training, where technical terms such as phishing can easily be mistranslated without careful review.
These details may appear minor individually, but collectively they determine whether learners perceive content as professionally designed or mechanically translated.
Artificial intelligence has dramatically accelerated multilingual content production, but Hewitt cautions against assuming AI alone can guarantee quality.
Instead, Real Projects has developed a production workflow that combines AI efficiency with multiple layers of human review.
Content is first processed using carefully selected language models. Native-speaking editors then review the output for accuracy, tone, technical terminology, and cultural appropriateness. Additional reviewers validate those edits before final publication, creating multiple quality checkpoints throughout the localization process.
This approach recognizes an important reality. Different AI models often produce different results, particularly in less common languages. Human expertise remains essential for identifying subtle issues involving tone, phrasing, sentence structure, or regional preferences that automated systems frequently overlook.
Rather than replacing localization specialists, AI has become a powerful productivity tool that allows experts to spend more time refining quality instead of performing repetitive translation work.
Equally important, Hewitt emphasizes that successful AI implementation begins long before translation occurs. Writers are encouraged to avoid colloquialisms, regional idioms, and culturally specific expressions that rarely translate well across languages. By creating globally friendly source material from the beginning, every subsequent localization becomes both easier and more accurate.
That philosophy reflects one of the recurring themes throughout the discussion. Artificial intelligence delivers its greatest value when organizations improve the quality of their inputs instead of expecting technology to compensate for poor content.
While artificial intelligence often dominates conversations about the future of learning, Hewitt argues that many organizations overlook something far more fundamental: metadata.
Every course contains information beyond its title and description. Learning objectives, topics, competencies, audiences, business functions, and classifications all contribute to how learning content is organized and discovered. Unfortunately, much of this information is often inconsistent because it has historically been created manually by different people using different interpretations.
Real Projects has invested significant effort into using AI to improve metadata creation before learners ever interact with content.
Hewitt compares the challenge to organizing a music library. A physical record store could only place an album in one section. A digital catalog allows the same album to appear in multiple categories based on genre, mood, artist, decade, audience, or countless other characteristics.
Learning content works the same way.
A single cybersecurity course might belong within compliance training, onboarding, information security, remote work, IT operations, and leadership development simultaneously. Proper metadata allows learning platforms to surface that content naturally regardless of how employees search for it.
Rather than relying on individuals to categorize every course differently, AI provides a consistent framework that applies identical classification rules across an entire learning library.
The result is more reliable search results, stronger personalization, and a learning experience that feels significantly more intelligent because the underlying information has been organized consistently from the beginning.
One of the most thought-provoking moments of the conversation centers on how organizations can learn from the search behavior inside their learning platforms.
Hewitt notes that internet search engines have spent decades understanding user intent. Whether someone searches for a product, a service, information, or a solution, companies like Google analyze those searches to determine what users are actually trying to accomplish.
Learning platforms collect similar information every day.
Employees constantly search for courses, policies, procedures, certifications, and answers to workplace questions. Yet many organizations ignore this data once the search is completed.
Instead, Hewitt believes search activity should become one of the most valuable sources of business intelligence available to learning leaders.
Search trends reveal knowledge gaps before they become performance issues. They identify emerging skills employees want to develop, recurring business challenges, and opportunities to expand learning libraries with highly relevant content.
Combined with well-designed metadata and AI-powered analysis, search behavior creates a feedback loop that continuously improves the learner experience.
Rather than guessing what learners need, organizations can begin making decisions based on actual demand.
Another compelling theme throughout the discussion involves learning analytics.
Within Learning and Development, some metrics are often dismissed as “vanity metrics.” Course completion rates are one common example because completing training does not necessarily prove knowledge transfer or business impact.
Hewitt offers a different perspective.
Instead of dismissing individual metrics, he encourages organizations to understand the story those metrics tell when combined with other information.
Completion data may reveal technical problems rather than learning problems. If one region consistently fails to complete courses, perhaps there is a network issue. If learners consistently leave a course after three minutes, that may indicate a design problem rather than a motivation problem.
Viewed in isolation, the numbers have limited value.
Viewed together, they become diagnostic tools that help organizations improve infrastructure, instructional design, learner experience, and operational performance.
This philosophy mirrors broader trends in data analytics. Individual data points rarely provide meaningful insights on their own. Their real value emerges when connected to larger patterns and business objectives.
Artificial intelligence is woven throughout the conversation, but Scott Hewitt offers a refreshingly practical perspective on its role in Learning and Development. Rather than viewing AI as a replacement for instructional designers or learning professionals, he sees it as a technology that enhances the work people are already doing.
Throughout the discussion, Hewitt explains how AI helps Real Projects accelerate localization, improve metadata, organize learning taxonomies, analyze learner behavior, and scale content production more efficiently. These capabilities allow organizations to accomplish in weeks what once required months, while maintaining consistency across hundreds of learning assets.
However, the technology is only one part of the equation. High-quality learning still depends on experienced writers, instructional designers, subject matter experts, and native language reviewers who ensure the content is accurate, engaging, and culturally appropriate. AI may automate repetitive tasks, but it cannot replace the judgment and expertise required to create learning experiences that genuinely influence performance.
That philosophy reflects the broader approach behind Real Projects. Technology alone does not solve learning challenges. Lasting success comes from combining thoughtful instructional design, well-defined production processes, quality data, human expertise, and the right technology into a scalable system that consistently delivers meaningful business results.
As the conversation draws to a close, Hewitt shares where Real Projects continues to invest its energy.
Cybersecurity remains one of the company’s fastest-moving content areas as threats evolve rapidly and organizations require continuously updated learning.
Compliance training continues to shift toward shorter, more engaging learning experiences rather than lengthy mandatory courses.
The organization is also seeing growing demand for de-escalation training, particularly within retail, hospitality, restaurants, and customer-facing industries where employees increasingly encounter difficult situations.
Rather than developing content based solely on internal assumptions, Real Projects continues to work closely with partners and customers to identify emerging learning needs before expanding its growing multilingual library.
That customer-driven approach reflects the same philosophy that has guided the company for more than two decades: solve real business problems through practical, high-quality learning experiences.
This episode of the Training Impact Podcast demonstrates that effective learning is about far more than producing content faster.
Scott Hewitt explains how Real Projects combines thoughtful instructional design, disciplined production processes, AI-assisted localization, structured metadata, and intelligent analytics to help organizations deliver learning experiences that scale without sacrificing quality.
Perhaps the most important takeaway is that artificial intelligence delivers its greatest value when it strengthens strong processes rather than replacing them. Organizations that invest in quality content, consistent metadata, meaningful taxonomy, and learner-centered design will be far better positioned to take advantage of AI than those searching for shortcuts.
As learning technologies continue to evolve, the fundamentals remain remarkably consistent. Clear communication, thoughtful organization, reliable data, and a relentless focus on learner outcomes continue to separate effective learning programs from those that simply deliver information.
To learn more about Real Projects and its multilingual eLearning solutions, visit https://realprojects.co.uk/.
For more from the Training Impact Podcast, follow us on Social Media:
https://t-sml.mtrbio.com/public/smartlink/trainingimpactpodcast
Jeff Walter (00:05)
I’m Jeff Walter and welcome back to the Training Impact Podcast, where we explore scaling performance through training infrastructure. My guest today is Scott Hewitt. Scott is the CEO of Real Projects. They’re an award-winning e-learning company that has spent more than two decades helping organizations transform complex knowledge into engaging digital learning experiences. Scott, welcome to the program.
Scott Hewitt (00:27)
Thanks very much, Jeff. Pleased to be here and have a conversation with you today.
Jeff Walter (00:32)
So as as my as listeners know, first thing I’d like to ask is, so Scott, how did you become CEO of Real Projects?
