ChatGPT For Trade Businesses

We'll think about AI for business now. Like, what are some of the different use cases? I showed you how I'm thinking about it because I'm a process guy and we got some examples at the start of how others are using it in the room. And I'll give you a few more ideas that you can kind of throw into the mix and then it'll just kind of get your juices going with what's possible. As I said, it falls into this process re engineering. The first thing that you can do is you want to start to change in your brain. You know how we used to say, oh, just Google it. If you didn't know an answer, you just Google it. Or maybe you just YouTube it. I know that's probably huge in the trades business. If you want to know how to change a washer or something like that or set up some equipment, just go to YouTube and tell your apprentice to find a video on how it's done. We want to change the thinking to can we just chatgpt it now? Sometimes video works, but there are times when ChatGPT can do a great result. I don't have my phone handy, but your homework will be to download the ChatGPT app. So just search download ChatGPT app. It's on Android and it's on Apple and I want you to start playing around with it. The voice recognition firstly is extraordinary. I think Trent mentioned before, he just talks into it, tells it what it wants and recognizes that voice really well. And it then it can translate that into what it is that you want it to do and start to ask it questions. Just get in the habit of how do I change that washer? Can you. I'm about to have a difficult discussion with a client who's not paying. Can you give me five tips on what I can say to them? Or I'm about to do a performance review with a team member and they keep on showing up late and they're not following process. Can you suggest three things I can mention in the performance review, you just need to get in the habit of of asking it everything. So you start to get an idea of what it can and can't do. So the first bit of homework is to download the app. The next thing is we start to think about what are some use cases here. So just like we did the MVs, you want to go through your business and think in each of the different departments because you can't help yourself, you'll default to what it is that you already know and you'll Go, oh, I will ask it to do this thing that I know well. But by using the mvs you might start to think about how could it work in finance, how might we use this in operations, how might we use this in sales? Like you want the whole team thinking about how might they use this. So in marketing, for example, that was another system that we re engineered. Now the takeaway again, I keep coming back to it. First identify what you're currently doing in your business. Then we look to go and how can AI do it better? So these are all processes that we do that then got me thinking, oh yes, I can do that. So the first one is our content machine. If you've been following my stuff, I am quite prolific with putting out YouTube content and emails and Twitter and all the rest of it. And probably in about the last six months we've probably about 5x'd the amount of content that I'm putting out. Higher quality stuff with less staff. And I've just been punching it out across all of the different socials and things like that because we re engineered the process. The whole idea is Dave makes a video. So one of the things whenever I go speak anywhere and I said to Rob, can I get a copy of the recordings for when I do this? So Rob's like, yeah, sure. So we'll take the recordings, I'll have someone watch it and they'll pull out five minute clips here, little clips there of me saying different things and then it gets fed into our content machine and we had a process where all of this was happening manually. But we'll take that video, then we'll get it transcribed, then we'll ask ChatGPT to write five Twitter posts for us and then we'll ask it to draft up a LinkedIn post and then we'll ask it to do Facebook and write our titles and descriptions and things like that. And enables us to turn one bit of content into 15 or 20 and then we have team members that can share it out. Now you'll develop your own prompts over time. So I have this prompt here that I tell ChatGPT I want you to act as though you're a world class content marketer. And then I give it a range of different direction on how I want it to act and then I feed it in my transcript and then we have a series of prompts that we ask it to do different things. And I have my virtual assistant based out of the Philippines go through and follow this step by step, put all of it into a Google Doc and then we pass it to my copywriter who's in Melbourne and say, here's your first draft. Can you polish it up and get it ready for publication? That's how we're able to do it more with a lot less. So that's our marketing example, a sales example. We have started transcribing sales calls, so we record them anyway because we do a lot of sales calls on Zoom. We'll transcribe it, then feed it in using the same process that I just showed you where we say, act as if you're a sales trainer and I want you to listen to or review this particular sales call and can you give me five suggestions that we can give to the sales team member on how they can improve their sales.

