129. Using AI In Your Building Company With Russ Stephens

AI is not going to replace your building company, but a building company that utilizes AI. Will.

Welcome to Professional Builders Secrets, the podcast for building company owners wanting to grow safely and securely. I'm your host, Will Blunt, and today I'm joined by APB co founder Russ Stevens. Russ, welcome to the show.

Thanks, Will. Great to be here.

Always love having a chat with you. And today's topic is Close to my heart because it's one we've been deep in for a little while now. We're going to be talking about artificial intelligence. Now, Russ, the reason this topic came up is because we're working closely with an AI consultant and APB and learning about implementing AI in our company so that it could eventually benefit builders. Do you want to just give some background on how we got into that?

Yeah, well, it's hit us quite hard and fast, hasn't it? It's one of those things. We knew it was going to be big back in December 2022, soon after ChatGPT was launched to the public, and we saw Builders in our members only Facebook group talking about it within a couple of weeks. And when that happened, like, we just knew. Wow. This. This time, yeah, we've been talking about AI for years, but this time it. It really is going to gain some traction really quickly. But there's still the challenge, isn't there? Like, what do I. We. We signed up to ChatGPT and, and I think there was maybe a concern internally that we could get left behind. This was, you know, we hear all these things on the news about what AI is doing, you know, what it's going to do, and you. You start to think, crikey, are we being left behind here? So that's when we made the decision to engage an AI consultant to ensure that we were on the cutting edge and ahead of the curve, really, so that we could be bringing our. Us so that no one got left behind. And, and I think we were pleasantly surprised that we weren't too far behind. Yeah. With what we'd already been doing internally. But it certainly made us aware of where the gaps were. And I think really what we got out of this, Will, was a framework, wasn't it? Yeah, we. We got some good guidance on a. On a structure because up to then we just been a little. I'd say a little ad hoc.

Yeah, exactly. And I think it's like drinking from a fire hose in a way. I think when we first went down this path, it was a lot. There was so many tools, there was so much information out there about AI and I didn't really know where to start and it felt like that for a little while. But now it feels like once we put this framework in place and started implementing it, all of a sudden now we're at a point where we are. It's a bit clearer. The path forward.

Yeah, yeah, that's right. Yeah, we can see the path. We've got a system in for staying on top of it. But I think probably most important as well, we've got this structure that touches everyone in the company and is now, as we're seeing, it's touching our members as well and bringing them along for the ride, which is pretty fun.

Okay, so that's kind of how we got into this AI journey. So what's the current landscape in AI adoption in residential construction then?

Well, I think a lot of residential building companies are already adopting AI by default. Really? Because it's already being incorporated in a lot of the tools that they're already using. If you look at takeoff software for instance, the use of AI has been a massive, massive time saver for builders with their estimates. Because of the way the AI can interpret plans and count items on a plan. It's literally saving builders hours and even in different aspects, you know, not necessarily construction specific. I jumped on a meeting with a builder just yesterday and it was a zoom meeting and I noticed he had a third party add on activated which takes notes during the meeting as well. And I recognize it because it's the exact same tool that we use. So again, that's a massive time saver. And I mean this particular add on is called Fathom Video. It listens so obviously everything's being said and it creates a checklist at the end of the meeting which emails through of the next steps. Also because we've actually got this integrated with HubSpot, even creates tasks for me on anything that I said I was going to do on the meeting. I've got a task in HubSpot to actually do it and it appends the meeting that's also in HubSpot with all the notes. So these kind of things, you're not having to go out and develop yourself or even develop the ideas. They're, they're kind of being presented to us and, and, and saving us time. But I guess the real big topic when we talk about AI and the AI landscape for builders, the big, the big topic is Chat GPT. And I think the reason that this is, you know, so big and so important is because with chat GPT you've got the flexibility to create your own solutions. You know, you can write a prompt and create specific solutions for your building company and then of course you can even take that to the next level by connecting API and actually automating as well. So huge, huge opportunities for builders right now. Yeah.

Fathom is really an example of the operational improvements that we've seen rust at apb. For example, using it in our weekly leadership meeting and referring back to the action points from the previous week. There isn't really anywhere to hide. It's like, hey, Will, have you done that from last week?

So that's what we used to have to remember, didn't we? Trying to take notes as you go. But you get so engrossed in these meetings sometimes that you kind of forget the action steps. And yeah, I think we are always heavily reliant on sky, who's very good at note taking and make sure that everyone's doing what they said they were going to do. But like you say, we get that email summary come through, we just paste that in the notes and the following week we got that ready made checklist for us.

Yeah, exactly, exactly. It's good. What motivated you touched on this before, but what motivated APB to explore AI in building companies and how it's impacting the industry?

