Nathan Hill: Managing a Team of Agents Now — What’s That Worth?

The business wants AI now — but who keeps it running once it’s live? Nathan Hill, Head of Telco at AWS and a 20-year telco veteran, sits down with James MacDonald to unpack what really happens when enterprises move AI from proof of concept to production.

They dig into the buy-versus-build decision most organisations skip, why so many AI projects stall with no path to production, and the operating-model questions — who maintains it, who fixes it when it breaks — that companies leave until it’s too late. Nathan makes the case that deep domain expertise is the asset AI can’t replace: you can’t grab a dev and say “now you know telco,” but you can finally give 20-year network engineers the tools to build.

They also get into human-in-the-loop for critical infrastructure, ICs becoming managers of agents and what that’s worth, deep-versus-broad careers, and how the go-to-market function is changing now that the salesperson has to understand the tech. Practical, grounded, no hype.

Building Tech Teams is produced by DayOne.

Listen to the episode

Transcript

Auto-transcribed with AI.

James [0:00]: I used to be called a principal software engineer. I was paid $200,000 a year. I’m now managing a team of agents. My productivity is 10x. What should I be paid?

Nathan [0:08]: Good question. Aren’t you in the recruitment business?

James [0:11]: Yeah, that’s it.

Nathan [0:12]: If your chatbot starts, you know, spitting out Gordon Ramsay recipes instead of the answer to common questions on your website, then who in your organization can fix it? So that in-depth knowledge is hard to replace. You can’t take 20 years of experience, grab a dev and say, hey, now you know telco, right? And you know, this concept of current ICs or individual contributors becoming a manager of agents, I think is a very real one. Yes, you’re going to be uncomfortable, but really back yourself that you can learn that new skill, you can adopt a new way of working, you can learn a new industry.

James (VO) [0:51]: On today’s episode, I have Nathan Hill, Head of Telco at AWS. Prior to being at AWS, he spent 20 years inside the telco industry. And we talk about the different career paths people can take, getting various experiences across multiple industries and domains or going really deep and narrow into one niche. We also talk about the challenges that he’s seen with AI being rolled out into enterprise. What happens when it’s rolled out? Who’s taking care of it? Who’s maintaining it ongoing? It’s a challenge that a lot of companies don’t actually think of upfront. It’s a really interesting conversation around the changing nature of go-to-market, the changing nature of those sales roles. Are you better set being a pure salesperson learning technology being a technical expert and then learning how to sell.

James [1:32]: I hope you enjoy this episode. Welcome to another episode of Building Tech Teams, Building Tech Teams in the Age of AI. And today we’re here with Nathan Hill, Head of Telco at AWS. Welcome, Nathan.

Nathan [1:43]: Thanks very much for having me.

James [1:44]: Mate, for those of the audience that don’t know who you are, give us a quick background on your current role and the journey to date.

Nathan [1:51]: So at the moment, I lead the AWS engagement across telcos in Australia. Prior to that, I spent more than 20 years in the telco industry, across a series of roles, operations starting out, engineering, pre-sales, sales, and ended up running sales and marketing for a challenger telco in market.

James [2:11]: Nice. We’ll get to breaking down your role and that skill set and background secondly, but start with day in, day out, you’re going into enterprise-level companies and rolling out AI on behalf of AWS, working with them, partners as well along the way?

Nathan [2:26]: Absolutely, yeah. A big part of our role, especially with the large enterprises, is working with partners, and they can be a mix of our business consulting partners, so the Bains and McKinseys of the world setting out strategy. It can be the system integrators that are doing a lot of the heavy lifting in and around these telcos, and also the ISV partners, so the software partners that are building on AWS. and really working sometimes in 2 or 3 different ways to try and bring together a solution.

James [2:58]: Nice. Big question for a lot of enterprises at the moment. Hey, I want to do AI. We need to do AI. The board’s pushing for AI. Question starts, do we have internal capability for it? Do we work with a partner vendor? How does that come together? In your opinion, where have you seen the most success? Is it building out some internal capacity first? Is it engaging a partner and then reverse engineering that? What’s the best approaches you’ve seen?

