Stop Losing Your Company's IQ

Welcome to Applied AI Australia. I'm Roman Rodriguez. This show cuts through the noise and delivers only on what matters, growth, margins, and time. Each week, I translate AI into practical outcomes leaders can actually use. Today, we're in conversation with Andrew McCarthy, general manager APAC at Notion and a former LinkedIn exec. Now, Notion is valued around 12 billion. And when teams at OpenAI and Enthropic need to organize their internal knowledge, Notion's the first one they call. Now, before we go any further, I want to ask you this question. Could a new hire understand why you made your Q4 decisions just by reading your documents? Not the actual decision, but why you made it? If your answer is no, keep listening because your workflow is broken. Tasks tell you what happened. Decisions explain the outcome. But if you're not capturing the decision logic, you're not capturing the value. We're going to talk about today how Heidi helped save 260 hours a month and how I saved over a million dollars in costs and took reporting down from 3 and 1/2 days to 6 hours. Capture the decision, capture the value, capture the why. Let's get into it. Thank you and welcome to the show, mate. Ramon, great to be here. Thanks for having me. So tell us a little bit about yourself. Yeah, so so personally, uh, born and raised in Sydney, uh, have three young children. they keep me busy. Uh related um I'm the GM for notion in the region. Uh notion is an AI workspace. We have over 100 million users globally. Um and we're growing very quickly. Uh a lot of the AI companies use us as their AI workspace. So open AI uh locally high Laurate relevance use us as well and also some large enterprises will wor like um to really help improve workflows and enable AI inside their business. So if we jump into it, you know what the data is saying is nine out of 10 executives in Australia, 90% according to wellsighted research say AI is the number one concern, number one opportunity, the number one keeping Australian executives up at night. that less than one in 10 are feeling equipped to handle it and you know capitalize on the wave that's going to hit corporate Australia in positive ways and challenge some organizations and industries for sure and over time you know the average investment over the last 12 months for enterprise AI in Australia is $28 million and not all of it is succeeding and I'm really confident and know based on my conversations that I'm having with executives every day and the data that it's not the models It's not the AI, but for AI to run succinctly. We've always known that data is critical, right? We've always known that since the internet came along and cloud computing, but how do you get to data? I think there's a bigger bottleneck that we're not actually talking about if we go into the topic of context and workflows. What's your thoughts, mate? I mean, being at Notion, you're obviously pretty familiar with this topic. Do you have a view? Firstly, I think you hit the nail on the head with what I call the AI aspiration gap. Um, I read a report recently in Harvard that spoke about 90% of CCed executives believe AI can transform their business, but only 3% feel like they're doing it effectively. Um, I can't think of another time in history where there's been that big a delta between aspiration and reality. Basically, everyone wants to do it and and not many people think they're doing it effectively. Um, I I think one of the challenges and the root causes about this is that AI like humans needs everything in the same context. And over the last 15 years, uh, a theme that has swept Australian organizations is tool sprawl. Um, so much of the knowledge and, um, ex, uh, context inside a business is captured in 10, 15, in some cases over a 100 different SAS applications across the organization. And it's very hard for the AI to get context to move effectively if that um, knowledge is fragmented across so many different sources. Um I also think we're entering a new challenge here which is if tool sprawl was the challenge of the last 15 years AI tool sprawl seems to be the challenge of the next 12 months where tools add concentating on more and more tools inside their workflows u that is preventing them from being able to leverage AI effectively and drive real business outcomes. When you say tool sprawl you know the thing that comes to mind for most Australian executives is maybe OpenAI license Gemini core you know the list goes on if you want to get into that. All you got to do is sign up to a few newsletters. But what is the right way to go? You know, should someone sign up to one vendor, all vendors? How do you eliminate the context switching? You know, because I know with adoption within organizations, switching between 10 different platform causes friction. Um, so I think there is a huge opportunity to to consolidate SAS back into a central place context and notion is an amazing place to do that. And when you do that, you're able to help drive AI inside the organization. Um, the other thing you mentioned is a lot of AI tools. There's so many fantastic AI tools out there. I use chat GPT a lot. I use Gemini a lot. I use Claude. Um, but it's also understanding like what was that tool designed for and what's the use case? And so if I think of the spectrum of AI tools today, on one side you have SAT GPT going after what I call the the Fortune 5 billion wanting to have a consumer product