WEBVTT 00:00.510 --> 00:02.670 -: Okay, thought leadership posts. 00:02.670 --> 00:05.970 This is my trick for writing good content 00:05.970 --> 00:08.160 that actually matches your opinion 00:08.160 --> 00:11.190 and you need to be the secret source. 00:11.190 --> 00:14.340 I think that's what a lot of people get wrong 00:14.340 --> 00:17.460 when they're trying to write content with GPT-4. 00:17.460 --> 00:20.700 So I'm just gonna type in here, 00:20.700 --> 00:25.700 write a blog post on value. 00:26.010 --> 00:28.440 Let's say the topic is 00:28.440 --> 00:33.440 the merits of value-based pricing for agencies. 00:37.380 --> 00:39.480 Actually, let's make this a little bit more interesting, 00:39.480 --> 00:42.963 the danger of value-based pricing for agencies. 00:43.800 --> 00:47.850 Okay, so if I just type, just let it type, 00:47.850 --> 00:52.320 it does a pretty decent job here. 00:52.320 --> 00:54.993 So it's just writing this out. 00:57.360 --> 00:59.070 And what you're gonna get 00:59.070 --> 01:01.500 is basically the average of the internet, right? 01:01.500 --> 01:04.877 So I'm just gonna stop this here 01:04.877 --> 01:06.870 'cause I don't need it to write the whole thing. 01:06.870 --> 01:09.180 But this is what a lot of people do. 01:09.180 --> 01:10.290 GPT-4 is really good at this. 01:10.290 --> 01:13.743 But it's saying it's increasingly popular. 01:14.670 --> 01:16.500 It seems like a win-win scenario. 01:16.500 --> 01:18.060 However, value-based pricing 01:18.060 --> 01:21.510 can also present significant risks, right? 01:21.510 --> 01:24.540 So financial stability, client relationships, 01:24.540 --> 01:26.580 et cetera, right, difficulty in estimating value. 01:26.580 --> 01:29.580 So it's picked up on some interesting things, 01:29.580 --> 01:32.280 but I would say that a lot of people complain 01:32.280 --> 01:36.570 about the writing here because it's kind of dull sometimes 01:36.570 --> 01:40.200 and actually a little bit too perfect 01:40.200 --> 01:41.430 and comprehensive, right? 01:41.430 --> 01:43.920 So how do you make it more like your own voice? 01:43.920 --> 01:44.970 How do you make it more interesting? 01:44.970 --> 01:48.870 So this is my sequence of prompts. 01:48.870 --> 01:50.310 I'm just gonna start a new chat. 01:50.310 --> 01:51.780 I'm gonna go GPT-4. 01:51.780 --> 01:54.330 And you need the browse model, so browse with Bing. 01:54.330 --> 01:57.390 If you don't have that, you can turn it on in settings. 01:57.390 --> 01:58.710 All right. 01:58.710 --> 02:01.230 So the first thing I'm gonna put in here is 02:01.230 --> 02:05.727 please look up topic and then provide a summary. 02:05.727 --> 02:07.110 So I'm gonna say 02:07.110 --> 02:12.110 the dangers of value-based pricing in agencies. 02:12.660 --> 02:13.710 And I'm choosing this topic 02:13.710 --> 02:14.880 'cause this is something I know very well. 02:14.880 --> 02:18.753 I used to run an agency and we did test value-based pricing. 02:19.950 --> 02:20.783 Okay. 02:21.780 --> 02:23.400 So now it's gonna browse the web. 02:23.400 --> 02:26.610 And the reason why we're doing the browse first 02:26.610 --> 02:28.680 is I found that it tends to hallucinate 02:28.680 --> 02:30.660 when it doesn't have at least a summary 02:30.660 --> 02:35.610 of what value-based pricing is, right? 02:35.610 --> 02:38.910 Like, and I'll say, it becomes less comprehensive 02:38.910 --> 02:42.030 if all you do is take your opinion. 02:42.030 --> 02:46.290 I also find that having the summary to look at first 02:46.290 --> 02:47.910 really helps you in the next section 02:47.910 --> 02:52.050 where we're gonna actually ask GPT-4 to interview us. 02:52.050 --> 02:54.510 So we're starting with this, 02:54.510 --> 02:57.150 it's coming back with a few drawbacks, 02:57.150 --> 02:58.410 a few different things. 02:58.410 --> 03:00.600 Actually, if you wanna see what it's doing, 03:00.600 --> 03:04.650 it found this Forbes article, found this Pricefx article. 