2905: How AI is Transforming Tech Companies from the Inside Out

[00:00:00] Have you ever wondered how the accelerated adoption of artificial intelligence is reshaping

[00:00:07] the landscape of the tech industry, particularly in terms of operational efficiency and workforce

[00:00:14] dynamics? It is a huge topic right now. And last year I had the pleasure of discussing

[00:00:21] the burgeoning role of AI with Thomas Law from TSIA. Given the vast changes since then,

[00:00:28] I think it's time to revisit this crucial dialogue. Because I've been to a lot of tech

[00:00:33] conferences where big tech companies all say that AI is a co-pilot, it will complement

[00:00:39] humans rather than compete with them, while at the same time cutting back on staff to

[00:00:44] buy more GPUs so they can invest more heavily into AI. So I want to try and bust a few myths

[00:00:52] today, address what everybody's thinking about right now. So Thomas is going to join me

[00:00:57] once again to explore the profound impact AI is having on tech companies, both good

[00:01:02] and bad, from reducing headcount to enhancing profitability, creating new business opportunities,

[00:01:09] exploring innovation opportunities too. And with AI making strides in both hardware and

[00:01:14] software sectors, understanding its influence, I think it is more relevant than ever. So

[00:01:20] buckle up and hold on tight as I beam your ears all the way to Ohio where Thomas is waiting

[00:01:26] to join me once again. Well, a massive warm welcome back to the show. I think this is your

[00:01:32] hat trick of appearances. But for anyone that's missed our previous conversations,

[00:01:36] can you just remind everyone listening with a little about who you are and what you do?

[00:01:40] Absolutely. You're right. This is the hat trick. I'm so happy to be back here, Neil. And I am

[00:01:45] Thomas Law, the Executive Director from the Technology and Services Industry Association.

[00:01:50] We're a for-profit research institute. We do a lot of deep operational benchmarking

[00:01:54] with technology companies to help them optimize their business model.

[00:01:57] Every single day on this podcast, I try and get people thinking differently about the role of

[00:02:02] technology and specifically areas around business and our world and our lives,

[00:02:08] technology impacts that we don't automatically associate with tech. So to set the scene for

[00:02:13] today's conversation, how significant is the role of AI in the operational changes

[00:02:19] within major tech companies that are part of the community there that you're working with?

[00:02:24] Because we hear a lot about AI, but not always in areas such as operational changes. So what

[00:02:30] are you saying here? Yeah, and I'm glad you're putting that lens on it because there's so much

[00:02:35] of the energy around AI, even for technology companies, is all about their customer-facing

[00:02:41] solutions there, right? We've got a really cool new Wicked AI tool here we want our customers to

[00:02:47] be excited about, but the reality is AI is just as much of a game changer for how technology

[00:02:53] companies are managing their business internally, right? Their internal workflows. And so when we

[00:02:59] look at that and that's what we've really had our research lens on very aggressively over the past

[00:03:03] year, that's the use case we're more interested in. And I can tell you that when you look at

[00:03:10] these companies, it's almost becoming this sort of world of have and have nots. And there are large,

[00:03:16] well-known tech companies that are aggressively leveraging AI for their internal workflows.

[00:03:23] We're doing these case studies where we look at, hey, when you deployed AI, how much did it improve

[00:03:29] a particular KPI? Whether that KPI was how long it takes to serve a customer or build a piece of

[00:03:34] educational content or whatever. And they're getting improvements anywhere from 15% to 70%,

[00:03:41] right? So massive impact there on their internal workflows. And then the flip side is I still see

[00:03:47] many tech companies that are lagging here. They're focused on AI externally, they're not as much

[00:03:55] focused on it internally. Some of them are even kind of, they got their arms crossed saying,

[00:04:00] well, could be a lot of hype. I don't know if I want to do this yet. So it is pretty fascinating

[00:04:05] because I'm sure that's playing out in every industry, right? There are people, companies

[00:04:08] leaning in, companies not. But even in tech where you think everybody would be leaning in,

[00:04:12] it's still kind of bifurcated.

[00:04:14] And one of the reasons I invite you back on the podcast was after reading your research that

[00:04:20] indicates that AI could reduce headcount by as much as 20% in tech companies. So one of the things

[00:04:28] that they struck me about is I go to a lot of tech conferences around the world and I see a lot of

[00:04:33] companies selling AI solutions and they're the ones saying that, no, it's not about reducing

[00:04:38] headcount. It's about being a copilot. It's about augmenting humans. It's about focusing on the good

[00:04:43] stuff, saying all the right things. Well, see quickly, we've seen those big tech companies

[00:04:48] reducing their own headcount. So what are you seeing here and how long do you see this happening?

