[Music] Welcome to another episode of Tech Unhinged, where tech gets human. I’m your host, Rabia Chave, and today we are diving into a topic that’s all about simplifying the complexities of tech and making it work for the people. Joining me today is Ian Tanet, the CIO of what three words. Before jumping into startup world, Ian spent over 15 years in leadership roles at big companies like Tesco, T-Mo, and the co-op group, working on everything from digital transformation to business operations. For the past six years at what three words, he has been leading teams across data, IT, automation, customer service, and sales systems. At heart, Ian’s a mathematician who loves solving tricky problems and making things work better for people. Ian, it’s awesome to have you here today. Thank you for being on the show. Delighted. Thank you very much. You’ve spent much of your career making tech easier for people. Was there a moment or project that really shaped that mindset for you? I think I would go back earlier than that. As you said in your introduction, I was drawn to mathematics very early on in my life. I like the neatness and the and the conciseness of being able to solve a problem, looking at something on a sheet of paper and having the whole problem in front of you and having to use the the tools at your disposal, the things you’ve learned, the techniques that you’ve developed to be able to solve that problem. I think earlier on in my my life and earlier on in my career, I really liked that there was one precise answer very often to those questions and I liked the neatness and completeness of that. But I think as I’ve had experience, as I’ve grown, I’ve really enjoyed learning much more about the the human side of the problem, the human solution, being able to solve something for someone else, to solve something for people, um to help them understand the solution or the patterns or the answer or the value within something and seeing them get benefit from it or their eyes light up as they do that. My first job after university, I work as an analyst but in a HR department and I looking at recruitment statistics, looking at performance of individuals, looking at things like sickness rates and attrition rates for people joining and leaving the business. It was the very very very early days of online job boards and people applying for jobs through the internet and looking at all of the numbers and statistics around that which was very interesting. But a big part of my job there was to find the best ways to communicate what the analytics team I was working in was seeing in the data to executives and directors within the people department. And typically a lot of those people weren’t like as data literate as comfortable with numbers or deep analysis or large data sets as the analytics team. And finding the right strategies, the right ways of explaining what we were seeing in those numbers, making those numbers talk to th those people and to see the opportunities. And it was actually really exciting to have to find a way to translate very like coming straight from university with very hard analytical academic approaches to talking about numbers and being very familiar with talking to everybody else who understood numbers the same way as I did and to the same level of academia to going to business place where actually I had to learn the language again and learn to talk to people from all sorts of different disciplines which was a really interesting challenge for me. when you got it right and you saw the aha moment drop in other people’s eyes, then wanting to work more closely with you or come back to you the next time they had a problem to solve. That was the thing that kind of really energized me in the business. No, I I think that’s that’s a great answer. So, you know, you you’ve worked in executive roles and then now you’re in a leadership role yourself. Ian, how do you you know, pick digital tools and how do you balance business schools with how people actually like to work on a day-to-day? I’d say it’s hard. It’s really hard. There is such a proliferation of tools out there and as soon as you start searching for one, the algorithms of the world will pick up on you and send you many, many, many others to look at and to be aware of. I think what’s important for me in picking a tool is to find one you can experiment with quickly. Uh where there’s very little barrier to seeing what it’s really like, getting your hands on it, being able to play with it, seeing what it can do, seeing what it can’t do, seeing how closely it fits with what you think you need or or what you think your business’s needs are going to be over time. I think it’s really important for companies ideally that you’re working with to have similar ideas to your own. So, we’ve made some mistakes in the past when picking a partner and then need growing quite fast, but the partner we picked wasn’t growing at the same speed of as us and they weren’t able to keep up with our demand. We really learned to understand when particularly working with newer businesses, are you going to be able to scale up? Have you got the appetite for fast growth? Have you got the resources for fast growth? Because if this thing goes big quickly, we need you to come on the journey with us. Definitely keep your eyes open. It’s really easy once you’ve picked a tool to focus on getting really good at that tool and your team getting really good at that tool and you become a little bit blinded to what else is new, what else is out there in the market, what other people are doing. Staying on those newsletters, staying like on those kind of information channels that you can find out what’s going on, going to conferences, talking to peers, talking to partners, finding out what they’re using to solve the same problem is a really good way of doing that. but also having the right account team at those partners, at those tools and those providers. Because having dialogue, having two-way conversation, knowing that if you’ve got a problem, knowing that if actually the tool could work so