Scott Hewitt (00:38)
Yeah, well,
Jeff Walter (00:39)
What was the journey for you?
Scott Hewitt (00:41)
yeah, it’s probably an interesting journey. I’d worked for a multimedia company that was part of a television company in the UK and had been made redundant. And then I’d had a couple of different jobs. I’d been made redundant again, having worked in actually in oil and gas, North Sea, oil exploration.
and having tried to get some employment in a number of organisations, I actually just wasn’t having any luck getting a job. So I decided actually I was just going to try and do it on my own. So I’ve got a background working in multimedia and e-learning. So I actually started the business by sending an email out to everybody that I knew.
And I managed to get some work and that’s essentially how it started really. And Real Projects was to begin with was a vehicle for sort of any sort digital work that I could get. And we were doing all types of things. We were doing project management, we were doing custom work, we were doing website work, intranets, we were doing all different types of custom work. And eventually, like a lot of people do, we sort of focus down into the areas that we were doing. it was…
Mainly it was website and custom development work. But over the last four to five years, we’ve just really focused down on our off the shelf learning development. And that’s the area that we’ve really focused in and I’ll probably touch you the most about today.
Jeff Walter (02:20)
Yeah, d so just going back for a second, ’cause it’s an i that’s an interesting background.
Scott Hewitt (02:24)
Yeah.
Jeff Walter (02:25)
You said well, I and it’s interesting and the reason I want to focus go back on the background just a little bit is yeah, there’s so many folks out there that they j they see somebody doing what they’re doing, whatever it is. And if they’re successful, it’s it’s there’s this kind of belief that you were born to do that, right? That that’s
Scott Hewitt (02:42)
Yeah.
Jeff Walter (02:43)
that that that’s you, that’s always been you, that’s who you are. And and and you said and you said it just in
you know, quickly, but it was like you were in multimedia. So okay.
Scott Hewitt (02:53)
Yeah.
Jeff Walter (02:53)
You know, I mean w you know, you were with a a broadcasting network. So okay,
Scott Hewitt (02:56)
Yeah,
Jeff Walter (02:58)
like that maps to e learning and the thing and and the stuff. But then there was this piece in the middle that was like and then I went
Scott Hewitt (03:04)
I
Jeff Walter (03:04)
to work in oil and exploration offshore in the South North Sea. That was kind there was a little cognitive dissonance in my head. So how did you go from being part of a you know a a broadcasting
Scott Hewitt (03:17)
Yeah.
Jeff Walter (03:18)
to oil explanation exploration. Those things just don’t seem to that doesn’t seem to be a natural journey.
Scott Hewitt (03:25)
Yeah, it’s not the normal journey that people take. Actually, the first role I did within the multimedia company was I was a picture researcher. So I actually didn’t really know what I wanted to do, but I’ve got a background in IT and tech. So I was a picture researcher finding pictures for the teams that were building. At the first they were called multimedia projects. They weren’t called e-learning projects. I remember sticking on to CD-ROMs.
and were doing three platforms, which was the Mac, the PC and the Acorn, which was a computer that was used in schools here in the UK. And the chip that’s in the Acorn is now used across the world. It’s used in most mobile phones. So the ARM chip is now powers most of the Mac computers and iPhones and iPads. It’s an amazing piece of technology. But as a set of
merger and acquisitions that business disappeared and I did the transition and one of the finance directors that was in that business, was part of a family business that was working in southern North Sea gas and he realised that it got made redundant and he said would you come and work for us in this business you’ve got some other skills and would you come and help us and do some work on business improvement?
time I was like yeah okay I’d been I’d done a little bit of work in a finance company doing intranetwork and I wasn’t really
Jeff Walter (04:58)
Yeah.
Scott Hewitt (04:59)
enjoying it and so I went into it and the thing was I actually learned absolutely loads in that 18 months that I applied to digital and e-learning so what they were really really good at was manufacturing and production and so
Jeff Walter (05:15)
Uh-huh.
Scott Hewitt (05:15)
all those things about project files version control
project costing. So the things you hear now, you know, hear
Jeff Walter (05:24)
Right.
Scott Hewitt (05:24)
lot about ROI, impact, behavioral change, know, delivering cost effectiveness. Those things I learned huge amounts because essentially they were really good at project costing, delivering version control, delivering impact. And I learned a huge amount there because it’s a, you know, essentially it’s a production environment. So yeah, so I went there, that business had another.
Merge and acquisition and then I
Jeff Walter (05:52)
Yeah.
Scott Hewitt (05:52)
realized I wanted to go back into digital so that’s when I went back into there.
Jeff Walter (05:55)
Yeah. I I mean
it’s an interesting transition and then and then I I I like I like the way you said it’s like well yeah, I figured I’ll just create my job. And so yeah, well I mean kudos ’cause you know, putting your shingle out and doing that doing doing what you did is you know, it’s it that’s not for the lighthearted. And
Scott Hewitt (06:15)
No.
Jeff Walter (06:16)
and it’s been over twenty years now, so congratulations.
Scott Hewitt (06:19)
Yeah, and to be honest, I’ve got my mother to thank me for that because I’ve been applying for jobs and I wasn’t really getting anywhere. I read a lot of posts now, you go onto LinkedIn and people have been getting, they say I’ve applied for hundreds of jobs and I’ve been in that situation and I wasn’t getting anywhere. And in the end, I might as well put the effort into trying to get my own work as opposed to…
you know, waiting for somebody to give me a job. So that’s essentially what I did. was like, well, I’ll just give this whirl and that’s what I did. I think that, not the difficulty, think, probably the mistake I made at the start after a little while was that I just did anything, any sort of digital thing. I got quite an accurate
Jeff Walter (07:07)
Right.
Scott Hewitt (07:07)
background. So that kind of prevented some of the growth in the first few years.
because we
Jeff Walter (07:17)
Mm-hmm.
Scott Hewitt (07:17)
were doing anything, because we could do anything.
Jeff Walter (07:20)
Right.
Scott Hewitt (07:20)
Whereas if I did it again, you would say, well, we’re only going to try and do one thing. it wasn’t the thing that started, I always said, it was not the desire to become a millionaire. It was the desire to ensure that I wasn’t poor.
Jeff Walter (07:34)
Yeah. Well it y well it i and I I don’t necessarily I mean, you know, I’ve talked to other entrepreneurs that have started and I don’t necessarily think that that being hungry to take anything that anybody would pay you for if you had the capability to do it and do it well is is a bad thing ’cause you’re looking for where is the need in the market, right? And it’s like a little bit over here, a little bit over here and then over time you kind of figure out hey, I’ve th there’s this deep vein right here that I can
go for, right? And and so to just kudos. And then and then you end up going like, okay, we narrowed in on our focus. And because we you know I we’ve had a a very a similar journey. you know latitude started because we were part of a you know the the next Microsoft you know back in the late nineties, early 2000s, you know, we were part of this company that was going be the next Microsoft just before anybody heard of Google, right? And
Scott Hewitt (08:29)
Yeah.
Jeff Walter (08:32)
I’m gonna
Yeah, I don’t this might be a surprise, but they did not become the next Microsoft.
Scott Hewitt (08:36)
Ha
Jeff Walter (08:37)
And and our office got nine you you talk about r redundancy. Yeah, we we laid off ninety five percent of the office
Scott Hewitt (08:45)
Yeah.
Jeff Walter (08:45)
until the powers of bee came to me as the last VP left standing and was like, Hey Jeff, how’d you like to spin the office off? And yeah, we went from like three hundred people down to sixteen.
Scott Hewitt (08:58)
Yeah.
