Home Friday Recording #55 Lauren Turner: AI & Customer Marketing

#55 Lauren Turner: AI & Customer Marketing

by Lauren Turner
0 comment

In this episode of CMAweekly, titled “AI & Customer Marketing,” speakers discuss various applications of AI, particularly ChatGPT, in customer marketing and advocacy. They share real-world examples of using ChatGPT in refining customer surveys, optimizing email campaigns, and generating tailored content ideas.

The conversation delves into ethical considerations, the need to protect intellectual property, and the importance of balancing AI with human input and creativity. The speakers emphasize the potential for ChatGPT to streamline tasks, but caution against relying solely on AI for critical decision-making.

Overall, the episode provides practical insights into leveraging AI, specifically ChatGPT, to enhance customer marketing strategies and drive engagement.

Table of Contents

00:00 Developing education and templates for Marketo segments.

Kristine Kukich – The Training Sherpa ((she/her)) (00:00)
Pay a fee. And it’s a private jet. So it’s from outrageously expensive. Right. Well, I don’t know if it’s outrageously expensive, but it’s not cheap for sure.

Lauren Turner (00:10)
He’s a salesman.

Kristine Kukich – The Training Sherpa ((she/her)) (00:12)
For Vice President of sales for a software company.

Kyle Tansley (00:18)

Kristine Kukich – The Training Sherpa ((she/her)) (00:19)
They have Airlines quite a while they went with the Striker, so they have a minivan that they cart around in.

Kyle Tansley (00:27)
That’s the airline you use when you are specifically taking your dog on vacation and you have enough money to take your dogs on vacation.

Lauren Turner (00:35)
Yeah. There was a small company. I forget it was called. It was like Pet Air or something. Like Pet Air or something like that. And I was so excited for it. But the Pandemic basically killed it before it ever really launched.

Kyle Tansley (00:49)

Lauren Turner (00:51)
They just never took off. It was for two years. Their website was like, Coming soon and then nothing.

Kristine Kukich – The Training Sherpa ((she/her)) (00:58)
Gone. Yeah.

Lauren Turner (01:04)
What’s going on in your world? Christine.

Kristine Kukich – The Training Sherpa ((she/her)) (01:07)
Oh, look everyone look.

Lauren Turner (01:15)
Nice logo. Okay.

Kyle Tansley (01:21)
It’s so considerate of you to include my name and caitlin’s name in there, too. There you go.

Lauren Turner (01:27)
Glad you liked it.

Kristine Kukich – The Training Sherpa ((she/her)) (01:30)
But Yeah. It’s moving and my website is indesign right now. And Business cards. Awesome stuff. A big AI speech next week to people outside of. My normal people that I talk to about AI.

Kyle Tansley (01:49)

Kristine Kukich – The Training Sherpa ((she/her)) (01:51)
Cool, Maryland Bar Association of all Places. Oh, interesting. Yeah. I’m excited to get to talk to attorneys about AI. So are they more interested in using AI or more interested in regulating my part? I’m on a panel that’s talking about how to use AI in your practice.

Kyle Tansley (02:08)

Kristine Kukich – The Training Sherpa ((she/her)) (02:11)
So. We’re focused on mostly using generative AI to design documents.

Lauren Turner (02:19)
That’s cool. Yeah. I’ve been playing around over the last couple of days with Chat gpt Four and. Their new uploadable.

Kristine Kukich – The Training Sherpa ((she/her)) (02:29)
Document stuff where it can work with images and things. And I’ve been pretty impressed.

Lauren Turner (02:35)
I put together.

Kristine Kukich – The Training Sherpa ((she/her)) (02:37)

Lauren Turner (02:38)
I sent it a couple of screenshots of slides of a deck that I put together internally about gifting use cases for customer marketing and customer success. And it was basically like, take a look at this slide and let me know what you would do to either rearrange it or add to it. Or are there any use cases that I missed? And it gave a really good analysis of stuff.

Lauren Turner (03:00)
Which is impressive.

Kristine Kukich – The Training Sherpa ((she/her)) (03:02)
We built one today in rob’s meeting. That he was playing around with some of it yesterday. And so we were playing around with the education, building out some training for some marketo stuff and. How we could leverage it to build templates or things like.