03:04.650 --> 03:07.350 So if we wanted to actually dig into this, 03:07.350 --> 03:09.090 we could look at those. 03:09.090 --> 03:12.120 And sometimes I do while I'm waiting for it. 03:12.120 --> 03:15.723 We're just gonna wait for this to finish thinking. 03:17.970 --> 03:19.890 It's clicking on links, reading. 03:19.890 --> 03:22.020 So hopefully it'll be done in a second. 03:22.020 --> 03:24.090 One thing I would say is that the browse model 03:24.090 --> 03:26.670 is sometimes a little bit funny, 03:26.670 --> 03:30.213 like status clicks fail, that sort of thing. 03:31.410 --> 03:33.390 Okay, so it's looked at three different articles 03:33.390 --> 03:36.090 and done some comprehensive research here. 03:36.090 --> 03:38.520 So we'll see what it comes back with. 03:38.520 --> 03:39.353 Here we go. 03:39.353 --> 03:40.770 Now it's finished. 03:40.770 --> 03:42.840 So let's see what it's coming back with. 03:42.840 --> 03:44.610 Difficulty in evaluation. 03:44.610 --> 03:48.240 Yeah, that's important for sure. 03:48.240 --> 03:50.910 Implementation costs. 03:50.910 --> 03:52.200 Yeah, it can be costly to implement 03:52.200 --> 03:55.680 because the perceived value of the service 03:55.680 --> 03:56.550 needs to be strong. 03:56.550 --> 04:01.550 So actually, so if you don't know about value-based pricing, 04:01.920 --> 04:03.720 it's where, instead of saying, 04:03.720 --> 04:08.130 I wanna charge whatever it costs me to deliver the service 04:08.130 --> 04:10.020 plus 10% plus 20%. 04:10.020 --> 04:12.240 Value-based pricing is I'm gonna charge 04:12.240 --> 04:15.150 whatever the value of the service is, 04:15.150 --> 04:20.150 even if that's significantly more than the cost. 04:20.940 --> 04:23.040 So if you're doing a big, 04:23.040 --> 04:26.070 if you're churning out logos for small businesses, 04:26.070 --> 04:30.360 you might do cost plus and say I'll charge you 30% margin, 04:30.360 --> 04:33.210 but if you're doing like a logo for Nike, 04:33.210 --> 04:35.040 it might be, you might do value-based pricing, 04:35.040 --> 04:37.350 you might say, even though this took me 10 minutes, 04:37.350 --> 04:39.900 it actually really took me a lifetime to be this good, 04:39.900 --> 04:43.320 and therefore, a new logo for you is gonna be worth 04:43.320 --> 04:47.460 like $50 million, I'm gonna charge you $5 million, right? 04:47.460 --> 04:50.910 So you get 1/10, I'm gonna charge you 1/10 of the value, 04:50.910 --> 04:53.340 but for me, it only took 15 minutes, 04:53.340 --> 04:55.920 so I'm getting a huge margin. 04:55.920 --> 04:58.710 Also, the problem here, 04:58.710 --> 05:01.140 so a lot of people talk about value-based pricing 05:01.140 --> 05:02.850 as if it's a holy grail. 05:02.850 --> 05:04.380 And of course, you should charge more, 05:04.380 --> 05:07.470 like of course you should try to, but there are some issues. 05:07.470 --> 05:10.740 And GPT-4 has done a good job of finding these. 05:10.740 --> 05:12.810 Implementation costs are difficult, 05:12.810 --> 05:15.210 hard to evaluate what the value is, 05:15.210 --> 05:17.340 and there's also ongoing effort. 05:17.340 --> 05:20.042 So yeah, that's something we found as well. 05:20.042 --> 05:21.960 It's alright with the first contract, 05:21.960 --> 05:24.600 but then over time they expect more and more value 05:24.600 --> 05:27.870 and you, it gets harder to deliver. 05:27.870 --> 05:29.940 There's also overpricing and underpricing. 05:29.940 --> 05:32.130 You could actually charge too little 05:32.130 --> 05:34.320 because they estimate the value 05:34.320 --> 05:37.980 as being less than what it actually costs to deliver. 