Jeff Walter (08:59)
And and I was like, and not deal with you turkeys, like, yeah, where do I sign? Yeah. And then
Scott Hewitt (09:03)
Thank
Jeff Walter (09:05)
but then it was like what you said, it’s like, okay, you know, luckily we had some book of business, but it was like, hey, I’ll w you know, we’ll do this, we’ll do that, we’ll do the other thing. Next thing you know, three, four years in, we got the rights to this learning management software because we were really good at implementation and one of our clients implemented it and then the company that
owned it was another conglomerate and they were shutting down that division and they were like, Hey, so we don’t get sued by our clients. How’d you like to take this over? And we’re like, sure. Yeah.
Scott Hewitt (09:35)
I think eventually
you work out what you’re going to do by trying lots of different things and I think that’s the sort of innovation part and so we’ve done a lot of web work as well and I think you’re taking on different clients and you’re taking on different projects especially if you’re sort of doing digital or customer any kind of business really
Jeff Walter (10:02)
Right.
Scott Hewitt (10:03)
and I think we’re also
happens after a while is that the…
your team wants a bit more clarity in terms of what they’re doing, I think. I think it helps them. So we are on one side, we were doing big web projects on the other side, we’re doing e-learning projects. And I think I reached a point where I was like, right, we’re not going to do the web work anymore. We’re just going to do the e-learning work. And everybody was a like, okay, good. And there was quite a bit of a surprise when we sort of gave up all this web work and we gave it away.
over time the focus is better because people are like okay this is where we’re going to be this is what we’re going to work on and the skills are the same a lot of the web skills are the same for the know for the e-learning work so I think that’s that’s a sort an important part of our journey is
Jeff Walter (10:53)
Mm-hmm.
Scott Hewitt (10:54)
you know it’s a bit like a funnel you’ve got all the ideas that you’re sticking at the side and I think that’s a classic thing for entrepreneurs isn’t it you sort of look around and go oh another idea I could be starting on and another idea
you
Jeff Walter (11:04)
Right.
Scott Hewitt (11:04)
can start on you know it’s like you know that’s
Probably my biggest thing is actually not starting new things and just focusing on the things that we’ve got. I’m sure you’re the same, probably got, we could do that, we could do this.
Jeff Walter (11:18)
You sound like you’re in one of my staff meetings. Well, but
Scott Hewitt (11:21)
Yeah.
Jeff Walter (11:23)
but but but it’s but well the thing is you you do that and and a lot of the skills are transferable, but then you find that thing that it’s like, hey, here’s where we can really add value. Here’s here’s
Scott Hewitt (11:34)
Yeah.
Jeff Walter (11:35)
here’s a a niche in the marketplace. And but with us it was you know, in the early twenty tens when
Scott Hewitt (11:43)
Yeah.
Jeff Walter (11:43)
everything went sass.
Scott Hewitt (11:45)
Yeah.
Jeff Walter (11:45)
And then and we’re sitting on this big honking piece of software, this LMS, that like with a little bit of and with this much effort could be sat you know, it was a it was built as a installed behind the firewall for large companies and and and with a little bit of effort we could sassafy it. And we’re like, Huh. Everyone’s
Scott Hewitt (12:04)
Yeah.
Jeff Walter (12:05)
buying and and we satisfied it and next thing you know, like people are like, Yeah, I’d like that and then we’re like, Hey, ’cause we’re doing, you know, lots of custom consulting and all that.
And and that market changed too, you know. and and and now next thing you know you’re s we’re we’re an LMS SaaS company and next thing you know we’re focusing on learning and for the like the last twenty three years. So it’s it’s fun.
Scott Hewitt (12:30)
And I think we had an opportunity really that came from the fact that I tried lots of different things and
Jeff Walter (12:38)
Yeah.
Scott Hewitt (12:38)
I was working for the broadcast company. We went into localization and nobody wanted to do it. We built loads of courses. The creative bit was interesting, it? You everybody likes it creative bit and we were localizing and nobody wants to localize and there was an opportunity to do the travel. Nobody really wants to do the travel.
We’d got an offshore team that we were setting up in India. There was an opportunity to go. I decided to go because at the time somebody dropped out of going and I took on board the localization team. And at the time localization just took a long time because we were doing it in CD-ROM. We were posting things over and we were actually localizing English content into American because we had a partner in the US. It was just taking a lot of time to do.
You know, we didn’t have the big network pipes that we needed for moving video across and, you know, the process just wasn’t really there. But I, a huge amount that I did it. And it wasn’t really until the last three to four years that I saw the gap, you know, in the market where we’re off the shelf and this is where we’ve really come into it. And I said, right, okay, well, we can localize our content now into multiple languages. And that’s what we’ve done.
And it’s been a mix of the localisation skills I learned probably 15, 20 years ago and
Jeff Walter (14:05)
Mm-hmm.
Scott Hewitt (14:05)
the production skills I learned from working in oil and gas.
Jeff Walter (14:09)
Yeah, yeah, it well it’s interesting from the oil and gas what what you said is because it’s it’s one of the things we we focus a lot on organizations that are trying to scale through distribution networks, you know,
Scott Hewitt (14:21)
Yeah.
Jeff Walter (14:21)
franchises and and and and dealers and resellers. And but it’s the same key for just every business. It’s like it’s that it’s you gotta get those processes down.
Right. That that like it’s gotta be repeatable and predictable, right? And and it’s and it sounds like your your stint in the oil and gas, like you saw that and
Scott Hewitt (14:42)
Yeah.
Jeff Walter (14:43)
and you’ve taken that and incorporated it to help real projects be successful over the years. So let’s let’s shift gears a little. Tell tell tell us tell everybody about more about real pro we really didn’t get into what real projects do we we we were an e-learning company and
Scott Hewitt (14:54)
Right.
Jeff Walter (14:55)
and and but what’s the focus? Why do clients engage you you guys?
Scott Hewitt (15:00)
Yeah, so we’ve got an off the shelf library. We’ve got 800 courses and we’ve got them in multiple languages. We’ve gone up to nine languages. We just recently launched in Indonesian, Thai and Vietnamese. And we’ve got German, French, Italian, Brazilian Portuguese and Latin American Spanish. So one of the things that we’ve really focused on is the languages because what we
Identified from speaking to customers really was the fact that languages were absolutely key and it wasn’t just about pressing translate You know, so it’s it’s localization not translate So, you know with with AI and with various different tools been able to do for a long time is that you know You can press translate and you can get the content, you know push that for you quite quickly and
But what we really identified was that really good quality content that had been localized and had been checked by native speakers and all the nuances were right and the audio was right and the content was right was what people were needing for people in various localization, sorry, for people in various teams in various countries, but also just because people have got
people working from all different countries and all around the world in different locations. So it’s just become a real requirement from people that we’ve spoken to. And we’re now getting requirements for languages all the time.
Jeff Walter (16:36)
Right.
Scott Hewitt (16:37)
It’s become a key differentiator for us.
Jeff Walter (16:41)
Mm-hmm.
Right?
Scott Hewitt (16:44)
Yeah.
Jeff Walter (16:45)
It’s it’s the impact. And you know, a lot of clients will go, well, univers you know, English is the universal language. I just need it in in English.
And it’s like, you know, and people can understand it. You know, our our company standardizes on you know English, right?
So talk about localization. Yeah, what I was saying was you know, so many
Companies, they because the standard language i within the company is English, there’s a tendency to believe that all you need is English. ‘Cause
Scott Hewitt (17:19)
Yeah.
Jeff Walter (17:19)
cause everybody within the company can operate with you know, it’s it’s the it’s the the but but what we talk about on this podcast a lot is impact. Yeah, like like, you know, if you want to get beyond checkbox training and you want to actually have an impact, the localization becomes high you know, it’s like yes, the other person can
understand it, but is it having the same impact as if it was in their native language? And
Scott Hewitt (17:45)
You’re gonna have some.