Kristine Kukich – The Training Sherpa ((she/her)) (03:24)
Cs Plans and Stuff Success plans. And onboarding plans and templates. And verticalize. Our segments. The contents more easily than what we can do now, right where it’s just time consuming if you’re dealing with.

Kristine Kukich – The Training Sherpa ((she/her)) (03:41)
Seven verticals and three breakdowns by Arr. Then it’s challenging to customize or tailor the content to those different groups in all of the permutations. Right. But using the chat GPT stuff. We can create the baseline that says, this is what our offering is, and then.

Kristine Kukich – The Training Sherpa ((she/her)) (04:06)
Let the tool go ahead and personalize. The content along the lines of all of those segments, and the cs team member with.

Lauren Turner (04:18)
A personality or tone based on the cst member. That’s pretty cool.

Kristine Kukich – The Training Sherpa ((she/her)) (04:24)
Doesn’t that sound exciting? Yeah, absolutely.

Lauren Turner (04:28)
It does sound exciting. Who’s your furry friend. Your dog so cute. Thank you.

Katlin Hess (04:35)
She never does this, but she decided to climb up on my lap today. And she doesn’t have any teeth. Which is why her duck eggs out like that. She is ten so she’s. But heard you talking about dogs on private jets. Needed to remind me that he was here and available for travel.

Katlin Hess (05:05)
Usually she’ll only do this.

Lauren Turner (05:08)
Derby Tongue is just the most adorable.

Katlin Hess (05:12)
It kills me. And she didn’t. Be like that. She had. Normal. Her tongue stayed in her mouth, and we went to pick her up after her surgery. And I said to the vet, I’m like. Is this.

Lauren Turner (05:27)
Brewery. She’s like, no, she’ll probably always be like that.

Katlin Hess (05:31)
Well, we love her anyway, but it is like some of the pictures are just so much funnier with her tongue hanging out.

Mary Green (05:40)
I think we’d all look funnier if we had our tongue hanging out. That could be a new LinkedIn personal branding approach.

Kyle Tansley (05:49)
Where you just do selfies with your tongue hanging out. Okay, that’s weird.

Lauren Turner (05:54)
I said, Take a picture.

Katlin Hess (05:57)
On LinkedIn and say, if you know, you know.

06:05 Lauren speaks about engaging customers, taking questions.

Mary Green (06:05)
Well, welcome, everybody. Lauren. Is on to share with us. Some ways to have fun with customers. She did. An imprompt call back with influid of maybe closer to the beginning of the year. I think and. It was very popular. So I was like, hey, come in and talk to us.

Mary Green (06:30)
And yeah, if anybody has questions, go ahead and ask. I see Adrian spayer. I know him from community. Work and. I think your first call with us. So yeah, we just talk about things and start out sometimes with a topic. Other times we don’t.

Adrian Speyer (he/him) (06:51)
I always miss this call, but for whatever reason, the gods decide to cancel all my meetings for the rest of Friday, so I was like I’m free. I can make this call. Yay.

Mary Green (07:03)
Awesome. I hope it lives up to your standard.

Lauren Turner (07:07)
I’m just happy to be here. I don’t have any standards. Yeah. I was actually prepared to talk. About chat gpt and basically go over the presentation from Customer xcon, but I’m happy to talk about improv and kind of bringing fun and entertainment into customer programs as well. I just don’t have a deck prepared. Oh, wow. Did I totally mess that up or what? Okay, you do you whatever you want to do today, we can take a vote.

Lauren Turner (07:36)
By a show of hands who wants to talk about chat, GPT and generative AI.

Mary Green (07:42)
I do love gpp. Let’s do it. Okay.

Lauren Turner (07:47)
All right, so it looks like half and half that’s not really helping here. Okay.

Cristina Levenetz (07:53)
And give me the designing row. I’ll do like chatty. Bt. Two. So five. Okay, cool.

Lauren Turner (08:02)
If we can somehow tie this in to improv and we’re entertaining content, I’m happy to do that too, or to speak about it at another time. Or you can also message me privately. I do workshops. About using improv skills for customer facing roles. And it’s one of the things that I love to do as my sort of side hustle, but also one of my favorite hobbies.