much better for you if they just did a small thing and knowing that they will listen to that and be able to prioritize or help. We’ve got some really good examples with um partners we work with, particularly where they’re they’re launching a new tool where they’re they’re so keen to hear our feedback. We’ve got a couple of AI tools that we’re triing right now and we’ll have shared Slack channels with people from the company that we’re using to come and say, “Oh, we’ve launched new feature. Can you try it?” And let us know straight away like, “Is it working for you? Is it good enough? Is it fast enough? Is it giving you good quality answers? Can you tell me your use cases so we can make sure it fits well for those?” Or we use a a low code automation tool called Zappia, which is super powerful. But they launched a new tool recently uh called tables where you can actually instead of having a spreadsheet or a Google sheet that it talks to, you can talk directly to a table on their cloud platform and that’s fantastic and we’ve had all sorts of like the latency on it is really down so it helps us a lot. Person who is actually the product owner for it when they saw that we were using it sent me an email and said is it working for you? I’d love to spend time with you and see how you’re using it and what you might want to use it for. It’s those types of things that you really value because you the human part of the technology is really important in building that trust and that relationship and going on that journey together. Yeah. No, that that makes a lot of sense. You know, it’s it’s it’s sort of one thing to familiarize yourself with a particular tool and then getting comfortable with it. But then given that you know how tech is everywhere, what it comes down to is usually that how is it turning into you know real business results for people you know usually at the top because that’s what they are looking at. And most of the times people would sort of take more time into you know experimenting a tool and then being able to tell about the kind of results that they can get out of it. So can you sort of share a time when you know a better data or a smarter um technology directly changed that business outcome or affected a business outcome? One of the things I say is when people come to my team and ask us to help on something is it’s great if they’ve been trying to do it themselves for a little while. It’s great if they’ve got their hands right in the process and they understand what’s good about it and what’s not good about it. So when we then try to come and help design a process that can change it for them or can implement something which is going to make their lives easier that they can tell us whether they think it’s going to work or not going to work. A good example would be so when um new people sign up to use what three words products, we try to understand who the best person from our business is that should be talking to that business that’s signed up. And we try to base that on things like the language we think they speak, the country we think they’re in, the kind of industry that we think that that company works in, the size of the company, and what type of uh use case they might have for our product. So, we try to figure all of that stuff out and then we have like a a way of say going down a table and saying, “Oh, who’s the right person to to flag on our internal messaging system, this business has signed up? We think you should take it.” And that process used to be a person’s job where every time a new API key or somebody filled in a form on our website or somebody created an account and indicated they were a business that would come onto a Slack channel and it would be somebody’s job to read it go and research online about who that company was find out what they could go and look across the teams we had in the business and said right that person needs to take it because we need somebody German speaking or this person is working on car companies or this person is working on career companies or with local government and that was quite laborious and quite hard work and as the business grew that became a almost a full-time job. And so we’ve been able to design using using matrices, using automation, using tools like Zappia, a way of establishing a matrix which the team the sales teams can control for themselves, but that which will automatically get the information we need and make the right choices and then notify somebody that they have that new lead or that new API key or if they’ve we’ve seen anybody from that business before across our digital estate, it will say, “Oh, this person will right person to talk to that business. they’ve talked to them before, they’ve spoken to three people, they’ve already got two API keys. Let’s just let them know that a new person from that business has has also now become known to us. That saves a huge amount of time. It gives us huge amount of background data. Uh and it’s also one which we’re improving even more with tools like uh linkup now which is an AI tool which can go off and actually find the information that we might need about that business from the internet for us as well in much more of a natural language query but fit it into our data structures and come back with you know only give me a response for which industry this company is in from our decided list that helps us determine something and that’s been a huge timesaver but also a really really scalable thing that will work. So what we find with those solutions is that they can uh using AI tools, using automation tools, they can scale with the business and it really doesn’t matter how many new people come into our kind of vision of awareness because we can just grow that process with those tools. So yeah, that’s something that’s gone from being a many hours a day job for one person to something which takes no time at all and we can trust in that process. Great. So Ian, we’ll sort of be putting more um light on, you know, um the kind of work at what three words you guys have been doing and you know while