Well, our goal at the association of Professional Builders is to improve the residential construction industry for both builders and consumers. So we look at absolutely everything through that lens and we all always ask ourselves, will doing this, will it help builders to deliver a better experience to their clients? And without question, AI is going to do that? Well, I mean, it is doing that without question, we're seeing it. And so from that perspective, when you know that this is going to have such a profound impact on the industry for both builders and consumers, you've got to start exploring every avenue where it can be utilized, but without the builders having to go out and figure this all out themselves. Because, you know, we've got 730 members now, it makes no sense for everyone to be making their own way through this maze. If we can figure out the structure and guide them as well on what the best solutions are and save them a ton of time. All builders want to do, they just want to implement time saving systems that are going to improve their business. So if we can guide them towards those solutions quicker, then that's what we see as our purpose here.

Yeah. One thing that drove me in this exploration as well, Russ, was just looking at some of the data from our recent source survey. So the state of residential construction and seeing how Closely marketing investment is linked to margins and, you know, the success and profitability of a business and how few builders out there are actually investing enough into marketing and how AI can actually help them invest more in marketing without spending so much more time themselves.

Yeah, it's been a big, big use, hasn't it? It's probably north of 50%, isn't it? Wouldn't you say? Yeah, the way ChatGPT has been utilized so far. Yes, maybe it's even higher. It has been predominantly marketing. I think we're going to see that kind of change in the coming years. But yeah, marketing has been a massive, massive opportunity for build, you know, to utilize AI and improve their marketing. Because like you say, margins are linked to marketing. We proved this time and time again every year in the SOURCI report, we proved this, that the more you spend on marketing, well, not necessarily the more 3% is the sweet spot, but you spend that 3%, you can optimize your margins.

And we can talk more about some of those GPT tactics a bit later on. We can get into the details of what's working because one thing to be aware of as well is if you use it the wrong way, it can actually have the opposite effect of what you're trying to achieve.

Yeah, there's certainly there's a lot of pitfalls out there, isn't there? Which we're going to cover that off a bit later or are we going to go into that?

Keep listening.

Nice open loop there. Nice little marketing trick.

That's it. Can you teach GBT to do open loops, Russ, or.

I reckon you can. I. Look, if you can teach a human to do it, I think you can teach AI to do it, can't you?

I think so. I think so. What are the implications on building company owners in terms of AI adoption?

Well, look, as we all know, it's going to be huge and it's going to touch everything, not just admin, but even it's going to touch on site as well. Because ultimately AI is going to enable robots to take over on site labor at some point. Now it's probably five to ten years away from happening at this stage, but it's something that we've got to be very aware of that, you know, the, the future in construction is going to be a combination of 3D printing and advanced robots and it's going to transform the way in which we, we build homes. But if we look more immediate in the more immediate future, then yeah, we can probably look at operational efficiency of every building company that is going to significantly improve now. Literally everything is going to be faster, it's going to be more accurate, and it's going to be done with less staff. So I think if you think about a custom home being designed, for instance, how that's done at the moment, just imagine a custom home being designed, something that has never been built before or even designed before with live pricing and of course, nothing getting left out of the estimate as well. I mean, how, how good would that be at the, the design stage? So, you know, there's scheduling, ordering, accounting, variations. Everything is going to feel the benefit of AI and you know, that that can be a little bit frightening, I think, for people when they, when they hear this. I mean, we've, we've certainly seen the surveys. You know, the individuals are a little bit unnerved by AI. Even business owners are, are concerned. And there is a saying out there that AI will not take your job, but someone that uses AI will. And we can, we can apply that to businesses as well. AI is not going to replace your building company, but a building company that utilizes AI will. So that's something we've got to be really aware of. It's really important to keep that top of mind and make sure that we are on the forefront, just like it's important to, to be on the forefront of systemization in the building company, to be more efficient. We now got to take that to the next level and make sure we're aware of the opportunities. And I think just to highlight this point as well, if we look back to 1980, I mean, a lot of people might not be able to do that, but I Unfortunately can. But 1980, it was a time of high unemployment. Computers were being talked about just like AI is being talked about now. They weren't actually in every business or yeah, they weren't even in the, the vast majority of businesses. But I remember people talking, saying, yeah, one day every business will have a computer. Yeah, I didn't realize that that was just touching the surface. You know, every person's going to have multiple computers was the, was the real truth. But back in 1980, high unemployment, first, you know, computers coming in. And there was a fear that computers were going to take away jobs. And yet, fast forward what, six, seven years, the economy was booming, unemployment declined, and personal computers were everywhere. So it's not something to be feared. Yeah, jobs will go. Lawyers, yeah, maybe, maybe the job of a lawyer might disappear, might get replaced by AI but not to worry, those guys can be put to work in society doing something that maybe contributes to society instead. So there'll be other avenues for people to work in.