Lauren Turner (08:29)
Okay. Can you see my screen? Yes, ma’am. Thank you. It can be a little complicated because I usually have something like ado tabs. So I know that most of you are already pretty familiar. But just assuming that this audience, that I was initially.

Lauren Turner (08:50)
Presenting to. Were pretty new to generative. Ai. They may have dabbled in it a little but didn’t really have. A firm grasp of it. I just wanted to kind of set the stage by saying, Well, there’s a lot of talk. Where’s my clicker? There we go.

Lauren Turner (09:04)
That okay generative AI is going to take over the world. The computers are coming after our jobs. They’re coming after our lives. They’re going to ruin everything. But I think. That’s probably pretty apocalyptic thinking. More likely we’re going to end up partnering with AI to get a lot of our work done more efficiently.

09:20 Partnering with AI for more efficient work.

Lauren Turner (09:28)
And this is going to end up creating more free time within our world. That granted, within a work environment, we’re probably going to need to do other things. So it’s not like we’re going to have extended vacations or three hour work days or anything like that. But it will enable us to become a lot more efficient with what we’re doing so that we can get more done within the course of a day.

Lauren Turner (09:48)
And that ultimately is going to create more value for what it is we do. So, christina, who I think many of you have met. One of my favorite quotes from her. Think of Chat gpt as the smartest intern you’ve ever had.

Lauren Turner (10:04)
I think it’s a great analogy. It’s one of those systems where it has a lot of great answers. Can give you a lot of detail, but you have to make sure that you have very firm guardrails around what it is you’re asking Chat gpt to do.

Lauren Turner (10:19)
Like a very enthusiastic intern who wants to impress you. They may also make up stuff that sounds good. And so like that intern, if you are. Getting. Citations from something that you’re asking chat gpt to do. You need to check all of those links to make sure that they’re actually real.

Lauren Turner (10:40)
There was a case a couple of months ago where there was an attorney who used chat gpt to analyze some sort of case that he was going to argue. And it turned out that. He used essentially Chad gpt hallucinations of cases that didn’t actually exist.

Lauren Turner (10:57)
He did not do his homework on that and then basically got called out and then fired as a result of that and sanctioned he was sanctioned as well, which means he had to pay money. Too. Yeah. So don’t do that.

Lauren Turner (11:12)
It’s a great tool as a starting point. You can have it really reduce a lot of the work that you’re doing, but you really do need to fact check. And don’t rely on it to do everything for you. You still have to complete the work. That it’s there. I’d say if you have some really good prompts out of the gate, you’re probably going to get around 80% of the way there, but you still need to edit it. You still have to put your own voice in there. And if there are citations or statistics or quotes that are used.

11:54 New technology used for text generation, voice recording.

Lauren Turner (11:43)
You need to really make sure that those are real. Just kind of showing this landscape of all these different ways that AI is being applied. I pull this graphic a few weeks ago, so probably within the three weeks since I downloaded this thing, it’s now probably twice as big. At least.

Lauren Turner (12:02)
Of all of the different tools for every imaginable thing with generation of text and video. Code creation voice to text. We’ve actually been using a couple of tools recently. Where you record your voice on a bunch of different things, and then you type in scripts. And it’ll basically use your voice to record those scripts which we’ve been using.

Lauren Turner (12:23)
Really well for our customer training. So we could basically put in the name of our customers in there, just slot them in. But we don’t have to create 18 different versions of the same video just so we can get the customer’s name in there. So it’s a really useful way to get some of that mass customization out there as well.

Lauren Turner (12:40)
Mary, I know you’ve been amazing at identifying and talking about all these different tools. So please jump in if there’s something that you see here, you’re like, oh, yeah. I’ve used that tool that’s really awesome or something that’s not here. And you’re like, oh, we need to check this out.

Lauren Turner (12:57)
Definitely jump in.

Mary Green (12:59)
Yeah. I mean if you think of anything specific. That you’re looking to use or kind of create something out of. I like Cast Magic, but I also bought the lcd for it back several months ago, so I get 600 minutes a month to do on there.

Mary Green (13:19)
And. You just upload your audio. It’s supposed to be for podcasts, but you upload your audio and it gives you the transcript, a whole bunch of prompts that are templates.

Lauren Turner (13:31)
And then a chat to chat with that episode.

