AI in Action: 5 Game-Changing Use Cases Marketers Can Implement Now

753
02:56:49.850 –> 02:56:52.270
Julia Nimchinski: Next up AI in action.

754
02:56:52.550 –> 02:57:04.530
Julia Nimchinski: We are going to delve into 5 game changing use cases for marketers with Jonathan Milne, Vp. Of AI. Marketing services at Demand Spring. Welcome to the show. Jonathan

755
02:57:04.980 –> 02:57:07.130
Jonathan Milne: Thanks for having me, Julia. Good to see you

756
02:57:07.550 –> 02:57:10.209
Julia Nimchinski: Good to see you, too. What’s the latest and greatest

757
02:57:10.590 –> 02:57:37.699
Jonathan Milne: What’s the latest grace? Well, I thought today we would have some fun, because I’ve been going to a lot of trade shows and conferences. And I keep getting the same question of like, you know, what tools should I use? Or you know, what use cases are people doing? And so I thought today we’ll do something a little bit more different in these settings and show the the viewers 5 real examples. And what’s fun about these examples? They’re all free tools, so everybody can follow along. Try it at home, try it at work, and see if they can apply it on their in their day to day lives

758
02:57:38.360 –> 02:57:39.699
Julia Nimchinski: Excited to dive in

759
02:57:39.940 –> 02:57:42.839
Jonathan Milne: Yeah, can I just jump in and

760
02:57:42.840 –> 02:57:43.189
Julia Nimchinski: Yeah, sure.

761
02:57:43.360 –> 02:57:46.840
Jonathan Milne: Sure screen. Awesome.

762
02:57:48.990 –> 02:57:51.999
Jonathan Milne: Alright. So hopefully, you can see my screen all right.

763
02:57:53.250 –> 02:58:07.610
Jonathan Milne: So we’re going to go through them live. We’re going to have some fun. We’re going to probably run into some demo challenges, but we’ll see how it goes. But I thought I’d share you one around SEO a fun one from Google around how to create a podcast review.

764
02:58:07.630 –> 02:58:31.270
Jonathan Milne: We’ll jump into some Facebook ads. If you ever want to do sentiment analysis that was kind of my Aha moment, and with Gpt. 3.5 a couple years ago, and then I thought we could also share people. I do a lot of reading. I’m sure you do, too, Julia. So I thought, maybe we can share. How do you actually take a book and put into practice? And then, if we have some time I’ll share one kind of little bonus fun, prompt at the very end

765
02:58:32.390 –> 02:58:36.199
Julia Nimchinski: So let’s jump on with SEO blogs. So

766
02:58:36.780 –> 02:58:42.970
Jonathan Milne: When most websites, if you’re not aware of this, most websites publish something called a sitemapxml.

767
02:58:43.140 –> 02:58:47.370
Jonathan Milne: So if I go to our website and type in Sitemap.

768
02:58:47.650 –> 02:59:03.560
Jonathan Milne: Xml, this is the same file that Google will use to figure out how to crawl your website. But what I’m looking for is this, I’m looking for something that’s going to share with us all of our posts or all our blog posts, and then a real quick hack is you can go file print

769
02:59:04.560 –> 02:59:12.419
Jonathan Milne: save as Pdf. And you can save this document to your drive for for us to use. So I already have it, which is great.

770
02:59:13.380 –> 02:59:28.089
Jonathan Milne: So the idea would be, go to your website, grab your sitemap and then pick 2, 3, 4. How many competitors you have, and go and do the exact same thing for their sitemap. So now it’s like, Okay, I have all of my blogs, and I have all the blogs of my competitors.

771
02:59:28.380 –> 02:59:51.130
Jonathan Milne: Then I’m going to go to Gemini. The reason I use Gemini for this is, I think Google has a little bit more data on keywords and a little bit more data around. You know what’s going on in the world of SEO. So I tend to always lean towards them. And then I’m going to write a prompt and the prompt is going to look something like this. And of course you can, you know, play with the prompt and reword as you see fit. But you want to say something as act as an SEO expert.

772
02:59:51.818 –> 03:00:00.170
Jonathan Milne: And I’m going to ask you to look at, you know. Insert instead of demand. Spring. Put your company name here, and I’m going to compare with some top competitors.

773
03:00:00.530 –> 03:00:08.750
Jonathan Milne: And then I’m going to ask it to do something. And this time in this example. I’m going to ask it to generate. You can pick as many as you want 5, 1015 blogs.

