It's been a big year from mergers and acquisitions. As a tough economy, competitors to join forces and try and achieve scale here and beyond Australasia.
But what does it mean in practical terms when two companies merge? How do you spice everything together in an orderly way?
Well? Two degrees is in the process of doing just that following its merger with Vocus to create New Zealand's third largest telecoms operator.
On the sponsored episode of the Business at Tech, Stephen Kerjier, two degrees is chief information officer, talks to us about the major job of bringing two large businesses together with two million users on the platform and what he's learned so far in the two Degrees Vocus merger.
And Stephen joins us. Now, Steven, thanks so much for coming on the Business of Tech. How are you doing.
I'm doing great. Thanks for having me. Yeah, gat to be here soon.
We're going to get into some really interesting things that you've been working on around integration of these telecommunications businesses from Vocus to two degrees. So lots of interesting stuff to cover there. But you're also joining us for our usual sort of run around the big tech headlines off the week and a Talco story that's a particular interest to you. The Commerce Commission is asking our big telcos, including two Degrees, One End, Zed and Spark, to start
showing coverage maps. Now I thought they already did this. You go onto your website, you'll see a coverage map there, but they're looking for something a bit more extensive. What's your take on this.
Yeah, I think we've all done it, but in different ways. So we have a coverage map on our website. I think it's actually a really good thing. It provides consistency to our consumers on coverage, so when they sign up, they know what they're actually getting. I think for two Degrees in particular, how we've always been the challenger in the market. One of our bigger issues is network perception. You know, do you actually have the coverage or not?
Given our history, and we've invested in our network significantly over the last three to five years and we have parity now, so we actually think it's a great thing for us to show. So, you know, is coverage poor, is it good? Is it great? That transparency I think is really important for consumers. How hard is to do is probably another question.
That's the thing, because the maps are there, the physical you know, your engineers go out and map this is where the coverage is. But then putting a rating on that is this is really good quality coverage, This is a bit miginal. I guess that's where it will be a little bit contentious.
And how do you actually measure it because I'm sure there would be fluctuations and it would depend on capacity of the towers and things like that. So how are you thinking about that nuance?
Yeah, in one way, you've got kind of network kind of measures, you know, signal strength, and those can be quite well defined. But I think ultimately its customer experience is the most important part of it. So I think coverage maps will give you a broad view, but it may not show kind of end to end network experience. So I think I'll give us enough of a good measure for consumers to make the right decisions, particularly in
the more rural areas. But our focus atally around customer experience and how can we measure that kind of that true end to end view, which is around collecting the right data from the handset, either crowdsourced or you can also get some really interesting telemetry from the network as well on the user experience and also independent testing of the network. So recently we've actually done really well in
some independent tests, so that's another measure. But from the Commerce Commission, it's really just get a baseline on what good coverage is and make it consistent, which I think is actually a great thing to do, but it is probably only one measure. It's not going to be the whole story to your question. Yeah.
Yeah, And the other part of it is that the commissioner is looking to have an exit right put into contract so that if the customer signs up and they find that, like I think it was thirty percent of the small businesses that they surveyed that the coverage they're not happy with what they're getting, that there is a clause saying I want to exit based on this issue. Now, two degrees is one of the tolcoes that's already that
in their contract for some time. Do you find that is something that's triggered a lot, or is it just gives you the confidence to bring on for customers to join you, Like, how does that work out in the field.
Yeah, well, I say bring it on. Yeah, we've already had it. You know, sort of money back guarantee network reliability. More recently we've been doing more advertising on that, like a network blind test. You can actually try the network out, give it a go again. We're really focused on our network's fantastic. Now it's parody across all the players, but we have this network perception issue so that when customers get on our network they actually love it.
So yeah, that was a big thing, getting to network parody as a third player, considering how long it took for both Spark and Vodafone to really flesh out coverage. There was a lot of complaints about those networks for a long time to even in semi rural places, even urban areas was patchy coverage. So I guess in some ways, coming from behind and as a third player, it was a bit easier because you knew some of the pain points they'd been through. But still a huge undertaking.