Nathan [3:22]: Yeah, I think it’s always, uh, it depends. I think, um, the organizations that are doing well, they’ve got a really good sense of how they want to operate. You know, there are organizations out there that have that capability. They’ve got the history of building. So that could be on cloud, it could be building out data platforms, now moving into AI. Or are they an organization that’s typically bought? And this is a concept we’ve spoken about a lot over the last 5 years, this buy versus build concept. So I think having a good understanding as an organization, which direction do you want to go? Do you want to go down the path of having teams of software engineers and that capability and building up that bench strength internally? Or you’re more focused on your core business, which it may be selling telco, it may be selling software, you may be a building company, and really then you want to either buy those types of solutions off the shelf or do you want to start to bring expertise in-house either through a partner or any other method?

James [4:30]: Yeah, it’s the whole thing with AI at the moment. Just because you can build it doesn’t mean you should, right? If it takes you away from your core capacity of what you actually do day in, day out. Where have you seen it go mostly wrong? Where have you seen the big disasters? Is it a traditional organization trying to then become a technology company? Where have you seen it go wrong?

Nathan [4:48]: Yeah, I think it’s— there’s a couple of things I look at. I think if we think about innovation, and it doesn’t need to be around AI, but how is an organization set up for innovation? Is this— do they have a singular function that’s looking at how we adopt this new technology, be it AI or something else, and how do we do it quickly and how do we drive business benefit, or is it done through consultation, 4, 5, 6, 10 different teams trying to have input? So that’s the first thing I’d look at to say, can they move fast enough? And a natural one in the cloud world is, could a software engineer create a new account relatively quickly, or is there 6, 7 hoops they need to go through just to create an account to create a sandbox? So I think establishing who’s driving innovation, who has that mandate, and ensuring they have control. And then sort of touching on my previous point, depending on the skillset, so, and also identifying what outcome they’re trying to drive. So is there something off the shelf that could allow them to move quickly, has the right guardrails, is secure, is gonna meet the sovereignty requirements depending on their industry? Or is the problem they’re trying to solve highly customized to their industry or their organization, which then lends itself to something you can’t buy off the shelf? That’s where you do need to look at bespoke and custom solutions to drive that outcome. So where I’ve seen companies trip up is where they get that mix wrong. They’re either buying a fairly standard solution and expecting it to be highly customized to drive an outcome, or they’re going the other way. They’re getting a very complex model, a very complex platform, a very complex architecture to solve a basic problem, which they could just go and buy off the shelf.

James [6:35]: For the companies that are trying to do it and doing it, you mentioned just before actually just using a use case to start with. So when you go into an organization, obviously you’re selling a big picture, you’ve got plenty of products there. Is there a successful avenue where you go in, find a use case for the new product, let’s say AI at the moment, get it into production, show some value, and then start to chip away at further parts of the business?

Nathan [6:59]: Yeah, absolutely. And it’s sort of that muscle memory. You know, can we find something that helps you build something end-to-end? And I think another challenge that’s been occurring across enterprises is finding a use case, finding a proof of concept, but building it in a way that it in no way can get to production. So sometimes there’s a trap of going, hey, we’ve done AI, we’ve proven something can work, But if the path to production is not built or is a long way out, the time to realize value isn’t there. So I think absolutely finding a use case that’s relevant to the organization, building a team, either internal or external, to go and execute that, ensuring that it has a path to production, and then finally that it can be operationalized. So is the organization you’re building it with, are they mature enough to manage this solution that’s interfacing with customers or becomes a key part of your organization? Or will there be a challenge if it stops working or if there’s a problem with that solution?

James [8:09]: I say there’s a lot of talk around becoming AI native as a company and the big valuations on the startups at the moment or scale-ups are AI native companies. There’s the trillion-dollar problem of becoming AI native service businesses, that type of thing. Is it a real thing for most companies to become AI native? Or do you think it’s just, hey, we look at it like cloud or like another tool and just improve efficiencies through the business and that’s a good use case?

Nathan [8:35]: Yeah, 100%. I think it’s going to depend. In the cloud world, we’ve had concepts of ISV motions. So these are cloud native businesses, Canvas, Atlassians, these people that were born in the cloud. So they only ever built on cloud. But then if you look at traditional organizations, banking, telco, they’ve got a long list of technical debt and legacy systems that have been in place for 10, 20, sometimes 30 years. So they need to think about a different way of adopting this type of technology. So I think, um, you know, the aspiration can’t be start with we want to be cloud native. It needs to think about How can this technology be relevant into our organisation? Does it differentiate us from our competitors? Does it make us more efficient? Does it allow us to reach a broader market to drive top-line revenue? And it’s those sorts of tangible outcomes that then drive the tech strategy rather than the other way around.