774
03:00:08.810 –> 03:00:34.960
Jonathan Milne: and again make sure you’re saying it. For whom? So you write your own name again as well. And then what I like to do with AI often is, not only do I want suggestions. I also want to understand the why behind AI so AI often gives us stuff. They’re getting better and better now, with a lot of their use cases, you’ll see they’re explaining more of what’s going on with AI. But I definitely want to push and say, Okay, explain to me your title, your target keyword and a little brief explanation why you’re doing this.

775
03:00:35.580 –> 03:00:59.609
Jonathan Milne: And so if we do something like this, I already reran this one just in advance, just to make sure it was working. So you can see it’s the exact same prompt. I only put in my sitemap so it doesn’t have my competitors. I didn’t want to do anything rude live. So we decided. And then here’s what Gemini provided me. So it goes through a little bit of analysis, explains what it’s doing it says, Okay, here’s some gaps I see, and then it pumps up these really interesting blog articles.

776
03:00:59.820 –> 03:01:17.589
Jonathan Milne: And the one that caught my eye it was actually near the bottom here, where it said, Okay, how about an essential guide on influencer marketing? Now, this is a topic that we don’t touch a lot at demand. Spring. I’m like, Ha! This is a great new area for us. You know. We’re not strong on it. Let’s definitely add this into our list of ideas.

777
03:01:17.770 –> 03:01:41.520
Jonathan Milne: and you know, explains why we should do this. And then the other one was video. And we just saw our friends from Vidyard. Right? So we’re we’re big big fans of them as well. So absolutely, you know, doing video doing avatars as a new key term, we’re all learning is definitely something that we’re seeing more and more of our clients using to help reach people with a more personalized outreach. So it’s definitely a really fun simple.

778
03:01:41.670 –> 03:01:54.000
Jonathan Milne: you know. Hack, I’d call it maybe that we can do, and then we’ll make sure afterwards, Julia, that we give everybody access to the whole doc right? So they can have access, and they don’t have to worry about, you know, screen captures and copying us down. But it’s super simple.

779
03:01:54.640 –> 03:02:04.210
Jonathan Milne: The next one. Let’s go into notebook. Lm, so if you’ve never used notebook. Lm, it’s a very fascinating tool. I actually was 1st introduced to it

780
03:02:04.210 –> 03:02:27.260
Jonathan Milne: through my kids. So what’s interesting is, I’m pretty sure that notebook Lm. Came out of their classroom tools, so it reminded me of these tools they had where used to be able to upload documents into your Google classroom. And then Google used to use AI tools to do things like turn into a quiz, summarize it, create flashcards. And all these kind of interactive ways to say, I’ve got all this content. How do you help your kid learn

781
03:02:27.300 –> 03:02:49.329
Jonathan Milne: what’s fascinating is they kind of open this up now to all of us to use for everything. And so one fun use case that I’ve seen quite a lot, and it’s really fun to do is put your own company’s URL into a notebook, and then you’re going to hit this thing called studio. And then we’re going to generate the audio. And I’m going to hopefully, this is going to play for you guys. But if I go into my notebook

782
03:02:49.870 –> 03:02:53.980
Jonathan Milne: and all I did was added my my URL for my company.

783
03:02:54.220 –> 03:02:58.410
Jonathan Milne: Let’s see if this plays through my speakers. And, Julie, give me a thumbs up. If you do this or not.

784
03:02:59.070 –> 03:03:24.489
Jonathan Milne: Today, we’re going on a journey into the world of AI and its impact on b 2 b marketing leaders. Okay, we’re diving deep into demand. Spring’s website. They’re, you know. Well, established Marketing Automation Company. Yeah. But they seem to be making a big push into AI, our mission to understand how b 2 b marketers are already using AI and what these experts at demand spring what they predict for the future.

785
03:03:24.700 –> 03:03:30.699
Jonathan Milne: So what’s fascinating about this is no prompt, nothing other than me giving them

786
03:03:30.760 –> 03:03:43.830
Jonathan Milne: the website, URL for our company. They created a 17 min podcast overview of our brand. And you can use this for anything. You can use this for documentation training materials.

787
03:03:43.850 –> 03:04:12.049
Jonathan Milne: competitive analysis, you name it. What’s fascinating is this, they have these 2 avatars digitally discuss your brand and have a discussion about you, and the accuracy. And the insights are always extremely fascinating. And it’s not just taking surface level information. So Google’s actually gone through a whole bunch of our website pages. And it’s picked up case studies, blog articles, and all kinds of information beyond the homepage

788
03:04:12.050 –> 03:04:39.009
Jonathan Milne: to generate this. And it does it literally in minutes. And so if you’ve never played with this, Julie, I don’t know if you’ve done this would. But definitely it’s something fun. And then what you can do with events, for example, is, we have a lot of videos coming out from from today. You could stick all of those videos and youtubes link them all here, and then have people listen to a summary of everything that was talked about today. So if someone wants all the highlights or all the insights from today’s things, you can do that.