Mary Green (13:35)
But I recently got. Airgram I O I think it is. It’s the one that’s recording this call. Okay. And. That is very similar but. It’s kind of like it’ll go into your call with you. And then do the transcript.

14:09 Replaced Cast Magic with Airm, Otter. Excited about using tools for church content.

Mary Green (13:51)
It does like a performance thing. It does a report that tells you pain points, action, steps. Questions asked like all of these things in a summary. And then they have a little hidden feature where it also turns every call into a chat so you can go in and say, what were all the questions list that were in this call and things like that.

Mary Green (14:15)
And that I’m using to replace. A lot of what I was doing on Cast Magic, even though it was free. I use it to replace grain.com, where I did a lot of the video stuff and that would just give me my transcript. I can now do that on airgram. And otter otter was just like you can use it immediately to start recording a call.

Mary Green (14:40)
And wherever you are. And you can do that with airgram, too. So from like $50 a month to $18 a month. But as you know. These tools will just keep evolving and getting better. So I like to play with a lot of them.

Mary Green (14:59)
And just see. What I can make up, especially if it’s like videos for YouTube. Or things like that. My pastor at Church gets super excited when I give him a transcript of something or make some shorts out of one of his sermons.

Mary Green (15:16)
But yeah, if anybody has anything specific they’re looking for, just let me know and I’m sure I can find it.

Lauren Turner (15:24)
That’s super cool. I would love something that shortens the steps. And I’ll go into this in a little bit more detail later in the presentation. But chat JPT has been really useful in helping condense the time to getting case studies out.

Mary Green (15:39)
And what my process has basically been is I’ll do a customer interview over Zoom.

Lauren Turner (15:45)
And then send the transcript from that record. Sorry not to get the transcript from that recording by sending it to rev. Ai. Take that transcript, put it in Word, swap out the names so that it’s anonymized and we don’t have to worry about proprietary data.

Lauren Turner (16:01)
And then I’ll upload that into chat gpt and then give it several different prompts about how I want to see it formatted as a case study. Within. A certain paragraph structure. I want the customer to be the hero. I want a pull quote here and things like that. And it’ll basically condense this process.

Lauren Turner (16:20)
That typically took up to a month into a few hours.

Mary Green (16:24)
Yeah. I think if you took. Airgram, for example, to listen in on your calls. It has an export feature that automatically exports things to Google. Okay. Google Documents and then there is a tool for Google Documents where you can add Chat gpt right into the document.

Mary Green (16:46)
So you would have to do the switching out of the name itself. But it would make the rest of it pretty short.

Adrian Speyer (he/him) (16:55)
What I also found is a data upload feature from Chat, GPT. Recently, we’re doing that survey that you saw, right? Lauren, the swag one. And I’m able to just go take all the results, throw it into chat. You be teams, and say, write me an executive summary.

Adrian Speyer (he/him) (17:10)
It’s amazing.

Kristine Kukich – The Training Sherpa ((she/her)) (17:15)
Absolutely, amazing.

Mary Green (17:17)
Yeah. I did some pretty cool stuff this week. I took the course that I built. On Maven and. Some feedback from the first couple of calls I’ve done, and I had it rewrite the whole syllabus and everything.

Mary Green (17:32)
Based on my course, like a new landing page and stuff. And I was like, Holy Cow, this is so much more. Quality than I thought it would be. But I have not played with any kind of. Database.

Adrian Speyer (he/him) (17:49)
The database stuff is crazy. Obviously with our community too. I don’t put people’s email, but I mean you can put. Their seniority level and other demographic information that’s non identifiable and you can get a picture.

Adrian Speyer (he/him) (18:06)
You would spend a couple of hours with excel formulas trying to figure that out which Chat Chibiki doesn’t in ten minutes.

Kristine Kukich – The Training Sherpa ((she/her)) (18:13)
The visualizations are really good.

Mary Green (18:17)
I’m going to die when I play with that. I don’t know if I’d be able to come down off that cloud.

Cristina Levenetz (18:25)
In the paid version. Yeah.

Adrian Speyer (he/him) (18:28)
It’s worth it, though.

Kristine Kukich – The Training Sherpa ((she/her)) (18:32)
She paid one. Christina just said $20 a month. The $20 a month one includes all of that. The enterprise level again, way more expensive and has a couple of lockable functionalities around it for apis and stuff. But the $20 the month version has.