Significant undertaking, and it keeps going right, it never stops. We're investing hundreds of millions dollars in our network every year. We've got a five G modernization program that we're well in flight. We started late for five G compared to our competitors, but we're well on track. I think we may be one of the ones to finish first, just in terms of our approach. So yeah, it never stops and it's never perfect, Like network coverage is always a
thing where you're always looking to improve. But yeah, coming from the origins of basically starting from nothing, right, two degrees is a challenger, effectively a startup. We're a really big startup now and we still think like a startup, but quite an incredible journey from what was a position of not owning a network to where we are now, where we own over two thousand cell towers, We've got
fiber assets that are four thousand, six hundred kilometers. The network's out of the equation for us, so our focus is now on customer experience and software and all that great stuff. It takes hard work and a commitment to just keep investing for a long term strategy.
And of course you've got the text satellite stuff comes as well, so you've got the coverage map, but then you're going to have where the big gray spots are at the moment, the ability to send text messages initially anyway via satellite.
That's right, Yeah, and that's the exciting part coming up, which we've partnered with link as I think you know and still trialing that. And the time for that to be effective is just how many birds you can have in the sky, how many satellites you can have in the sky. So yeah, we're really excited about that to really plug those other areas of coverage. And I'll also
say it's an industry effort as well. You've got the Rural Connectivity Group that has been looking at how we do things more effective, so it's not just three players going out it themselves. We actually do collaborate significantly and that's going really well as well. RCG went over five hundred sites in terms of the real connectivity areas, so that's been really good. And complement that with a low
Earth orbit for the celt satellite. So I think you have a really great story about more ubiquitous coverage in New Zealand.
Yeah, you said you were thinking you'll possibly be completed your five G roll out ahead maybe some of the others. What does that actually look like. Would it be would have equitable coverage to for G or is it separate like more specified area rollout.
Yeah, it's a great question. I think everyone starts with the higher population areas, right, that's where you start with terms of just needing more capacity, but there's also being a program of work to get five G in rural towns, working again with the government, so that's been really successful. Again that's part of investment strategy to get better coverage into rural towns. But yeah, our approach has been basically
to modernize everything on that cell tart once. So not deploying just five G but modernizing four G and five G simultaneously means to actually get better four G experience and a five G experience at the same time, which I think is a really good strategy for us. It does mean probably it's a little bit slower initially because you're having to do a bit more work, but in the long run it's a great strategy for us to move forward on.
So we'll keep an eye on those coverage maps to ComCom wants them within a year.
Yes, yeah, the team are busy already working through it timeframes.
Yeah, so we'll keep an eye on out. Another big story making headlines this week, which is sort of relevance to us, the Australian government has proposed mandatory guide rails for use of artificial intelligence. They want require testing, transparency about how AI is being used, and accountability. There's a ten point checklist that they have floated. This is out for public consultation for the next month or so. We haven't moved to this step yet, and I think the
government will look quite closely at this. NBA has laid out sort of the roadmap that it's following, and Collins has said that she would like a risk based, proportional and light touch regulatory regime around AI. Here they'll only create AI legislation if there's a real need to, So there's a lot of parity sort of going on with the Australians. What's your take having looked at this already using AI at two degrees and the practicalities of trying to conform to something like this.
I think what I've seen was really doing is actually really good because it is a risk based approach and actually focused on the higher risk areas, but also holistically, they're looking to invest in AI and actually accelerate it, particularly in government, So I think that's a really great thing. Yeah, for us, we've already been doing a lot in this space. So when we compare our AI policy and responsible use
and all the great things around governance for AI. It very much aligns to what we're already doing at two degrees trust and transparency, having a human loop. Yeah, the explainability aspect, which is hard for a black box. But I think it's also important to recognize AI of the past is very different to AI now with large language models.
So regulation previously was effectively based on like machine learning, are you applying decisions to HR process or to a criminal activity or churn predictions and how are you using it? But large language model is actually very different to that, although I wouldn't use them for actually making autonomous decisions yet, but decision support is really great. So I think a lot of the regulation is effectively the AI of the
past that I've actually seen trying to control that. But the AI of the future is very different, And what the Australian proposal is is actually taking more of a holistic view of that, kind of breaking down what is a developer of a model, you're training a pre trained model, and what you need to do because that's a higher risk area. You know, if you don't have the guardrails on, you could create an unmitigated AI l em that could do anything versus someone who's deploying LM into an organization,
which is effectively where we're at. And then also just users of AI. You know, we're all using it every day. I think nearly everyone is, So I think what they've done is actually broken it down quite pragmatically into those different areas and focusing on the higher risk components. So yeah, I think it's actually really good. It aligns with what we're doing. I don't like more audits, so I think there was a mention in the proposal around actually auditing
entities around it. So having a more integrated approach where it's actually part of existing audits and compliance, I think is a better way to go. And I think they actually proposed a few options have it more integrated as well versus separate.
