Measuring What Matters: How to Drive Real AI Transformation — Conny Ploth, VP Global AI Transformation

0:00So measuring, I think, is super important. And I think we need to bear this always in our mind. What we are not able to measure, we shouldn't even start doing it, right? Because in the end, you need to measure impact, you need to measure your return on investment. So if you don't even have a baseline, if you haven't measured your baseline, if you don't even know where you're starting from, you better get started with measuring your baseline before you jump on AI.

0:22Welcome to Data and AI Mastery, the podcast where we bring you cutting-edge insights, practical advice, and inspiring stories from the leaders shaping the future of data and AI across the globe. I'm your host, Raoul Gabriel-Urmer, founder of Cambridge Spark, the leader in transformational data and AI upskilling, career development, and progression. In each episode, I will be diving into real-world case studies of companies harnessing the power of AI to drive innovation, reduce costs, and create new business opportunities. So whether Whether you are an aspiring data scientist, AI engineer or seasoned executive, this show is designed to give you the tools and knowledge to stay ahead in a world where data is transforming every aspect of business.

1:07Stay ahead, stay inspired, stay masterful. Welcome to Data & AI Mastery.

1:16Hey, Connie, how are you? I'm very good, Raoul. Thank you for having me here today. Hey, it's a real pleasure. I can't wait for our conversation. You have such a great experience in the financial sector, VP of Global AI Transformation. So it's going to be fascinating.

1:35So, Connie, I mean, to kick us off, I'd love to hear your story, right? What was your path to the data and AI leadership, you know, journey and where you are today? Absolutely. It was definitely not the straight path. So it took quite a few turns. So I think my main key drivers have always been curiosity, you know, to explore new things and being up for new challenges, but also at the same time to create a positive impact. I don't think I could ever do something that doesn't have like a bigger meaning behind. So it has really taken me through, you know, the path has started actually in microfinance.

2:10I was setting up some microfinance banks, you know, in Latin America and Africa. And from there, I moved on to consulting. And I've kind of ended up in financial sector consulting. I had done some different corporate roles and in the financial sector for the past 10 years, different roles really from in-house consulting, digital transformation, operations, but always kind of at the edge of trying something new, you know, and bringing value to the teams and to the people. So I think that's kind of my biggest driver. and last year um when i was in a in a role that was more related to esg we started really experimenting a lot with ai and how to personalize our experience you know our clients experience more with ai and this is how it kind of got stuck in there and i think it's not because of the hype it was really because it's a new technology or it's a new you know it's a new world that's opening up which i think sparked my curiosity and yeah i am right now yeah oh fantastic uh well what a rich experience and it's great to hear curiosity as a common thread.

3:10Yeah.

3:13Well, so I'd love to get your perspective, you know, there's a lot to talk about AI, transformation, you're clearly the expert. So to kick us out, how do you define AI transformation for the audience? Yeah, so it's a really big word, right? Transformation. It means actually transforming doesn't mean just, you know, putting some AI layer on top of something or it's just having another IT project running. I think it's really a change of a mindset, a cultural change that needs to happen. I think we heard it as well this morning, right, in the conversation. So I think this cultural change needs to happen with the people.

3:43It starts with the people. And I think if the organization, especially, you know, kind of more legacy organizations, they want to drive this change, they need to bring the people on board and they need to make sure that AI becomes an AI first mindset so that people, whatever they do, whatever kind of process, whatever customer interaction that they face. It's like, how can we actually drive and enhance this and make it better and be actually, you know, better at what we're doing with AI? So I think it's really this change of mindset of leaving behind, you know, the legacy systems and processes and really thinking, how can we reinvent ourselves with AI?

4:19And so I think transformation really happens in all levels of the organization, but especially it starts with the people. So the people need to be on board. They need to drive this. Great. I like what you said about reinventing yourself with AI and it's all about the people. So what's in it for organizations? Like, why should they care? Why should they think about, hey, like, you really need to think about your AI transformation? Yeah. So it very much depends on the sector, but probably there are some common themes that I see. So when it comes to processes, I think the very common theme is always process automation, right?

4:49So automating processes, you know, getting rid of the spreadsheets. There are entire organizations that are built on spreadsheets. There's no PowerPoint. points. So I think obviously, you know, that's the easy one. However, you need to connect as well, the different dots and different initiatives in order to create an impact. So I think automation is always a big topic, creating efficiencies through automation. But also, actually, how can you increment the revenue, you know, of your business and of the organization? How can you serve your customers better? So it doesn't matter if it's like a B2B or B2C organization, how can you serve your customers better?

5:22How can you make their interaction with you know more of a better experience and also what kind of products and data and insights can you actually offer your customers so how can you hyper personalize this offer so i think for me there are three common themes and probably depending on the industry and probably as well some risk and compliance perspective how can you make it more robust right but i think the three common themes that i see really is creating efficiencies through automations but also So incrementing basically your business growth and generating more revenues, unlocking revenue streams that probably were not possible before.