that's single player mode that's delivering great results for individual users. And then on the other side you have very large organizations with multiund million budgets doing huge transformation projects. And then you have what I call like the Fortune 500,000 which is companies in the middle, right? And what they really need is a collaborative layer to be able to help the team work. Um the single player mode AI tools don't have access to the right company knowledge. Um they don't have huge budgets to be able to deploy multiund million dollar projects to to like redesign those workflows. And so there's a real need for um uh to be able to create create a collaborative space inside organizations to enable AI to work for teams. And that's the challenge that that we're really focused on helping solve. The numbers are pretty clear. Leaders that are using AI effectively are seeing between 1.7 to 1.8% increase in revenue growth, 3.6 uh shareholder returns, and those who use it 88% of the time are deemed to be a high performer. So many businesses that we're speaking to are just capturing every bit of data they can, but they're not actually focusing on having it structured and organized so the AI can actually do what it needs to do. I was going to say I think a really good way to think about this is what's knowledge infrastructure look like? Yeah. What is what is the decision making infrastructure your organization needs to be able to be successful and um and thinking about what information that's most important where information is captured today. One of the biggest challenges is most knowledge in organization is actually in the heads of the employees. Yeah. Right. uh the context on decisions, uh the experience, the knowledge and that walks out the door each day with with leaders and team members and it creates a lot of single points of failure inside organizations. Um and so like a really simple place to start is to say how can we effectively get knowledge uh from inside the heads of our team members and be able to capture and organize that in a really meaningful way um to provide a a a knowledge asset for the business to be able to grow and build on. Um, and so there's a lot of tools around like there's a in notion there's an AI meeting notes where you can turn it on during meetings where a lot of knowledge is shared and it can capture information and organize that and then make sure it's stored in in a central place. Um, and that's just a really simple way to help uh a frictionless way to help get the knowledge from inside people's heads into a structured organized knowledge infrastructure that can provide um a huge value. Uh something else I've noticed as we're going down this thread is the best companies in the world today are viewing their knowledge base as an asset. I think if you if you play out the conversation around um LLMs becoming commoditized and a lot of um great LLM providers being out there the where is the value captured if in that world where intelligence and LLM are commoditized it's captured inside the knowledge in an organization and that knowledge becomes a huge asset for the business to grow and build and so I think this mindset of how you leaders are thinking about their their company asset being a knowledge base and making sure you're capturing all the decisions and the context and the experience and being able to have that compound and grow over time um is creating huge advantages for some organizations to seeing this and moving first into it. I think this can be a huge competitive advantage for leaders in the next 5 years. So just to close out, it's not about having more data, it's about having the right data, it's not about the models, it's about better inputs, right? I think I think we can both agree on that. And in terms of, you know, knowledge infrastructure, often we see many organizations pointing the fingers at, you know, the engineers, but often we we record decisions we made. You know, we invested in X, we did XYZ. But often for AI to actually make correct decisions for your business, it's actually understanding the logic that layers underneath that. It's 93 94% of company files in Australia have got inaccuracies and it's not because of the bad systems, but it's due to workflows not actually documenting what happened and why it happened and what was decided. So yeah, everyone obviously has taken meeting notes. I think everyone can say, you know, for the last hundred years, people would walk into a meeting with a pen and paper and you would write something and often we would just write down what happened, but not the why, when, the logic. But I've been in so many meetings over the years and we're trying to work something out and you know, time goes on and we decide something and then I think back why did we actually decide that and no one knows. And to your point, when people walk out the door with that information, it's more than just replacing them. It's that knowledge that you can't necessarily put a monetary value on all the time, but but it is critical. So, Andrew, you were speaking to businesses every day, right? How do you know if their key problem, their bottleneck in the organization in terms of workflow is actually decision capture and not something else? Like what are the telltale signs for you? I think business leaders have a huge opportunity to rethink the endto-end process um and think about what business outcome they're looking to to achieve. Um most