[00:04:54] Is there any kind of timeline that you're seeing here?

[00:04:57] Yeah. I mean, and this is, I'll be honest with you, I mean, this is a very touchy subject for any

[00:05:05] leadership team of any company, much less these tech companies because everyone is using this

[00:05:11] vernacular around AI is going to augment, AI is going to get rid of the really laborious mundane

[00:05:17] work. There's nothing but upside here for our employees. But the reality is it fundamentally

[00:05:24] changes the labor equation. There's just no doubt about it. And so again, I go back to these KPI

[00:05:30] improvements. Let me just pick on one. Let's say you look at how long it takes your educational

[00:05:35] services department to create content around your products. Well, if you can reduce the amount of

[00:05:42] hours involved by 50 or 60%, that means one of two things. Either you need less employees

[00:05:49] developing content or, and I think a more likely scenario, you can basically just freeze the amount

[00:05:55] of content developers as you grow, right? As you add more products or whatever, which I think is

[00:06:00] going to be a common scenario. So I think this question we put out there 20% and I think that's

[00:06:07] ultimately going to be conservative. And if you say, okay, if tech headcounts are going to be

[00:06:12] reduced by 20%, there's one of two ways that happens. One is companies literally say, look,

[00:06:17] I just don't need these employees anymore. I'm going to downsize. We are already seeing that.

[00:06:21] Google is one of the few companies that's been transparent when they've done these reduction

[00:06:25] forces. They've said, hey, because we've deployed AI in this use case, we don't need these types

[00:06:30] of employees anymore. So that's one. I think the other scenario, which is also already unfolding,

[00:06:37] but it takes a little bit longer to see it, is tech companies basically just freeze or

[00:06:43] dramatically slow down their headcount growth, even though the revenues are growing. So if you

[00:06:49] look at Google, Microsoft, Amazon, Salesforce, those are all companies that have more revenue

[00:06:55] under management now than they did two or three years ago. And all four of those companies have

[00:06:59] less employees. So it's already happened. Yeah, I completely agree with you. And if

[00:07:05] I look back though, at the same time on the flip side of that, computers, the arrival of the

[00:07:10] internet, the arrival of the smartphone and now AI, they've all promised to give us more time,

[00:07:16] more free time. I don't know about you, but I don't have any more free time. I'm doing 10 times

[00:07:22] the amount of work that I did do more than a decade ago. You seem to fill up those gaps with

[00:07:26] more work and other stuff, don't you? Yeah, true, true. Yeah. On a more positive note, what kind of

[00:07:34] AI use cases have you seen show more financial or greater financial impact on businesses within the

[00:07:41] tech sector? What are you seeing there? Yeah, because it is lumpy in a sense. And so what we

[00:07:47] do for our research is we track AI use cases across seven very specific areas. So we look at

[00:07:55] the service activities. We say, how's AI being used in professional services, support services,

[00:07:59] education services, customer success, managed services. We also look at how AI is being used

[00:08:05] in revenue management and offer development. And what we're seeing right now, there's a very clear

[00:08:11] heat map here. There are massive ROIs for AI in both education services, as I mentioned,

[00:08:18] developing content and support services. So this is helping customers be more self-sufficient.

[00:08:23] This is automating some of the technical outage workflows, all that kind of stuff.

[00:08:27] So those two areas, really big ROIs. And then the opposite end of the spectrum is it's fascinating.

[00:08:34] If you look at revenue management, so if you look at sales organizations using AI,

[00:08:39] let's say to qualify leads, to prioritize opportunities that they've changed, to do

[00:08:44] propensity and renewals, all those types of areas where AI could be applied, there's a lag there.

[00:08:50] So I think that they're going to catch up, but right now those use cases aren't as common.

[00:08:55] And on behalf of a lot of business leaders that could be listening to this conversation all around

[00:09:00] the world, finding everything a little overwhelming, that breathtaking speed of

[00:09:04] technological change. What are the major barriers that companies are facing when

[00:09:09] integrating these new AI capabilities and how can they overcome these obstacles? Because again,

[00:09:14] I'm seeing a certain amount of anxiousness and nervousness about knowing they need to do

[00:09:19] something, but equally it's quite challenging, isn't it?