Interesting that you spoke at the start there about the building of houses might be different and maybe potentially for building companies now the differentiator for them might be the way they build a house. But if that becomes less of a differentiator because technology is doing a large amount of that process, then the only real thing you have to differentiate yourself is your business acumen and your marketing.

Of course, because there's numerous examples of people selling commodities as, as products. But it's the way you talk about your product as well that can be the, the, the, the difference. I think Schlitz Beer is a great marketing example where a very famous marketer was asked to take over promoting this, this beer company and when he walked around the factory they were telling them their process and he said that is incredible, that is amazing. We should talk about this in our marketing. They said no, everyone does that, it doesn't matter. No, was talking about it and he actually took Schlitz Beer. Was it number one or number. He took him to number one or number two just by doing exactly that. So exactly what you're saying will. Yeah, we might find that the process is going to become very similar for a lot more companies. But yeah, the way we can differentiate ourselves is in our marketing. And of course one thing that won't change is relationships as well with more opportunity for all these automated video creation, automated emails, etc. Etc. Authenticity and in person meetings that is going to have a premium importance I think going forward. The more we start seeing deep fakes online, the less we're going to trust what we see on our screens. Therefore we will naturally place more value on face to face meetings. So it could be a good thing for professional building companies.

Yeah. One thing I've started to see is that OpenAI and other AI companies are really becoming aware of, you know, the deep fake concern and the security and safety concern of the public. Like for example, I was using GPT earlier today and asking it to write something for me and it framed it up by saying due to copyright and plagiarism laws, I can only, I can't copy exactly what you've given me. I need to give you, you know, a form of that, a summary of that or something like that. So they're already starting to be aware of that and I think when it comes to deep fake videos and things like that, there's going to be a lot of legislation that comes in.

Yeah, for sure. We're starting to see A lot of it being rushed through already and, and I think some of it is just kind of, it's not even new legislation. They're just adjusting the, the wording that they're, they're using like outbound phone calls. I think the FCC just said recently that it's illegal to use AI because it's been deemed artificial, therefore it comes under current legislation. It's illegal to use AI to make scam calls. Well, you know, it's always been legal to, to do scam calls. You know, nothing's, nothing's changed here. No laws have changed. But yeah, they're, they're scrambling, aren't they? They're, yeah, scrambling to, to make sure that people are aware that this stuff is illegal.

Well, one of OpenAI's new tools, Sora, which is not available to the public yet, but it's going to, going to allow you to bas videos from a prompt. It looks amazing.

It does, doesn't it? That's going to be great for B roll, isn't it?

Yeah, it's phenomenal for marketing. But they're actually already talking about using the metadata within that video so you can determine whether when a video has been created using the software to try and remove that fear of people creating, you know, fake versions of other people or themselves and things like that.

And, and the quality of this is off the chart as well. It's high definition, isn't it? Which is something we've not seen up to this point. And there's talk now because I believe it does 60 second clips from a prompt at the moment, but now they're talking about people literally sitting in their bedrooms able to create these feature length movies by, by using this tool when it hits. So yeah, the, the world is, is changing very, very fast at the moment.

It's quite amazing. It's interesting as well, like how we were speaking about the implications for building company owners and we've spoken a lot about marketing and the opportunities to actually get more content out there and market your building company. But one thing that actually affects is the fact that there will be a lot more builders marketing out there because creating a Facebook post or a blog article takes a fraction of the time now than it used to. So you still need to get your positioning and your strategic marketing really on point to cut through because there'll be a lot more people out there marketing. Right?

Yeah. And, and this is where it's going to come down to quality, not quantity, isn't it? And I think we've spoken about this numerous times, but over, over time. You know, if you think back to what SEO was. Yeah. Maybe 15 years ago, it, you know, there was a bit of keyword stuffing going on. It was less about the quality and more about the, the basics. But you know, that's, that's obviously changed very much now and that is going to continue on that path for sure. So, yeah, there's going to be a lot of builders. Their eyes are going to light up, aren't they? You know, I haven't, I've never done content marketing, but now I can press a button and get all these articles out there. It's not, probably not going to have a massive impact on their business because, yeah, they're just going to get lost in all the noise. The quality is going to stand head and sh. Shoulders above the noise.

Yeah, I can give some tips on how to get the quality a bit better in a moment. Russ, we'll open another loop there and come back.

I'm looking forward to that because, yeah, I know you and the marketing team have been putting a lot of time into this, Will. So, yeah, I'm looking forward to hearing about that.