James [9:36]: Yeah, nice. Mate, earlier on you mentioned security, you mentioned scalability, guardrails. I think the big challenge at the moment for a lot of organizations, especially at enterprise level, again, all the hype around startups. I can build this so I can scale it quick. I can get a website live in a day. There’s a recruitment company I’ve, I’ve seen online that have talked about building their own ATS system.

James (VO) [9:58]: Hmm.

James [9:59]: Building your own ATS system then has ramifications once you start to put candidate and client data in there and you start to put ID in there, and then you actually become a software business and you have legal ramifications around actually securing that, making it stable, secure. For enterprise, I think it’s a bigger challenge, or it’s that same challenge again, right? You have to actually, you’ve got that fight between doing it quickly, showing some form of value, because again, boards are looking for, shareholders are looking for, what are we doing? How are we progressing? Versus the technology team, which can sometimes be seen as pulling the company back or putting these guardrails of security in place. How have you seen that? play out? Have you seen companies do that well? It’s a very real challenge.

Nathan [10:40]: Yeah, I think if we look at taking ultra-conservative approach, so some organizations are, we’re not going to adopt AI until we’re 100% certain that we’re fully compliant, secure, governed. But then there’s an opportunity cost of doing that way. So the customers, especially in finance and telco where they’re heavily regulated, the way to adopt this is what can we do now safely, right? So that might mean not trying to automate one process end to end. It could be we’re comfortable and we’ve got confidence that, you know, half of the process can be supported by AI. The second half, we’re going to use our traditional methods. But that doesn’t mean that, you’re not getting value. You may still realize 50% of the benefit, whether that’s in effort, whether that’s in time to release, whether that again touching or getting access to a greater client base. So breaking that up into what’s achievable now and then building a path to that fully governed, fully secure, fully sovereign style of environment. So getting that balance right.

James [11:56]: You mentioned before, Obviously you work with a lot of partners and you, you yourself as an external vendor, you’re working with these companies. The big challenge is as you’re completing the project and starting to finish that, do the companies you’re working with have internal capability, capacity to actually maintain the products ongoing?

Nathan [12:13]: Hmm.

James [12:13]: And the big challenge is that handover moment. Where are you seeing things go well? How are companies best managing it? Is it hiring that team in conjunction with the project actually going on? Obviously there’s big costs involved in either building internal capacity. We’ve touched on it before.

James (VO) [12:28]: Are they going to maintain it? What does that look like?

James [12:30]: look like?

Nathan [12:30]: Yeah, it’s a really good point. I think when there’s strategies built outside an organisation, often the people writing those strategies aren’t involved in the delivery. So I think that’s a common challenge. So getting that connection between strategy and the delivery teams through a partner, obviously embedded within that organisation. I think the second part is, again, looking forward towards that end state. As, you know, these solutions are being built, what does my operating model look like when in production? Now, for some organizations, that is, well, we’ve now got access to amazing tooling. We’ve got the ability to use tools for natural language programming and software development. So we want to embrace that and bring the capability in-house, right? So we might have domain experts, let’s say I’ve been a network engineer for 20 years. It’s now very possible to upskill them to have a software aspect to their role, as opposed to 5 years ago, that was very, very difficult, or there was a very small subset of those resources that could make that bridge. Or do you say, no, we’re going to focus on our knitting, we’re going to focus on our core business, which is doing XYZ. and we’re going to bring a partner in to manage the operational elements, right? So I think, again, getting a clear vision on what that end state is. But what’s worked well is if a customer’s looking to leverage a partner through a program to deliver XYZ outcome and they want the capability in-house, is really ensuring that those teams are side by side. So the partner is, teamed up with the internal teams, that the internal teams have enough time to be able to absorb this new information. Often what we’ve seen is enterprises are pulling people out of their day jobs to do this certain project, but their time is limited. They can’t absorb the new capability. And really ensuring that, you know, the question is, if we’ve put an agent into production, if it’s engaging with our customers, if it becomes a core requirement for us doing our finance or our end-of-month reporting. If your chatbot starts spitting out Gordon Ramsay recipes instead of the answer to common questions on your website, then who in your organization can fix it? Do you pull it out of production? Can you adjust it on the fly? I think there’s various models that can work, but asking those questions upfront, asking the questions of your partners, really having a think about what your operating model looks like post-project is really key in getting it right.

James [15:21]: Nice. The way the teams are set up, is it any different at the moment with rolling out AI as opposed to when you were rolling out cloud many moons ago?