knowledge work can be broken down into three simple buckets. The first is knowledge, the second is intelligence and the third is a workflow. And so a simple example here is uh a journalist uh speaks to someone. they bring the knowledge together from the person they're interviewing and the journalist. Then the journalist uh takes that intelligence and produces an article and then the article has a workflow at the side. And so the way I like to to think about and what I've seen successful organizations do is to start thinking from the outcome you're looking to and then understand what are the inputs that go into that. And so in most organizations the question would be um what are all of the knowledge sources or the context that we need to make better decisions. So what is our decision making infrastructure and then how can we look at those points and understand what is being captured and what isn't and there'll be a variety of of of uh differences across organizations and across people. Um and then the second part is where can we imply apply the intelligence the AI models effectively in that workflow to help produce an outcome. And the third is how can we create multi-step processes that help deliver that outcome in an effective way. And so a simple example of this is we have a lot of conversations across our business. My team is talking to clients and customers. They're having meetings. Uh there's one-on- ones internally. There's there's customer conversations. All that information is captured in a central place which is notion which is our company memory, our brain. And that enables me to automate a report to be written every Monday morning to say what were the key themes that came out of the conversations last week. And I've identified partners that we should be working with, identified challenges in our business. I've identified things that we should focus on that I previously would have missed because all that context would have been lost with the individuals. And so I really think about this is what is the endto-end process you're looking to achieve and then how can you break that down into the components. What's the knowledge that goes into that? What's the intelligence that's applied on top? Then what's the workflow component? And if you can start breaking that down across an organization, you can start delivering much more efficient results. Then the next part is how do you encourage and get team member buy in as well. And I think focusing on things that everyone agrees are painful like reducing internal meetings like reducing the need to spend an hour each day searching information across different systems. It doesn't matter if you're have been in a role for two weeks or 22 years. Too many of of those things can can very quickly um that like impact productivity, impact morale. And so focusing on on on the the the benefits for the end users and how this is translating to better business outcomes is actually really inspiring. Um, and that's why I think it needs to come from the top when they talk about what we're doing as a business, how we're transforming is setting up this next evolution. Um, and making sure that businesses continue growing and and thriving as the world evolves. You know, top down is the way to go. AI implementation and strategy failure often isn't the tech. And I don't think it's an engineering failure. It's a leadership failure. And we're talking about organizational change and how we're doing business and how we're going to market. I think we're we're so early in this evolution. Uh Microsoft Word was released in 1983, Excel in 1985. We're probably 1989. It's been quicker than than the PC evolution, but we still have 10 to 15 years of transformation in front of us, which is hugely exciting. Um, and I think the other thing I would say is is there are some small easy wins that you could do today that could get going. So, so things like capturing your own meetings and conversations with meeting notes and then being able to review those to understand what the key action items were. Um, I take pen and paper still when I do notes. Uh, but having an AI meeting note in the background um reminds me of things I missed. I wouldn't have thought of and helps me better follow up and gives me better memory and I think about um notion as being my second brain that place where I can capture information and that works both individually but also collectively across teams um and and gives everyone sort of additional uh capacity to to deliver good results. I was with a large enterprise top 10 in Australia last month and I was just having a bit of a chat to their CFO and I asked them an interesting question. And it comes to mind because we've sort of been talking about it. Can if you had to bring in a new executive today, would they be able to understand why you made your Q4 decisions just by reading your documents? And before she responded, she had to really think about it. And the answer was no. And the answer was no because the workflow was broken. Like they had their business plan, but the new executive coming in didn't actually understand, wouldn't be able to understand why they made those Q4 decisions just by reading it. And I think that's a really good way to think about if workflow is a problem and decision capture is a problem. If someone walks into the role, a new employee, are they going to understand why you did what you did, not just the outcome and, you know, how it went. I think