Absolutely. And I know everyone out there wants to hear about how they can successfully integrate AI into their building company. So why don't we, we start talking about that. Do you want to give a bit of an overview, Ross?

Yeah. Well, when you think about it, it's literally, it's no different to how a builder would go about systemizing their building company in the first place. And you need to look at your business first of all and identify the biggest bottlenecks that are currently happening in your process and then you tackle those. And first of all, by implementing a system that's going to solve the traditionally would be a case, we'd always try to automate that system first. And if we couldn't automate it, we would develop this manual process which we would delegate to someone to execute. Now, of course, with AI, we now have the opportunity to automate those things that previously we felt weren't possible and we were having to delegate. So that's the big opportunity. But in terms of the process, it's exactly the same. I mean, first of all, you do have to systemize your business first before you can start bringing in AI. Unfortunately, if you've got a building company that is not systemized, you can't just go and, you know, hit the magic button and have Chat GPT systemize it for you. That would, that would be a nice, cool trick. But yeah, that's, that's, we're kind of not at that level yet. So yeah, you do need to systemize your building company first. You then identify the, the bottleneck that you're going to tackle and you can address that bottleneck with ChatGPT prompt creating your own prompts to start breaking down the state. You know, the stages of this problem and when you refine that. And of course you don't have to do this yourself. Yeah, if you want to have a bit of fun, you can, but you can always contract people to do this for you. But once that has been done and the prompts are being developed, then you would connect it with API to automate the whole process, which, yeah, you probably definitely want to bring some people in to do that for you. But in a nutshell, that is kind of the process that you're going to be following in order to successfully integrate AI into your building company.

Yeah, that's a great point. I don't think there's any AI tools out there at the moment that can just plug and play automate a whole system within a business. It's like if you don't understand all the steps within that and you can choose certain steps to try and automate with AI, then you're probably not going to see a lot of improvement. Right? Yeah.

And of course the other thing is, in order to implement solutions, you've got to be aware of what's possible because if you have no idea about what is possible, then how would you even know that you can automate it and use AI as a solution? So you do have to be spending or investing some of your time into learning about what is becoming available. I mean, obviously it's not just chat GPT is it where we're hearing all sorts where AI is being used for phone calls inbound and outbound? We're seeing chatbots being developed that can communicate with us and put their finger on a needle in a haystack. Really. You might have a 3, 400 page contract, but you know, rather than read through that contract, you can ask a simple question and get a very quick answer, but you've got to be aware of what the possibilities are so that as you're looking at your problems, you can kind of connect them with a potential solution. So yeah, is about learning, but I think it comes down to the same as systemizing. Again, you've got to have this growth mindset because if you feel that there isn't a better way of doing things, then you, you're going to keep doing the same things and expecting a different result.

Yeah, exactly. So when we engage with this consultant that we invested a lot of time and money into over recent times, Russ, they had a framework for implementing AI within our company, which we used and then adapted. And you've refined this process so it's more suitable for builders and also just a bit more robust for the way we wanted to use it. Do you want to just give an overview of that?

Yeah, well, look, this was a combined effort, wasn't it? We, and, and to be fair, there was a bit of fluff, wasn't there, sprinkled in from the, from the consulting company after all. Yeah, that's right. Yeah. So, so we removed the fluff because we really, we just wanted to get down to the nitty gritty, the. Yeah, maybe it's our personalities, but yeah, we just wanted the functional steps but you know, step one really, and, and yeah, this was a, a great tip was to get AI on the agenda. You got to it on your agenda. But I guess before you, maybe you even do that, you've gotta, you gotta have a plan. You've got to have a bit of a roadmap and also a hub for sharing the, all this information. So it might be the different tools that you come across, the different prompts, how it can be used in different departments, but you need some central hub to actually document that information. Yeah, we started this off by using Google Sheets and using different tabs for sections. You know, we just, we find Google Sheets, you know, useful for a lot of things. Yeah, we might use it as a spreadsheet one day, but yeah, it's been, it's been pretty good. Yeah, but that's just the start. Who knows what we'll evolve to, you know, in the, the coming months and years. But really important to have that central place for the information but also have that high level plan which, which, and I think the, the roadmap wasn't as complicated as maybe we thought it might be before we embarked on this. It really was just having some clear steps on the journey as to, you know, what we were going to put in place and how we were going to roll it out. But I guess getting back to putting AI on the agenda, you got to have it as a talking point because if it kind of gets ignored, you know, you will get left behind. So we've put it on all our meeting agendas, haven't we? So we got it on the leadership meeting, which is probably the most important because it gets our leaders talking about it. You know, the leadership team are obviously responsible for strategy as well, and forward planning. But then when you go down the level, each of the leadership teams have it on their agenda to bring up with their teams. So now we've got information flowing up from, you know, from, you know, the guys that are actually using it day to day in their role. It's flowing up to leadership and then cross pollinating different departments as well. And then yeah, of course, I mean another basic, you know, you've got to set your team up with access to Chat GPT. I think we'd probably recommend setting up a, a team account or a, a workplace I think they call it, so that you know, you've, you, you've got all your users on one account but not just chat GPT. You know, you've got to be looking at who's going to benefit from using these different advancements. Like, you know, like we mentioned earlier, Fathom Video, if you're using Zoom or any team using Zoom, you know, for what it costs, that's a good, that's a good add on to, to make your team more productive. Because really, I mean I believe when I look at the costs of this stuff, when you compare it to what the hourly rate that you're paying your staff, it's a small amount to pay on a monthly basis in order to make them more, more productive.