Nathan [15:29]: Yeah, a lot of conversations with CIOs, CTOs around the concept of a cloud centre of excellence has been around for a long time. We’ve now evolved that into data centre of excellence as you went into data warehousing and lighthouses and, and other concepts. Now it’s AI centers of excellence, um, and how does that work across an organization? You know, there’s 2 parts. There’s one, okay, what’s your resourcing model? What’s your organization structure look like? Um, but then it’s how is your organization adopting AI? And I was lucky enough to spend some time with a guy by the name of Tom Soderstrom. He, he spent some time as the CTO of NASA, right? And he had a really cool analogy when rolling out new tech. And, and at NASA it was about adopting cloud for the first time. And when you talk to customers, they’re concerned about risk. He would just say, well, it’s pretty risky putting a space shuttle up in the sky. So we thought about risk, right?

James (VO) [16:29]: Yeah.

Nathan [16:30]: But he spoke a lot around the people aspect, which is if you centralize innovation, at least initially, and a lot of organizations are doing it, you stand up an AI team, right? So you can really focus that capability and upskill them, making sure they understand the tech, the security, the governance. But the way to get real business value is you need the lines of business leaning in. You need— they’ve got the insights in what can drive value, right? They may not understand at a technical level the AI component, but they’re the ones that know where it can have the biggest impact. So how do you create that bridge?

James [17:05]: Yeah.

Nathan [17:05]: And his concept was around, you know, you don’t want this, this cool kids aspect where there’s this centralized team of cool kids that get up in front of the town hall and get to talk about AI and have all these cool aspects, but they’re isolated from the business. So I think what we’re seeing is organizations start with a central model, but they’ve got a very clear plan to then think about either how do we seed this capability into the various business units or lines of business within an organization, but the end state is that it should live there.

James [17:40]: Yeah.

James (VO) [17:40]: Right.

Nathan [17:41]: And that’s the only way to tangibly continue to drive value, ensure that if it’s James that has a great insight into how to improve XYZ process or how to make a customer experience more effective or how to drive greater sales, that they’re close to the technology that can enable it.

James [18:03]: Yeah, and that probably comes back to the point made before where consultancies potentially own the strategy. And if that strategy is not owned internally, not driven internally, the opportunities for that to fall away or get buy-in throughout the organisation falls apart.

Nathan [18:15]: 100%. Or you’re leading into another engagement and another engagement and another engagement, which is not sustainable. In the long term.

James [18:26]: No, I like it. Now, you mentioned before the network engineer who’s now potentially more capable, potentially network engineer is a pseudo software engineer to an extent. From the internal teams, how have you seen that adjustment work best than the companies you’ve worked with? Is it we’re working with hiring those specialists in or is it just taking those people that have that internal knowledge, internal knowledge of how the business works? Now you upskill them with the right the right skills, you potentially help them on that engagement, on that learning path. How have you seen that evolve?

Nathan [18:56]: Yeah, 100%. I think if you looked at telco networks 10 years ago, you might say 5%, 10% of those resources knew anything about software. What we’ll see in the next 5 years is that 90% of those resources will know a lot about software. These networks will become more software defined. I’ll be leveraging AI to drive better outcomes. What I’ve seen is that you can’t beat domain expertise, especially in networking. We’re talking about real things. We’re talking about copper in the ground fiber. We’re talking about satellites. We’re talking about subsea cable. We’re talking about physical exchanges. So that in-depth knowledge is hard to replace. You can’t take 20 years of experience, grab a dev and say, hey, now you know telco, right? But you can with, with the tools available now, and you’ve, you’ve shared this online on how you can vibe code and, and build agents, that we, we can bring that knowledge to those domain experts. And I think that’s the path that organizations want to take. Now, not everyone’s going to be on, on, on that. Not everyone’s going to want to do that. But I think as that evolves over time, that’s how telcos or any organisation is going to have the biggest impact.

James [20:20]: I like this angle and there’s multiple parts I could pull apart. Let’s go start with you’re a network engineer, strong domain knowledge, new tooling now helps you understand software, even be able to write something, as opposed to the role of a software engineer who maybe doesn’t have that deep knowledge but understands how to actually build software, what’s under the hood, how to architect it, right? Who’s in the better position going forward?

James (VO) [20:45]: Is—

James [20:45]: do you have an opinion on that?

Nathan [20:47]: Not really. I think we’re still working it out. I’d say the industry’s working it out. You know, there’s various models where you have the domain experts now with software capability, but to ensure consistency, compliance, Then you have a deep software development expertise sitting above so that we’re checking the coding, we’re ensuring it’s aligned, we’re ensuring that it meets all our criteria. That’s one model I’ve seen. But I think it’s, it’s going to evolve.