789
03:04:39.010 –> 03:04:57.410
Jonathan Milne: I also use this tool. For with my clients that after we do a 12 week AI 1st mindset training, I put all the documents in here, and then everybody has access to all of the training materials, the best practices. If they forget something. It’s a really fun way to kind of educate people. So so please check it out and it’s free.

790
03:04:59.237 –> 03:05:01.320
Jonathan Milne: Let’s get into another one. Then.

791
03:05:02.567 –> 03:05:12.292
Jonathan Milne: So yeah. So I just share with you. It looks like they’ve already changed the the ui a little bit since I created these slides on Monday. Believe it or not. So always be.

792
03:05:12.740 –> 03:05:22.439
Jonathan Milne: be aware that these tools keep changing, so you might not see the word generate. It might be something new, like publish, or something like that. But but go ahead and you’ll figure it out pretty quickly

793
03:05:23.110 –> 03:05:43.279
Jonathan Milne: next one. This one’s fun. So this one is ads. So a lot of us are, you know, not ad experts. We sometimes hire agencies to to create ads. We sometimes don’t get the best results. And so you know, why not use AI to guide us. And what’s really fun about this is almost every ad platform now provides

794
03:05:43.350 –> 03:06:05.929
Jonathan Milne: transparent visibility into the ads that are being published on their ad set. So you could do the same thing for Google. But I’m going to just use Facebook today. So if you actually go to Facebook’s library, I thought we’ll do some fun ones around food. Clearly I’m hungry. It’s lunchtime, so let’s do some fast food. So I am going to pretend I am Burger King.

795
03:06:05.930 –> 03:06:14.689
Jonathan Milne: And so again you look up your your ads of it could be yourself or your competitors again, I would suggest going up to the file and print.

796
03:06:14.790 –> 03:06:43.839
Jonathan Milne: And then what happens again is you get to pull all of the ad sets. And AI does a very good job of taking Pdfs Ocr in it. So that’s using the technology to turn that into text, and then you get to use it. So I’ve already did this in advance, just to kind of save us a little bit of time. But I did, Burger King a few others. So what we’re going to do is we’re going to run this. I have not clicked go yet. So I’m going to learn just as much as you. But what I did is I said, okay, act as an advertising expert.

797
03:06:44.050 –> 03:06:53.239
Jonathan Milne: You’re gonna put the the industry that you’re doing. So we’re gonna specialize in the fast food burger industry. I’m gonna pretend I’m from Burger King.

798
03:06:53.540 –> 03:07:04.009
Jonathan Milne: and we’re going to analyze and compare our current ad strategies and content against our top 2 competitors. And then we’re going to add Wendy’s and Mcdonald’s. So I grabbed the other 2 just for fun.

799
03:07:04.610 –> 03:07:18.260
Jonathan Milne: And then I want to look at the strengths, weaknesses, and opportunities. And then I’m going to ask the AI to provide a breakdown of the key differences in targeting, messaging and creative approaches. And then, based on this analysis, suggest 3 highly targeted ad campaigns.

800
03:07:18.490 –> 03:07:24.589
Jonathan Milne: I’m just going to add for you here for fun for us at Burger King to remind it on what role

801
03:07:25.160 –> 03:07:44.310
Jonathan Milne: I wanted to play. And then these suggestions should be backed by advertising best practices and data driven insights. So what I’m trying to do there at the very end is saying, Hey, AI, you’re quite clever. You’ve seen a lot of things. You have a lot of information about. Ads, I want to pull that out of you and see if you can apply it to to what I’m trying to do here today.

802
03:07:45.060 –> 03:07:46.460
Jonathan Milne: So what we’re gonna do is hit.

803
03:07:46.590 –> 03:07:47.430
Jonathan Milne: Go

804
03:07:48.010 –> 03:07:57.969
Jonathan Milne: and we’ll see what magic chat gpt. Now you can change different models if you want, see the different comparisons, I just kind of use the default Mini.