Jeff Walter (17:46)
and and because we’re trying to communicate knowledge and and explain skills, you know, oftentimes, and I th it sounds like this is what you were saying in terms of the importance of it, is yeah, you can do it in a sec in somebody’s learned langu second language, but if you do it in their primary language, it’s just
that much more impactful. So if it’s really if it’s defensive, you know, if it’s it’s that defensive training that I won’t get sued because I I I I I train people in this thing and you can’t sue me for, you know, whatever, you know, that’s one thing. But if you’re actually trying to have training that has an impact in in production and and operations and and sales, i i i it’s always better to speak in a person’s native language.
Scott Hewitt (18:33)
Yeah, I think for organisations it’s there’s a number of things as well that it sort of says to employees. I think. For an employee that’s maybe just might be looking through a platform. If it’s in multiple languages, it’s like, good, it’s not just in English. Yeah.
Jeff Walter (18:55)
Right. Right. It’s like they they value our what we’re doing here.
Scott Hewitt (18:58)
yeah, so there’s just there’s the optics of it, which is.
great it’s in multiple languages you know and there’s different ways of doing it even if it’s even if it’s closed captions you know it doesn’t even need to be you know there’s different ways of doing it you can just have closed captions you can have it fully localized but just having that option to begin with so there’s not the struggle because like you just rightly said is the there’s the understanding of the content you know very fortunate speak English when I travel the majority of people will I’ll be able to have a conversation you know and
have an understanding of some languages but by and large a lot of people I don’t take it for granted that everybody speaks English but it’s incredibly I think it’s incredibly difficult and challenging
Jeff Walter (19:46)
Yeah.
Scott Hewitt (19:46)
you’re having to you know I was thinking my mind some people who are working with three languages that the translate it’s incredible they’re translating in their heads and working through the you know especially complex you know models or
systems or different things that they’re working with and things that we’ve picked up when we’ve been working with our translators. So for example, AI is a popular subject. So in German it’s KI. So
Jeff Walter (20:15)
really? c n n knowledge intelligence was or or
Scott Hewitt (20:19)
it’s KI not AI. So that’s a case of us going through and changing all of the icons, all the graphics, all the content and bits and pieces.
And then other things like so fishing in cyber security, is pH quite easily gets changed to fishing as in the literal of fishing for catching for fish. So there are those things as well about making sure that you’re not doing literal translations and you’re getting all the nuances of correct. I remember a conversation we had with some of our
that were working on Brazilian Portuguese and they just had a debate for about a week about homeschooling because we had a remote working course about homeschooling and it was just about getting those things right so that the optics are right for the people who are using that course and you’re
Jeff Walter (21:10)
Right.
Scott Hewitt (21:10)
not just whizzing it through and pushing it out.
Jeff Walter (21:14)
Yeah.
Scott Hewitt (21:15)
the conversations we’ve had with organizations as well is that they want to make sure that they’re getting the language courses
It’s not just the organisations that have got offices in multiple locations. It’s
Jeff Walter (21:31)
Right.
Scott Hewitt (21:32)
to support their work, you know, their multinational workforce. And that’s been that’s been a real driver for people, you know, and how they do it. Now, AI has enabled that to happen at a much greater speed. But what we’re trying to make sure is that the quality of the product is still really good. So when I’m looking at it.
I’m reading it and thinking right okay this has got care and attention to it and it’s you know I’m still looking at a really good product not something where it’s been press translate to go.
Jeff Walter (22:05)
Well quite question about that, as long as we’re talking about AI is you know, that’s a hot topic. And I r I r I
Scott Hewitt (22:10)
Yeah.
Jeff Walter (22:11)
remember talking to somebody about it recently and and one of the things they were say I’m I’m curious your experience with with you know these other languages. ‘Cause
Scott Hewitt (22:22)
I’m
Jeff Walter (22:23)
the the assertion that they’re that they had made is all it’s it’s AI is most sophisticated, most knowledgeable on on English.
and that it’s you know, to s a certain degree struggles with other large l or or the most mature large language model is English
Scott Hewitt (22:38)
Yeah.
Jeff Walter (22:39)
and and other language models are you know not as mature, right?
Scott Hewitt (22:43)
Yeah.
Jeff Walter (22:44)
And therefore it it tends to struggle. And you know, when a lot of times when we think about language, we think of, you know, the world’s major languages.
Scott Hewitt (22:54)
Yeah.
Jeff Walter (22:54)
But you don’t get into some of the the the languages that have
many fewer people, like like you mentioned Vietnamese, right? And
Scott Hewitt (23:01)
Yeah.
Jeff Walter (23:02)
and Thai and and and and I mean I I yes, there are millions of people, but it it it it’s how do you you know in using AI to help you with the translations and localization, which you drew a really important comparison between, how do you you know, it is it are those s you know, other languages, those less popular languages, do you see the the
Then i is it reliable, mature, or what’s been your experience with that?
Scott Hewitt (23:34)
That’s a really really good question and I think some of those questions have been around for a long time whether you’ve used a localization expert or not because I’ll go back to something that happened to me a long time ago I was doing a Japanese translation before AI and this one I worked for the broadcast company and we did a Japanese translation we sent it over to Japan and the Japanese partner said it’s not right it’s not reading properly and
The thing with kanji is unless you’re fluent in Japanese you can’t actually sort of see whether things are breaking in the right place. if I’m looking at if I listen to Spanish or listen to French or listen to German and I look at the words on the picture I can phonetically I can sort of see if it’s breaking in the right place or you know I have a
Jeff Walter (24:25)
Right.
Scott Hewitt (24:25)
reasonably good guess and actually we just had no idea so we had to get somebody in
then had to check the Japanese and there was absolutely nothing wrong with it, it was absolutely fine and that’s the challenge that you get with some of the other languages where the character sets or the alphabets are basically different, know phonetically you can’t hear it or you can’t match it up with the text of the fonts and bits and pieces and so that’s one thing. I think some of the models are slightly better than others, it depends what you used
we did a very detailed procurement process in terms of what we were going to use. So some of the models were better than others in terms of the output that they got. The only
Jeff Walter (25:10)
Right.
Scott Hewitt (25:10)
way that you could do that was by using native speakers. Now, again, what was interesting was that some of the results that we got from the speakers were different from others. Some of them would go, yeah, it’s fine. Another person was much more detailed and said, no, this is not great. This output is not good. So
Jeff Walter (25:28)
Uh-huh.
Scott Hewitt (25:29)
that
reason we actually had a lot of people check all the content output and then we would have them check against each other so for example we would put the script through an AI model more
Jeff Walter (25:42)
Right.
Scott Hewitt (25:42)
than one we would get the output then we would get it edited by a human you know that language and then we would get that checked by another human who understood that language and then we could get them check each other’s so that we could
You know, humans in the loop is an often, you know, the phrase that’s often used, but essentially we’re checking the AI model and then we’re checking the edit that the human has made of the AI model, if that makes sense.
Jeff Walter (26:10)
Yes.
Scott Hewitt (26:12)
So some of the models are better than others. think what you have to be really careful or conscious of a couple of things that are really interesting is the tone and voice of the model.
something that we learned about using German. So the German translators would say the tone was quite formal and we didn’t want a formal tone so that a lot of the models were doing a very formal tone which wasn’t right for our audience.
Jeff Walter (26:45)
Right.
Scott Hewitt (26:46)
I’ve worked with some other languages whereby the breaks are in the wrong place so therefore when you then print it it doesn’t read right.