going through and researching up a bit on it, I I figured that it has been used in cars and deliveries, but you know, it has also been sort of used in critical areas like emergency response. So you know, for our listeners, how do you see AI making location tools even more life-saving and accessible in the future? A really good example very recently uh that we’re seeing where due to open APIs, due to MCPs with AI tools being openly accessible and being able to communicate with one another is that we’re able to join what three words up with other services. We had a really good example recently where a lady fell in the woods and her Apple Watch detected her fall and that then alerted the emergency services and they were able to combine that with using what three words to get the emergency services directly to that lady. But yeah, I think we will see things combine in ways that we we can’t predict yet. I think when we have seinal moments and technology shifts, everybody assumes it’s going to make the thing we already do better. Whereas actually what we see with like mobile data with everything becoming digital like you don’t end up with the existing companies today, but with a website from before the internet, you end up with different types of companies, different types of services. Nobody would have predicted Airbnb. Nobody would have predicted Uber as a result of us having data on our phones or having access to this type of collaboration. We’re not sure, but there are definitely areas where we think that location and AI will become very clever. Things like uh like driverless cars. So, if you’re in your Uber and you want to say to your Uber driver, I’m sorry that the address that you’ve got on the for the pin on the map isn’t quite right. Can you drop me another 100 meters? Like, you can say that to your driver. You can’t say that if you no longer have a driver in your taxi because your taxi is driving itself. That’s a really interesting thing with like Whimo or when parcels are no longer delivered by people in vans and they’re being delivered by drones. How do you ensure a very very accurate address for something like that? If you’re hanging a box outside your window on your apartment block on the seventh floor and that drone needs to just drop it in, how do you tell it precisely where it is? But more simple examples for just like customers and people just wanting to have better experiences with their own addresses. So things like if I’m talking to a customer service agent or if I’m using a chatbot on a website and I need some support that I can type in my delivery address and that can get automatically verified straight from that chatbot through to my through to my account. I don’t have to log into my account. I don’t have to, you know, worry about talking to somebody or if I am talking to somebody, maybe when they say, “What’s your new address?” they can have an AI voice detection on the call that transcribes my address and then has some validation of it. And so then instead of having to say, “Oh, I’m sorry. Can you repeat that zip code or repeat that postcode or can you spell that street for me?” That actually they don’t have to worry about that and all of that simplicity and all of that that confidence that you can give to a a customer by saying, “I’ve got your address here. It’s validated. I’m going to read it back to you.” and that to happen really really quickly. Or even something like um say you’ve just moved house and you want to say to your agent, your AI agent, please go and update my home delivery address for my online accounts right now. I don’t want to have to log into everyone and tell it I’ve moved address that for me. And it’s that type of thing where I think we’ll find location and AI really just speeding it up and making something which should be simple much more simple. We we are sort of already living in the future I believe because um the things that we once saw in the movies they’re already here and um the time is moving so fast the clock is ticking and I’m sure that most of these ideas that you just talked about they are you know already in place probably that you guys are working on or maybe you know people like you are already working on within the tech which also sort of leads me to my next question. We see that you know when working on complex systems it can sort of get overwhelming you know so how do you make sure you know that it doesn’t sort of become overwhelming for the end user at the end of the day right what are certain tricks that you have which help you ensuring that you know this particular um service or tool is more human focused rather than just rooted into um the tech side of it because you know um you have one idea in the mind But most of times it’s very disconnected with how it’s going to come across for the user and that is where a lot lots of changes go down. So how do you navigate through that? I would give a big shout out to the product team that I work with at what three words. They are masterful at this. Um particularly when thinking about user needs of different types of people using our app and using our services. They really focus hard on the accessibility of the app. had some wonderful comments just this week from somebody who’s blind who uses our app to help them navigate to many specific places and the fact that it works so well with like screen reading tools and things like that. So I think being aware of your all of your users, not just your average user is really important. Um the the philosophy that I try to carry here is something like build it for them, not for yourself. It’s very easy when you’re building a tool or building a process or a service to kind of get invested in what you want from it and how you want to build it and the thing that you want your name attached to. And actually that’s not the point at all. The point is to solve the problem for the user and to not let go of that. Um start with user needs. So really understand why somebody needs this thing at all. Why is there why if it doesn’t exist today, why should it exist? What