805
03:07:58.100 –> 03:08:02.099
Jonathan Milne: so you can see it’s taken some time to to analyze the documentations

806
03:08:02.260 –> 03:08:27.960
Jonathan Milne: a little trick. By the way, if you ever are using documentations that are prompt, always put your documents in 1st and then ask AI to work on it don’t act as something. Say, I’m going to share something next and share the documents. It turns out for some strange reason it likes to get the documents. 1st read them, analyze them, and then look at your prompt, and then try to act on it. So it’s just a little little hack to get a little bit better results.

807
03:08:28.460 –> 03:08:35.550
Jonathan Milne: So it’s looking at the target. Audience. Awesome. So, Burger King, we at Burger King here are young digital, savvy consumers.

808
03:08:35.810 –> 03:08:38.190
Jonathan Milne: Wendy’s. They’re saying, they’re a deal focused.

809
03:08:39.020 –> 03:08:45.300
Jonathan Milne: And Mcdonald’s, they’re going broad looking at family students and budget conscious diners, which is interesting.

810
03:08:45.660 –> 03:09:13.580
Jonathan Milne: So here it’s giving me some ideas of my messaging and creative style. It’s talking about my strengths and weaknesses like we asked it to. It’s comparing the strengths and weaknesses or opportunities such as opportunities here. So that’s interesting. So it realizes. Instead of saying the weaknesses about Wendy’s, we’re going against them. AI has decided to share me opportunities, which is means. These are Wendy’s weaknesses, but my opportunities to take advantage of right? So it’s a very kind of cool little switch they did.

811
03:09:14.260 –> 03:09:23.889
Jonathan Milne: And then, okay, says, here’s some information, and then here we go. So it’s giving me kind of this idea called beyond the Whopper a flavor adventure.

812
03:09:24.230 –> 03:09:28.420
Jonathan Milne: It’s giving me some execution, and it’s giving you some best practices.

813
03:09:28.600 –> 03:09:32.199
Jonathan Milne: This is an interesting idea. Vibes is a very

814
03:09:32.490 –> 03:09:56.060
Jonathan Milne: term often today by the younger generation. So I think this is cool bk, value vibes. I kind of like that one, right? So going after a demographic, I’m getting a smile out of you, Julia. So that’s a good sign. And then this idea of frightastic moments. So you know, like how cool it is that you know I’m not an ad expert. But here I am using AI asking it some information.

815
03:09:56.060 –> 03:10:22.210
Jonathan Milne: pounding in some some extra data, some examples of what we’re at. My competitors are doing, and very quickly it’s starting to give me some fun ideas to to play off of. And then I could take these ideas and stick it into my favorite design tool and start to create the background designs. I could also use other AI tools to do quick editing of videos. So you know, this vibes one, I can see myself doing something funny with like burgers, and maybe like sound vibration vibes or something that kind of makes it

816
03:10:22.210 –> 03:10:32.980
Jonathan Milne: like shake and move. I just you can start to get some fun ideas very quickly and execute it right? So one thing I always love to tell people is that for the very 1st time in all of our careers.

817
03:10:33.050 –> 03:10:56.649
Jonathan Milne: All of us have this. I like to call AI my companion, who’s constantly available to me, who allows me to get as much support as I want. I can ask as many dumb questions. It will never hurt my feelings. It’ll continue to try to help me succeed, and if you don’t get your answers or the results you want, then I always say it’s us making the mistake. Not AI. So we have to add more context.

818
03:10:56.660 –> 03:11:08.459
Jonathan Milne: provide more details, perhaps provide more data, and then ask AI to say, Okay, how can I, you know, get get better at it. So so that’s ads. So that’s a that’s a 3rd one.

819
03:11:09.350 –> 03:11:14.959
Jonathan Milne: So again, here’s the the prompt and a little reminder of going to to the library

820
03:11:15.990 –> 03:11:27.640
Jonathan Milne: next is perplexity. So perplexity is a very cool tool, if you’ve never used it. What’s super fun is you can do stuff like deep search on it now for free as well as you can do that now with Gemini. I’ve noticed

821
03:11:27.770 –> 03:11:49.789
Jonathan Milne: but one thing perplexity does that I thought was fairly unique is you can do something called social search. It looks like today. The data is limited to just reddit, which is fine. It’s a very great resource. Reddit has all kinds of data again, for, like blog ideas terminology your customers might be using. But one of the use cases I like to use for it is sentiment analysis.

822
03:11:49.790 –> 03:12:11.779
Jonathan Milne: So what I mean by that is, I want to try to get a kind of a sense of what are people saying around my brand, or about my customers brands in these reddit groups and kind of get an idea. Are they talking about it? Positively, negatively. What are they saying? And it’s very a lot of work for us to go through a whole bunch of subreddits and and kind of track it, and also get a vibe of is this positive or negative language?