So there’s some technical details of language that unless you’re an expert, you just wouldn’t get it right. And then there’s some other things about them when you use audio, goes to a different level again. There’s things it just doesn’t work with.
Jeff Walter (27:10)
And and so if you were to grade the models like do you see d like in because you now you’re getting into the nuances, right? Like tone and and you and and also
Scott Hewitt (27:18)
Yeah, yeah, yeah.
Jeff Walter (27:24)
you know, just you know, turns of phrases you know, that that like you know that
Scott Hewitt (27:29)
Yeah. Yeah.
Jeff Walter (27:31)
when you th on the face of them don’t mean anything but but they mean something in that language.
You know, like you know, the f the full nine yards, that type of stuff. You know, colloquialisms and and other
Scott Hewitt (27:40)
Yeah, yeah.
Jeff Walter (27:43)
things. Do you d d like i it so i is it would it be s I’m just looking to what you say, would it be safe to say that it’s like well technically it does a de you know, if you’re using AI you’re getting a decent thing so that if you translated it a person who only speaks that language would
be able to decipher the meaning. But
Scott Hewitt (28:10)
Yeah, think it does a good job.
Jeff Walter (28:12)
but but when but you know, but it’s like but it may not be the right tone, set of colloquialisms, way in which a person would ease more easily absorb that information had they actually just been talking to a native speaker. It’s not the way their their neighbor would have explained it to them.
Scott Hewitt (28:31)
Yeah it’s a good job.
Jeff Walter (28:33)
Is that
Scott Hewitt (28:33)
think if you’re thinking about cyber security or something like that, personally it’s not the way that we work anyway but you wouldn’t want to push it through without it being checked by your expert and your language
Jeff Walter (28:46)
Right.
Scott Hewitt (28:47)
expert anyway. What we do is that our writers ensure that we don’t write with colloquisms and things like that anyway so it doesn’t create a problem because
People even in the target language don’t understand it. So we wouldn’t write things like, know, people in glass houses don’t throw stones, whatever, you know, because it doesn’t translate anyway, because sometimes it doesn’t translate across language, because it doesn’t translate across languages. It doesn’t translate across an AI. So
Jeff Walter (29:19)
Right.
Scott Hewitt (29:19)
before you even get to the AI, don’t write it anyway. So then you don’t get, you don’t have an AI problem. So I think that’s the thing to be thinking about is.
if you’re struggling to translate it anyway so if
Jeff Walter (29:34)
Right.
Scott Hewitt (29:34)
someone in Mexico is not going to understand it or if you wrote it in El Salvador and they’re not going to be able to understand it in Hungary then don’t write it anyway.
Jeff Walter (29:46)
Interesting. So all right, so so f folks are working on that, then it I get I I guess where I was going is like, okay, if you if you use some rules like don’t write in colloquialisms and use
Scott Hewitt (29:57)
Yeah.
Jeff Walter (29:58)
this, it’ll it sounds like and and you always want to keep a native speaker in the loop. but right
Scott Hewitt (30:03)
Yeah, yeah, avoid eating those types of things.
Jeff Walter (30:07)
and
And and you’ll get something that’s technically correct but not as impactful as it could be that than then when you start tailoring it with tone and other things to that particular native speaker. Is that a good way of thinking of it? Right.
Scott Hewitt (30:22)
I think you still get the impact because what
you need is a really good writer to begin with.
Jeff Walter (30:28)
Right.
Well yeah, I
Scott Hewitt (30:33)
I was fine.
Jeff Walter (30:34)
yeah, I was referring to, you know, w like I I I was just thinking of how you had gone about that and you had all these you you you checked and rechecked it with not one but many local speakers
Scott Hewitt (30:46)
Yeah.
Jeff Walter (30:47)
and then had them check each other to get a very high quality product.
Scott Hewitt (30:51)
Yeah, so
it goes back to my time working in a production process.
Jeff Walter (30:57)
Right.
Scott Hewitt (30:57)
One of the other reasons of doing it is essentially audit. It sounds quite robotic and it’s removed the creative element from it. It’s not created. We haven’t removed the creative element from it. And it’s not a case of checking everybody. What we wanted to understand is if we’re going to use AI in the process, we want to be able to understand
you’re going to scale which we did so we did 500 courses in four months yeah
Jeff Walter (31:27)
Uh-huh.
Scott Hewitt (31:28)
so the first thing that we did is we understood we built the engine essentially so we built the process so if you’re going to scale you’ve got to make sure it’s right because if you scale and it’s wrong you’ve got to go back to do it
Jeff Walter (31:39)
If
you’re scaling it strong, you’ve you’ve you’ve multiplied the yes.
Scott Hewitt (31:49)
You’ve created a monster. So
we made sure that was right and so by doing it we basically broke everything down into small pieces like a production process. So then by going in and saying right how do we do it and the first thing is going back to the start and saying what we need to feed it with at the start is really really good. Now what you want to make sure is that you’re just not it’s just
haven’t stripped everything away so it’s so dull that there’s nothing
Jeff Walter (32:16)
Right, right.
Scott Hewitt (32:17)
in there you know because otherwise it’s it’s just dry but some things like I say they just don’t work anyway in a different country you know so we try and avoid that so we spent a lot of the time with the writers doing a really good set of writing guidelines so this this is just going to this is going to work across all of them content on the you know focus on the content and then the rest of it follows through because then
it doesn’t really matter. Some of the models work well, it’s a
Jeff Walter (32:47)
Uh-huh.
Scott Hewitt (32:49)
classic IT, garbage in, garbage out. People
Jeff Walter (32:53)
Right.
Scott Hewitt (32:53)
have been saying it for years and it still works.
Jeff Walter (32:56)
Yeah, well it it i I I laugh because I I I I’m hearing I haven’t heard that expression used frequently for decades. Right?
Scott Hewitt (33:08)
Yeah.
Jeff Walter (33:08)
Like like it was a really popular expression in the eighties and nineties when we were like computerizing accounting systems, you know, all these back end systems that were getting computerized. And then kind of the internet came and it kind of was you know, it wasn’t a thing.
And then all of a sudden you’re I’m I’m hearing that expression more and more which is the same thing. It’s like, look, if you’re gonna feed it garbage, it’s gonna give you garbage. I it’s more eloquent garbage, but
Scott Hewitt (33:35)
Yeah.
Jeff Walter (33:36)
it’s still garbage.
Scott Hewitt (33:38)
The thing that we work with our AI and I see this all the time a little bit because we’re working, you so we’re working in off the shelf content, okay? And so content
Jeff Walter (33:47)
Right.
Scott Hewitt (33:47)
at the minute, there’s a lot of people like, don’t need more content, we’re working with impact and business, well, that’s fine, but there’s a place for content and it’s gonna be good content. Well, if we’ve got an LMS, what we’ll do is we’ll put an AI layer in there or put something in, you know, the AI will serve the…
scorn files and whatever and we’ll get some output and we’ll get some personalized pathways or bits and pieces
Jeff Walter (34:10)
Mm-hmm.
Scott Hewitt (34:12)
but what that misses is that if your knowledge layer is not right or if your metadata is not right or if you haven’t done the process that we’ve done at the start you’re
Jeff Walter (34:22)
Right.
Scott Hewitt (34:23)
just serving junk.
Jeff Walter (34:25)
Right. Well, you
know, that’s interesting because I I I I have had conversations with people about that and learn you know, the and and learning and development, which is you know, th there’s kind of a there’s this one thought that, well, you know, you don’t need to do all this. You just need all the knowledge put into a you know, a knowledge base and then, you know, basically you just need a chatbot, right? Like
Scott Hewitt (34:51)
Yes.
Jeff Walter (34:52)
I a and they’re more sophisticated than that. I’m oversimplifying
Scott Hewitt (34:54)
Yeah.