James [21:19]: Yeah.

Nathan [21:20]: We don’t know, you know, with new models, new, new tooling available, you know, who would have thought we’re going to be here 2 years ago, let alone where we’re going to be in another 2 years?

James [21:30]: You’ve mentioned before online about agents not replacing engineers. You can have a look at, hey, we’ve got more capability now internally because our team’s got better tooling. Do we just do more with that team now as opposed to replacing them?

James (VO) [21:43]: Yeah.

James [21:44]: You want to flush that one out at all?

Nathan [21:46]: Yeah, no, absolutely. I think especially when it comes to things like telco, which is critical infrastructure, you think about the telecom days when we had copper in the ground as a government-owned organisation because it was essential. People needed to use the telephone. Now we have critical infrastructure, but it’s, it’s run by privately owned companies, right? So I think given the compliance and regulatory aspects and we’ve seen, you know, the impacts that, that can happen when there’s challenges in these telco networks, that the human-in-the-loop process will still be there. But in the networking space, there’s examples globally where it is about aggregating insights. These telco networks are extremely complex. There’s multiple vendors, there’s environmental challenges. So being able to aggregate that information, so whether that’s aggregating alarms and proactively providing a resolution that an engineer then takes context into and takes action, That’s one we’ve seen. Or it’s around generating those insights at scale. So when you think about all the information that can be captured across mobile and fixed networks, what insights can we get from that?

James [23:03]: And that’s where this tooling’s just given capability that we’ve never had before, right?

Nathan [23:07]: Absolutely. And you think about forecasting, we look ahead to Brisbane Olympics, you can build out this digital twin of an entire telco network and start to run simulations on the amount of tourists in a given suburb and think what that impact that would have on capacity. So definitely human in the loop. Obviously this will evolve, and this concept of current ICs or individual contributors becoming a manager of agents, I think is a very real one. Agreed.

James [23:41]: If we’re talking about that role in particular, again, controversial online. If I’m one of these people that manages a team of agents now, I used to be called a principal software engineer. I was paid $200,000 a year. I’m now managing a team of agents. My productivity is 10x. What should I be paid?

Nathan [24:00]: Good question. Aren’t you in the recruitment business? Yeah, that’s it. It’s difficult. I think again, everyone’s working this out and I think a big challenge at the moment is although intuitively we know that you know, whether you’re using Claude or whether you’re using ChatGPT to write recipes or plan your holiday and in a business context, clean up your emails, write strategy documents, is quantifying the exact value of that right now. Is it minutes in a day? Is it hours in a day? Is there more accuracy? Is this driving certain benefits organization-wide? So I think as you know, we can quantify the benefit, then that’ll have a natural flow on to things like remuneration and scale.

James [24:46]: And I think we’re starting to see with some people offering million-dollar-plus salaries and they’re doing that on the basis of not necessarily they’re going to provide million dollars today, but as they evolve, as the agents get better, as they orchestrate, architect this solution, that they’re going to easily provide that as a positive ROI, right?

Nathan [25:03]: Yeah.

James [25:03]: Nice. Mate, your background obviously very strong in that telco. You mentioned before domain knowledge. You’ve stayed in a vertical, which is not common these days. I would say, I think people will jump around industry to industry and you can be agnostic across industries. Do you think obviously it’s allowed you to excel and go deep to domain knowledge? If you are giving advice to a younger version of yourself, same again, would you just stay deep and get that domain knowledge? Obviously you’ve learned different things across the way. Thoughts, advice for a younger version of yourself?

Nathan [25:38]: Yeah, I think there’s probably 2 ways to answer it. One is from the individual and one is when you think about the organisations you’re working with. So if I look at how I’ve engaged in industry, say in a sales role, the horizontal versus vertical conversation is based on the maturity of that customer. If you’re walking in, it’s the first engagement, let’s say 10 years ago, they’ve never adopted cloud. Um, you’re not going to go in with a heavy industry discussion, right? They’re not, they’re not ready to execute on anything industry or really deep in that domain. It’s more horizontal, right? Um, the other side, when they’re really mature, then you go deep on industry because that’s where you can drive value, you can earn trust and whatever else. And then when I think from a career perspective, it’s similar. If you’re at the early stages of your career, I think horizontal is perfect. Get exposed to as many different industries too. It broadens your knowledge. Sorry, firstly, it broadens your knowledge. But secondly, I didn’t know where I wanted to be. I didn’t know if I wanted to be in telco. I didn’t know where I wanted to be at the early starts of my career. So the more exposure you can get to different organisations, different industries, different verticals, Yes, it broadens your knowledge that will help your career long-term, but I think it’s an awesome opportunity to find out what you’re really interested in, and then you can choose to really focus in on a particular industry.