It seems like when you look at AI risk, it doesn't immediately jump out that Talco would particularly be one of the more risky areas we are using AI. But then when you add that large language model factor into it for things like customer service or whatever else you might use for decision making support, that does kind of enhance the risk a step up by default of using
those models. So how are you thinking about that? At two degrees in terms of maintaining your level of I guess soft compliance because we don't have that regulation, but compliance with your own internal guidelines.
So we did deploy straight away quite early on, obviously the explosion of chat gipt and everyone got really excited about it, an AI responsible use policy with the intent to actually embrace the technology, not to kind of stop people using it, but put some guardrails around it so people can use it safely. So that was kind of immediately done and create more awareness around use of AI. So a lot of our work has been just creating,
just doing lunch and learns and awareness training. So we do want to invite people to use it for their daily work. I think that's the best way to test and learn get ready for what I think is actually a huge you know, we're in a new revolution of technology.
We're just we're trying to get ready and taking many different stakeholders on that journey, your board who are probably very risk adverse, you know, the use of AI just in simple things like if you're putting Microsoft Copilot on the teams meeting and the people that you invite into it will automatically get all the summarization notes and you may not want them all to so just being aware of, you know, the technology and what it could do in
terms of transparency of information, we may not want it, and I think ethically we've just our main purpose is to put the customer first. So when you think about your customer experience as your north star, that's your guiding light for anything. It doesn't necessarily have to be AI, right, it could be any technology. When you put the customer first, that's the most important thing for us. So that's kind
of our main principle under the use of AIS. You know, is it going to actually make a better experience for our customers.
You've got to foster that trust, ye You've it's got to be a high trust system and part of that, you know, this is an interesting point that they've floated across the Tasman be transparent with other organizations across the entire AI supply chain. So being able to explain to your customers, Okay, we might be using something of AWS
or Azure, which is underpinned by open AI. We need to be able to tell you a little bit and have confidence that our suppliers are giving uugh enough information to be confident that we're not going to breach privacy or that. So I guess that'll be something that increasingly you'll be looking at through the entire supply chain. What AI are we using and do we know enough about it?
Yeah, definitely, And how using our data? We use a lot of third party systems that probably hasn't changed in risk profile. You've got to effectively trust. If you're using a cloud provider like Microsoft or Amazon, you have to trust what they're saying around their data use even something like open ai GPT four, which may be a misconceptionist because I think people might think of AI or as
a entity. So when you put data in there, it's just learning everything, but it's still just a computer program in some respects, or a data program. And there are features in there where you can specifically say don't train on my data or use it privacy features like you would get with anything like Dropbox or anything. How using your email Gmail, for example, you may not trust Google to be training on your data either, but there are
ways to actually opt in or out. So I think that's you do have an element of trust in the supply chain definitely, which we've already had I think in the past. So whether organizations trust these big tech companies to do that is another question. You've got themes around more sovereign AI and private AI as definitely building momentum
as well. You're seeing that globally, seeing what France is doing that EU are doing, So I could see a future actually where we do have more sovereign AI use, which I think is actually not a bad thing, because the future of AIS, I think will be actually a zoo of them. There'll be LMS for various things. You've got your obviously your big foundational frontier models, but you've
obviously got a huge amount of open source models. So I think it's going to be more of an ecosystem and more of a hybrid approach going forward, which I think we have to be ready for and regulation needs to make sure that is safe.
And small language models for very specific purposes, and of course we've got the hyperscale is setting up here, so the option of keeping all your data within New Zealand's borders will be attractive as well from a privacy point of view.
Absolutely, yeah. Are we using Amazon bedrock for a lot of our kind of experimentation and what we're doing today which has the LMS based in Australia, but you also don't get the latest models that often either. That's also a challenge with these entities. They do deploy them in their main regions first and they come down to our region. So we still want to be on the bleeding edge
as well and experimenting. So yeah, we're also are minting with open source models, creating kind of those foundations to be able to have optionality. Earlier in the year, we released a chatbot, which was a generative AI chatbot to our customers in a test and learn approach. We actually deployed it into Slingshot, which is one of our brands still rather than our bigger two Degree space, just to really understand it's actually quite a dangerous thing to do.