Yeah, I think putting it on the agenda has been one of the things that's had a huge impact on me and my team, Russ. Like it's opened up the conversation more. Like every Monday we have AI, we talk about it in our marketing weekly huddle and there was numerous examples where we went to the AI hub and initially added tools in there and use cases for those tools. And there was people within the team, including myself, who didn't realize that other people in the team were actually using it in that way. And it opened up new opportunities for us and ways we could improve our efficienc. And even like this week, for example, two new examples of where the team were utilizing either new AI tools or AI tools in a new way improved the efficiency of the whole team. So I mean just sharing that information has been enormous.

Yeah. And even almost non structured, I would say. But once you put that agenda point in, you know, something AI related comes up. And yeah, one of our team mentioned that they'd been working on a presentation and they'd actually used ChatGPT to do a lot of the, you know, the, the reference work, the research. And that then prompted a discussion about, well, you have to be careful when you're using Chatgpt for research because it hallucinates. And I'm not sure if the guys listening to this podcast would be familiar with the, the term hallucinations. But we have to remember that ChatGPT is generative AI and what that basically means if it doesn't know the answer, it will make it up because it's creative. That's what it's supposed to do. It's creating content. So you could, yeah, you could ask it. And there's been examples of this recently. Wasn't there just before the, the Super Bowl? Who won the Super Bowl? And it will give you the answer? Well, no, it will give you a answer. The super bowl hadn't even been played, but it would quite happily tell you who won, what the score was. And yeah, it's incredible. So this, this, this conversation came up in our leadership meeting because we were talking about AI and it gave us the opportunity to talk about an alternative solution, which is you dot com.

Yeah, that's the one.

Yeah, you dot com, which is an AI tool, but it references all of the information that it's kind of presenting to you so you can cross reference. So yeah, rather than ask Ch chatgpt, you know, who was the first person to fly solar across the Atlantic and just take its word for it, you know, u.com actually comes up with a URL showing where it extracted the info so you can cross reference. So in terms of research, yeah, that's a very good tool to, to be utilizing.

Yeah, that's a great point. Because AI, apart from, you know, producing content, like when you talk about chat gbt, it's a great research tool. It can do a lot of manual research for you, but, but you got to make sure you are fact checking.

Yeah, I heard instances lawyers were using it and it just made up previous cases to prove a point.

So if you're getting it to write.

Lawyers do anyway, I guess.

Yeah. So what you're saying, if you're getting it to write your contracts, you better be careful?

Yes. Yeah, absolutely. I think it's great for creating a contract template. I probably wouldn't use it for a building contract because there's too many gotchas I think at the moment in terms of local legislation. But if you think of things like simple contracts like concept agreements or prelims or maybe even employment contracts, a general level, you'll probably get to the point where you want a more complex employment contract and you're going to pay money to a lawyer to get that done properly. But I think it's a great first step to get these really high level contracts done.

Well, like anything, even if it's a Facebook post, you should be doing your due diligence on whatever content you're putting out there or whatever contract you're creating or whatever you're doing with your business to make sure that it's aligned with what you're trying to do. Right?

Yeah.

Okay, so AI is now on the agenda of the building company. What are the steps you take next, Russ?

Well, you, you want to identify these processes that are currently in being done manually. Maybe they're creating problems or maybe they're just very time consuming, in which case they're expensive if they're being done manually. So identify those that can be replaced by AI automation. And when I say those, I mean we might identify a few, but we really want to identify one. We only want to take on one at a time really. So you're going to need to do an automatic audit of your current operations and then you're looking for a combination of is this going to be relatively easy to automate? And what's the bang for buck, the time saving. Because no sense in spending a lot of money automating something that doesn't really cost you that much money to have it done manually. So you want to identify that number one thing that you, you can knock over and, and then you start, start just testing really. I mean, you can, you can do this with a consulting company that, you know, give them the brief and they will solve the problem for you. Or you might want to start trying to knock it over yourself, like testing ChatGPT prompts. But yeah, you want to test that process again just like you would if you were automating a new system. System, make sure you're happy and then you adopt it and then you just repeat the process. So you just start working through all this stuff that's being done manually.