Mary Green (18:48)
Some pretty cool visualizations and the new stuff that Lauren was just talking about. There are a lot of tools too, christina, that if you can’t pay for chat gpt per month, like some people can’t right now go look on like there’s an AI for that or one of those AI tool websites. And you can a lot of times sign up for something.

18:51 Free AI tools available for data analysis.

Mary Green (19:08)
For free, where they already have the ability to do something, and now you can run your data through it.

Cristina Levenetz (19:17)
Yeah. That’s amazing, man. The applications for that.

Mary Green (19:22)
Wow. It is super cool.

Lauren Turner (19:25)
I used it when we did. Our Large Product survey. So a lot of the survey questions were multiple choice and ratings and things like that. But a lot of them would have follow up open fields. And so it was really great at taking these hundreds of open field responses and then summarizing them and grouping them and finding the themes.

Adrian Speyer (he/him) (19:49)
I’m seeing huge success.

Kristine Kukich – The Training Sherpa ((she/her)) (19:54)
Having for customer education. With using it to build out assessment questions. Right. Like when you’re taking a course and it halfway through, it asks you three comprehension questions.

Lauren Turner (20:06)
Right or Building certification exams.

Kristine Kukich – The Training Sherpa ((she/her)) (20:09)
Yeah to create the questions and particularly to create distractors because as people who have to build those things all the time. That’s just the most hateful job ever. We can go ahead and generate the questions themselves if we want. And then let the system generate the distractors and they come through much more professional. They take way less time. Yeah, absolutely.

Lauren Turner (20:34)
I know there was a big deal when I was at user testing, setting up certain screening tests because you had people who just wanted to take the test. And so in order to keep them from just clicking click and clicking.

Lauren Turner (20:45)
We would throw in fake data, like which of the following have you purchased in the last three weeks? And it would be like Enterprise software. Groceries, a clown tent. And you’re like, okay, if they click on that, they’re clearly just.

Lauren Turner (21:00)
Trying to get in, and it was a good way to weed people out. But you had to basically manually look through all of those responses to be like, okay, this one’s done.

21:11 Potential for enhanced features in ChatGPT.

Adrian Speyer (he/him) (21:11)
I didn’t mean to go on a tangent, though, but in your slide prior you were saying about Chat gbt as being a really good intern. What I’m wondering, though, because with their gpt that they just launched, I mean, technically, and I just did this on actually a website. You could upload.

Adrian Speyer (he/him) (21:27)
A book. Not that I did it, but you could. I know leia did it with her community book, which you could upload a book about ux design principles. And you could say go look at this website and tell me how to improve it. And you could probably get expert level UX.

Lauren Turner (21:43)
Recommendations from it.

Mary Green (21:46)
Or I’ve been playing with it for some hobby type things.

Adrian Speyer (he/him) (21:51)
And able to. Like if you had your own work, if I want to upload my own book. Or obviously you don’t want to take non copyrighted material, right.

Lauren Turner (22:04)

Adrian Speyer (he/him) (22:05)
I do think that it may get a bit better than just a straight intern because if it’s able to go and read books and have guidelines outside of the chat, gpt, web. You were saying, like when you take a case study and you throw it in, if like, Seth godin was to load all his books into it and then create the Seth godin. Ai right. Then you’re able to kind of basically talk with Seth.

Adrian Speyer (he/him) (22:27)
I could not see that happening very shortly.

Kristine Kukich – The Training Sherpa ((she/her)) (22:31)
I think that’s possible. Yeah.

Cristina Levenetz (22:34)
Didn’t they just announce like you can build your own GPT.

Adrian Speyer (he/him) (22:37)
I was playing with it. I’ve made an AI cmo that I’ve loaded a whole bunch of stuff.

Lauren Turner (22:43)
Super cool. And I said, hey, can you look at our website and tell us how you would make it better.

Adrian Speyer (he/him) (22:48)
And then the cmo gave me. It’s okay though, because all the ideas we’ve already covered, but. I was joking with my marketing manager that, hey, we did a good job. Our list covered all the things that were there, but it basically went through it and said.

Adrian Speyer (he/him) (23:03)
Here are some things that I would improve.