05:37.980 --> 05:41.220 I think that's why accountants and lawyers charge hourly 05:41.220 --> 05:46.220 because people always expect that accounts 05:47.190 --> 05:48.900 will take less time than they do. 05:48.900 --> 05:51.870 People always expect that law, legal cases 05:51.870 --> 05:54.120 will be settled quicker than they actually are. 05:54.120 --> 05:55.440 So I think that that's, 05:55.440 --> 05:59.220 in situations where perceived value is lower 05:59.220 --> 06:01.530 or the expected value is lower, 06:01.530 --> 06:04.710 then hourly pricing might be much better. 06:04.710 --> 06:06.780 And then same thing with the fifth one, 06:06.780 --> 06:08.190 limited applicability. 06:08.190 --> 06:10.170 It's not always suitable for every business. 06:10.170 --> 06:13.050 Actually, one thing we found was, 06:13.050 --> 06:15.840 when we were working with really big businesses, 06:15.840 --> 06:17.280 Fortune 500 companies, 06:17.280 --> 06:19.050 they just didn't let us get away with it. 06:19.050 --> 06:21.120 We had to quote the hourly rate 06:21.120 --> 06:22.740 for everyone working on the project, 06:22.740 --> 06:26.040 otherwise, they would have said no 06:26.040 --> 06:28.530 and we wouldn't have been able to pass procurement. 06:28.530 --> 06:29.736 Okay, cool. 06:29.736 --> 06:32.160 So, but that's what GPT-4 is gonna ask us. 06:32.160 --> 06:33.690 So the next prompt, 06:33.690 --> 06:37.050 and I'm just gonna bring up my prompts here, 06:37.050 --> 06:39.170 the next prompt is... 06:41.973 --> 06:46.973 we'll say, ask me five interview questions on the topic, 06:47.250 --> 06:48.303 All right, so, 06:50.430 --> 06:53.610 on, and the topic is the dangers 06:53.610 --> 06:56.190 of value-based pricing in agencies. 06:56.190 --> 06:57.760 And now, I'm just gonna relay 06:59.430 --> 07:00.930 some of the things that I just told you 07:00.930 --> 07:01.983 when we were talking. 07:03.030 --> 07:06.330 So the goal of this, I mean, this is really the meat 07:06.330 --> 07:10.680 of the technique I'm using here. 07:10.680 --> 07:15.680 We are trying to get like my professional opinion 07:15.930 --> 07:18.390 on these things, my shared experiences. 07:18.390 --> 07:21.810 I think people use ChatGPT the wrong way around. 07:21.810 --> 07:25.200 ChatGPT is an incredible editor, it's a really good writer, 07:25.200 --> 07:26.760 but it doesn't have life experiences. 07:26.760 --> 07:27.810 It's a computer, right? 07:27.810 --> 07:29.520 I have life experiences. 07:29.520 --> 07:33.000 And I wanna talk about those things, 07:33.000 --> 07:34.650 but I never have the time. 07:34.650 --> 07:36.510 So if I can just be interviewed, 07:36.510 --> 07:38.850 and we're gonna do this in a few minutes, 07:38.850 --> 07:42.450 then ChatGPT can write it up as a professional post. 07:42.450 --> 07:43.770 That saves me a lot of time. 07:43.770 --> 07:46.320 It's basically what ghost writing is, right? 07:46.320 --> 07:47.610 That's what ghost writers do. 07:47.610 --> 07:51.960 A lot of really popular social media accounts and blogs, 07:51.960 --> 07:54.360 especially if it's the founder talking, 07:54.360 --> 07:56.250 like the founder doesn't really have 07:56.250 --> 07:58.320 that much time to write that stuff. 07:58.320 --> 08:00.780 They get interviewed and then the interview notes 08:00.780 --> 08:03.930 get turned into a blog post by a professional writer. 08:03.930 --> 08:05.035 Cool. 08:05.035 --> 08:05.868 So it's asking me a few different things. 08:05.868 --> 08:08.310 I'm just gonna give really brief answers. 08:08.310 --> 08:10.770 Normally, I would give a bit more in depth answers 08:10.770 --> 08:12.840 'cause I actually really enjoy talking about my experience. 08:12.840 --> 08:16.260 But, and I think actually it might be even better 08:16.260 --> 08:17.700 if you could just talk back and forth. 08:17.700 --> 08:22.620 So if you're using the OpenAI app for iOS, 08:22.620 --> 08:23.520 you can actually do that, 08:23.520 --> 08:26.940 or if using Google Bard for this, you could do that. 08:26.940 --> 08:28.530 So can you share a specific instance 08:28.530 --> 08:29.430 in your professional experience 08:29.430 --> 08:31.110 where a value-based pricing strategy 08:31.110 --> 08:33.660 posed challenges for your agency? 