Jeff Walter (34:55)
obviously. But but I was you know, and the com and and the the the pushback I always had is like, yes, but that’s assuming that the student knows what they need to know, right? Like there’s their I know what I know, I know what I don’t know, and I don’t know what I don’t know, right? And like, okay, if it’s something I know but I forgot and I want to get back, you know
Or I know that I don’t know it and I know the right ask the right questions to get it. It’s like, yeah, then the chatbot works fine. But what about I don’t know what I don’t know? And like that’s the essence of learning. Is it’s being introduced to things that you don’t know you don’t know.
Scott Hewitt (35:31)
And
you’re essentially you’re a cycling rubbish.
Jeff Walter (35:35)
Right.
Scott Hewitt (35:36)
So one of the other things where we used. Where we used AI and built a model. And I think this is been something that’s been a success for us is that so we’ve got metadata around courses and people talk about how they use data and bits and pieces. You know you your normal stuff, you have title description, how long the course is not kind
of stuff.
Jeff Walter (35:59)
Right.
Scott Hewitt (36:00)
But then we’ve also got taxonomies that we use for different customers and you know, so learning objectives,
Jeff Walter (36:06)
Right. Right.
Scott Hewitt (36:07)
what it belongs to and all that kind of stuff. Well, the thing was, was that two years ago, that’s done by humans. You’ve got to decide what it goes into. And the analogy that I use is it’s a bit like going and working in a record shop, yeah, or a bookshop. Well, it’s a bit easier, you know, so.
You know what the Rolling Stones goes into. It doesn’t go into jazz. It goes into rock and roll. know, so does the Beatles. You know, it’s easy what category it goes into, you know. And,
Jeff Walter (36:34)
Right, right.
Scott Hewitt (36:35)
you know, but it’s a physical thing. It only goes into one thing, but now it’s digital. You can put it into, know, on Spotify. It might sit in two or three or four things. Yeah.
Jeff Walter (36:45)
Right, right.
Scott Hewitt (36:46)
It’s got, it’s got those multiple categories. Same for courses. Yeah. But if you do it manually.
Yeah, and you’ve got different people doing it. The taxonomy and the output that you get might be completely different. Yeah. But what you try and get to get the best output, and this comes from being a researcher and tagging photographs and bits and pieces and working with big databases is you want consistency. So if you can build a model within your AI and this is right, I’ve got this course. This is its attributes.
I want it to be in this. This is the taxonomy and now categorize it for me. And it goes, these are the things it’s allowed to be in and you do for all your courses. What then happens is that when people then serve content in things like LMS is RxPs, they get better search results. Cause you’ve got better data and then you don’t need an AI level that’s going through and having to search through school and stuff because your data is better to begin with. And that’s the bit you use the AI before.
it gets to the content chucking in the, you know, through
Jeff Walter (37:55)
Well.
Scott Hewitt (37:55)
the LMS.
Jeff Walter (37:56)
Well you know you know, it’s I I hadn’t I hadn’t thought of that from a taxonomy standpoint because and I had a couple of thoughts that popped in my head as you were talking. You know, when you talk about the Beatles and Rolling Stones, it’s like well, you know, the the rock and roll versus jazz was almost kind of hierarchical taxonomy.
Scott Hewitt (38:13)
Yeah.
Jeff Walter (38:15)
It’s like you you only fall into one bucket or the other. With tagging and everything that’s evolved over the last you know, t it’s more of a network
taxonomy where a thing because you you’re you’re looking at multiple different dimensions, so a thing can be in multiple different dimensions, right? Yeah, when you’re looking at from, you know, that perspective, it’s jazz or rock and roll. If you’re looking at from a a lyrical standpoint, it’s a different there’s a different taxonomy. Like and you get these networks where that that the beatles can be in six different places or a song can be in six different places, depending on the perspective that you’re looking at. And then
As I and then when you jump to the AI, it it it’s interesting because then the I and I never thought about this before, so thank you. well because one the langu it because taxonomy is a language
Scott Hewitt (39:08)
Yeah.
Jeff Walter (39:09)
and and words have meaning and but they words don’t have any inherent meaning. They just have the meaning that we as a group decide they have, right? And you know, like that sound that and and
And it’s interesting you said, okay, I’m I’m doing this taxonomy disco and and so I want, you know, when you s when you talk about picture, my mind went to like, I want a ocean sunset. Well, if I’m categorizing pictures with ocean sunsets and you’re categorizing pictures with ocean sunsets, we’re going to have different interpretations of what ocean sunsets mean.
Scott Hewitt (39:40)
completely.
Jeff Walter (39:42)
Right? And so and now you throw, you know, and then there’s twelve others abo of us and we’re categorizing things and and
And we all have slightly different definitions of what an ocean sunset is. And so you get the conglomeration of all of that, which is like a meta definition, which is not as sp as precise as what you and I would do had we done all the ocean sunsets, right? And it’s interesting. And then I’m like, then all of a sudden the tag ocean sunset becomes a much more specific thing if you apply the same rule over and over. So I really
Like i if I’m if I’m getting what you were saying, if I if I can train an AI model to do the taxonomy based on everything, then I might have some differences on the edges and you might have some differences on the edges. But what we but more importantly, if we train it right, it’s some definition of ocean sunset that is consistently applied so then
Scott Hewitt (40:44)
Absolutely.
Jeff Walter (40:45)
so then the user learns
That when I type ocean sunset, I’m going to get a certain type of photo or yeah, or I’m going to get a certain type of response. And right. Interesting. I never
Scott Hewitt (40:56)
you get much more consistency in terms of what you know.
Jeff Walter (41:00)
I never thought of that from the ta from a taxonomy. A taxonomy is really a language. The the the elements of the taxonomy have to have specific meaning. And that’s that’s really interesting. Huh?
Scott Hewitt (41:11)
And then what you get, and then what
you get is that the talk is, you know, personalization and concert or anything. But if you, if you develop that bit, right. Then you solve the problem and there’s your impact because
Jeff Walter (41:25)
Right.
Scott Hewitt (41:25)
you’ve got the metadata to begin with. Yeah. And you’ve put things in well, then it’s, it’s properly personalized. And then what I’m really interested in. So my background was where, but if you look at, L and D and know, bits and pieces.
There’s so much data, people are just obsessed about things like completion states and things like that. completion data is good if you use it in the right way. If I was working it with, I used to work in IT strategy, I’d be interested in, if a course didn’t complete, I’d want to know because it might mean that the browser didn’t work or I’ve got a problem with computers not working. That’s just an aside. But all that data is really, really important.
because I can grab that data and I can build personalization. can, you know, I can be putting that out to API’s. I can be doing all that kind of different bits of personalization myself. And you’ve got all the things that you need. And that’s where your impact comes from. If you look at search where they’ve been really great is look at Google. They’ve got the four types of intent that they’ve discovered, you know, that they use. So you’ve got.
know, commercial transaction, informational and you know, and there’s the other one that they’ve got. And so if you look at LMSs and LXPs, they’ve got loads of different information that they’ve got around keyword searching.
Jeff Walter (42:51)
Mm-hmm.
Scott Hewitt (42:51)
So why don’t you grab it? Why don’t you find out what people are searching on? Link it to impact, put it in an AI model, link it to a text on me. There you go. You find out what type of search intent people have got. So I think it’s quite straightforward.
Jeff Walter (43:05)
Mm.
Scott Hewitt (43:06)
You
don’t really even have to change your LMS. And then you find out what people are looking for. Then you can also find out what content they want. And then you can find out what they’re trying to do. You also find out business problems. Then you find out business problems. Then you can start going back to impact. It’s just from a keyword search.