Lauren Turner (23:07)
Kind of interesting. I wonder as continues to evolve, what that impact might have on the average tenure of a cmo in SAS. They’re already pretty small.

Mary Green (23:20)
Well, I mean, hey, I’m just thinking like, you want to try it out. Here it is.

Adrian Speyer (he/him) (23:26)
It’s not publicly available.

Lauren Turner (23:29)

Mary Green (23:30)
Cool. Yeah. I mean, there are chap bought like tools you can use now to upload. Tons of documents. And videos from YouTube. You can take a channel from YouTube and start talking to the channel. All of these. I haven’t done all of them. Well, I’m just wearing.

Adrian Speyer (he/him) (23:52)
Sorry kind of crazy. No, go ahead. No. I guess my thought was, do you think that right now people are saying, Treat it like a smart intern. And I’m just wondering if. That as well might.

Lauren Turner (24:05)
Be temporary. Is why I’m getting that. Yeah.

Mary Green (24:10)
I think it is temporary because. They say I’ve heard that it’s just going to keep doing better and better. So I do question, like, ooh. So where does that leave us? But I think there’s a lot still to be said.

Mary Green (24:26)
For the Human interaction.

Adrian Speyer (he/him) (24:29)

Lauren Turner (24:31)
And yeah, we get into that too. But just as a summary. And this is still evolving, too. I almost have to update this daily just to see, to make sure that everything is aligned with the evolution of this tool. But just quick summary. It can analyze data. It can create copy. It can help you with strategy and brainstorming. It can refine your language. And I’ll get into that in a little bit more detail to show how you can.

Lauren Turner (24:57)
Make it fun. What it can’t do is protect your IP on the free version it cannot produce graphics. It can’t really produce it on the paid one either, but it basically can use graphics and help interpret them.

Lauren Turner (25:09)
And give advice based on that. It still right now cannot reliably cite sources. It does have a tendency to hallucinate. More importantly, for customer marketers like us, it is not human. It can approximate human conversations. But it is not going to show real empathy. It’s not going to connect with your customers.

25:26 Connect with customers in a meaningful way.

Lauren Turner (25:30)
In a really person to person meaningful way, it’s not going to pick up on subtext. So if you’re talking to a customer and you might have it on your agenda that you want to really push this new feature that your company is coming out with, and that’s really where you want to guide the conversation. But you’re talking to this customer, and then suddenly they burst into tears because their mom.

Lauren Turner (25:50)
Was just diagnosed with something awful. And. That’s when as a human, you know, you are not talking about your feature that day, like you are talking to this other human being about the crisis that’s going on in their life and the business stuff can come later and the eai isn’t going to be able to pick up on that.

Lauren Turner (26:08)
This is where when you have these great tools, you still need to really fall back on your humanity and what’s going to bring to the table. Something that is never really or hopefully never going to really be duplicated through AI.

Mary Green (26:26)
That’s a good point. Last the other day, my kids and I were stuck in the car for a half hour. And so I started playing with Amazon music. I found wow of the World, or wow to the world, world of wow, which is like a family podcast.

Mary Green (26:44)
And the most recent podcast on There was about these kids turning. A ventriloquist dummy into like using AI. To implant into spantrilliquist dummy. That is so creepy.

Lauren Turner (27:00)
I’m sorry, but like a program.

Mary Green (27:05)
And it was so funny because they were trying to get it to laugh at the right times. And. They struggled with it a lot, but it was really cute because. Of explained AI to kids. In a fun way that even my husband was like, oh, wow, that’s what it can do.

Lauren Turner (27:26)
Interesting. There was a South Park episode where they had funny bot that was supposed to be like. The robotic stand up comic and. It was absolutely terrible. And got booed off the stage. I wonder if they were to update that episode because it was a few years ago.

Lauren Turner (27:45)
Like what the improved AI enhanced funny bot could do.

Adrian Speyer (he/him) (27:50)
That’s kind of where I’m not sure about stuff. Like, for example, I would love to see an AI upload all of George carlin’s material and then have George carlin do a 2024 version.

Lauren Turner (28:03)
The watermelon of the words you can’t say. Yeah.

Adrian Speyer (he/him) (28:10)
There’s a lot of debate on the ethical nature of but then at the same time, it’s pretty cool.