In some respects it can hallucinate. We put a lot of effort in testing guardrails and making sure it was defined in its own context so it couldn't go out of the context. My own team we're testing it. You know, you know, what would Steve buy from two Degrees or from Slingshot? Would you like a pair of shoes with
that or something? And we had a lot of learnings from that, so we've built kind of an orchestration guardrail ecosystem around it to ensure that the chatbot doesn't stray off too far, which is a great learning for us, and it also taught us many other things like our knowledge bases need improving data is a huge aspect to get value out of these things. So it's accelerated our view on data modernization and knowledge management along the way.
So it's I think every organization should go on this journey and actually test learn scale, just try it out, take a few risks, but make sure you've got the guardrails in place. That one is an example where it's actually customer facing. I think a lot of organizations are doing internal based chatbots as well, which is what we're doing, which is more is a safe way to do it as well.
Did you have any moments of like real shock when a customer tried to do something or found a way to break the bot in any kind of way that you suddenly you had to deal with.
I think we got most of the guardrails pretty well sorted what it did highlights because we do have the human and the loop still and looking at the actual responses and actually measuring how effective it is. It just really highlighted our knowledge management wasn't as good as it should be. So our knowledge base is our help desk articles, which is what the model is using to refine its answers.
So when we had some results that we thought weren't quite right, it's actually we're already saying that to our customers. So it's actually helped our feedback loop. So not necessarily anything too wrong. Yeah, I think there are some cases where people tried to say, you know what I like some would you like frise with that or you know a few extra things, and the bot did actually go off and try to help. It really wants to please people,
you know, that's the thing. So we had a few instances where it did stray off a little bit to try and please our customers because that was you know, it was kind of told to do that. So yeah, it's never perfect. I think. I think these things do hallucinate a little bit, so that's why it's effect. You'd still need the human and the loop, which is part of the Australian proposal and human oversight. Do you need
to pull the trigger as well. We've got a within our own policy like a something does terribly go wrong with it we can just turn it off immediately. So we've got, yeah, the big red button, the control loop.
Yeah, So that's what the Aussies will decide on. They've got three options in terms of how they do this from a framework point of view. They might have an AI act like the EU, or they might just tweak existing legislation. Does it sort of fill you with trepidation the idea of dedicated legislation being introduced. Would you prefer to see using existing legislation or is it much of a muchness to you?
I don't think there's any right or wrong answer here on this one. I generally would prefer it integrated into legislation, which I think will need to happen anyway. You've got the privacy Acts that we comply with today, so there's maybe amendments to that. But it is such a transformational
technology that may need some different aspects to it. But I'd like to see, you know, this kind of happening in parallel to the kind of scientific development improvements, rather than being kind of at the end of the process and being preemptive and blocking technology actually going together on the journey, which would be a completely different way of regulation. I think technology is going to move so fast anyway,
and regulation is going to catch up. So yeah, I do fundamentally believe that would probably need a different way of regulating in the future given the speed of this. But yeah, I think getting a few smart people together to work it out would be really important in New Zealand context. I think jud To Collins is already kind of working on that, bringing the experts together. So I don't think there's a right or wrong there, but I'd prefer integrated. Yeah, And if we.
Can harmonize where possible with the Aussies, that would be ideal as well. We have a lot of businesses operating both markets, so it's good. Thanks for your views on those two topical stories. And this really speaks to what we wanted to talk to you about today, which is, you know, there's a lot of mergers and acquisitions happening and tough economic times. We've covered at Business Desk a
lot of takeover attempts of New Zealand companies. Private equity out there is looking at New Zealand going Some of these businesses are relatively cheap at the moment out of economic necessity, Competitors are looking at each other going would we be better off have more critical mass merging. You've been through this process over the last couple of years
with Vocus merging with two degrees and really fascinated. First of all, if you can tell us where is the merger at at the moment thousands of people have been brought together, but also all the systems underpinning these customer bases coming together. Where are you at in that transformation.