823
03:12:11.880 –> 03:12:17.609
Jonathan Milne: So what you can do is you can go to perplexity as the search tool, and I’ll share this in a second. Here

824
03:12:18.140 –> 03:12:40.440
Jonathan Milne: and again you can. You can do a search a little bit like a prompt and say, act as a social media manager. Put your company name here. I’m keeping this for now, because I’m very hopeful that perplexity. Like all these tools, we keep adding new features very quickly. It’s going to do more than just Reddit. And so I’m hoping they pull in more data very quickly. But you could just use this for Reddit.

825
03:12:40.610 –> 03:12:47.920
Jonathan Milne: And then I give it a task. And I say, Okay, your task is to identify both positive and negative mentions of my company name.

826
03:12:48.400 –> 03:12:58.790
Jonathan Milne: I want you to prioritize and share links to negative mentions. Because I’m curious what the negative first, st because I want to see if there’s a way I can address it. Maybe I can jump into Reddit and answer the questions or give another perspective.

827
03:12:59.250 –> 03:13:22.989
Jonathan Milne: And then what’s fun about AI is? And if they find something negative, not only do I want to respond to it. I’m going to ask AI. Why don’t you give me some suggestions on how to respond to it, so that I don’t have to think about it as quickly, or give me a starting point where you know what. Maybe AI is not perfect at responding to this, because it doesn’t really know. My brand doesn’t know kind of the way we we speak, but at least gets my kind of creative juices flowing again, and do that. So

828
03:13:23.140 –> 03:13:27.090
Jonathan Milne: in perplexity when you’re playing here.

829
03:13:27.492 –> 03:13:50.340
Jonathan Milne: There’s this little button here, called set sources. So what I suggest is, you click off the web and you click on social right? And so this is where you would stick your prompt in it. I’m not going to do it now, because sometimes this can take a couple minutes to run, so I don’t think anybody wants to see our perplexity run. But I did do one already that I can share the output. So I did one on

830
03:13:50.860 –> 03:14:02.779
Jonathan Milne: what’s it called Tim Hortons. So Tim Hortons is a beloved brand here in Canada. But for a lot of users they’re just starting to discover that it’s no longer owned by a Canadian organization.

831
03:14:02.840 –> 03:14:07.580
Jonathan Milne: And so I’m curious to see then what’s what’s happening and what’s going on with the brand.

832
03:14:07.610 –> 03:14:23.619
Jonathan Milne: The one thing I did in my in this example prompt is, I did not give it subgroups. So if there are subgroups in Reddit that you are aware of, then absolutely add that into the prompt. What I suggest is, let perplexity find the 1st groups for you.

833
03:14:23.620 –> 03:14:39.389
Jonathan Milne: and then, when this follow up, add in your subreddits that you know about, and ask it to go in deeper right? So kind of let it explore, see what it figures out. Let it start to think about its neural network, and how to solve your problem and then give it more instructions is the best way of using these tools.

834
03:14:39.390 –> 03:15:01.429
Jonathan Milne: And what I found fascinating is that I’m here based in Ottawa, Canada. So one of the things it did for me which I didn’t ask it for. It went to our Ottawa. So it went to my local one and said, Hey, I’m curious to see how your local people in your local area are talking about this brand for you, which I thought was super interesting. Right? So again. These ais are learning more and more about us.

835
03:15:01.726 –> 03:15:18.609
Jonathan Milne: You know. It probably knows where I am based on my IP address, but it’s fascinating to me that it’s pulling in these extra data of contextualization. So you talk about a video personalization, right and contextualization, you can see the AI agents are doing the exact same thing. They’re trying to grab extra data about us

836
03:15:18.610 –> 03:15:44.960
Jonathan Milne: to try to give us a better answer, so that if you know, Julia’s doing this prompt. And I’m doing it. We’re gonna get obviously different results. But it’s also gonna say, Hey for Julia, you might have these interests. And for John, you might have these ones, and that makes it even more contextualized for us, which is which is fascinating. And yeah, so it pulled up some some interesting information it highlighted the fact that you know, address the ownership concerns which I knew about are actually something being talked about which I thought was interesting.

837
03:15:45.070 –> 03:16:06.739
Jonathan Milne: And there’s been a lot of changes recently. And so we definitely see a quality challenge here. The donuts are no longer as good as they used to be. They used to be made fresh on site, and they’re now all warmed up from frozen. You can definitely see there’s a trend here of what’s happening to the to our beloved brand here in Canada, called Tim Hortons. So I thought that was, that was quite, quite interesting.