problem is it solving? How do we what need is currently being underserved that could be met with digital tools or digital technologies? I would say don’t fall into the trap of saying oh we’ve seen a problem like this before so the solution is probably similar to how we solved it before where you can where you’ve got time where you’ve got the resources to do so try and start again. There may be things particularly about this problem that means it needs to be solved differently or there may just now be better solutions out there since you built it last and trying to use new methods better methods. We had a bit of a clean out in my team because our kind of automation suites were getting very very cluttered and things that we didn’t really use or things that were only used a little bit were still in there. And I the advice I gave to the team was be brutal. Delete first or delete more than you would think or don’t do this sensitively because actually if somebody comes to us and said, “Oh, you deleted that thing and actually I still need it.” We would probably build it again better than we built it 2 years ago or 3 years ago or four years ago. And I think remembering to kind of like keep your solutions modern, keep them up to date can be really important. Um, but also partner with your customers, partner with your users, involve them right throughout the process. Um, there’s a group of people who I used to work with who previously were at the UK government digital services and they had a big poster on the wall that said, “Show the thing.” If I’m giving my customer or my user an update on how my solution is going that I’m building for them, I don’t give them a status report on it. I don’t show them a video of a demo. I don’t like, you know, it’s not it’s not a representation of the thing. Show them the actual thing. Show them the beta. Show them the prototype. Show them the demo that like let them interact with it. Let them see how you’ve interpreted their needs and you’re starting to solve it. And they’ll give you so much better feedback than you bringing up a status report for the project and saying everything’s amber for an August launch doesn’t help anybody get you to where you need to be. I think you’ve you’ve beautifully put it all together because I mean if if you’re actually starting with the problem and not the technology, you ultimately get there, you know, where you show the thing rather than just sort of talk all about it and not really make your user be able to engage with that thing. So even some companies dream of building one big AI that does it all, which we we are facing a lot these days, right? But others are moving towards um using lots of small specialized AI agents or different tools that work together and probably you know focused on one or other task. Have you seen this sort of many agents approach in action under yourself or at companies who are working in parallel and if yes then how would you sort of compare it when it comes to speed and flexibility? Yeah, I think it’s something like many organizations that we’re way more on the experimentation end of rather than implementation right now. I went to the the Salesforce World Tour conference which was rebadged as agent force this year uh where they’re heavily leaning into the agents within side sales platforms. And for companies that use Salesforce in a very very large way and use all of their different products and services, I can imagine that having those agents joined up within one platform and able to talk to each other and access the same data could be hugely powerful. We’re at more of an experimentation stage and we’re seeing really good kind of test results in terms of what might be possible in this arena, but nothing that I would strongly commit to as saying I think will work tomorrow or in a week’s time. I think that the problem is edge cases. The problem is about I think the the specificity of individual companies needs and trying to build an agent that works by referencing existing data or kind of best practice data. You have to tune it very well to your individual business’s needs to get it to work in all cases. agent to agent to agent can be great but yeah I think probably also knowing where this question is too hard to answer or I’m not sure enough about this answer and I probably before passing it to another agent that’s going to onwardly interpret it and probably compound the error you know it’s you don’t want to you know if you whisper in somebody’s ear and they whisper in somebody else’s ear and they whisper in somebody else’s ear the chain of communication can get lost and the person at the end hears something very different to what the person at the the start thinks they said and I think when you’re constantly ly interpreting through AI agents, you’re at risk sometimes of bot at the very end not understanding the initial principles, the initial core message. And certainly with very large businesses and we’ve seen a lot of CEOs putting messages out there about encouraging their teams to experiment about wanting to be on the front foot with AI about taking advantages of the efficiencies available and where big companies are building many of those things and perhaps they’re not necessarily aware that an AI agent in one department has taken over a process and an AI agent in another department has taken over another process and now they’re talking to each other whereas it used to be two people. Are those things being built and understood in precisely the same ways to be able to manage that handoff in the ways that the people who were doing that before were and like anything with AI today is the worst it will ever be. It will only be this good or better as we go into the future. So yeah, it’s a great space to experiment in a great space to kind of watch what’s happening and what’s developing and to take advantage of that. Um but I think there are a few things that I’m I’m cautious of right now. Yeah. And since you know we we are at that point or stage where we are seeing that AI can automate a lot or have automated many