Yeah, when you ask me that, I reflect back. It's been a wild journey. So it's been two years into effectively what is a three year strategy for us to integrate. Merger happened first June twenty twenty two, and we've made a huge amount of progress, and I think it's important to kind of reflect. You know, both organizations have been built up in different ways from different mergers and acquisitions.
Like you mentioned, if you look at the history of the organizations, we're actually built up over sixty different service providers and entities. So in a way, mergers and acquisitions are in our DNA, but we're also there's a common thread through all of it, which is that we're a challenger. And Yeah, if you look back through all the different entities, they've had their own stories around challenging the market, disrupting
the market. You look at two Degrees and the two Degrees effect that occurred in two thousand and nine on the Vocus side, there are many similar stories. So I think when we came into the merger, we had a lot of knowledge and how to do it, and we had a blueprints and what we needed to achieve. But every time it's different. Fundamentally, you can have a great vision, but it's how you achieve it that's really important. So, yeah, we're two years in and yeah, effectively nearly at the
end end of our three year integration strategy. It's going really well. The main things we've got left kind of at a high level from a system's perspective, we had these two probably a large business support system which is the two Degrees of Origin that's got most of our mobile customers in it. We've got this other one called Tahi, which is our target state. We rebranded it as part of the merger as well, which is part of kind
of getting people on the journey. And we've been building all the mobile journeys in that system, and we've done a few trials in the last actually three weeks through those mobile journeys and it's night and day. The difference. It's not to say the other one was bad necessarily, but we've invested so much in the customer experience aspect. So while we're integrating trying to consolidate, we're trying to
improve customer experience. And to give an example of the difference, kind of signing up a customer in the retail shop that used to take probably around ten minutes on the old system is now under a minute, so that's a significant change. And the digital experience as well for our customers to sign up online and also our digital experience for our colleagues to actually interact with our customers is
night and day. So it's very exciting. So now we're focusing on our migration aspect to actually migrate customers without impacting the customer experience. So last year we've actually been working on a range of techniques. This is more of the geeky technology side of it in terms of transitional architectures so that we can actually support customers through both stacks at the same time, so building effectively transitional overlays,
so it's cross tax support. So from a customer experience, if you've got your mobile app or your web experience. It's seamless because it actually is supported across both systems. So there's a range of little techniques that we use to try and make it as seamless and effective as possible. Our main focus, which is, don't treat it like a big band kind of big project. Get incremental delivery time
to value is one of our key principles. So, you know, the traditional digital transformation efforts that we go two years in, I'm sitting here quite happily. I'd normally, you know, I'll be gray by now as a CIO or CTO, and normally the half life of a CIO is around that two to three year mark. I'm sitting here reasonably comfortably touch would. But it's because we've been so focused on
the incremental delivery. We've got that value and learning often we don't want to do big bang approaches, and we've seen that over the last two years we've actually had value delivered. We are brought in the fixed broadband from one system to in to TAHI quite early on, within four months, and now the focus on the mobile journey. But underpinning that, you know, you've got this business support system you've also got operational support systems, you've got ERP systems,
you've got the billing charging. Yeah, all of that, TAHI if it does a lot of that for us, But there are hundreds of different systems. Got your data systems started warehousing, range of programs. So it is quite a diverse thing and it's very very hard to do and you have to empower your teams to do it. So I guess that's one of our key principles is actually empowering our teams. Focus on the outcomes, focus on the why, but empower the teams to deliver it. Yeah.
How do you start to make decisions about which parts of your tech stack you've got? So you've got two CRMs from different different entities that have come together. How do you start to think about which one you're going to wind down, which one you're going to build up. What is the strategy that you think through there? Do you go to your customers and ask them, do you just take other metrics to see which are delivering better outcomes?