Yeah, I think that testing process is a really important one. There's a certain amount of time you want to spend on testing a tool before you need to decide whether or not it's actually helping with efficiency or not. For example, I mean, I always think through the marketing lens because that's the team I work in. But, but we have tested tools out in marketing because there's so many AI tools in marketing and some of them just don't actually work that well. For example, we were trying out a tool to automatically create snippets of the podcast and what we found is for our podcast editor, it was actually creating extra work because reviewing the snippets that weren't quite right would take longer than going through and choosing the snippets yourself. But we adjusted that process. Rather than using that automated tool, we would download the transcript of the podcast, upload that to ChatGPT, and then give it a set of prompts to pull out the quotes with timestamps and then get the snippets that way. And we found that process workaround was much more efficient than using the automated snippet tool.

Yeah, look, that's a very good point, Will, because you don't want to get married to the idea that this is the solution that we're going to do and then just continue down that path regardless. You've got to be prepared to adapt and even in some cases, abandon. We've done this numerous times where we've explored an idea and we've gone quite far down the track with it, invested a lot of time and money, but realized that it's just not going to be the solution we'd hoped for and we've had to completely abandon. But obviously I like your solution a lot more where you've been able to adapt and come to a solution anyway. But I think there's a lot of landmines out there, really, that builders probably need to be aware of at this stage. We hear a lot about automated phone calls, for instance, both inbound and outbound. We hear these AI solutions are able to dial half a million people at a time, which is a bit of a concern. But in our experience, we've tested quite a few of these things as and when they've been released. They're still not quite good enough. I think we're quite close now, probably about 12, 18 months away, but I would probably say at this stage, I would be waiting before committing to something like that. So, you know, before you commit to an idea, obviously a limited amount to testing to see is this even going to be viable, because, you know, in the term, in the case of, you know, phone calls, it, in my, in my opinion, anyway, it's just not viable at this stage. It's just not good enough. The bots was another one you can, can, you can upload your standard operating procedures to a bot. And our idea was that anyone could query a bot. So it takes us weeks to onboard new employees and they, they might, in some departments, they might have 3, 400 page SOP to go through and reference whenever they got a question. But what about if they could type their question into a bot? Well, that sounded great in theory, but when we tested it we found that the bot was, was summarizing the procedure, rather delaying the procedure now. Something that we hopefully can overcome in the future. But yeah, this is why testing's so important, isn't it? Because you know, a new employee is asking the bot, what's the process for doing this? And it just kind of gives them a summary rather than every step of the process. Yeah, that's, that's not the ideal solution there.

The standard operating procedures, that's, that's probably a great example of where builders could utilize AI to, you know, if they don't already have a number of SOPs in place in their business, they could create those. Right?

Yeah, and, and I think what's exciting as well is as you get new bits, you can just tell it to add it, you know, to your existing sop. I actually did this with a contract that I drew up recently with someone and it was, it was based around personal guarantees and you know, what the deal was all about. But I got chat GPT to drop the contract. But then every time I thought of something I thought, yeah, add this clause as well. And you see, just like add this clause and it puts it, and it redoes the contract and it puts it in, you know, in the right spacing with the clauses, etc. So yeah, it's going to work exactly the same with SOPs as well. You can, yeah, you can add to it very, very easily and build something else. But yeah, having it, having a bot then that you can like chat to or ideally ultimately even talk to and give you the answers, I mean, how good is that you think how much time you spend onboarding people and you can automate, have someone there answering all the questions of a new employee. Fantastic.

Yeah. Okay, so AI is on the agenda. You've identified some manual processes within your building company that you can replace with AI or at least improve the efficiency of. Started to test a few different tools out to see what works for you. What happens next? Ross?

Look, we're just repeating the process, aren't we? At this stage? Yeah, we're just going around in this continuous loop, never ending improvement, continuous learning.

And making sure that the ones that you're actually implementing at scale in the building company are the ones that are working.

Yeah, absolutely. And probably on the that point as well. Another thing that I always like to implement, especially if you're putting in something new and especially if that is automated, is exception reporting. So whenever you create something probably doesn't apply to the example you used earlier, Will, because you kind of, you were using it for snippets and you didn't need exception reporting for that. But if you're going to automate a process of sorts that's going to do a repetitive task, then you need to think about, well, what could go wrong. So can you then design a report to report on the exceptions? You know, the things that kind of stand out. And I think by focusing on those exceptions rather than the things that were done correctly, you can then further tweak these improvements and processes that you're putting in place.