James [27:08]: I like that. You mentioned sales. Obviously, you’re really strong in that go-to-market. Go-to-market, I think, is increasingly important role. I’ll go back to more startups, scale-ups, right? If everyone’s more technical these days, building products easier, the sales aspect of it actually becomes the bigger challenge for an organisation. How have you seen the role of go-to-market change? Because I think back in the day, again, if I go more startup scale-up, I can whip up a POC pretty quickly and put that in front of a client as opposed to talking about a PowerPoint presentation or something conceptually. How’s the role of go-to-market and sales changed from what you’ve seen over the last couple of years?

Nathan [27:45]: Yeah. It’s changed a lot. I think so much 5, 10 years ago was just relationship selling, and that’s still important regardless of the tech relationships, especially at the large end of town, are key. You need to earn that trust. But I think customers expect more in those first types of engagements. And the analogy I always talk to with my teams is, you know, when you go to buy a car, you can go and talk to a sales guy in a showroom and you can say, you know, what you’re looking for in a car, the size of your family, you know, what you want to spend, you know, do you like red, do you like it fast, do you like slow? And the sales guy can tell you why that car would be good for you and why it’s better than other brands and how much, how many kilowatts it’s got and how much torque it’s got. he doesn’t grab a mechanic to walk you through that, right? So I think that sort of mirrors what’s expected in our industry now. As a salesperson, you’re not just the relationship person, you’re not just picking up the tab for lunch. You need to understand the technology that you’re selling, you need to understand your customer, you need to understand your industry and where that tech is having an impact. And I think that’s just going to continue, and especially over the last 2 years, When I look at salespeople in the tech world, you know, when, you know, it was a given that you needed to understand cloud fundamentals and things like operating model security. But now the, the requirement to deeply understand AI, deeply understand guardrails, deeply understand the competitor landscape, and probably more so that bigger aspect on people and how they’re going to adopt this technology is where it’s really changed.

James [29:39]: I like this because back in the days, obviously definitely more relationship driven. You would open the door and then maybe take your solution architect, your enterprise architect along with you to actually have that technical conversation where the expectation these days is you actually have at least a depth of knowledge on that. You’re still going to have somebody that’s more technical.

James (VO) [29:57]: 100%.

James [29:59]: Is it easier for somebody who’s a charismatic, typical salesperson to learn the technical side of things or an engineer to increase their sales knowledge?

Nathan [30:10]: I’m heavily biased because I came from engineering background. So, you know, and look, I’ll openly say that I’ve hired many a salesperson that’s come from the solution architect world.

James [30:24]: Yeah.

Nathan [30:25]: Having a background either in operations is that you see the tangible impacts of if a solution works or it doesn’t, and that’s important and it makes you very pragmatic and you think about how these impacts when you go to actually sell. Obviously, if you’re a solution architect and you’ve got an interest in moving into sales, it’s a massive massive benefit. You earn so much trust with a customer when you can understand the technology and understand the business. Not saying it can’t be done, but I’ve seen that sort of career progression as a, as a consistent sort of launchpad into sales.

James [31:08]: Yeah, nice. And I think the expectation goes further than go-to-market anyway. If you have a look at software engineers of today, the expectation back in the day is you’re sitting in a dark room, you’re banging out code. The expectation these days of software engineers actually understand the stakeholders, understand the customers. Who are you building the technology for? Why are you building said technology? So I think it’s starting even as technical as a pure software engineer right through to—

Nathan [31:31]: Absolutely. And you’ve seen just how software engineers need to engage across business now, where they are sitting within business units, they’re sitting within the contact centers or the customer experience teams. They’re sitting amongst the sales teams, they’re sitting amongst the operations teams. They’re not siloed away centrally and only get engaged when a project gets raised. So that’s a huge evolution over time.

James [31:57]: So if there’s more and more overlap of the roles, how are team structures changing going forward? Is that go-to-market role a specific part? Is there more overlap between the roles going forward? How do you see that evolving?