5:58And then this comes very much linked to how do you improve your customer interactions? What kind of service do you deliver and a hyper-personalized offer? Amazing. So if I play this back, organizations should care because there's a few levels that are important. There's cost efficiencies. There is obviously revenue growth. Every business out there wants to, like, you know, drive top line. but it's also something around customer experience and innovation that you mentioned, right? Absolutely. And possibly risk management. That's super interesting. Yeah, it makes sense. Every organization should care about all of that.

6:27Yeah. And well, in this context, what do you think is important to measure, right? Let's say an organization and you're investing in this AI transformation. What are the key things that would be worth thinking about? Yeah. So measuring, I think, is super important. And I think we need to bear this always in our mind. What we are not able to measure, we shouldn't even start doing it, right? Because in the end, you need to measure impact. You need to measure your return on investment. So if you don't even have a baseline, if you haven't measured your baseline, if you don't even know where you're starting from, better get started with measuring your baseline before you jump on AI.

6:58So I think what are the KPIs that you're measuring in terms of process automation is obviously the rate of automation, the number of processes that you have automated, the time that you have saved, resources that you have saved. So basically everything which is efficiency related. So I think this is and it comes in terms of, you know, savings, but also I think when it comes to revenue growth, it is are you expanding properly to new markets? Are you incrementing your customer base? Are you incrementing your conversion rates in your funnel? So I think that's also a very interesting measure, especially when it comes to hyper personalization of offers.

7:36So the more insights you have about your customers, the more bespoke the offering and the value proposition is that you can make to your customers. Do you have higher conversion rates in your funnel? Probably yes, because customers, you know, they are more drawn into a properly personalized offering. But as well, the way how you interact with them. So if you have talked a lot about contact center, you know, automation, there are a lot of new technologies right now that help contact centers. So if you get a very good customer interactions through, you know, VoiceBots, for instance, and you have a 24-7 self-service available, is your NPS getting better?

8:13It's like, is your customer service, you know, and your customer satisfaction higher? So it really very much depends, you know, on what kind of initiatives and what importance you give. But you need to be able to measure it. The baseline and the impact that you're producing, the ultimate impact. And in between is the adoption rate. So I think if you don't adopt it, you will not get to the ultimate impact. You need to know always what impact you're measuring, but as well in the middle is the adoption that you do measure. So how well is actually your organization adopting AI? And this starts with the individual productivity, individual use of those tools, but also in terms of teams and the overall organization, basically what's the adoption rate.

8:53And I think this is a very important KPI to have in your mind. Super interesting. So really important to keep track of operational metrics, marketing and sales funnel, and of course, like on the people adoption side. So it sounds like you need some real science to make the most out of the transformation. So does that make the case for you should invest in data infrastructure even more so? Or what's the right balance, you think? Data is key. Data is key? Data is key. If you have messy data, you'll get a messy output or probably a suboptimal output, right? Data is always a challenge in organizations, especially if they sit in silos, if they sit in unaccessible databases.

9:34Not so much the fact if they are structured or unstructured, but if they sit on a spreadsheet or someone's local drive, then probably they are hard to access, right? Or worse, even if they sit in someone's mind and they are not even sitting on a spreadsheet. So I think data is always the biggest challenge. You always get the best results if you have your data super well organized and accessible data links, right? Now, organizations, especially, you know, the large and complex ones, they usually have their data sitting in various different sources and they are not always accessible. You should definitely invest in data.

10:05Absolutely. I think this is a no-brainer. You can do things. Don't get me wrong. You can already do small things with the data that you have accessible or even on your SharePoint, you know. But look, it won't move the needle. You really need to invest in a good data infrastructure in order to have access to those data. Great. And you need to as well invest in guardrails. So only because the data are accessible doesn't mean you can always use them for every purpose. So I think especially in highly regulated industries, it's super important, you know, that there is the right level of data protection in place and that very sensible customer data, personal data are not used for the wrong purposes.

10:46Right. And especially when we talk about open language models, you know, that are open to the public. So I think it's very important to make sure that they are safeguarded. They are, you know, they have the right protection layer, but not in a way that they are totally unusable. It comes to this extreme as well, you know, where they're totally cut off. So it's striking the right balance, actually. Yeah, super interesting. It sounds like you can do with, you know, sort of suboptimal data assets. But if you want to get the flywheel, you know, you're going to have to increase the quality because that really gives you better outcomes.