08:33.660 --> 08:36.420 So I'm gonna say, one is, 08:36.420 --> 08:41.420 yes, we used to sell tracking implementation 08:45.510 --> 08:50.510 and we always under-priced when using value-based pricing 08:54.180 --> 08:59.160 because customers expected tracking 08:59.160 --> 09:04.160 would be easier than it actually is. 09:06.480 --> 09:07.313 Cool. 09:07.313 --> 09:08.280 All right. 09:08.280 --> 09:10.470 In your view, are there any particular types of agencies 09:10.470 --> 09:11.820 or sectors where value-based pricing 09:11.820 --> 09:13.980 is more dangerous or less appropriate? 09:13.980 --> 09:18.980 Yeah, I would say, sectors which, 09:18.990 --> 09:23.990 where the value is, well the perceived value is high, 09:25.230 --> 09:28.533 or the perceived value is high, 09:30.180 --> 09:35.180 higher than the actual cost will get away with it, 09:39.750 --> 09:43.563 especially if they have a monopoly. 09:45.502 --> 09:50.502 If you don't have a well-known brand name, 09:53.910 --> 09:58.910 or other people can do what you do, 10:02.280 --> 10:06.933 then you'll have to charge market rate. 10:09.180 --> 10:13.170 Monopoly profits only accrue to monopolists, right? 10:13.170 --> 10:15.000 Given the potential dangers of value-based pricing, 10:15.000 --> 10:17.350 why do you think agencies prefer this approach? 10:19.050 --> 10:24.050 I think the ability to charge more is exciting 10:28.200 --> 10:32.073 and you feel stupid for not charging more. 10:35.760 --> 10:39.280 There are also a lot of gurus out there 10:41.250 --> 10:46.250 who haven't actually run scaled agencies 10:47.760 --> 10:50.763 and talk a lot about this stuff. 11:01.440 --> 11:04.890 Some agency owners listen to their detriment, right? 11:04.890 --> 11:07.320 And I don't think that there's no value 11:07.320 --> 11:08.550 in value-based pricing, 11:08.550 --> 11:11.850 but I would say it's much less applicable 11:11.850 --> 11:13.470 than most people expect, 11:13.470 --> 11:16.680 or at least less than the gurus say. 11:16.680 --> 11:18.960 Okay, so next one is how do you approach the issue 11:18.960 --> 11:21.960 of potential overpricing or underpricing? 11:21.960 --> 11:24.450 I would say the way we did it 11:24.450 --> 11:29.450 was make sure your costs are covered in any scenario, right? 11:36.990 --> 11:39.720 So if at least your costs are covered, 11:39.720 --> 11:43.143 then everything else is upside. 11:47.400 --> 11:48.233 Okay. 11:50.130 --> 11:55.030 I would say, also try to differentiate your service 11:56.280 --> 12:00.840 so that the perceived value is higher 12:00.840 --> 12:05.840 and there's less of a market to compare you to, right? 12:08.680 --> 12:13.680 If they wanna work with someone who does Facebook ads, 12:14.700 --> 12:17.580 that's one thing, but if you're someone 12:17.580 --> 12:20.370 who specifically who's done like Facebook ads 12:20.370 --> 12:23.100 for Uber before and this is Lyft calling, right, 12:23.100 --> 12:26.640 like if you're the person who used to run it for Uber, 12:26.640 --> 12:28.260 you're gonna be able to command a lot of money 12:28.260 --> 12:30.990 from someone like Lyft, right? 12:30.990 --> 12:33.720 In their minds, there is no other person, right? 12:33.720 --> 12:36.000 Like nobody else is available, 12:36.000 --> 12:38.520 that their market is one person, and that's you. 12:38.520 --> 12:41.820 So you can command, you can set the market price. 12:41.820 --> 12:46.050 If you're just doing generic Facebook ads, you can't. 12:46.050 --> 12:46.883 Right? 12:48.090 --> 12:48.923 Cool. 12:48.923 --> 12:51.947 And then number five, it's getting hard to read this, 12:51.947 --> 12:55.497 so I'm just going to copy this, get rid of it, 12:55.497 --> 13:00.240 and just, I've just pressed Control X to cut it. 13:00.240 --> 13:02.010 It's a little bit of a UX issue. 