838
03:16:07.810 –> 03:16:20.380
Jonathan Milne: all right. And one last one is books. So this is a fascinating one that I kind of got a ha moment. If you look behind me, I’ve got a shelf. I do a lot of reading, I’m sure, Julie, you do a lot of reading as well.

839
03:16:20.450 –> 03:16:38.500
Jonathan Milne: But AI is the 1st time I realized that we can do a couple things one is, a lot of people are obviously can use AI to give you summaries, insights, quick takeaways from these authors, and I would challenge you not to do that. And the reason for that is I’m a big believer in. If you don’t have all the context of reading a book.

840
03:16:38.500 –> 03:17:04.600
Jonathan Milne: you’re not going to get its true value. You’re going to get a shortcut to something, but you’re not going to get all of the real takeaways and the kind of the why, you’re gonna get a lot of like the hows right? And so sometimes you want to understand the why and the what before the how. And so I just challenge us as humans that we should be careful of using AI to shortcut learning. But one of the cool things you can do now is, you can take something you like from a book and apply it.

841
03:17:04.710 –> 03:17:22.849
Jonathan Milne: So I’m going to use a fun one from April Dunford. She’s somebody I admire and love reading her books, and she wrote a book a little while ago called obviously Awesome, which has this really cool framework inside of it, on how to position and write clear positioning and messaging as A. b 2 b. Marketer.

842
03:17:22.910 –> 03:17:36.020
Jonathan Milne: And so I came up with this as a simple, prompt. Again, you can modify these prompts for any book. But the idea is like for her, because it’s about marketing, messaging. I’m going to act as a marketing messaging specialist. I’m going to ask the AI to use the book.

843
03:17:36.420 –> 03:17:49.130
Jonathan Milne: I’m going to give it a little context of who I am, what industry I am, what is my primary objective. And then I’m going to ask it to create the statement that they talk about inside of this book. And I want to target this audience.

844
03:17:49.860 –> 03:18:05.560
Jonathan Milne: And with it you’ll get this very cool output that basically says, Okay, from the best practice of this book for the use case, you share with me for the context you gave me. I’m going to give you now a great messaging framework and a way you can use to position your company moving forward.

845
03:18:05.700 –> 03:18:26.039
Jonathan Milne: And I’m like, Wow, like, like, how cool is this that I can read a book, turn around and action it right? So another one I had recently I was talking to a Cmo at an Hr. Tech company, and there you were reading stuff about emotional IQ. So one of the things I said to them is like, this is awesome. So you’re doing emotional? IQ. Sorry, Eq. Not IQ. Eq.

846
03:18:26.150 –> 03:18:36.299
Jonathan Milne: And and I said, great. So are you trying to learn how to apply it. And she’s like, Yeah, but I’m struggling with it. So I said, if your employees are open for it, record a couple of your one on one sessions

847
03:18:36.410 –> 03:18:57.730
Jonathan Milne: and put that AI, that take that content, put it into an AI tool and then use this book, and a prompt similar to this, and asking it to coach you on how to elevate the way you talk, and use Eq. In your conversations with your employees. Right? So 1st of all, you want to make sure your employees are cool with the idea of recording and sharing information. But again, you’ve read this entire book.

848
03:18:57.730 –> 03:19:11.340
Jonathan Milne: you know. It’s a couple 100 pages long. You’ve got all these great ideas coming from it, but you’re kind of stuck on like, how do I actually apply it and get going with it. Well, for the very 1st time in our lives you can now take a book and put it to action, which I think is super super powerful.

849
03:19:12.750 –> 03:19:36.010
Jonathan Milne: All right. We still have some time awesome. So I’m going to give you one last one. This is a fun one. If you’ve never seen this one. You’ve never done it. Please try it today. If you’re writing any type of prompt and you’re not getting good results, or are you getting good results. Still, ask AI to roast your prompt. It’s a very fun way actually to see into the mind of AI to make fun of you what actually happens by making fun of yourself.

850
03:19:36.010 –> 03:19:49.110
Jonathan Milne: you actually start to learn things that the AI could use to help improve it. You can also, on any time ask AI. When you write a prompt before you run it, say, please review my prompt and ask me 5 more questions or

851
03:19:49.110 –> 03:20:01.999
Jonathan Milne: X amount of questions or more details that would enable you to give me a better response, and very quickly AI will come up with some stuff that says, Hey, if I knew this, this, this and this about you. So obviously, it’s more context.