things at this point where do you draw the line Ian for that you know as as a leader yourself what is your thought process on that I think it’s still really important for the people who are building or running these agents to to still be fully embedded in the processes themselves to ensure what it is they’re building and what they’re designing for. I think there are also efficiency points within this where if the agent has to work really really really hard on an edge case or on something which is just difficult. Where does it actually stop saving you money or saving you time versus flagging to a person who says yes or no from their experience or says you know red or blue or whatever the choice is or actually this is a special case for a highv value customer where I no longer want the agent to take that decision for me because it’s my accountability and I don’t want my accountability to fall on a decision that an agent has made for me and understanding those thresholds of value and kind of like your risk appetite. Um the how likely is it to go wrong? What’s the worst thing that could happen if it did go wrong? And how willing am I to accept that risk versus it defaulting to to something which is more human-managed? Because humans might make the same mistakes, but at least then you’re doing it outwardly, knowingly. Obviously, I think there was an airline in Canada that had a an AI chatbot on his website and a customer asked about the refund policy and the chatbot just made up a re refund policy. The airline said, “I’m sorry, I don’t want to honor that because that’s not our policy.” and they had to honor it because you know if you’re putting somebody in charge of giving communications to your customer and you’re saying that that is autonomous and correct then you have to live by what that thing does. You know it’s still there’s a lot of debate going around the idea of keeping humans in the loop right and there there has been fear which is ongoing that there might not be a human in the loop ultimately because with every passing day AI is becoming more smarter. It is now even using words that could sort of be on you know an empathetic level with you would be trying to buckle you up in saying things which you’re not asking but then there would be things you know that it would just be saying with every new version of chat GBT there are new wins that you could see and you know how people rely more on AI chatbot therapist than the real therapist in person you might have heard you know the news you know what do you think about that have you yourself used chip in a way where you would just um give more credits to AI in comparison to a human being. Different AI tools are good and intentionally good at different things. We see like the the measures and scorecards when new models are released and we see that they score higher and lower for different things. I think ensuring as a user that you have a particularly demanding use case or kind of requirement of an AI tool that you ensure that it is a good tool for that type of query. But I also think that in terms of keeping the human in the loop, as I said earlier, there are lots of individual company peculiars that you can train an AI agent for, but that it’s not going to know straight away. It’s not going to learn or have a good answer from its training data because your the internal workings of your business is isn’t going to be inside its training data. And so particularly as we start to implement these things for the first time, making sure it is kind of like a shared learning that we’re learning how the agent is trying to work, but we’re we find the right ways to constantly input from ourselves and to refine and to iterate. Yeah. And also as a having those thresholds, having those limits where if the impact is bigger than this, then ask me first or if it’s bigger than this, then you know send to a person before proceeding and provided you know there might be time considerations. One of the great things about agents is they can be up all hours of the day and night. you can be a 24/7 kind of global service operation because opt will always be on and be able to work for you. But sometimes the um the advantage of speed isn’t isn’t worth the risk. Yeah. So Ian, having worked at different companies and then now being in a startup, how did your perspective shift or change? Some of the big changes for me were around comfort level. I’ve spoken about it already, but comfort levels with risk, comfort levels with speed and comfort levels with taking decisions with not enough information and personal accountability. So before I started working in startups, I was very used to having to go through a chain of approval before I could do something even if I was reasonably sure. And that was great for reassurance and great for certainty, but meant that things moved slowly and sometimes the opportunity that you wanted to seek had gone by the time that you could get yes. you know, it’s like, I really want to do this thing straight away. It’s like, great. Well, can you fill in these 30 slides of PowerPoint and we’ll get you a a spot on the agenda for 3 weeks on Thursday. And then they might have some more questions. You might have to come back the following week. And it’s like, ideally, no, because we’re not going to get the thing done or we’re going to lose the momentum with the opportunity that we’re seeing from the data that we’ve got. Whereas coming into a startup, it’s the other way around. I remember very early on I was looking after some of our B2B advertising what three words and a campaign that we were doing was performing particularly well and I wanted to put some more money into it and I sat about two desks away from the CEO and I said so what’s the process for me asking for more money for this bud like more more budget to get into this campaign cuz I think it’s working really well and he said how much more do you want and I told him the number and he said great go for it took me a few moments to realize that that was the approval process and that was okay and I could now