13:02.010 --> 13:04.440 From your perspective, how can agencies mitigate the risks 13:04.440 --> 13:06.363 associated with value-based pricing? 13:07.710 --> 13:09.960 I would say that's probably the same, 13:09.960 --> 13:12.900 so same answer as above. 13:12.900 --> 13:15.390 I don't wanna hold you guys up. 13:15.390 --> 13:16.710 Okay, cool. 13:16.710 --> 13:19.080 All right, so there we go. 13:19.080 --> 13:22.230 It's now summarizing what I've talked about, which is good. 13:22.230 --> 13:25.080 I think this is actually pretty helpful just in general. 13:25.080 --> 13:26.760 I find that if someone asks me, 13:26.760 --> 13:29.310 hey, what do you think about value-based pricing? 13:29.310 --> 13:31.320 I find I can talk a little bit about it, 13:31.320 --> 13:33.930 but even just going through 13:33.930 --> 13:37.710 these ChatGPT-generated interview questions 13:37.710 --> 13:40.860 really gets me interested in the topic again. 13:40.860 --> 13:42.990 It makes me think, oh yeah, I have done this before. 13:42.990 --> 13:44.463 This is interesting, right? 13:45.360 --> 13:46.463 Cool. 13:46.463 --> 13:48.270 So, that's, now we have the summary 13:48.270 --> 13:50.790 and we have, so we have the summary, 13:50.790 --> 13:52.530 that's from like everyone else's opinion, 13:52.530 --> 13:55.620 and then we have my specific opinion. 13:55.620 --> 14:00.620 So the next prompt is to generate the outline and the title. 14:02.940 --> 14:05.913 All right, so let's do this. 14:07.860 --> 14:10.113 And I'm just gonna grab the topic. 14:17.070 --> 14:17.903 Okay. 14:17.903 --> 14:22.903 Now, what this prompt is doing is it is asking for, 14:23.400 --> 14:25.110 to generate an outline first, 14:25.110 --> 14:28.080 and also 10 very creative title options. 14:28.080 --> 14:30.900 Now, the reason why we don't just say "write the post" 14:30.900 --> 14:32.550 is that, well, two reasons. 14:32.550 --> 14:35.310 One is it tends to not be as coherent, 14:35.310 --> 14:36.870 like it starts writing and it stops, 14:36.870 --> 14:38.100 and then when you ask it to write again, 14:38.100 --> 14:41.730 it writes and kind of starts to duplicate old stuff. 14:41.730 --> 14:45.160 Two is you want the post to be relatively comprehensive 14:45.160 --> 14:47.910 and cover all the bases. 14:47.910 --> 14:51.120 But then, I would say that 14:51.120 --> 14:56.040 this is generally a good principle 14:56.040 --> 15:01.040 for working with complex-type AI projects or systems, 15:01.200 --> 15:04.590 when you need something relatively big to happen, 15:04.590 --> 15:06.540 like the 1200-word post, 15:06.540 --> 15:08.850 in general, use the least to most principles, 15:08.850 --> 15:11.460 so start with the smallest thing, like the topic, 15:11.460 --> 15:14.400 and then expand out to the next smallest thing, 15:14.400 --> 15:15.540 which is the outline, 15:15.540 --> 15:18.327 and then you can kind of write each section in the outline, 15:18.327 --> 15:19.560 and then you have the outline, 15:19.560 --> 15:21.870 then you can kind of write the conclusion, 15:21.870 --> 15:23.583 and then, from the whole post, 15:23.583 --> 15:25.320 then you can kind of write out 15:25.320 --> 15:27.000 like all the other promotional stuff. 15:27.000 --> 15:27.833 Okay, cool. 15:27.833 --> 15:30.000 So, I'm just gonna check the outline. 15:30.000 --> 15:32.370 So it's say understanding value-based pricing, 15:32.370 --> 15:34.860 appeal of value-based pricing, real world challenges, 15:34.860 --> 15:36.003 a case study. 15:36.840 --> 15:37.950 Yeah, okay. 15:37.950 --> 15:40.653 And then, dangers of value-based pricing, 15:41.880 --> 15:44.070 sector-specific considerations, 15:44.070 --> 15:47.190 mitigating the risks of value-based pricing. 15:47.190 --> 15:51.420 I would say this looks fine to me. 15:51.420 --> 15:55.573 And then, in terms of creative title options, here we go. 16:01.147 --> 16:04.083 "Navigating the Pricey Waters", I like that, it's funny. 