852
03:20:02.000 –> 03:20:19.389
Jonathan Milne: I can give you a much more in-depth response to give you a better result. Right? So we often find people who are kind of new to AI and prompting. They try to kind of do a 1 shot prompt, and then they kind of blame the AI for not getting results. I’d say, no. It’s kind of like a co-OP student.

853
03:20:19.390 –> 03:20:31.519
Jonathan Milne: you know. Make sure you’re working with the Co-OP student, giving lots of details on what you’re trying to do. Make sure your outcome is very clear, and also one extra nuance is, make sure your task is one thing at a time.

854
03:20:31.520 –> 03:20:44.730
Jonathan Milne: although we have deep reasoning, and the reasoning models are allowing ais to do a lot more, I still would tell you. Try to make sure your prompt is not trying to have like 3 different things. I wanna I want you to to, you know.

855
03:20:44.850 –> 03:20:56.810
Jonathan Milne: create an outline, a blog post. And my social media posts all in one. Go because I heard these things are called AI agents now, and as an agent you’re supposed to take over my job. Therefore you should be able to do all these things.

856
03:20:56.810 –> 03:21:16.670
Jonathan Milne: It’s a really poor way to get good results out of AI. Start off with like, let’s do the outline. Agree to the outline. Okay, that’s good. Now let’s use that outline to create my blog post. Let’s work on the blog post. I don’t like your intro. Let’s change some things. Let’s evolve the blog post. Okay, I like that. Now that I have a blog post, let’s do some social media posts. Right? So always chunk up

857
03:21:16.670 –> 03:21:23.530
Jonathan Milne: and to multi-steps your your activities or your jobs to be done with AI, and that will definitely help you

858
03:21:23.650 –> 03:21:25.019
Jonathan Milne: with what you’re trying to do.

859
03:21:26.040 –> 03:21:29.849
Jonathan Milne: And with that, Julia, I think I’m pretty good on time

860
03:21:30.600 –> 03:21:35.430
Julia Nimchinski: Super insightful session, Jonathan. Thank you so much. Just so

861
03:21:35.580 –> 03:21:45.159
Julia Nimchinski: quite funny that we spend we humans spend so much time just and resources training. You know the algorithm and AI, the Lms

862
03:21:45.755 –> 03:21:49.664
Julia Nimchinski: to think like us. And now we’re trying to the hack.

863
03:21:50.719 –> 03:21:54.310
Julia Nimchinski: The AI, and you know to adapt the thinking

864
03:21:54.650 –> 03:22:04.319
Julia Nimchinski: to the machine. I’m curious and our audience is which AI use case is delivering the fastest roi for marketers right now.

865
03:22:05.230 –> 03:22:17.499
Jonathan Milne: Yeah. So the the one, the one thing I like to to kind of address is, I often find when I do a lot of leadership coaching around AI that we we read a lot of stuff around effectiveness, right? Sorry. Sorry efficiency.

866
03:22:17.680 –> 03:22:33.539
Jonathan Milne: And what I mean by that is, we keep reading stuff like AI is going to take over our jobs. It’s replacing this. It’s making a lot of things more efficient. But what I find, especially in the 1st kind of 9 months to 18 months of experimentation. For most teams, you’re actually improving effectiveness. So you actually find yourself

867
03:22:33.540 –> 03:22:49.630
Jonathan Milne: things take longer, not shorter because you’re building out your prompt libraries. You’re building out your custom Gpts. You’re discovering that things you do. You can become deeper and more experts in than before. So if you’ll think about, I’ll go back to blog writing in the past.

868
03:22:49.890 –> 03:23:19.670
Jonathan Milne: you know, for a lot of companies. They write a blog post. They go through a little checklist, and then they publish it with AI. Now I can. I can write about 5 or 10 custom Gpts that’s looking at stuff like interlinking. I’m looking at keyword density. I’m comparing my blog article to the top 5 to quickly see if I’m missing something. You can go so much more deeply into that blog article to make it more effective, right? That we’re spending more time on that. So the one thing I’m finding is that effectiveness is kind of the way we need to think

869
03:23:19.780 –> 03:23:24.419
Jonathan Milne: next is, everyone’s obsessed with tools. And my my biggest advice around tools is.