What does that look like? Yeah, a really good question, and again, yeah, you've got to face it on its merits at the time actually making these decisions and when you look back at the different entities, each one has their own legacy as well. So fundamentally, our approach has always been probably to choose the best based on certain metrics, but not build something new on top of that. So
not a green fields approach, actually an evolutionary approach. So you're looking at your ecosystem of systems on its merits. You do obviously have to factor in commercial constraints as well, but basically capability and the people around it that can support it. It may not be the system that has the most amount of customers or usage or adoption, it's really the system that is going to be future proof
for us going forward. So when I mentioned TAHI, it's actually in house built system, which is quite key for us because it's basically our intellectual property, our unique proposition. I've got a team around four hundred and fifty staff in the technology team that crosses digital, cybersecurity, data, network engineering, the radio access engineers is quite a big team, but we have over one hundred software developers. I consider ourselves
actually a software company with great Taco assets. So one of the decisions that rarely is you know, for TAHI, it was a decision around we have people around it, we support it, We can innovate every day. You know, we deploy features tens of times a day, launching darkly into our system. So we have an ability to iterate
and have really fast feedback on that system. So that was one of the key decision points, was actually time to value, time to learning versus other systems we might be dependent on third parties or vendors, or even the system itself might just have constraints. So one of our main factors was again back to that time to value. May not be the best system in terms of capability, but we could back ourselves to build certain capabilities on that system due to our ability to deliver quickly. So
I think that's actually unique for us. Is nothing I don't think in the New Zealand markets or even when I look globally in terms of that software approach, and that means it turns up to customers in terms of customer experience.
That is yeah, I think quite quite unique because we've heard so much. A lot of the big software vendors have tailored a version of their software for the Talco market. They see that as an important market to serve, whether it's SAP or salesforce, for instance, you've gone to bespoke route is really part of that your legacy as a challenger brand. You're one of the originals way back in the in the early days of Slingshot and all that,
with Malcolm Dick and Mike Callender. You know you're still there. Is that The mindset really is that if we want to be competitive, more competitive than the incumbents, we have to be bespoke. We can't just buy something off the shelf and have a great experience. We've got to build ourselves.
Yeah. I thinks, yeah, there is a mindset to do things differently. You know. One of our values is say no to the status quo. So we like to innovate, do things differently Bespoke Yeah, I think I still think of it like a product. So it's actually what we're building is a product in its own right. So it's not a kind of customized you know bespoke aspect. It's actually something that's strategically important for us. And we don't
build everything. We use a lot of open source, We do use a lot of other software to complement an ecosystem, but the core of it we do actually build and own, which I do think is unique. To give you an example of you know, we don't necessarily you know, build versus by aspect. Right, Building's not always great because it can take a lot of time. You don't want to reinvent the wheel. So when we were looking at we built this actually a few years ago to do network
as a service. So when a customer, particularly business customers, we have a portal called Flex, they can actually log in get networks on demand, so you can provision a network service from a data center to the cloud as your AWS in real time. Typically the approach would be to go to a vendor for your network orchestration layer.
We actually looked across what other big tech companies were doing and we saw what Uber we're doing in terms of their workflow orchestration to book a taxi we go, why don't we look at using that technology for the network and we did and that was actually a huge game changer for us. So I'm just looking wider than kind of a traditional talco. And then we've used that same workflow engine that we Uber open sourced it to
do mobile provisioning as well. So since the merger, we've actually been using the same workflow engine to do mobile workflow orchestration. So that's one example where you're leveraging technology not necessarily building all of the building blocks, which is pretty exciting. The team are excited about it, and they're all just across the road actually from where we're recording this podcast. Most of them they're all in New Zealand software developers and they're really bought into the vision of
two degrees. So I think that's a really important part is they're there to fight for fear. They know how important it is for two degrees to exist in this market, and they come to work every day to actually provide value to our customers, and that quite often can be you know, as I've been a previously a software developer, a network engineer, you can be stuck in the on your keyboard, not really understanding why you're doing it. So a big part of what we're doing is the cultural aspect.
It's actually eighty percent cultural. When you kind of break it down.
When you're doing such a big transformation integration project and you really hyper focused on trying to go from one thing to another, how do you balance just focusing the resource on kind of the transformation aspect versus bringing in new features upgrading, adding on to create new experiences. How do you decide where to go big and new and where to just stick to your bread and butter.