11:20But also I like what you say around you also need to be prepared to invest in the governance, the guardrails of it, depending on your sector. Absolutely. Yeah. Super interesting. That leads me nicely to the next question, which is what are some of the misconceptions maybe that you're hearing about, you know, AI transformation that, you know, you want to debunk a little bit? Yeah. So probably there are three years out. There are a lot. There are plenty, right? Probably top three. Treating AI as just another tech project. Yeah. It's really, it needs to be embedded in your organization in order to produce the desired impact.

11:56And this really starts again with the people and with the change management. Like, don't expect just because you give them a tool, they're going to use it. Even though they might use it, you know, for their personal life. If it's not really deeply rooted and embedded in your systems and super well integrated in your processes, otherwise it just becomes another tool that's isolated, swivel chair approach. It's like just another one. So it really needs to be, it's not just a tech project. It's really an entire shift of mindset, right? It starts with the people, but they need to have the access to the right tools and they need to be embedded in your system.

12:29So it's not an isolated project. Okay. Then in order to get the people on board, you cannot just expect them, as I said, just to use it. You need to make sure that the adoption rate is growing exponentially because you need them to use this and to produce the outcome. So how do you do this? There are several approaches. I think what's always kind of working is very gamified approach, you know, where they just experiment, they play around, they can use challenges, solve challenges, do hackathons, run events. So something, give them something to be excited about, you know, and something they can play around and experiment with.

13:05So I think this kind of playful approach usually works very well. And then for me, it's another thing, just expecting AI to do the magic. so it's like you know we are sitting here in a very beautiful setting but just imagine you know it was just a facade that was like painted you know very beautiful but then you enter actually this building and you see that the pipes are broken heating is not working and everything like if you just put an ai layer on top of something that is broken it's still broken it's just probably a bit more quicker here now and a bit more shiny from the outside but inside it's still broken so you need to make sure to fix your processes you know and you have everything else in place as well because then ai can actually do some more magic and not just it's not just a shiny facade from the outside great great i love it so you really have to think about the technology the people and the processes and i guess we live in a world where there's a lot of excitement about ai so it's important to manage expectations and do the hard work if you want to get the best of it i hope you're enjoying today's conversation if you're finding the insights useful please do take Take a moment to subscribe to the Data and AI Mastery podcast and leave us a review on Apple Podcasts, Spotify or YouTube.

14:17Every new follow helps us reach more people and shed incredible work being done by today's Data and AI leader. All right, let's go back to the episode. Here's a scenario, right? Like you're, you know, an organization, a large organization, and you're thinking about your AI transformation. Can you walk us through how would you go about, you know, helping set the vision for AI and, you know, the North Star, right? Because you mentioned we need to get people on the journey. So how do you achieve that? You know, where do you start? I think it needs to align with your ultimate business goals. So what do you want to achieve as a business?

14:55So really do you want to serve your customers better? You want to achieve, I don't know, high turnover rates. So I think it needs to align, first of all, with your business. It cannot just be another thing that's isolated from your business goals. It must really fit very well into your overall business strategy. and then what's super important you need to buy in from your executive from the c-suite basically and why is it so important because it all starts with the tone from the top as we call it if you don't have buy-in from the c-suite then even though you know people will do some bottom-up work but they will still stuck somewhere in the middle so it will never produce actually the real outcome if you have got sponsorship from the top you'll have sponsorship in terms of budget in terms of commitment, in terms of resources, in terms of priorities.

15:39You get a lot more things done if you've got this, you know, high level buy-in. How do you achieve sponsorship from the top? I think they need to be very clear about the results and the impact it can produce. And if the philosophy is really of using AI to transform the organization, to transform the business and becoming AI first and embed this in everything we do, then this needs to be properly communicated. it okay and it needs to be communicated from you know our like the ai team that we are doing all this work to our top level so that they understand actually you know what are the benefits what's in for the organization how can we benefit from it and then it needs to be communicated to the rest of the organization so that everybody is on board so i think communication is absolutely key and buy in is absolutely key great so you know let's say that we we've got a way to make sure the you know the execs are brought in, right?

16:33And they want to make things happen. Is there like a process or a series of workshops that help define the vision and the North Star that you recommend to go through? I think every business line basically needs to have a very clear view on where AI can produce its results, right? And I think it's kind of a three-dimensional puzzle. It all needs to link in very well in order to have a global picture. So I said before, right? So AI can give you very good results if you have a very clear vision of what will happen with your customer interactions, what will happen with your business, what will happen more with the back office, right, in terms of risk.

17:12It really depends where you set your priorities. But I think it's very clear that especially every business is very clear about what results do they want to achieve, you know, and how does this fit in. Does it start with the people again and with the workshops? Absolutely. I think there are different ways how to run this. I think a very interesting way as well to see where the real need is. If you give people the tools and you let them play around and start POCing, you know, and piloting, you actually see where the real need is, where the real pain points are that you want to resolve, right? Because people, they will start using this tool in order to resolve their major pain points, the boring repetitive task and so on, right?