16:07.717 --> 16:09.180 "Between Perception and Reality: 16:09.180 --> 16:10.830 The Hidden Dangers of Value-Based Pricing." 16:10.830 --> 16:11.663 I like that. 16:11.663 --> 16:12.496 Number 10. 16:14.060 --> 16:17.580 I like title number 10. 16:17.580 --> 16:20.523 And you could actually just come back and say, 16:21.420 --> 16:23.610 actually I want it to be something else. 16:23.610 --> 16:25.980 But I find this really helps me 16:25.980 --> 16:29.010 come up with a few different ideas. 16:29.010 --> 16:30.427 I really like number one as well. 16:30.427 --> 16:31.830 "Golden Goose or Hidden Trap?" 16:31.830 --> 16:34.140 Yeah, this is some really creative stuff in here. 16:34.140 --> 16:35.284 Cool. 16:35.284 --> 16:37.320 And then, I'm happy with the outline. 16:37.320 --> 16:40.983 And I'm happy with the outline. 16:43.920 --> 16:45.300 Cool. 16:45.300 --> 16:48.603 All right, so now the next prompt is, 16:50.460 --> 16:52.350 write me one section at a time. 16:52.350 --> 16:55.083 Incorporate my feedback for the next section. 16:57.870 --> 16:58.703 All right. 16:59.580 --> 17:04.260 And then, let's see what it comes back with. 17:04.260 --> 17:06.780 Now, the reason why you wanna do this 17:06.780 --> 17:11.520 and kind of get it at least two paragraphs per section 17:11.520 --> 17:15.000 is that I find when you tell it to write it all in one run, 17:15.000 --> 17:18.000 what happens is you end up being in a situation 17:18.000 --> 17:20.640 where it like, it gives you just like one paragraph 17:20.640 --> 17:24.540 per section and it doesn't end up being long enough. 17:24.540 --> 17:26.860 Then also, the other thing I find is that 17:27.900 --> 17:30.990 it's not that comprehensive, it's not that interesting. 17:30.990 --> 17:31.823 Cool. 17:31.823 --> 17:34.923 So the introduction's the most important thing. 17:36.090 --> 17:39.810 Yeah, so it stirs debate, lucrative rewards, it's alluring. 17:39.810 --> 17:41.100 Yeah, okay. 17:41.100 --> 17:42.573 It's not that simple. 17:43.537 --> 17:44.370 Okay. 17:44.370 --> 17:47.302 Attractive facade, lie a host of dangers. 17:47.302 --> 17:48.135 (chuckles) I love that. 17:48.135 --> 17:49.713 A nightmarish reality, this is really cool. 17:51.000 --> 17:53.910 Maybe just say like, mention in here 17:53.910 --> 17:58.910 that I'll be sharing my experiences with Ladder, 18:00.690 --> 18:05.690 the 50-person agency I founded, say marketing agency. 18:10.950 --> 18:12.100 Cool. 18:12.100 --> 18:12.933 And it's just gonna rewrite it. 18:12.933 --> 18:16.443 So this is one of the real benefits of ChatGPT. 18:31.170 --> 18:35.520 Okay, and then I'll say, yes, this is great, 18:35.520 --> 18:37.473 continue to the next section. 18:38.940 --> 18:41.670 Right, and you just keep going through like that. 18:41.670 --> 18:43.680 And I'm not gonna go through the whole post, 18:43.680 --> 18:46.680 just because I think you get the gist 18:46.680 --> 18:48.630 and it's starting to get pretty long. 18:48.630 --> 18:50.740 Now, the next thing you wanna do 18:51.750 --> 18:54.450 is you actually kind of wanna get it rewritten 18:54.450 --> 18:56.460 in a new style, right? 18:56.460 --> 19:01.230 So first thing we're gonna do is, 19:01.230 --> 19:02.760 I've actually run this prompt before, 19:02.760 --> 19:05.940 but briefly describe the writing style of Harry Dry 19:05.940 --> 19:08.130 from marketingexamples.com in bullet points. 19:08.130 --> 19:10.320 Now, this is un-bundling, right? 19:10.320 --> 19:11.850 If you haven't seen that lecture, 19:11.850 --> 19:14.310 we have that elsewhere in the course. 19:14.310 --> 19:16.470 But essentially what I'm trying to do here 19:16.470 --> 19:17.370 is get an understanding 19:17.370 --> 19:20.100 of like what style I want this written in. 19:20.100 --> 19:22.800 And the one that I like is this style 19:22.800 --> 19:25.260 of Harry Dry from Marketing Examples. 