Cool, Great advice. We did promise that I'd give some more insight on ChatGPT, so maybe we need to dive into that a little bit.

Yeah, look, and I think you're probably the best person to do this, Will, because I know you and your team have been spending a lot of time developing prompts which you have very kindly shared with the rest of the company in our own AI hub. But yeah, I would imagine you've got some good insights onto prompt writing.

Yeah, I mean, what a big light bulb for me with ChatGPT. Originally, when I started using ChatGPT to write content, whether it was an article or a social media post or something like that, that I wasn't getting good content back. And so I started thinking, this is not good enough, like I'm not getting what I want, therefore, you know, AI is the problem. I don't know how to use it potentially. And then I went to see a presentation from a business expert here in Australia and gave me an insight into some of the prompting processes that he was using to get the best out of ChatGPT. And I started to build out this process that we use in marketing now. Now basically where I've found that we get the very best results out of ChatGPT and we use ChatGPT plus, for example, to the paid version to get the best outcomes from it. So you can upload content to that is using multi stage prompts. So what I mean by that is you literally just entering something once, waiting for a response from GPT and then entering something again, waiting for a response and then entering something again and again, like going down this row. So it's like having a conversation with someone. If you wanted to teach someone about your company, like apb, for example, how would you educate them so they could write something for you? And that's the process that we go through with ChatGPT every time we, we want to get a piece of content created. So it's really simple. You can download a Chrome extension to save prompts. Therefore every time you want to use that, it's literally click a button, copy it in, wait for a response, copy the next one, one, add it into the chat GPT wait for a response and just rinse and repeat until you get the output and the prompts that we've developed that we use repetitively. The most important prompts is the business profile. So we've created a prompt, it's probably about 300 words, which explains an overview of APB, the industry we work in, our target market, the business model we operate, the competitive landscape, our brand, our brand, voice and tone. And then we end that prompt by saying, before I ask you to complete a task, do you understand what I've said here? Or do you have any questions? And that is one of the most critical steps in the prompt. Because what it does is it actually because otherwise GPT would automatically start doing something, but instead of them doing that, it just says, no, I don't have any questions. Or sometimes it does have a question. It depends on what time, what time of day it is. So that's always the first prompt I use that business profile. So it gives us that, that context. And then we layer that with a customer avatar. So a detailed customer avatar on an overall description of our customer avatar, the problem that they have, the solutions they're looking for, the dreams that they've got. And then again we end that by saying, before you complete the task, do you have any more questions? So that is like a really, really important step to use. And once, once you've entered your business profile and your customer avatar, then you can have a prompt which is specific to the task that you're asking it to do. So for example, if we were asking for a blog article, we have a blog article checklist that we then upload and then we have a series of writing styles that we upload depending on the task. For example, acting as a copywriter if we wanted the blog article, or acting as a customer specialist if you wanted something to communicate to a customer and things like that. So that's probably as far as I need to go on this podcast. But I think the importance is, and I could talk about this all day, so I'm getting a bit carried away.

But I think I love it. Will you putting in 2,000 words in prompts to get out a 600 word article?

Well, that's, that's it. But once it's written once, then it only takes, I mean, for example, I like, I come from a writing background and to write a, you know, a good 7, 800 word article would take several hours to get the initial, initial, you know, draft out, check it and all that kind of stuff. I think the other week I sent you an article in 8 minutes which I asked you how it was and you said, yeah, it was good.

So, yeah, that was pretty, pretty incredible. And yeah, it wasn't that you'd managed to knock out an article in eight minutes. It was, I guess it would be in eight months in the making, isn't it really? Yeah, I guess so. About eight minutes from here on in.

Yeah, that's it. One other, one other tip I would give. If you have any content you can upload to GBT for, to help it understand the task you're asking for it to do, it gets a much better outcome. For example, I could get a transcript of this podcast and upload it to GPT, use those prompts that I just explained, and then ask it to write an article about the content from the interview. And that would get a much better outcome than if I just asked it to write an article about AI invest building.

Yeah. And I guess this is really leveraging up not just yours, but your team's time as well as you have, you know, gradually built up this library of prompts and it's just making a lot of the repetitive stuff like, you know, summarizing podcasts, etc. It's just made it so, so fast for you now.

Yeah, and my team has just benefited significantly from being able to use these things as well. Because, you see, I mean, another example is uploading a podcast transcript to get quotes. And then my team can upload that directly to Canva and Canva generates, you know, 50 to 100 images off those quotes straight away. Then you can use them on social media with a template. So there's some amazing stuff going on in marketing, which, which is. Yeah, pretty.

And we're taking it all for granted so quickly.