It's a great question, and it's a yeah. We do have this discussion a lot. Yeah, So we laser focused on some key deliverables and that integration and getting onto a single stack is such a game changer for us. So we're trying to be really really focused, but we see opportunity, we may take it. I guess one example we've had of that when we were building some of the new mobile capability in TAHI, we saw an opportunity to launch a new product out as kind of a
relatively simple product, a traveler some package. So when overseas people come in and they want to roam around New Zealand and have the great experience, they could easily just get on boarded and have a traveler prepay like a thirty day unlimited use, and we built that on the new system. In some respects, it was to both test the system, but also there was a gap in the market we saw and we can actually provide a really
great digital experience you can sign up with eSIM. So there was an example of where it was a win win. You could actually test your integration thesis while also building something something unique in the market. But we don't do that too often because we're trying to stay trying to stay focused on what we're trying to achieve. But yeah, it's a daily discussion. But I think that's the important part is you get the right people together to have
that intentional discussion around these decisions. I think that's the most important part. So where we may deviate, it's for a good reason, we're following some sort of value, or it's giving us ability to test and learn for where our future is. But yeah, there's many many aspects of that. You can't really break it down to a recipe or playbook. It could just be an opportunity that is in front of you. And I think, yes, again, it's really important you have a plan, but you have to be able
to adapt. And quite often over the last two years, it's been the actual constraints and the impediments or the oppertunities that actually have created our journey for us. So not being rigid on a plan, which I think is also learning that I've had in my history, and I think many other CEOs and many organizations where you do follow a plan rigidly, that's where things can go wrong.
You have to definitely adapt. You know, AI wasn't really a thing when we started off on the merger either, so that you know that all you know became blew up in October November twenty twenty two. Actually, can we leverage this technology to accelerate our integration strategy And we've
been looking at aspects in that area too. One of those has really been around software development, which I think is a massive game changer for anyone using AI to actually improve productivity and software development, which we did invest a lot of time in. That's you know one example, things change and you've got to adapt to it.
Yeah, that's a massive one. Andy Jesse, the CEO of Amazon, said a couple of weeks ago that using their in house sort of copilot for software development, I think it's called Q, they were able to save four and a
half thousand developer years just on upgrading Java apps. At Amazon they have tens of thousands of Java apps, so this is they said, it's going to save about a quarter of a billion dollars a year, not just in getting rid of developers, all the infrastructure the apps run more efficiently on their big server farms and ad as well, so the energy use is lower. So are you seeing that you say you're a sort of a software firm in many respects, are you seeing big efficiencies and software development?
Yet we are seeing some maybe not claim so many hours in productivity. Yeah, I think early on, yeah, we saw around twenty percent as we measured kind of by hours, and it's probably got to about ten percent now in terms of productivity. But it's hard to measure software development productivity. It's not an easy thing to do, and there are other constraints. So this is something I think about a lot. I've been on the journey for too long. And yeah, you've got to think about how you can figure your
teams and how you can figure your overarching system. An organization is actually a complex adaptive system and it can be immune and you can poke it and it can do things. But you've got to look at the end end value. So software development can improve by maybe even fifty percent, But if you've got constraints to go to market, or you've got constraints at the front end and your business casing, your finance modeling, how you actually do business cases you won't get the true end to end value.
So we do think a lot about that idea to value and measure that flow. That's a really key part of what we look at. So while you can optimize and that wedge in the middle for software development, you've got to optimize the full end to end, which I actually think is the whole thing around generative AI. There
is a huge software is completely revolutionized. I believe from what happened kind of in the nineteen seventies, software actually hasn't changed too much and in many respects you're still kind of was very traditional, but with the advent of large language models, it's turning more into a natural language. We have co pilot assistance next to our software developers. I really think the multi modal aspect is going to be huge. So you can actually have a voice assistant
while you're coding. That's going to be significant. So yeah, it's going to be a huge game changer, and there's a whole lot of I did not necessarily threaten my team, but I provocated that, you know, software development profession will be dead in five years, you know, and think about the evolution of it. I don't think that's actually a case, but it will definitely adapt and change. So the teams are you know, they're just getting ready, getting the skills
to adapt, but it is the next step. Can you have these we call them value streams. Can you have value stream agents that can connect the requirements to the teams that are doing the work. Can you then get strategy agents over the top. So if you can get a X improvement and software development, can you get the team working better at the value stream? That's another ten X. Can you get the strategy connected to execution better. That's you know, could be a thousand X improvement, which I
think would be massive. And then you've got the other case where AI is now I play with it all the time. You can build apps and software so quickly as an individual contributor, So you'll see the advent of these, you know, one person companies, which I think it's going to be great for startups in the future as well. Yeah, but our team's assisting them day to day to write better software, also for testing software, doing our reviews of
software as a complementary like a copilot. And that's something we've actually talked about a lot, is around the co pilot nature of AI. Actually being an assistant next to you, saying, with our care team actually having a co pilot next to our agents that are answering the phone to support our customers, they've actually got a co pilot that can coach them and work with them. That's our next step
in our journey, particularly in the care environment. So that's I think that's the approach I think we'll move forward on.