19:25.260 --> 19:27.660 And it actually knows who this person is. 19:27.660 --> 19:31.950 If you want to kind of learn from a style 19:31.950 --> 19:35.205 that you, that it doesn't know about, 19:35.205 --> 19:37.350 then you can actually just post in 19:37.350 --> 19:40.050 a bunch of someone's writing or your own writing 19:40.050 --> 19:42.660 and then ask for a similar style. 19:42.660 --> 19:47.160 Now, I won't use exactly say Harry Dry's style here, 19:47.160 --> 19:48.990 I actually have my own style that I use, 19:48.990 --> 19:50.370 but I just wanted to kind of show you, 19:50.370 --> 19:51.810 like you can actually kind of have it 19:51.810 --> 19:53.490 in the style of the HubSpot blog, 19:53.490 --> 19:54.420 you could have it in the style 19:54.420 --> 19:56.850 of Seth Godin or Malcolm Gladwell, 19:56.850 --> 19:59.940 you can do whatever it is you need. 19:59.940 --> 20:03.540 So I'm just going to show you, I've run this prompt before, 20:03.540 --> 20:04.440 describe the writing style 20:04.440 --> 20:06.300 of Harry Dry from marketingexamples.com. 20:06.300 --> 20:07.800 It's concise and to the point. 20:07.800 --> 20:08.633 One thing I would say 20:08.633 --> 20:10.260 is you need to ask it for bullet points, 20:10.260 --> 20:11.610 'cause otherwise it blabs on 20:11.610 --> 20:16.260 and it gives you different, I guess, formats. 20:16.260 --> 20:17.580 But this is quite useful. 20:17.580 --> 20:22.580 Now, once you have this, you can basically copy and paste, 20:24.240 --> 20:25.620 and then we're going back to here. 20:25.620 --> 20:28.410 So I have the first two sections and I'm just gonna do this 20:28.410 --> 20:32.133 and you're gonna say, I'm just gonna paste in this style. 20:33.600 --> 20:36.540 Actually, let me just paste, 20:36.540 --> 20:40.353 I wanna copy it with all the bullet points. 20:41.310 --> 20:42.693 So let me just find that. 20:45.420 --> 20:48.810 Okay, if I copy here, it brings some bullet points with it. 20:48.810 --> 20:52.710 So yeah, here we go. 20:52.710 --> 20:57.420 All right, so now, let's just say this is the final prompt. 20:57.420 --> 21:00.843 This is like how you get it done at the end. 21:01.920 --> 21:05.940 So I'm just gonna paste that in here. 21:05.940 --> 21:09.483 Rewrite the article in the following style. 21:10.440 --> 21:13.380 So once you have the article written, you could do this. 21:13.380 --> 21:14.760 Now, you don't necessarily have 21:14.760 --> 21:16.800 to use specifically these styles. 21:16.800 --> 21:19.830 If there's some things in here you disagree with, 21:19.830 --> 21:22.650 engaging and conversational, concise and to the point, 21:22.650 --> 21:24.990 relatable and accessible language. 21:24.990 --> 21:26.913 So maybe I don't want humor and wit, 21:28.320 --> 21:31.200 'cause then maybe that's a bit too much for me. 21:31.200 --> 21:32.973 Yeah, I think this is good. 21:34.440 --> 21:37.223 Yeah, it relies on examples and, okay. 21:39.390 --> 21:42.150 I'm not gonna have informal and casual tone, 21:42.150 --> 21:47.150 I'm gonna say professional and knowledgeable tone. 21:53.220 --> 21:54.053 Cool. 21:57.608 --> 21:59.520 And I'm gonna get rid of engaging and conversational. 21:59.520 --> 22:00.681 Cool. 22:00.681 --> 22:04.500 So now I've kind of edited the memes of Harry Dry, right? 22:04.500 --> 22:06.270 Like Harry Dry has his own style, 22:06.270 --> 22:08.820 I've kind of remixed it a little bit and changed it. 22:08.820 --> 22:11.040 And you could actually work on this quite a bit. 22:11.040 --> 22:13.320 You could bring in some of your own elements, 22:13.320 --> 22:15.330 like you can see what it says about your writing. 22:15.330 --> 22:18.003 You can mix two different authors together. 22:19.470 --> 22:21.630 Really the world is your oyster. 22:21.630 --> 22:25.350 But I find this is a really powerful way 22:25.350 --> 22:30.060 to basically get it in a good format. 