870
03:23:24.660 –> 03:23:54.340
Jonathan Milne: if your organization has allowed you to get a tool. Don’t worry about all the other tools. Learn how to master this tool as quickly as possible and as deeply as possible. And so you know whether or not you have Gemini, or Chatgpt, or copilot. You know they all have their strengths and weaknesses. They’re definitely not equal in the sense of output today, because they have their kind of best practices and focuses, but every single month, every single quarter, one does something better than the next one kind of copies it into something else better. And it’s kind of like this constant battle.

871
03:23:54.340 –> 03:24:20.189
Jonathan Milne: So yeah, so for a couple of months you may not have the coolest feature available. So you know, for example, I noticed that Gemini’s now launched canvases, which is something that was inside of Chatgpt right and then. Now they also have deep search, which is something you can get for free in Gemini, which you have to often pay for an open air. So they just keep offering more values and stuff like that. So if you are blessed to work in an organization where you can use multiple tools, then I would definitely say, keep playing with multiple tools.

872
03:24:20.190 –> 03:24:46.900
Jonathan Milne: It shows that there’s different things like I mentioned before, I use Gemini anytime. I’m doing like SEO keyword Youtube video stuff. I feel like it’s it’s power around that data source. And that knowledge is greater. But you know, I might use something like mid journey now that has a great Ui experience to do images, even though Gemini and Openai both have their own image. Creation, I still find, like mid journey, is just not much better.

873
03:24:46.900 –> 03:24:52.530
Jonathan Milne: but for how much longer? I’m not sure. It’s a very tough world to be in these

874
03:24:52.530 –> 03:24:59.380
Jonathan Milne: Lms. Who are constantly spending a lot of money to invest in their technology stack. They just keep kind of

875
03:24:59.550 –> 03:25:02.040
Jonathan Milne: meeting equilibrium. If that makes sense

876
03:25:03.230 –> 03:25:05.390
Julia Nimchinski: Absolutely a question from Andy

877
03:25:05.660 –> 03:25:12.189
Julia Nimchinski: for the book reviews. Have you found that the AI has enough information about the book I also love April.

878
03:25:12.410 –> 03:25:15.439
Julia Nimchinski: can we see an example of the prompt you gave

879
03:25:15.910 –> 03:25:23.829
Jonathan Milne: Yeah. So that’s a very good statement. I can share back the prompt I gave as well while we’re talking. So one thing you have to recall

880
03:25:23.960 –> 03:25:45.729
Jonathan Milne: is that the Llms are trained in historical data. Right? So what I mean by that is, the the Llms have data up to a period of time. So I think Openai is about 6 to 9 months old now. So if a book has been published in between that period, it has very little information, but it will try to fill in the gap by using search.

881
03:25:45.730 –> 03:25:59.749
Jonathan Milne: So if you go to Openai, for example, and you remember there’s a little button on there called Search. When you click on that, it’s like I’m going to go try to find things on the web today that exist, pull that in and and give you summary. So what it does there is. If you have a new book, it’s going to quickly look at reviews in other sites.

882
03:25:59.810 –> 03:26:08.660
Jonathan Milne: older books, or any book kind of prior to that 9 month window. It doesn’t have a copy of the book. It has not read the entire book. What it’s

883
03:26:08.660 –> 03:26:24.369
Jonathan Milne: gathering is data that people have run summaries, discussions, videos. It’s kind of correlating all this other data it has around the book because it doesn’t have actually the book. So it’s very important to understand that it does not have all of the context of the book.

884
03:26:24.370 –> 03:26:41.440
Jonathan Milne: So if you really want to play with it. You should add more into your prompt around the context of the book that you like, and even give examples, or pull out some paragraphs and share that in your prompt to help the AI give you a more succinct answer to what you’re trying to do. So that’s a phenomenal question.

885
03:26:42.460 –> 03:26:43.999
Julia Nimchinski: I’m gonna say something bad about it.

886
03:26:44.120 –> 03:26:48.299
Julia Nimchinski: You can use deep sea. I’m pretty sure

887
03:26:48.620 –> 03:27:02.350
Julia Nimchinski: the information about the book is quite complete there, I heard. But yeah, for clear reasons, Jonathan, this is amazing, very practical session. Where should our community go to support you and learn more

888
03:27:02.720 –> 03:27:30.320
Jonathan Milne: Yeah. So please reach out onto Linkedin. So, Jonathan Milne, my full name on Linkedin. You’ll see my, my beautiful red hair on there as well, or you can always reach out to me through a demand spring reach out to us on contact us. We are very much a non traditional agency. We’re really about helping people advance their best practices, their teams, their skills. And we can really, you know, have a conversation. We are super easy to talk with super available. So so please reach out