go and type in a new number for our budget and it would just be all right that there wasn’t some kind of purchase order. There wasn’t some kind of procurement process or exchange of contracts or emails or things I received from an SAP system that told me it was then okay to spend that money. But also I think yeah saying what’s the worst thing that can happen and saying I’m more comfortable with that thing happening and knowing that you can correct it. Um having much more human relationships with the people you work with and then being understanding of mistakes or new learnings or errors. You know, if I have a process which is pushing out a number which is telling us about how something performed yesterday and actually I build it really quickly because we want it really fast because it’s urgent and the next day it pushes out a number and it’s wrong. I can put a message out and go sorry that number’s wrong. and we’ll send you the real number in an hour rather than in traditional corporate days where I might send out that would have been an email to 300 people and then you have to you know reply all reply all reply all you know that looks a bit strange for the figure we’re expecting and it just spirals into a thing that’s not not great at all. So yeah, I think the the single point of accountability is great. It’s a double-edged sword because instead of having that backup, it’s on your shoulders. But if you thrive in that environment where you can really get things done because you know you can make the decisions yourself, then it allows you to be much more agile and to move much faster, which is obviously something that we we see in a lot of startups. Yeah, I I think startups have have this tendency to put you a bit out of your comfort zone and because they run at a rather faster pace. So you know, on a day-to-day there’s something and then there’s more room. there’s there’s a lot of individual growth you know but then in the corporates you rather stay within the comfort zones because everything sort of comes easy or at times not easy but then there is a process built up for everything but in startups you’re rather building the process yourself you know another thing that I really really enjoyed about startups I mean even in my my corporate life the thing that frustrated me the most was when I was being asked to make a change or to build some some tool that would try and improve something but could only improve the bit that I was allowed to impact that was only this part of the business and it’s like well you’re not in that team you’re not in that department so don’t do anything that changes that but then you would kind of map the process or map the system and you’d go well actually we can do some stuff here but the bigger benefits that we could get at are actually in this team oh well that team aren’t bothered about changing that thing they’re happy with it or they haven’t got the time to engage with us at the moment or they haven’t got the skills in their team to change that thing and I’d be quite frustrated that I couldn’t go and change the thing that I thought was going to make the biggest difference and I think many fewer barriers in startups and the ability to easily go and engage with your colleagues from different departments and to get things done cross functionally in a much faster way and being given permission to do that for people to say it’s okay that people are are typically less hierarchical and less siloed is really helpful and just yeah collaborating and doing stuff together yeah but then I also believe that there is always you know another side to the picture because coming from setups where everything is set you know from the go here in startups you have to at times do things from the scratch that you might have been doing in the first year of your career Right? So that at times gets frustrating because now you’re building the process. So how did you navigate through that in your time? Remember what you’re there to do. Remember that actually startups so often are created to solve a pro like solve a big problem to say this thing isn’t done well enough. We’re going to be a successful startup. If we’re going to be one of those those companies with a massive growth trajectory, then we need to solve this problem really well for customers. And actually that’s what we’re here to enable every day. and my job title or how big my team is or how big my budget is does not take away from what needs to be done such that that thing can happen, that problem can be solved. And if today that means carrying the water up the stairs for the water filter because it’s been delivered to the wrong floor, or if that means, you know, diving in and fixing a simple piece of code because per the only other person who does that is on holiday and normally in a bigger company, you might have a team of many people with the same skills, awareness to be able to do that. or if that’s, you know, sitting with somebody who’s new in the business and taking them through slowly and inducting them into what’s going on and how things are going and what you’re trying to do. I think all of that is actually really important. It keeps you grounded. It can keep you kind of humble and remind you what you’re there to achieve every day. And I think again in terms of people’s willingness to work together and to collaborate and just kind of muck in and just get it all done together, that can be really helpful because then when a different problem comes up, it’s much more about who are the people we have, what are the skills they have, how can we deploy those skills rather than which roles have we got in which teams which just become too fixed in your ways of thinking about how can you quickly do something to solve problem or to meet an opportunity. I get it. you know reminds me of a lot of things which are applied there because some sometimes the line is straight but they would just make it zigzaggy for no reason and hence make it more frustrating so you know you’re just stuck in the process but then you know I sort of ask you these