Jeff Walter (43:24)
Very interesting. Huh. That’s interesting. I’m gonna I’m gonna have to I’m gonna have to noodle that. Yeah.
Scott Hewitt (43:27)
But it’s just some of it’s just yeah, it’s but you
can build it because it’s just what it’s just what Google uses. It’s just so if you you know if you do if you use Google Search Console, it will tell you what your website you know what keywords people using to find your website. Most people have got an LMS or LXP. It’s got a free search you know it’s got a free text search. You know how are people using it? What you know what search words are people using? Run a model is it?
what’s it linked to? Build a taxonomy, then build a personalized, you know, it’s just, that data’s really straightforward. So that’s where I would work from that from. And that’s the sort of stuff we’re interested in because there’s a lot about giving people the right content, giving people the right things at the right time. But to do it, you’ve got to build the right models.
Jeff Walter (44:19)
Yeah. Well I I never thought of taxonomy the way you laid it out. And that was that was really interesting. And it and and it and it goes back to the model and then I never thought of looking at the intent data on a search as a feedback loop into what content should be available.
Scott Hewitt (44:42)
Yeah, so.
Jeff Walter (44:43)
Right,
like like it’s yeah, I’ve been in this business for a couple of decades and and and I and I own an LMS
Scott Hewitt (44:50)
Yeah.
Jeff Walter (44:50)
and that never occurred to me. So that’s really interesting. That’s fascinating.
Scott Hewitt (44:52)
Yeah. So if you look at people how you
use the web, so if you’re looking for a flight, most people who are looking for a flight are looking to book a flight. If people look for car hire, they’re looking to
Jeff Walter (45:02)
Yeah.
Scott Hewitt (45:03)
book a car hire. So it’s a commercial and it’s a transactional search.
Jeff Walter (45:07)
Yeah.
Scott Hewitt (45:07)
And that’s how search engine optimization works. And if you go on those websites, what you will normally see straight away is a booking form because they know that people in those particular positions and those keywords
You don’t see pictures of cars and you don’t see pictures of planes. You don’t see like here’s the A380 or whatever it is. You see booking forms. If there’s a different search term which is I don’t know how do I remove B’s from whatever it is. You might actually see a bit of both. You might see some informational stuff and you might see and so L and D and LMS’s don’t really they’ve not really made that jump yet but that data has been around for years.
Jeff Walter (45:47)
Yeah, yeah, I know. I I well I I mean, as soon as you said that, I’m like on both of those, the taxonomy and the search data. Well and the interesting thing is like some some of the conversations we’re having internally is we you we’ve we’ve started putting we’ve got the the the learning assistant, the chatbot, build sources of knowledge and all that with within the the learning platform.
And then it’s like the interesting thing is you y you can use n you can use all that data now to provide use the AI to provide the analytical insights to do that analysis, right? Like like you know, which is really interesting. It’s really that’s really interesting. Yeah.
Scott Hewitt (46:24)
Yeah, Data is everything. my final
thing was like we worked in football, like soccer. And
Jeff Walter (46:32)
Yeah.
Scott Hewitt (46:32)
so every year I’ve been on it. Last three years I’ve been on a sort of football tour of various clubs. So I’ve just been to Espanol in Barcelona and their use of data is just incredible. They just have huge amounts of data. And so it’s interesting, I always think going to see other sectors and other businesses and how they use data.
So how they use data is really interesting. There’s somebody that I know works in Syria in Italian football. And his thing
Jeff Walter (47:01)
Uh-huh.
Scott Hewitt (47:01)
about data is that he’s a data analyst. But basically he gets 100 % of the data, but by the time he delivers it to the manager, the manager only sees five to 10%. So he has like a pin number basically. It’s like he gets 100%, the other analysts get 50%.
coaches get about 25 % and the manager gets like 5 % so it’s just an interesting breakdown about what you actually get.
Jeff Walter (47:33)
Yeah, and when you say get, it’s not because they don’t want the manager to see it, but but but it’s been distilled for what’s important to the for what’s important to the manager. Yeah.
Scott Hewitt (47:38)
It’s been distilled, yeah, because they’ve got so much, what’s important? So he gets the headline
and then he would say, well, you if you want more, we’ll give you more. But this is, we’ve given you this in granulated formula. So these are the key bits
Jeff Walter (47:53)
Right.
Scott Hewitt (47:53)
that you need to see. We’ve got so much data that you haven’t got time to see it, but these are the key bits. But if you want more, we’ll give you more. so because
Jeff Walter (47:59)
Right.
Scott Hewitt (48:00)
they’re capturing so much data, it’s essentially KPIs.
Jeff Walter (48:03)
What
Yeah, well I I I f you know, like the whole moneyball effect in in sport, I think, you know, where everything has been because it’s a controlled environment, right?
Scott Hewitt (48:15)
Yeah.
Jeff Walter (48:18)
you’ve been able to digitize so much of the experience that you’re you’re well washing data and and and and there’s it’s also high stakes, right, in terms of economic stakes, right? Like if you can
increase your win percentage a couple of points. That’s the difference between a championship and relegation, right? And you know, it’s it’s so it’s interesting. It it’s interesting to go there and look at how they’re using data analytics. That’s that’s fascinating. And that
Scott Hewitt (48:45)
interesting they capture
tons of stuff
Jeff Walter (48:48)
Yeah.
Scott Hewitt (48:49)
and what I find really interesting is that they just don’t have this concept of vanity metric which you see in L &D they just don’t dismiss anything they might not use it but they don’t dismiss anything because they might use it in a model for something else
Jeff Walter (49:07)
Yeah. Van vanity metric. I haven’t heard that term, so what do you mean by vanity metric?
Scott Hewitt (49:12)
It’s something I hear a lot in the UK. it’s a metric that’s viewed as measured a lot, but having limited value. So completion
Jeff Walter (49:21)
Okay.
Scott Hewitt (49:23)
metric is used, a metric here in the UK, which
Jeff Walter (49:29)
Yeah.
Scott Hewitt (49:29)
people dismiss quite a lot because it just shows you how many people have completed a course, for example, but what does it really show you? Well, you know, on its own.
It’s not really going to show you a lot, it’s just going to show you how many people have completed a course. So for a lot of people, it gets a lot of posts and articles about it. my view is that most pieces of data on its own doesn’t show a lot, but you have to mix it with other things. the example I used before is, know, if I was working, you know, in IT or data, you know, you might have a course and I’d be looking at it and I’d be going through and saying, well, actually, if everybody’s gone through the course in
know one minute 15 then there might be a course design issue. If I’ve deployed this course in let’s just say six different regions and one region has got no completions then I’m
Jeff Walter (50:25)
Mm-hmm.
Scott Hewitt (50:25)
less worried about the at this stage I’m less worried about the actual people completing the course have I got a network or a technical issue that I’d want to look at the know the CTO with so I might again look at times of people using things.
So my view is also look at it slightly differently, not just about people being in the course. I like look at the infrastructure and how the business are using things. So that’s the stuff that we, you know, that we would look at. So, you know, you might look at, you might have built a course, don’t know, let’s just theorize it’s 30 minutes long and there’s a drop, you know, there’s a first drop off at three
Jeff Walter (51:03)
Right.
Scott Hewitt (51:03)
minutes or whatever. So I’d be wanting to go in the course. look at it, you know, why are people looking at it for only three minutes?
That might just be a course design issue.
Jeff Walter (51:12)
Uh-huh.
Scott Hewitt (51:13)
So I’d never dismiss any particular bit of data. I’m more interested in what’s that data telling us.