Nathan [32:10]: Look, the sales and solution architect roles haven’t changed too much. I think the expectations on technical capabilities obviously changed. I think what I’ve seen is there’s probably more a shift from the pure pre-sales engineer handing over to a production engineer or to someone that’s going to build a proof of concept or build a solution, I’m seeing that blend more where customers want to see the tech, they want a proof of concept that’s relevant to their business. So we’re seeing more and more that pre-sales function is getting hands-on and sort of helping bring to life the value that some of these solutions can build.

James [32:58]: So if you’re architecting a go-to-market function for a new company in 12 months’ time. How do you go about building that go-to-market? So if I’m an enterprise company and I want to start to build a go-to-market function, how do I go about, you know, who’s my first hire, who’s my second hire, what skill sets am I looking for, how does that evolve?

Nathan [33:14]: Yeah, I think a big part, if we’re still focused around the data and AI piece, is what’s your channel? You know, we’ve got model providers that have direct channels to consumers. And that’s generating a lot of revenue, a lot of interest. Sometimes that interest bleeds over to enterprise. If we go back to the iPhone, the iPhone wasn’t an enterprise device, but what brought it into enterprise was CEOs saying, hey, my kids are talking about this cool thing called the iPhone. It was really good. I now want to use it at work. So we’re seeing a bit of that. So I think as a go-to-market business, you’ve got to think, what’s my channel? Am I direct? Am I targeting organizations that are buying off the shelf? Am I targeting organizations that want to build? Or am I supporting another go-to-market via a partner and I’m essentially providing the motor, they’re providing the chassis? So I think getting real clarity on that. So am I hiring a go-to-market sales leader to build a strategy in order to engage directly? And am I hiring a a channel or a partner leader to go and build up the channel capability. So I think that’s where I’ve seen organizations challenged, not having clarity. And it could be where are we at today, where do we want to go, or what’s our roadmap look like from a capability perspective. Then when it comes down to people, it comes down to obviously being able to engage with customers, but then build your awareness. And you can’t do that without people on the ground. Yes, we can do marketing. Yes, you can use LinkedIn. Yes, you can do this type of thing. But being able to engage directly with organisations, share your message and then amplify it is going to be the first key piece.

James [35:05]: So if we go buy versus build, same with technology with your actual team. In that instance, it sounds like you’d find somebody with that industry or vertical knowledge, pick them up from somewhere else, build a capacity around them and then build internally.

Nathan [35:16]: 100%. And I think If you are going down an industry approach, then importing industry knowledge, industry relationships is critically important just to get you in the room so that you can put forward your proposition. If it’s a more horizontal play, then that’s not as critical than it is around scaling.

James [35:37]: Made mention that you’ve got an obvious bias to go-to-market or salespeople coming from that technical background. In a world where we’re going more and more online, more and more technical, we’ve got AI writing spam, getting hit up. Your inbox, I’m sure, is shredded. My inbox, LinkedIn, hammered every day with messages probably not written by a human. I think there’s that importance of still being able to shake a hand, build a relationship with somebody. You mentioned definitely at the senior end of town. Where do you think that core sales skill function still plays out in the market?

Nathan [36:14]: Yeah, it’s still, you know, and I’m working with some of Australia’s largest organizations, some of the that are adopting AI at significant scale. But there’s still a huge relationship aspect and there’s still a huge trust aspect. You know, Tech doesn’t sell itself. The capability can be there, but I think the role of go-to-market is to bring that to life, bring it, make it relevant to that organization, ensuring that you’re presenting the technology through that customer’s lens. You know, what value is it going to drive? But also having honest conversations on their ability to execute and consume that type of technology. where I’ve seen some challenges in the past where you could have the world’s best solution for that customer, but if internally or organizationally or at an executive level, they’re not aligned to take advantage of it, then it’s never going to be successful. But relationship is one of the durable pieces within our organization— sorry, within the industry.

James [37:25]: Yeah.

Nathan [37:26]: is still so important.

James [37:27]: You mentioned there also the trust piece on like you might have the best piece of kit for them, might be the best piece of kit just full stop. But if they’re not able to actually execute on that and you hand it over and it’s done and they’re not set up well or it wasn’t designed for them, then that trust is broken straight away, right?

Nathan [37:45]: That’s right. And I think there’s a lot of new entrants to market and a lot of the new entrants are using a lot of the same technology under hood as some of the more established players. But having a track record of delivery or a track record of resiliency or a track record of security still means a lot.