questions to establish the context where you know we can dive into certain failures or lag behinds we we talk a lot about you know what’s what’s done and what’s going to be done but when a company is growing fast it’s easy for tech systems to lag behind or break under pressure, right? So, what’s what’s been your playbook for scaling systems without, you know, losing control or how do you deal with those kind of failures? I think it’s really important to understand your process and understand your business case and the sensitivities within that. So when you’re trying to build something, say, okay, if actually the volume of activity that we think is going to come through here 10% more or 100% more or a,000% more, how much does that change what we would do? And how quickly do we think that change could happen? And how long would it take to change what we’re doing? And also prioritizing is kind of like, I know a bunch of platforms are burning, but which one’s really burning first? Which one’s going to cause the most damage if it goes? And trying to understand that kind of dependency map. And quite often if process could be manual to start with and then it might move onto a spreadsheet and then we might kind of add some things that talk to the spreadsheet to automate things out of the spreadsheet and then we might move it into a database or or whatever it is how that process grows. Sometimes you can make those things last a lot longer than they would naturally do by just applying a little bit more thinking, a little bit more time and effort to them and understanding the limitations of not what something was designed to do but what something can do and finding ways of making things stretch beyond their originally designed capacity while you fix the really important thing. So sometimes sometimes starting out as a manual process is actually the best way because you learn what really matters about it and finding ways of when you turn something new on doing it slowly and then grow start small and iterate fast uh is kind of the mantra that you would go for but also partnering with other people within the business. So what is our cut over point? If it gets bigger than this and we can see something going past this line at what point can’t we do it the way we’re doing it today? and being eyes open. Set up the monitoring, set up the reporting. That means that you’re really aware of when that line’s coming up and how fast it’s coming up, you know, so that if I really need to hand this thing over to a much bigger technical partner, do I need to have a third party that can do this because our internal skills aren’t good enough, do I need to work with our our tech teams and our data teams to spin up a database or whatever it is, being eyes open about that from the start. So rather than reacting or somebody else telling you something’s getting slow or something’s getting problematic, just have something that checks every day, every 2 days, once a week, depending on the growth rate, it just gives you a number and says, you know, you’re currently 80% of the way to the number that you think is the problem and you know it’s changed by 5% over the last week. So if it keeps growing at this rate, you’re going to be in trouble in a month, so you better go and talk to somebody versus a you’re at 80% but actually you’ve been at 80% for the last 6 months, so it doesn’t look like it’s growing that fast. You’re okay. you know, don’t don’t always just look at where it is, but look at the rate of change as well. Ian, you know, you’re very active on on your LinkedIn as well. So, I you recently shared a post about your team building and you know, um knocking down wooden block bridges at work and you know the the collaboration with the Lego. So, how important do you think moments of play and creativity are for you know building strong connected teams especially in fastmoving technology environments? Relationships are so important in terms of having an understanding of people that you work with, who they are, what’s important to them, um, both inside of work, outside of work, what’s going on in their lives, whenever you need to talk to them with it, whenever you need to communicate with them. Being sensitive, being aware, being kind of just being correct in the way that you phrase things, you can really get so much more from those relationships if you believe that you can. Don’t don’t pay lip service to it. Don’t do it skinde because if you do it as a tick in a box exercise, people are smart. They know you’re doing it as a tick in the box exercise. And so then when you go and engage with them when you do really need something, they won’t feel like you’ve invested in them when you need something from them. And I think building up that I think I’ve read something in the past calling it relationship capital, but kind of building up that kind of all of the time that you’ve been great and engaging and thoughtful and really deliberate with the way that you’ve engaged with somebody. Then when you send them a message that says we’ve got to do this now, they know that something is important and they will help you and engage with you. If you say we need to do this now, or just do the short version of the message all the time, it’s kind of like the boy who cried wolf. They don’t really know when it’s really important and when it’s super important versus when you’re just being somebody who always asks for something in a short kind of concise way. Yeah, we’ve talked about work life balance before um at what three words and how like since particularly since COVID people’s return to work, people’s kind of hybrid work style, people people’s lives and everything that’s outside of what we would traditionally consider their work life is crossing over with their home life. And I think finding a way of getting your job done just through the week rather than segmenting this is my hours in the office, this is my hours in the office. And that’s great and that works for some people, but for a lot of people being allowed to be flexible in how and where and when they get