Jeff Walter (51:23)
Interesting. Interesting. So so just looking the future for real projects. well one one thing we actually one thing we haven’t covered, so you you’re focusing on on on libraries of of courses, f focused on low high localization of of the course content. Is there a particular set of subjects that you guys are focused on or what do you what do you
Scott Hewitt (51:51)
Yeah,
Jeff Walter (51:51)
focus on there?
Scott Hewitt (51:52)
yeah, just recently we we’ve actually been doing AI courses for about four years. We just released some new AI courses with a with a focus on cyber security. We’ve
Jeff Walter (52:03)
Uh-huh.
Scott Hewitt (52:03)
got two brilliant writers who’ve got a background in AI and cyber. So cyber is an area that we continue to work on because despite people saying that, you know, we don’t need any more content, cyber is just an area that continues to keep moving and
you need to keep aware of. So that’s an area that we’ve got an interest in. actually compliance. Again, it’s another area that people are looking for content in and an interest in different ways of doing it. think the days of somebody wanting a one hour course on how to lift a box or manual handling of sort of I think they’ve gone. So it’s short.
Compliance courses that we’re interested in. they’re the things that we’ve done and we’ve done, you know, we’ve done lots of different interesting things as well. we’ve done courses on.
Jeff Walter (52:58)
Okay.
Scott Hewitt (53:07)
That’s okay. You all right?
Jeff Walter (53:10)
Just a tickle. Sorry.
Scott Hewitt (53:11)
I’m It’s okay.
We’ve done some interesting courses on travel security which has been interesting and actually an area that we’ve seen some real growth and interesting has been de-escalation and so that’s an area that we’ll continue to develop the library and specifically in the retail areas so we’ve done a lot on de-escalation in retail shops, hotels, restaurants and so on.
Jeff Walter (53:45)
Is that like
the de-escalating customer issues?
Scott Hewitt (53:49)
Yeah, difficult
customers, difficult environments, difficult situations. And actually we’ve done them on chats and phone as well. And they’ve just been proved to be really popular. So we’ve continued to do those and they’ve been really popular. And then a few things that we’ve over here that have followed. So we’ve done things like text-based phishing, which is for cell phones receiving.
messages on yourself, phishing messages and bits and pieces like that. So that’s worked really well. the main thing really though comes from it’s partner led. So we speak to customers, we find out what are the things that they’ve got. Because otherwise you’re just putting things out that you think might work. And really as we back to the thing about impact, we spend a lot of time speaking to customers, finding out what their problems are.
and trying to deliver the content that supports them in their businesses.
Jeff Walter (54:51)
Right. And and and Scott, as as our time is almost over here, have is there anything else you you want to share with the audience before we have to go?
Scott Hewitt (55:06)
What a great question to end with. Yeah, I think it’s an interesting time for people to be working with. think one of the things I actually find is that we go to a lot of exhibitions and lot of conferences and we speak to a lot of people. you know, the key thing is that I just think there’s some really great people out there doing some really interesting work. And what I would say is it’s not always the huge massive companies as well. I think some really
really interesting small new innovative companies doing projects and I think they’re always worth speaking to when you know they’re some of the best conversations that I have when I’m at exhibitions and conferences you know the people on the exterior on the perimeter of the conferences you know and
Jeff Walter (55:55)
Yeah.
Scott Hewitt (55:56)
they’re just doing the stuff that you know that unless you just stop and give them two minutes to have a chat with and you know and you find out you know they’re the start-up and I think
That was the position that we’re in and sometimes still are when we go to the huge conferences. You just got the small stand and those people have got the great ideas. think it’s always good just to stop and have a chat and find out what people are doing. think that is the thing. Not everybody’s trying to sell something to you. think that’s the thing. You just have a great conversation where you can share things and that helped push the sector, push the industry.
Jeff Walter (56:33)
Yeah. Well, I I do feel I and I I agree with you wholeheartedly, especially in the last couple of years with with AI, I there are so many people taking this new tool and trying to think of how to use it in the way the it’s it’s really going back down to the to me, the foundations of learning development and what is it we’re really trying to do, right? And there are some people that that are like,
Well, you know, it’s gonna change everything and everything we learned in the past is, you know,
Scott Hewitt (57:04)
Yeah.
Jeff Walter (57:05)
you know, useless. And then other people are like, No, it’s yeah.
Scott Hewitt (57:08)
Yeah.
Jeff Walter (57:08)
But it and the reality is always somewhere in between, right? It’s like here we are twenty five years into the e learning revolution and there is still instruction, you know, classroom based instruction. Right? Like
Scott Hewitt (57:18)
Yeah. Yeah.
Jeff Walter (57:20)
w we didn’t we didn’t get rid of it. But you know, it’s so it’s gonna be
It’s gonna be interesting to see how it pla like like we were talking earlier, like some people are like, Well gotta load up the information and put a chatbot. And it’s like, yeah, no, that that’s not gonna do it either. Right? Because
Scott Hewitt (57:31)
Yeah. Yeah.
Jeff Walter (57:34)
you don’t know what because the student doesn’t know what they don’t know. And you know, most learning is teaching them something they don’t know they don’t know. And and so they don’t know what questions to ask, right? And but then but then it can be a very useful tool in part of the learning process, right? And and and can answer very specific questions to very specific
you know, as they’re going through the learning process. So it’s really interesting. So it’s fun s so I agree with what you’re saying. It’s like it’s an interesting time. The wh I think the whole industry is trying to figure out how to maximize human potential with this new technology and how it fits within what we’ve been doing. And it’s it’s a fun journey. It’s a fun journey
Scott Hewitt (58:10)
Yeah, yeah,
I think there’s plenty of work out there as well. There’s plenty, you know, there’s plenty of to do and there’s, you know, there’s plenty of
Jeff Walter (58:17)
Yeah.
Scott Hewitt (58:17)
interesting people. I always think, you know, speak to the people. I often find, you know, people are always willing to, you know, give you 10, 15 minutes for their ideas or, know, and have a chat. I think, you know, don’t be afraid of just asking people what they’re doing, you know, because
Jeff Walter (58:34)
Yeah.
Scott Hewitt (58:35)
people will always sort of say, well, you know, we’re working on this because most people are in a similar position.
Jeff Walter (58:40)
Yeah, one and and I and one of the more interesting things along those lines was I was at a conference and there was a gentleman, Jeff Peavy, speaking and he his thesis was we are migrating from the knowledge worker economy to the learning worker economy, where y you know, because of AI c can can apply known knowledge repetitively relatively easily.
We’re going to you know, the the worker of tomorrow where economic value will become is is the constantly learning worker. And all and it’s and I I and he and I’m doing his argument huge injustice, but he I was like, huh, that’s a really you know, like
Like i it what is what we’re talking about, it’s like, huh. It’s like all of a sudden this stuff goes from you know, if that thesis holds out, what we’ve been doing from learning and development starts shifting into a core critical asset for a company. It’s just intro it’s very interesting. So but hey Sat, if if folks want to get a hold of you or real projects, where would they go? How do they get a hold of you in real projects?
Scott Hewitt (59:54)
Yeah, so we’re real projects dot code at UK and I’m on LinkedIn. You can find me Scott Hewitt and you can you can connect with me there and I’ll you know, accept your connection requests and we can we can chat there as well. So yeah, absolutely fine.
Jeff Walter (1:00:11)
So, Scott, thank you so much for joining us today. It was a pleasure. And I learned a couple of new things, which I always love. This is my favorite part of my job, is talking to guys like you and learning their things. It’s it’s it’s it it’s so much fun. So thank you for your time. Yeah. All right.
Scott Hewitt (1:00:25)
I appreciate the invite. It’s been great. I really enjoyed it. Thank you.
Jeff Walter (1:00:29)
Yep. And to everybody out there, thanks for joining us. Have a good day.
Scott Hewitt (1:00:31)
Thank you.