[00:09:22] Yeah. And you see this with management teams, you can be frozen by this, right? I mean,

[00:09:27] you're like, gosh, I know there's something going on there. Maybe nobody in the company

[00:09:30] has any direct experience with AI and how to deploy it or where to deploy it.

[00:09:35] So there's a couple of things that I can put on the table there. When we do these industry

[00:09:40] case studies, we're looking for where companies have successfully deployed AI,

[00:09:44] in particular in this case, they have a clear ROI they can talk about. And when you bring back

[00:09:50] the lens and you look across all of those different use case studies, you start to see

[00:09:55] some common what we call success attributes, right? Things that happen again and again and

[00:10:00] again when people successfully deploy AI. And so we've defined nine of those. One of them I'll put

[00:10:06] on the table here, which I think is helpful is if you're thinking about deploying AI,

[00:10:12] you want to start, and this is, you might not be intuitive, but start with workflows that are

[00:10:18] workflows that are already well-defined within your company. So as opposed to just saying,

[00:10:23] I'm going to take AI and grow out like co-pilot online employees and kind of see what happens.

[00:10:28] We're seeing if you want an ROI, it's the opposite. So again, I'll pick on content management. Most

[00:10:33] education service organizations have a very well-defined workflow for generating their

[00:10:38] content. You look at that workflow and you say, okay, where could I insert AI?

[00:10:43] You look at support tickets, right? And we have a very well-defined workflow for how

[00:10:47] the support incidents come in, how they get escalated, where in that workflow could AI

[00:10:52] make a difference. And when you start there, it's so much easier to get the ROI.

[00:10:58] So I highly recommend that you think again within your company about the processes you feel you

[00:11:04] already understand pretty well, but I still bet there is huge opportunity to apply AI to that

[00:11:10] workflow. And again, you're going to get 20, 30, 40, 50% reduction in effort if you apply it in

[00:11:15] the right place. And if companies do reduce their workforce or slow down getting new people into

[00:11:22] the business due to AI integration, I know this is a huge question again, it's very touchy subject,

[00:11:28] as we said a few moments ago, but how should they handle the societal and ethical implications of

[00:11:34] these difficult decisions and maybe invest more in bringing people along for the ride and

[00:11:39] re-skilling people? How do you see this playing out?

[00:11:43] Yeah, I mean, I think this is going to be a massive topic for multiple industries here over the next

[00:11:51] couple of years because I think, and this is not a first time technology issue mentioned, right?

[00:11:56] Technology has come along and it has disrupted us. And I was just reading a fantastic book about

[00:12:02] the admin of the steam engine and how many workers who were really guild workers, right? They were

[00:12:09] doing everything by hand, were displaced. But I think that the issue here is I think about this

[00:12:16] is first of all, this initial job displacement is going to happen, right? So it's not like you

[00:12:22] can look at companies and say, oh, you just deployed AI, now 50% of the employees in that

[00:12:27] particular function are required, but you're going to just hold on to them. That's not going to

[00:12:31] happen, right? So companies are going to go after those cost savings. I think there's no way to stop

[00:12:39] that. I think the question that gets on the table for us as a society is once that starts to happen,

[00:12:48] there are two things you have to react to. Number one, how do you invest in retraining that workforce,

[00:12:53] right? So in part of that could be, okay, if AI is going to be here to stay, how do you make sure

[00:12:58] people have the right skills to work with AI or you train them to do something completely different,

[00:13:03] right? Because we obviously have places in our society where we need employees. So who's going

[00:13:08] to pay for that, the retraining? Or if we truly have ultimately employees that are displaced and

[00:13:14] not required, how do you basically fund those people? And I think that the one thing that we

[00:13:22] don't have our heads around because we dealt with manufacturing, right? So the blue collar

[00:13:27] workers, some of those jobs on the way, they never came back and there wasn't a natural place for

[00:13:32] some of those people to go. How do you support them? This is really the first time in my lifetime

[00:13:37] where we're talking about the cost of supporting unemployed white collar workers and how are we

[00:13:42] going to deal with that? So I think that is on the table. I think it's going to unfortunately move

[00:13:48] fast for us, faster than the blue collar ship that happened. And I think we're going to have

[00:13:54] to get our arms around it. And unfortunately we did not learn a lot of good lessons with the

[00:13:59] previous displacement on the blue collar side. So I think we have some new muscles as a society

[00:14:04] we're going to have to build here. Yeah. And I would also say if we did look back 20 years,

[00:14:09] there were roles like social media managers, data scientists, cloud architects, mobile app