22:30.060 --> 22:34.170 One, actually, one trick you can use, by the way, 22:34.170 --> 22:35.850 is if you're writing a guest post 22:35.850 --> 22:38.970 or you're kind of wanting it to get it past someone, 22:38.970 --> 22:41.340 like an editor, like say you wanted to publish 22:41.340 --> 22:45.540 on the HubSpot blog, you could actually have it rewritten, 22:45.540 --> 22:47.670 like write it yourself, have it rewritten 22:47.670 --> 22:49.200 in the style of the HubSpot blog 22:49.200 --> 22:52.470 so it's like much more likely to get published. 22:52.470 --> 22:55.320 You could actually even, if it's a smaller blog 22:55.320 --> 22:57.690 that ChatGPT doesn't know about, 22:57.690 --> 22:59.520 you could actually then just go 22:59.520 --> 23:02.910 and copy a bunch of the recent posts 23:02.910 --> 23:05.100 or the ones that rank the highest on SEO 23:05.100 --> 23:07.983 and then paste them in here and then ask it, 23:09.000 --> 23:12.180 tell me, briefly describe in bullet points 23:12.180 --> 23:13.383 like what the style is. 23:14.250 --> 23:15.383 Cool. 23:15.383 --> 23:16.216 So, here we go. 23:16.216 --> 23:18.900 Now it's been rewritten and I think it's much better. 23:18.900 --> 23:20.760 It's much more concise actually. 23:20.760 --> 23:25.760 So in the dynamic world of agency pricing, its appeal, okay. 23:27.630 --> 23:29.220 Yeah, I've had my share of encounters. 23:29.220 --> 23:32.010 Now it talks about Ladder, which is cool. 23:32.010 --> 23:34.050 And yeah, I think this is much better, right? 23:34.050 --> 23:37.860 Like it's much cleaner, it's, 23:37.860 --> 23:39.450 yeah, here we go, it gives a really good analogy, 23:39.450 --> 23:41.553 it's like trying to price a painting. 23:42.600 --> 23:43.800 Yeah, this is brilliant. 23:43.800 --> 23:44.640 Okay, cool. 23:44.640 --> 23:46.620 Now we're at a much better post. 23:46.620 --> 23:49.980 And, now the, I think the ethics of this 23:49.980 --> 23:51.090 is much better as well. 23:51.090 --> 23:53.460 Like I really don't like the idea 23:53.460 --> 23:55.860 of just having ChatGPT write all my posts, right? 23:55.860 --> 23:59.010 Like I actually have interesting stuff, I think, 23:59.010 --> 24:02.880 to share with the world, and now using it as a ghost writer, 24:02.880 --> 24:04.230 I think is much more ethical 24:04.230 --> 24:07.170 because I'm contributing to the training data, right, 24:07.170 --> 24:08.640 I'm not just consuming it. 24:08.640 --> 24:11.820 I'm actually giving ChatGPT a lot of direction here, 24:11.820 --> 24:14.190 saying like here's what I think about the topic 24:14.190 --> 24:15.960 and here are my opinions. 24:15.960 --> 24:18.600 I think this is much more likely to rank on SEO, 24:18.600 --> 24:20.970 but it's more importantly 24:20.970 --> 24:24.720 gonna be more uniquely interesting to other human beings, 24:24.720 --> 24:27.540 and it's not gonna be just like the average of the internet, 24:27.540 --> 24:30.930 like you would get if you didn't follow this process. 24:30.930 --> 24:32.158 Cool. 24:32.158 --> 24:33.240 So hopefully that's useful to you. 24:33.240 --> 24:37.470 And I'm using it yeah, a fair bit now 24:37.470 --> 24:41.850 and finding that it saves me a ton of time. 24:41.850 --> 24:45.390 This post, it's taken us 24 minutes, 24:45.390 --> 24:46.560 but I've had to explain it all to you, 24:46.560 --> 24:49.110 but I find that I can actually write 24:49.110 --> 24:52.230 like a four-hour post sometimes in half an hour, 24:52.230 --> 24:53.583 which is pretty amazing. 24:54.510 --> 24:57.510 Normally, it'd take me half a day, or even a day, 24:57.510 --> 24:58.770 to write some of these things, 24:58.770 --> 25:01.260 and now, I can just get it out of my head and onto paper 25:01.260 --> 25:03.360 and use my professional writer 25:03.360 --> 25:06.000 to make it professional for me 25:06.000 --> 25:09.093 and help me move on quicker to the next post.