17:49But they will also become very creative and say, actually, what do we want to resolve more for, you know, globally as a business? So I think these visions, they combine in the end, you know, the overall strategy and say, this is where we want to go. Okay. And this is where we set our priorities in. We cannot tackle everything at the same time. All the technology is probably not yet there to tackle all the problems at the same time. But I think it's clear the direction of travel is very clear. And right now it's about setting the priorities. What do we want to achieve first? Which kind of benefits do we want to unlock first?

18:19Or where's the business case behind it? It all depends on the business case. without a business case you know there is no initiative yeah so it sounds like you know there's a step of ensuring we're lying with the the business goals absolutely you find some commercial priorities defined obviously by the board and the execs but it sounds like it's useful to also complement that with a bottom-up approach because sort of use cases pain points get suffocated out bottom-up and then top-down and then you combine it with top-down okay top-down bottom-up i think this combination will work the best because then you see what the real needs are, you know, from bottom up and you will get the people on board.

18:58But this needs to be combined absolutely with a top-down message and buy-in. Great. And how do you glue it all together? So it's kind of consistent that you suggest that what should be the accountability for like gluing all of that together, top down and bottom up and make progress? On all levels of the organizations. Accountability doesn't just sit with your CEO, right? Or just doesn't sit with your chief AI officer. Sits in all the levels of the organizations. And it starts with the end user, literally. It's like, if you manage, if you're a process owner, you know, if you're a product owner, starts with you because you're managing this product or this process.

19:33So think about how you can embed AI and make it, you know, beneficial for you and the organization and your customer. Accountability is on all levels. It cannot just be delegated, you know, to an AI team or a tech team that does the execution. They will deliver the product, but actually the adoption, you know, and the change management, it happens on the product owner and the process owner side. So accountability is there. Great. I like that. So you're ultimately accountable for the business goals. A hundred percent. If that's clear, you know, you have something to work towards and AI can support delivering on that.

20:08Really like that. So which leads me, I guess, to the next question. any recommendations on how to identify opportunities where AI might deliver quick wins? You know, there's a balance of short term versus long term, especially with change management. You want to show kind of quick results. So how do you think about this question? Yeah. All right. I think the best way to go about this is a combination of certain quick wins that produce very quick results and also test it to see if it actually works. And I'll tell you in a minute, right, how you think you go about it. And also really more the bigger strategic initiatives that take longer.

20:41that take certain investment resources and take just longer time to build and to deploy right the combination of those two things i think works the magic why because you give people already a product and you know something to test and to pilot and to see if it actually works and just by gluing basically together you know if you just give them a tool that helps them just to fix get rid of a stupid spreadsheet to be honest you know just a vlookup you can replace it with ai or just instead of going through hundreds of pages of contracts and documents, just have this. So I think if you give people already the tools for certain task automation, you will start already seeing small results.

21:20It doesn't just give you the big impact yet, right? But a big impact, you really need to have a transformational approach and need to have the bigger bets, you know, the bigger initiatives. So I think it's a combination. And you always have this feedback loop because what work or what might not work, you know, into smaller initiatives will feed back into the longer term and or the medium term initiatives. So how do you identify them? I think OpenAI, I think it was in July or so, they published an interesting playbook around this and say, just ask your people where their major pain points are. What's most painful for you?

21:52When I ask this question, you know, to a group of people and we do a little hackathon or so, and they say, oh, spreadsheets, PowerPoints, you know, it's like, just imagine a world without spreadsheets and PowerPoints. I think we would all live way happier, wouldn't we? Obviously, that's the first thing that comes to their mind, right? documents, PowerPoints, spreadsheets, just get rid of it. It will definitely help you, but will it produce this big transformational impact? Probably not, but it's a first step. So identify this boring, repetitive, painful task where a human doesn't really add that much value.

22:24This can be done by AI, where does human add value, customer interaction, human interaction. So I think that's the way how we really add value to our environment. And this being said, if you then connect the dots, you know, and have the bigger initiatives running end-to-end, transforming a process end-to-end, and especially if you're reimagining this process end-to-end, just starting on a blank sheet of paper, this is then actually how you drive the real bigger impact. And bet on this. Rather do fewer bigger things, right, than just getting lost in the millions of pilots and initiatives. Interesting.

22:58So if I play this back, it sounds like there's a combination of general productivity, task automation across the wider workforce yeah which makes the job more fulfilling and one approach is to do a surveying of what are the pain points what you're struggling with but what i think you're saying is you need to combine that with maybe a bit of a deeper transformation focus on workflow and processes that are aligned to to your business and they'll take that'll take longer and if i just may add to this because so basically you have the quick wins right and you have the more strategic initiatives i think what's always important to have in mind are two key drivers, which is the impact that you want to produce.