[00:14:15] developers, SEO analysts, podcast producers that didn't actually exist 20 years ago. So there is

[00:14:22] new roles that will appear. Looking ahead though, how do you envision AI shaping the future landscape

[00:14:29] of the tech industry over the next decade? And I realise how ludicrous that is. I'm trying to make a

[00:14:35] prediction 10 years into the future with everything that's happened in 18 months. But

[00:14:40] what do you see on the horizon here? Yeah. I mean, I think... I mean, I'm... And I'll listen

[00:14:45] to your... I guess you're an optimist. I think like I am. I mean, I think that there are,

[00:14:50] again, there's some societal issues we're going to have to deal with here. But ultimately,

[00:14:55] do I think that is AI going to be a positive impact on the industry and for the customers

[00:15:00] of the industry? And I think the answer is absolutely yes, for sure. And because I actually

[00:15:06] believe that the two big benefits of AI for tech companies in the long term here is number one,

[00:15:13] and we're already seeing this, is it's going to enable technology companies to deliver business

[00:15:19] value more consistently. So when you look at some of these AI enabled solutions, instead of telling

[00:15:25] your customer, hey, I've got a great piece of hardware here or a great piece of software that

[00:15:29] has all these features, you're really getting into the customer's workflows, you're really moving

[00:15:35] the efficiency to the customer, which makes everything better.

[00:15:38] So I think the value propositions of protect companies goes up,

[00:15:42] it's much more anchored on business outcomes. And then also these technology companies are

[00:15:46] going to be able to do that way more efficiently and effectively. And we're talking about displaced

[00:15:52] employees, but you and I know that for most of the history of tech, it's been an issue of a talent

[00:15:57] shortage. Right? And so we're good... So that's the big benefits. I think the other two things I'll

[00:16:04] put on the table here is for tech professionals, what does this mean to you? I think everybody's

[00:16:10] workflow changes. Everybody is going to be leveraging AI in some way. Even myself as

[00:16:18] a researcher or writer, I'm leveraging AI. I'm sure you're leveraging it in what you do for a

[00:16:23] living and what you weren't a year ago. Everybody's going to have to get comfortable with that.

[00:16:27] And then the final observation I'll make is I do believe in what we'll call the short term.

[00:16:33] So maybe that's the next three to five years. AI is going to create some huge market winners and

[00:16:41] losers in tech, because I think the companies that are AI aggressive, again, two things are

[00:16:46] going to happen. They're going to have better value propositions for their customers and they're

[00:16:49] going to be way more operationally efficient. And that is going to allow them to grab a lot

[00:16:54] of market share compared to the laggards. So I think a lot of moving parts here, but again,

[00:16:59] ultimately I am an optimist on the impact on the industry.

[00:17:03] Yeah, as an optimist too, I can't help but think that those businesses that use AI solely as a

[00:17:09] cost saving exercise, they'll only make a few short term gains, but those who leverage the

[00:17:14] best of people and technology, they're the ones that will thrive. So what advice would you give

[00:17:19] to leadership teams looking to leverage AI, not just for cost cutting, but for doing other things

[00:17:24] like driving innovation and new business opportunities too? Because for me, that's

[00:17:29] where the magic happens, right? Yeah, I totally agree. And I think the first, because let's be

[00:17:35] honest, I mean, AI as a capability, as a technology capability is such a new thought. So if you're not

[00:17:43] having been studying AI for the last decade or whatever, and you're an executive, you're like,

[00:17:47] whoa, whoa, whoa, what is this thing? And I think the first job for leadership teams,

[00:17:53] I tell them you need to go get an MBA in deploying AI capabilities. You need to understand

[00:18:00] what's practical, what are the use cases, how can I make a difference in workflows that you really

[00:18:07] need to lean into that, right? You have to get almost like a new education. So I think that's

[00:18:11] for all management teams, you should be just voraciously studying and reading and talking

[00:18:17] to anybody you can on this topic so you feel comfortable that you understand how it fits

[00:18:23] to your business. And then the second thing I would tell management teams is you do need to

[00:18:28] have a sense of urgency. And I know we always go through these hype curves and everyone's like,

[00:18:32] oh my God, this is revolutionary and everything's changing. But I'm already seeing,

[00:18:38] I get nervous about the difference now between how companies that are leaning into what they're

[00:18:43] already learning, what they're already deploying, how they're already adding new value to the

[00:18:48] customers versus the ones that do have this wait and see mentality. I think that's going to be a

[00:18:54] losing posture. So I think if you're an executive out there, a leader in tech, and you're still not

[00:19:00] sure how serious you should get about understanding AI, I think that ship has sailed. You got to get

[00:19:05] on it, you got to swim on it, get back on it and have a sense of urgency.