So it would be in quotes, listening to the conversation as you're having it and say, oh, I noticed that you've talked about this. You know, don't forget to tell the customer that this is an option that kind of thing.
Yes, definitely, Yeah, that's it.
Yeah, ye.
And we've already got a AI auto summary working with our care team, so is actually it's got the real time transcription happening. It actually summarizes the entire call for the agent. It used to be the human and the agent used to type it all out, so they get a little bit distracted while they're trying to talk to a customer while they're writing notes. So now our humans can actually have a real conversation, and then the AI is actually writing up the notes in real time, which
gives us really good data. It actually captures sentiment and intent. You can then put that into our feedback loop and actually improve the situation. We haven't quite got to the point where we actually have a really true co pilot, but that's our next step is to have a co pilot alongside our human agent to give that sort of coaching aspect and go okay, you should maybe think about talking about this. It actually a near term for us, it's not far away, which is pretty exciting.
It's pretty cool.
Yeah.
So just finally, Stephen, what's your advice to executives in New Zealand companies that may have just acquired or been a process of merging with another company that might be a retailer or a construction company. You've talked about the sort of the gradual approach versus big bang. It's nice to be on one system from an efficiency point of view, so there will be this temptation to get everything onto
one platform. We've seen where that has gone wrong. Some projects have gone off the rails have been paused because that's too complex. We've seen Jason Parris, you know, your competitor from One Ends, had been very upfront and honest about the pain that that organization went through with all these old systems hanging around. Even if you can bridge through APIs and that you can get data shared across those systems, how do you approach what sort of thinking
do you go through? Everything? Is every situation as different as you sort of said, But what are some of the fundamental questions you need to ask yourself as a leadership team when you're in this sort of merged position.
Now, yeah, yeah, you've got to have belief in a commitment to do it. So it's not easy. So integration in particular and legacy has to be a priority. It has to be the forefront of your strategy. So one of our strategies is that one company strategy. So integration is everything that's fundamental. I think sometimes leaders can get a little bit maybe just focus on the next thing they think it's done, But there's a long term commitment and you have to reinforce it, and I think leaders
go first in that respect. Build the belief, build the vision of the future. Don't inflict change on your own teams because part of it is also you're not just changing technology, you're changing how people work. There's so many different aspects to you making change happen. It's like it is eighty percent of human endeavor, So you really need
that cultural aspect. I think that's the critical part is having a team that are willing to have each other's back and actually, you know, it's okay to make mistakes. Actually having a psychologically safe environment where it's okay sometimes maybe not do the right thing, and amplify those signals quickly. I think that's a key part of it. But you
have to move fast. You've got there's a pace to it, and that can be you know, people are stronger on loss of version right, two times more impactful losing something than gaining something, So you do have to build that belief. You have to create a really safe environment and empower people to go on that journey. So I think you
focus on the outcomes and why incremental is critical. It's really easy to follow a plan and might be a big modernization plan, but it may feel painful at the beginning, actually questioning things, really do we really need to do this? You know, is that requirement really necessary. It's quite easy for people to add in all sorts of things as you go on the journey, so you need to be deleting things removing things. The best way to achieve it
actually not to build anything. So you have to be really really focused on it and really ask those tough questions, and yeah, focus on the value, focus on the customer. That's ultimately why we're doing what we're doing. Yeah, why we're doing what we're doing is get onto a single stack, be the customer experience for the future for our customers. So yeah, easy as that, so simple.
Right, Well, that's it for this week's episode. Thank you to Stephen Kajerv for joining us and to two Degrees for its ongoings. Abort the Business of Tech.
Subscribe to get all of the episodes of The Business Off Tech in your favorite podcast app or from iHeartRadio, where you can stream every episode. Show notes for the Business of Tech are in the Tech section on the Business Desk website.
Get in touch with your feedback, ideas, topics, and guest suggestions email me on banat Businessdesk dot co, dot z or find both of us on LinkedIn and sometimes on.
X and another dose of The Business Off Tech coming your way next Thursday.
Catch you then,