Yeah.

It wasn't long ago you'd be sitting there, create handcraft in each of those 50 little images, wouldn't you?

Well, yeah, exactly. In Photoshop or something. So you need those skills. Anyway, that's, that's probably a, A conversation we could spend all day on. So why don't we dive into, wrap things up a little bit. What mistakes are builders making when adopting AI that they could do better?

Well, aside from the ones that we've, we've covered earlier, which is, I guess, number one, believing what chat GPT says all the time when it'. When it's hallucinating. And the, probably the, the other thing I Think the, the biggest mistake maybe that the builders or any business owners can make is a similar mistake that they may have made when systemizing their. Their building company. If you, if you just start pushing this out to your team and asking them to document their role, you. You might get a bit of pushback because people are naturally afraid that the only reason you want them to document their role is because you want to replace them. And if you push a. Same way, without a narrative around it, you know, bringing them along with you, then again, you, you might. Yeah, you. You might struggle with this. So that's probably the, The. The. The biggest, most fundamental mistake. The others I think we touched on as well, it. It might be jumping onto the. The shiny object, which sounds pretty exciting, but it's not quite there yet with AI so, yeah, we want to limit our expectations to a certain degree and jump on stuff to. That's been proven. We don't want to be the pioneers, really. We want to follow a proven path.

Yeah. The way I framed it up to my team, Russ, which I actually was inspired by, the way you framed it up to the organization was that we want to be leaders in this space. It's something that's coming. It's something that's really already integrating within the industry and across all businesses. So why not be at the forefront of it? And especially in marketing, if you're not on top of all of this kind of stuff, then. Then you're basically going to be obsolete at some time. So you need to understand it. You need to actually be the ones that are using it.

Absolutely. It's an exciting time, isn't it? And it's much more exciting to be leading from the front. And I guess we see this as a very important part of our role for builders is to be the ones that are leading and showing building companies how they can utilize this. But you'll inspire your own teams. Imagine bringing. Bringing the team together in your building company and talking about your vision for the future and how things are going to change. And this is the journey that we're going to go on together. You're not. And this kind of sticks in my mind. This was one of the questions the AI consultant asked right at the beginning, if you remember this will. But one of the first questions was like, are you looking to reduce headcount by introducing AI? It's like, no. Yeah, we're probably going to continue taking more people on, but we want to make everyone more productive. And to me, it was a strange question to ask. It's like, why Would you want to lose people? You know, put so much time and energy into, into building this great team? We certainly don't want to lose anyone.

Yeah. I mean, you might be able to get by with not hiring new people because the output's much higher.

Yeah, absolutely. Yeah. Because you've made everyone else so much more efficient. But it's just an exciting, great opportunity for everyone in the organization.

Also, if you, if you train someone within your team to understand the processes and implement AI at scale, they're highly valuable.

Yes, yes. It was Richard Branson, wasn't it? I think that said. Yeah. Train them well enough that they could go and work for anyone, but treat them well enough that they never want to.

It's great advice. Great advice. Okay, so what are we doing at APB to increase AI knowledge in the building industry?

Podcasts.

Yeah, exactly.

Look, I, I, I think AI is getting like spread along all of our communication outlets. I guess so. Yeah. Obviously the podcast blog articles. Internally in our members only Facebook group, we've been talking weekly about the, the latest things happening in AI and raising awareness amongst our members. Other members have been joining in as well and sharing what they've been doing. So that's been fantastic. Building an action plan. So I know we've kind of touched on the structure within this podcast about how you can successfully implement AI into your building company, but the action plan obviously has so much more detail, not just in that structure, but on the actual prompts that you can be using as well, which, yeah, we've alluded to here. And yeah, it's, it's just everything that we already do, we're just bringing AI into all those channels.

That action plan is available for members now.

Yeah, absolutely. And I think we'll, we'll put a link in the, in the show notes for the action plan that the members others can dive straight in on. Hopefully most of them have gone through it already, but if not, that's something you definitely want to make time for. Otherwise you will face the prospect of being left behind.

Well, Russ, thanks so much for coming on the podcast today. I've had a lot of fun chatting, as you probably found out halfway through when I went off on my tangents. But I think AI is super exciting in the building industry and the way that builders can make huge leaps ahead and basically use it as a competitive advantage. But is there any final advice you'd like to share?

Absolutely, yeah. Sleep with one eye open, guys, because the future is coming.

The robots. Well, thanks, Russ. Appreciate you coming on.

Thanks.

Will also a Big thank you to our listeners wherever you are in the world. If you like professional builder secrets today, please subscribe to the podcast on your platform of choice. And if you're feeling gentle, generous, leave us a review. But until next time, have a great day.