[00:19:10] 100%. And we have covered some tricky points today, addressed the elephant in the room or

[00:19:16] the third pan gorilla in the room, whatever you want to call it. But I'm curious, in your work,

[00:19:21] how are you supporting your members in navigating the rapid adoption of AI technologies? And what

[00:19:26] kind of resources do you provide to help them stay ahead? Because for people listening, it's not a

[00:19:31] journey that you should be taking alone, is it? No, it's not.

[00:19:36] It's a time that everybody should be leaning in and get to learn from each other. We're all learning

[00:19:40] together. As I mentioned, we do a lot of case studies and document where our member companies

[00:19:45] are having success so people can understand that. But the way I think about this research in AI

[00:19:50] and the artifacts that we create, they really end up showing up in three buckets and we just

[00:19:56] keep updating this. So the first one is we do research on why. Why is it critical for a technology

[00:20:02] company to lean into AI? Because we find that if leaders don't have a strong why,

[00:20:09] if they don't understand the why, they don't lean into, okay, I have to understand how to go do

[00:20:15] this. So we do research on the why. We do a lot of research on what we think good is going to look

[00:20:20] like. That's why we think about headcount modeling. That's why we think about value propositions in

[00:20:24] the future. And then last but not least, we do the research on how to operationalize. So we just

[00:20:28] keep refreshing content in those big buckets to help our member companies move forward on this.

[00:20:33] Matthew Wallace Well, it's been a huge pleasure

[00:20:35] to have you back on the podcast. I always like to finish on a lighthearted question and ask you to

[00:20:42] leave the guest, ask you to leave everyone listening with one final gift. And that is

[00:20:48] a book that you would recommend or has inspired you that we could add to our Amazon wishlist.

[00:20:53] So what book would you like to leave and why?

[00:20:55] Yeah, well, your time is great here because I literally just finished this book. It's called

[00:20:59] Da Vinci's Mental Models. It's by Peter Hollens. And he walks through basically how Da Vinci

[00:21:07] lived his life of continuous learning and what tactics he employed. And he translates that into

[00:21:12] how we can all leverage the same kind of tactics. So I think it's a great book for anybody who

[00:21:17] considers himself a lifetime learner. I think in the world of AI right now,

[00:21:20] you better be a lifetime learner leaning into it. So I think it's a fantastic book.

[00:21:25] Awesome. I'll get that added straight to the wishlist. And for anyone listening,

[00:21:29] just want to find out more information about you, your community, the work you're doing.

[00:21:33] What's the best starting point for everything?

[00:21:34] Yes, simplest. So someone just go to tsia.com and launch from there.

[00:21:40] Well, as I said, man, it's been what 12 months since we last spoke. I don't want to leave it 12

[00:21:46] months until we speak again, even though it has gone super quick. But with so much value in our

[00:21:51] conversation today, and I'm so glad we were able to talk about some of the areas that a lot of

[00:21:56] people are thinking about but don't want to discuss, especially in an open forum. So I'd love

[00:22:00] to invite everyone listening to share the challenges that they're going through at the

[00:22:04] moment. But more than anything, just thank you for starting this conversation today.

[00:22:08] Yep, my pleasure. Thanks for having me back.

[00:22:10] So as we wrap up today's conversation with Thomas, I think it's clear that artificial

[00:22:14] intelligence is not just a technological advancement, it's a transformative force

[00:22:19] that's reshaping the very fabric of the tech industry. From boosting operational efficiencies

[00:22:25] to altering employment landscapes, AI's role is both promising as it is challenging.

[00:22:31] So how are you seeing AI influencing your work environment or the industry that you're a part

[00:22:39] of? I'd love to hear your thoughts and experiences on this good, bad, indifferent, I want to hear it

[00:22:44] all. We can keep this conversation going, so please join the discussion by sharing your views,

[00:22:50] your insights, your experiences. You can do that by simply emailing me techblogwriter

[00:22:55] at outlook.com, Twitter, LinkedIn, Instagram, just at Neil C Hughes. We will keep this conversation

[00:23:02] going. And that's it for today though. I'll be back again tomorrow with another story of how

[00:23:06] technology is transforming our life, business, world and everything in between. So thank you for

[00:23:13] listening today though. Hopefully you'll join me again tomorrow and until next time, don't be a stranger.