AI Won't Fix a Messy Business w/ Kieran Lee

Jordan: A recent MIT report found that 95% of generative AI projects within a business, they’re failing to deliver any return on investment right now. All that time, all that money for basically nothing. So, what is actually going on here? Because it turns out there’s this massive difference between the 5% of projects within a business that’s driven by AI that succeed, and the 95% that don’t. And it appears, according to that study, that the failures are almost always caused by businesses deciding just to grab a generic off-the-shelf AI tool, which may look slick in a demo, and sure, they can get adopted for small, trivial tasks. Things like basic uses of ChatGPT chatbots and things like that.

Jordan: But the moment that you need to work within your specific high-value business process, they seem to break. So, the 5% of businesses that are actually getting return on their AI investment at the moment, they appear to be doing the opposite. They’re actually trying to think about how they can embed AI deeply into their most valuable workflows, and create systems that actually have memory and actually learn. And that’s where the real value is.

Jordan: Hi, my name is Jordan McFadyen, and welcome to For The Daring Growth Seekers podcast, a podcast designed to help you navigate this cluttered world of marketing in the AI era. And today, we’re chatting about how every business owner and marketing leader is really starting to feel the pressure to adopt AI and essentially have this AI strategy. But there’s obviously a massive rush, and that’s exactly what is causing so many businesses to actually stumble. And the problem, obviously, isn’t the technology. It’s more around how businesses are trying to build on a messy foundation. Their processes could be clunky, their data’s most likely all over the place, and there’s just no clear structure for AI to even work with.

Jordan: So, to help understand how to fix this, I actually chatted to Kieran Lee. So, Kieran runs a business called N16 Consulting, and I chatted to her on a recent episode of the Most FM Business Hour radio show that I host. What I’ve done is I’ve condensed down our hour-long interview together, as there was a specific section that we spoke about in terms of getting your house in order first as you build out your AI systems. And that really struck a chord with me, so I wanted to share it with you today.

Jordan: So, in the chat, we cover the real reasons why AI projects are actually failing. And also, we do go through a couple of practical steps you can actually take to map out your business operations. You’ll hopefully, at the end of this podcast, learn how to identify some of those opportunities for simple automation, and also just to structure your information, so that when you’re ready to invest in AI technology, you’re actually going to be set up for success. It’s a bit of a foundational conversation, but I really do hope that you find it valuable. Let’s get into it.

Jordan: Kieran, thanks so much for joining us and being on the show this morning.

Kieran: Kia ora, Jordan. Thanks for having me.

Jordan: So, let’s just take it back a little bit. The business consulting role, what was the main part of your role as a business consultant?

Kieran: My background is in engineering, which is, I think, the key skill that I take from my engineering background is problem solving. And also, the basis of the work that I do is around processes and systems, and how businesses fit together. In terms of, I guess, the first thing that, particularly looking at it, that operational excellence lens, is looking at an end-to-end process of a business. And whether businesses are aware of it, everyone has an end-to-end process, from the customer orders through to delivery, or whatever your business might be, there is an end-to-end process. So, understanding that process, and you can look at it at lots of different levels depending on the detail you want to go into, and you see how information flows through that process, is a really important place to start.

Kieran: And by doing that, you can identify where some of the challenges are. And a good way to do it is to ask, interview people, maybe in a workshop or at a one-to-one, talk about that process, get them to map out what they do, and identify where are the frustrations, where are the challenges that they face? And that is really gold information, that you can then identify what the root causes might be of that. And sometimes the work that I do, it can simply be, let’s make this process more efficient, let’s remove the waste from this process, let’s redesign it, streamline it, standardise it, make sure everyone’s doing the same thing every time they do this process.

Kieran: It could be, actually, this part of the process is very repetitive, very manual, this is a good opportunity to automate this process. There’s lots of that. And then, increasingly, there’s opportunities to look at those processes and bring in AI, so they could be manual and repetitive but require slightly more complex logic. AI is a good opportunity to bring AI in there, as long as your data and information is well structured and AI can pull on it.

Kieran: The other stuff that you get from looking at the end-to-end processes is often that the problems aren’t necessarily with the process. There’s always some things you can improve in the process, but sometimes it kind of leads to looking at other parts of the business. It could be a cultural issue, it could be about the way that your business is structured and understanding. And the key thing that I always do is look at what’s the business strategy, what’s the value that you’re adding to the customer, what’s that value proposition, and make sure that the way you’ve structured your business is actually designed to deliver that strategy. And I think until you take a step back and look at it holistically, you don’t often realise that actually the way you’ve structured your business isn’t actually designed to deliver that. Certainly in the most efficient, effective way.

Jordan: How good. You’ve touched on just there around technology and AI. So even the way that you’re doing the work obviously has changed a lot over that time. So you mentioned there, sitting down and interviewing people. That’s obviously a core part of it. But I guess there’s technology now to help process all that information that you’re gathering.

Kieran: Oh, absolutely. And there’s definitely huge efficiencies in using AI. So one of the things that I, to be honest, I have ChatGPT or some AI open on my desktop all the time, and I’m constantly working with AI. And one of the biggest efficiencies in consulting is getting it to analyse and synthesise data and information. So, for example, a great example are those interviews or those workshops where you’re gathering all of that information. I’ve always got some sort of AI transcript on for those, provided the company are happy to share that. And then simply using AI to pull out those key themes.