their work done, which can work really well for digital and kind of companies that are based in many countries can be way more powerful and respecting that, being aware of that, being understanding what works for people. People feel valued and feel trusted and feel invested in. They will pay you back for that. I think that yeah, it kind of pays dividends over the long long term. So Ian, when we look ahead, what upcoming trends in technology are you sort of most excited about, especially ones that will help, you know, businesses cut down certain complexity and make life more easier for the users, for us? Yeah, I think certainly from smaller technology businesses is just given the scaling opportunities through AI tools is that you’ll be able to compete as though you are a much much bigger company. You don’t have to scale internal resource and people with with the amount that you’re trying to do, the amount you’re trying to process. We’re seeing with bunch of AI startups now, these are 10 20 person companies and that they’re able to move fast and do things very very quickly because of the tools that they now have available to them. So I think that’s super exciting. I think for what three words the industries we work in, I think the partnership of the mobile phone and the car is really interesting. So very very few people get into their car now and use their car’s satellite navigation system. They will use their phone plugged into their car or they will talk to their their phone to unlock their car or ask their phone where did I park my car. And I think for or um again with deliveries and courier systems, all of our couriers now have a handheld device, but that the parcels are still physical and the delivery vehicles are still physical, but the digital technologies that join those up and the ways in which we’re we’re seeing those kind of make everything more efficient. Not just we did a um we did an experiment with um DPD and Mercedes-Benz in uh Germany a few years ago where we were looking at the the time that could be saved by using what three words addresses for deliveries as opposed to typical traditional street addresses. And the time that was saved was was really strong. I think it was more than 15% in a day. But also what the leadership of that business was saying to us was this allows us to invest back in drivers. This allows us to say to our driver, you’re going to genuinely have a lunch break today, you know, and actually like because the deliveries we give you are going to go fine and we’re not going to give you too many and you’re going to be able to be efficient for us and profitable for us in terms of the number of packages you can get delivered today. You’re not going to leave next month because you hate your job because you’re too busy all of the time. And actually having experienced drivers that know their routes, know their deliveries, look after the packages and aren’t kind of constantly churning out business, having to hire and train new people is another huge benefit to large companies where those margins on deliveries are so razor thin. And so yeah, finding the right ways to kind of embrace that technology to give value back to people rather than just value to large businesses, I think could be really exciting. Yeah, know I think that that’s one of the major majorly important points today um in the fastmoving tech world that we are living in. So, uh moving towards the end of our conversation and now with one of my most favorite questions that I love asking my guests. So, Ian, looking back, what’s one hard lesson you’ve learned when leading big tech or business changes that others could learn from? I think one of the hardest things for me is becoming a more senior leader in businesses is you don’t have to be an expert in the thing your team is doing for you to be able to help them and to lead them. Quite often earlier on in your career, you’re doing a job and then you become a more senior person in that job and a more senior person in that job and then you typically have the people reporting into you who are just doing the job that you used to do and you’re like great I can be brilliant for these people by showing them how brilliant I was at the job that I used to do and that them then learning that think you’re great as well because they’re they’re going oh great I can see where I’m going I’ve got this path and I think um in startups you have to cover a much broader area but just in general as you become a more senior leader finding the coaching strategies ies, finding the mentoring strategies, finding the listening strategies to be able to really help guide people and help help them to help themselves in overcoming hurdles. Um, finding getting help for them that isn’t you telling them the answer. And being patient with yourself in that and learning to trust others delivery and to learning to kind of scale your own delivery through others um is a really powerful thing. believing that as a leader, your teams are going to do way more if you can get them doing things by themselves and without you keeping an eye on them rather than you having to be in their jobs and doing their jobs with them. And it’s uncomfortable. And again, it goes back to relationships and building that trust, building that understanding, knowing who your people are and knowing how best to help them and what enables them, what gives them that spark. Um, like for me, like that problem solving thing, what what’s the work that they need to be doing? and what’s the topics they need to be working on that’s really going to bring their enthusiasm right up and give them that drive to deliver and how do I get out of their way so they can get it done. All of the words made a lot of sense and that is sort of a great advice for our listeners as well and especially the people you know who are sort of navigating their way in startups and at leadership positions. Well Ian thank you so much for you know bringing such a down to earth and thoughtful perspective on making tech work for people. Thank you so much for having me. [Music]