And what might have taken a sales trainer an hour to listen to a sales call. They get pretty good insights like that. And then when they meet with the team member, they can jump straight to the insights. You can go a step further and get your best salesperson, get a transcript of them doing a sales call, upload that in and then say, hey, this is an example of what an ace, you know, gold standard sales call sounds like. Then upload the other team member and then say, can you tell me the difference? Or what advice you would give for this junior, newer person to be able to model the person who's doing it really, really well. Like, there is so much you can do with this if we think about it from an HR perspective.

Position descriptions. How painful is it to write position descriptions? Pretty painful. Use ChatGPT and say, Here is an example position description that I love. I want you to follow this same model and write me a position description for a bookkeeper that we're about to recruit using the same position description. Use it for OKRs and KPIs and staff handbooks, Use it for reviews, annual reviews.

And use it in your operation. So this kind of speaks to what the gentleman was saying earlier about training a bot on your data.

So in ChatGPT you can use.

What they call custom GPTs and we can upload our own data and then have that as an extra layer in addition to all of the other knowledge that ChatGPT has. But we can train it also on our data as well. So we've done that over@systemology.com systemology bot and I've created a bot and I'll just show you how this works. So this is the Systemology bot. So I've uploaded, I run little master class trainings and I've got process and I've got my systemology book and I'll get all of that data and I save it into a folder on Google Drive. We're doing things like customer support tickets and just whatever knowledge we can pull together that I can save into a PDF. And then I create this custom bot and I upload all of that information in here and then that gets added into this large language model which then tailors the responses. So a lot of my customer service team have access to the Systemology bot, where if they've got questions, if a new person asks, what's a critical client flow? What's minimum viable systems? So here, what's a critical client flow?

Actually, yeah, please define. Now, if you just go to normal ChatGPT and ask it what the CCF, DRT, SAS and MVS is, it's going to give a pretty average response. But because I've now trained it on my data, it's now telling me what the critical client flow is, what the DRTC is, and it'll keep moving as that's doing that.

I'm going to then go here.

And then come back to the bot and then I'm going to say, I run a plumbing business. Can you help me craft out a critical client flow based on best practice? My target audience is large residential builders building new homes and my service is general plumbing services for new homes. So I've hit enter.

So it's going to go off now and map out what it thinks is best practice when it comes to a plumbing business. Now I'm going to go in here. What's it suggesting? It thinks that, you know, referral programs we talked about. I think a lot of you work on word of mouth, so it's picked that up. Engagement. There's some sort of presentation, then there's some sort of discovery session. Now I'm going to come back in here and then I'm going to go here and I'm going to say, I've attached my ccf. Can you please make some suggestions on how I can improve it based on your best practice of what I have, I'm going to attach here is a plumbing business ccf. Is that the one? Look, I think that's the one. So I'm going to go ahead and attach that.

And it's telling me that there is text because it was handwritten on there. There is. Let me just see. I've emptied my recycle bin. I've saved the wrong one. If that didn't have the text on here, it would then be able to read it. It's because that's handwriting and it's having trouble reading that. But if it didn't, I would upload that PDF and then it would be able to.

Give me suggestions on how to improve my critical client flow. So I think, I mean, the takeaway here, and we'll just switch back to here. How does this work inside your business? So then how does this work inside your business? As the gentleman rightly suggested? Do we really want to be uploading all of our sensitive information to OpenAI? Maybe not, but there are a lot of other LLM solutions. I just do that as an example for you. There are private deployments, there are other solutions. But this idea of, as we said over here, taking all of the data that you've got, uploading previous quotes and proposals and sales discussions and job specs and information about the materials that you use, like uploading everything into this LLM and then starting to engage with it and then get it saved as an app on all of your team's phone. So if they've got a question, the first thing that they can do is ask the chatbot, what is this or that? Now, again, you sometimes have to be careful because it'll sometimes guess when it doesn't know, but that is incredibly powerful for your customer service and really your whole team.