Kieran: And people are probably listening going, oh, but it could hallucinate. It could give wrong information. And that’s absolutely right. And I guess the way that I use it is kind of like if I’m working with a junior consultant, maybe someone fresh out of university, and I’ve got them to go through and pull out key themes from the information, the interviews, I don’t automatically assume that they’ve done a perfect job. And I’ll obviously review it. I’ll ask questions to interrogate their logic, their thinking. And I’ll do the same thing with AI, because you’ve been in those interviews, you’ve been in those workshops, you have a sense of what the key themes are likely to be. And if it comes out with some wild things, you can go, hang on, where’d you get that from? And it’s definitely, just in the last few years, it’s got so much better.

Jordan: Interesting. It’s crazy how much better it is. Really keen to get a bit of an understanding around automation, and some of the ways that you think that it is changing businesses from a process standpoint. And I guess a lot of small businesses are thinking, when they think automation, they think, oh, that’s far too difficult for us. Which I don’t know whether your thoughts on that. And then from the larger organisations, sitting there with a directive from a board of how we’re using AI, it’s kind of that loaded question. So keen to understand, how do you see technology, automation, AI?

Kieran: But I think the first thing I say, so there’s a lot of talk, there’s a lot of work going, a lot of money being spent on AI. And as yet, I guess the jury’s out on how successful it’s all been so far. There was a report out of MIT a few months back, which said, for gen AI projects, 95% of the projects weren’t getting a return on investment. So essentially 95% of projects were failing.

Kieran: And what was interesting, so obviously that’s a really alarming and interesting stat in itself, but what was really interesting was talking about why that was the case. And one of the key reasons is that business processes and data and information is not structured in a way that AI can actually draw on that information and make decisions. So that’s a really key point. It’s not actually that AI can’t do it, it’s just that our businesses aren’t ready for it. And so that’s a really key thing that I work with businesses with, and making sure that those processes, the information is ready for it.

Kieran: And this is something that I do, and also something I recommend all businesses do. You’ve all got so much knowledge and information in your business that’s stuck in people’s heads, or in documents. Structure that information, because you can use AI, using real language. You can ask basically any question that you might have of your business, of any project you’ve worked on, or for any sale you’ve had, you can ask AI. And as long as you’ve got that information structured somewhere, it can pull it out and give it to you immediately. Yeah, very careful that all the work that I do is really well structured, so that I can build that ongoing.

Jordan: So kind of like having a really clean and tidy digital file system.

Kieran: Yeah, yeah, yeah, that’s exactly it. And now AI can pull on it really, really simply. In terms of what businesses can do now, I think there are things that small businesses can do around automation. It’s about being really clear on what is the value you want to get out of it. It’s not just about automating for automation’s sake. It’s about, actually, this particular process is repetitive. We spend a lot of time on it. Maybe it’s prone to human error. Those are the types of things that you can put automations in.

Kieran: And it doesn’t matter how small your business is. I have automations running in different parts of my business, around emails, around just repetitive processes. Obviously I talked earlier about transcripts, making sure that they are being recorded and sent through into the right folder, so that, as I say, it’s structured in a nice way. Those types of automations. And I think also things are changing so quickly.

Kieran: One of the things that I hear businesses say is, I’m a bit reluctant to invest any money now, because six months’ time it’ll all be obsolete. I empathise with that kind of sentiment, because things are changing really quickly. So there’s a couple of things. Getting your information and your processes in order is worthwhile, because you’ll get benefits anyway, but it actually means that when you are ready to invest, then you’ve got the infrastructure ready to go. So that’s cool.

Kieran: But also I think it’s worth just taking a step back and going and kind of thinking about how AI might change your business. Even just doing a simple SWOT analysis: what are your strengths, weaknesses, what are the opportunities, what are the threats, just to look at your business and think a few years out, if AI continues on the current trajectory, what are the parts of the business which could be automated? What are the AI-native competitors going to come in and do? So how can you make sure that you’re, I guess, set up secure, so that you’ve got a sustainable business in that future environment?

Jordan: I really like that. One of the key bits that I’ve taken away from that, as well as that comment you made around if a business thinks that, yeah, it’s all going to change, and they don’t want to invest too early, that type of thing. But I don’t know, my mentality, and I think what you’re describing here as well, is that it’s probably worth trying to think about some of those processes now, and actually starting to embed them and learning, and seeing about how they can embed.

Jordan: And I guess it’s similar to what we’ve done at my business as well, it seems like the same for you: spotting different gaps, and then trialling different automations, and then going, oh, this worked, this didn’t work, but we can change this. And then once it does get amazingly good, then you’ve got it set up, you know?

Kieran: Exactly, exactly.

Jordan: So it’s kind of, instead of just sitting there watching. Yeah, I kind of get that sense from what you’re saying. And yeah, I’m definitely needing to have a bit of a folder file cleaner, I think.

Kieran: Yeah, most people do. And before you start cleaning it, think about the structure would be like, my most important advice. What’s the right structure for that, for your business, and the layers, the areas that you work in.

Jordan: And I guess that’s the importance of getting a consultant through to be able to help with that structure. Because a lot of the time we’re thinking about our businesses 24/7, and getting that external view. And I think a future-ready view, I think would be really important as well.