James [38:03]: I completely agree. And that’s why I disagree with the SaaS apocalypse. People are still going to use Salesforce because when you buy Salesforce, you have an expectation of the 99.99% uptime. They’re going to be safe, secure, et cetera. They’re going to continue to evolve their product. Enterprise isn’t moving away from that to vibe code their own solutions.

Nathan [38:23]: And I think it’s always an asterisk, the concept of headless SaaS. I think it’s real. And through my experience in using agents and leveraging MCP, there’s many platforms that I don’t directly log into anymore. But there’s organizations out there that have made their business on having awesome UX, having an awesome— that someone with no technical background can go and make changes to, whether it’s an HR system or whether it’s a contact center or whether it’s a CRM. So I think that is real, but it’s going to be dependent on the organization. So, you know, if you’re, you know, running, you know, 20 dry cleaners around Australia, Are you just going to log into a nice shiny UX? Probably. You don’t have a team of devs. You probably don’t even have an IT manager, right?

James [39:20]: Yeah.

Nathan [39:21]: So you’re still going to consume things in a traditional way at the top end of town. If you’re one of the large financial organizations and you’ve got hundreds, if not thousands of developers, could they then create their own CRM, manage it and do it securely? Probably. So I think it’s going to be a bit of a mix.

James [39:40]: Yeah, nice. Back to the go-to-market function. One last question there. Obviously, if somebody is traditional, let’s call them salesperson, not overly technical, but they want to evolve to the new go-to-market function, easiest, best way, how would they go about it? Upskilling themselves?

Nathan [39:56]: We’ve got all the information. We’re so lucky. We’ve got all of the internet’s information literally in our pocket through our phones. There’s so much free knowledge available. So about being disciplined in upskilling in those different areas. And I think we’re just lucky as a tech industry in Australia that there’s so much, whether it’s community enablement to get access to certain training, whether it’s industry enablement, there’s so many opportunities to come together with industry or like-minded people to upskill. But it’s just having that motivation to learn and to be curious and to continue to look at this technology. And I think we all need to be disciplined about it because it’s moving so quickly.

James [40:43]: On that point, you’ve become head of telco at AWS, one of the biggest global organisations. Your journey along the way, what have you done to continue to jump those steps along the way? Is it how long do I stay at the one organisation before jumping? How long do I jump multiple times? is that, hey, if one organization continues to provide me opportunity for growth, do I continue to stay there or do I need to be seen to be jumping to not be a plotter? Based on your career, how, and somebody wants to grow a go-to-market, that’s what they want to do for their life. That’s what their vision is. Give some last point, some advice for people to grow that function, to grow their career in around GTM.

Nathan [41:21]: Yeah, I think it’s just back yourself. You know, I think I was lucky. I’ve had a long tenure at a few organizations, but along that time I was changing roles sort of every 2 years, sometimes less. And probably through naivety, I thought, yes, I can do that next role. Yes, I’ll put my hand up. And I really just backed my ability to learn. And that’s what I’d encourage everyone to do. Yes, you’re going to be uncomfortable, but really back yourself that you can learn that new skill, you can adopt a new way of working, you can learn a new industry, and look for opportunities to expand your knowledge. And I think whether that’s at one organization that can continue to provide you that differentiated experience, or whether it’s changing organizations. Yeah, I’ve just been extremely lucky to probably mostly by luck to get those opportunities over the last 25 years.

James [42:24]: Yeah, you put yourself in the right place. Luck tends to have a way to find you though, right? Mate, appreciate your time today. Appreciate your time in looking at how enterprise can actually implement AI, what to avoid, best ways about going about it, and also just that changing nature of that go-to-market function. So appreciate your time.

Nathan [42:39]: Thank you.

James [42:40]: Cheers.

James (VO) [42:42]: The thing I’m taking from today’s conversation that I really agree with Nathan is the domain experts that learn how to build, that learn the technology, are going to be the most valuable people going forward. If you understand how a business works, how an industry works, and then now you can overlay these new skills on building automations, on looking how different technologies can be applied to existing problems, you become really valuable. I think that’s the thing I’ll be taking from today. Use your domain expertise. Overlay some technical skills and you can become one of the most valuable people within your organization. This was the Building Tech Teams podcast. We release new episodes every Tuesday. Really keen for feedback on guests you’d like to see, more topics you’d like to see in depth. If you’d love to provide any form of feedback to me, find me on LinkedIn, James McDonald AU.

James [43:26]: You won’t miss me.

James (VO) [43:27]: This episode was produced by Day One.