it's just a very good way to look at it and compare it to, you know, the upstream, the cause of capturing the task, the missing context, the gap on why, and, you know, the downstream impact. Because when you don't, that's when you get to things like hallucinations and the AI slop that, you know, we've been speaking about before. Yeah, I I I agree. And let's let's build on that analogy. Um, a new executive walks in, can they answer the questions? Can they understand the context decisions made from the last board meeting? Um, how do they get answers to that? Well, what they do is they book meetings with the most senior people who are in that and ask them for context and they start to collect fragmented pieces of information. But the flow and effect of that is it's actually taking up future time from your executives who then are in more meetings to help onboard new people to help them understand the previous decisions. So there is a lag effect here both in terms of um the impact for the new employee walking into that role but also the future impact and requirements of other executives to spend more time on boarding this person to help them understand the context as opposed to everyone starting from the same vantage point and be able to think about I value future focused initiatives that can drive the business forward. So for the CEO listening mate to the call, what is the one thing that you would recommend them doing within the next 48 hours? What would be your number one top tip to take away from today? From today's conversation, I would say I'll go to the very start, which is most decisions, context, and knowledge walks out the door with your team members each evening. And that creates two problems. It creates multiple single points of failure but also creates a fragmentation problem of knowledge where knowledge is stored across a lot of different people's heads and then past information is captured in a lot of different systems. And so thinking through that the process of of the aspiration of our knowledge is an asset to the business and we want to make sure we can capture the right information and make sure that includes the context of why decisions were made the alternative that were considered why we chose what we chose and making sure that's stored in a central place a second company brain um which is notion um is a really good place to start um in terms of like a practical thing that you could do individually. Start using an AI meeting notes tool and recorders uh inside one-on ones and inside um meetings. Use that to capture information and then share the output of those with all of the team members that were there. That's going to make sure that you're you're reducing that workflow problem and the risk of knowledge being stored just in people's heads and leaving the business. and you're going to start putting the the first foundations in place of capturing knowledge and capturing decision context to ultimately help build that asset that's going to drive your business going forward. Yeah, love it. Love it. And always like to end on just on a more of a personal front, what are you most excited for in AI, mate? What are you what's your favorite tool outside of Notion? What do you love doing? You know, where do you see this going? Uh I was in Singapore recently with my family and uh we wanted to go see the light show at Marina Bay Sands and so we asked uh what are the AI tools? Um what's the best path to get there? What's the best vantage point? Um how can we watch the show? Like are there any ice cream vendors nearby? And actually designed the whole itinerary what the evening was. Um and that resulted in a really great family experience. And so I I think that knowledge is power and and and and being able to organize that knowledge at massive scale um can have really profound um brilliant impacts on on people. And for me that was having a great evening with my family where we got to see amazing light show from a great vantage point, get some ice cream and enjoy the evening and AI uh you wouldn't expect as front and foremost in that. It was actually the relationships and the experience with the family. uh but the insight AI provided and helping us walk the right path uh get their most efficient way and then find the best vantage point was hugely valuable. So I'm excited for that. Thanks very much for coming on the show mate. And how do people get in contact with you? How do they get in contact with the team at notion? Yeah, so uh email is amnotion.com. Um you can also go to my LinkedIn profile which is Andrew McCarthy. uh anyone interested in in in partnering or getting a job or um being a customer, I want to speak with everyone. I'm a strong believer that um our story is best told the voice of our customers and really listening and understanding what's important to them is really important to me. Um so reach out, let me know and uh and let's connect. Thank you. Thanks for coming on the show. See you. If this was useful, make sure you subscribe. We publish regularly, one podcast, one newsletter each week. All you need to stay informed and execution ready. If you want to understand where your AI reader sits, head over to appliedustralia.com.au/schoolcard and take the 5minute assessment. Now remember, you can't control the speed of change, but you can control how quick a gap. Thank you. My name is Ramon Rodriguez. This is Applied AI. Thanks for listening and I'll catch you soon. Heat. Heat. N.