Lauren Turner (28:17)
Yeah. It is also that’s a really good point about.

Mary Green (28:22)
People who are dying and passing away. I can see people try to use this to replicate some of their personal.

Lauren Turner (28:34)
Thoughts and things like that. Be able to talk to it later. I mean, that whole TV show upload is basically that where the person’s uploading their. Into this VR world. And all of the things that happen as a result, and wacky hijinks that ensue as a result.

Mary Green (28:53)
Oh, boy, yeah, I’ve not heard of that. Okay, then. Yeah. And by the way.

Adrian Speyer (he/him) (28:59)
It’s also happening at the moment. I remember there was a story about a woman whose dad had passed away. And I think they had, like all her voice was replicated by eleven labs.

Lauren Turner (29:10)
And there’s some psychologists, actually they’re quite concerned about.

Adrian Speyer (he/him) (29:14)
The potential of people not being able to let go and move on because.

Mary Green (29:18)
Of well, we already have that.

Lauren Turner (29:22)
Yeah, but making it even harder. Well, then you can talk to your AI therapist in order to help with your inability to move on because of AI.

Mary Green (29:32)
Oh boy. Getting a little too dark for a Friday afternoon.

29:41 Using data for career advancement in marketing.

Lauren Turner (29:39)
Sorry. Okay. So we can quickly go through this. I asked Cat gpt okay, if you’re a customer marketing and advocacy person. How can you use this to advance your career? And we’ve already discussed a lot of these. It’ll help with market research. Content Generation helps you sort through a lot of data and get insights from that. It can help you create and refine your customer surveys. So if you’re not a trained researcher, it can help with best practices. So that way.

Lauren Turner (30:07)
You’re getting better data from. Your surveys. It can give you suggestions of software packages that you can use that will help in your job. It can give you ideas for programs and content that would feed those programs helps you optimize your email campaigns. Come up with ideas of ways that you can better engage with your customers so that they can be more adept at using your product and.

Lauren Turner (30:32)
Then eventually drive loyalty. And can also help you with your skills development as well. So if you can figure out where there might be gaps in your own knowledge, it can suggest different courses that you can take or books that you can read in order to help you stay on the cutting edge of the field.

Lauren Turner (30:47)
So what I was looking to do here. Just a second. To kind of give the real world example. So in early September, I launched alice’s Customer Community as a pilot called The Alice gifting gurus. And I used Chat gpt in basically every stage of that development.

Lauren Turner (31:12)
So in order to get approval for the platform itself, we use influid of I had to basically put together a business case that was going to show very, very solidly what the roi of investing in a community platform was going to do. And I used chat gpt with a couple of prompts to help me put together a solid business case. And this is where it was giving me CIT.

Lauren Turner (31:33)
Citations that I had to check to make sure that they were real about the value of community. Typical rois, different use cases. Any quotes from very, very well known business people about the importance of building community.

31:56 Used AI for pitch deck, negotiation, organization.

Lauren Turner (31:49)
Adrian. Actually, one of yours came up in there, too. And then basically used that to build out my pitch deck for the executive team. That eventually got the approval so that we could buy the platform. And then once we were in the negotiating processes with Influid to get the platform.

Lauren Turner (32:13)
It helped me with coming up with the name for the community, giving a series of parameters that we wanted. I wanted alliteration I wanted something that ties into the gifting theme. I wanted something that was easy to abbreviate. So we call it agg.

Lauren Turner (32:28)
Or Gigi if we’re just going to use gifting gurus in our discussions, I then asked it to give me some key talking points. So that way, when I evangelize the community across the different stakeholders of the company.

Lauren Turner (32:42)
To keep in mind what was going to be most important to them and how I can get them on board and bought in so that way everyone feels like they have skin in the game, in the building and success of the program.

Lauren Turner (32:53)
Once I got that pilot set up and we launched it. It helped me organize. The different content that I was going to do. I already had a pretty good idea of how I wanted to organize that, but it was always good to be able to have that kind of double checking of okay, well, here’s the way that I’ve typically done it. Is there anything that you would recommend that would be different? And it did come up with some really interesting suggestions.

Lauren Turner (33:16)
Helped me find different ways that I could both segment and prioritize the different customer groups. And I’d say the most valuable one was I wanted to have a seven email drip campaign so we would send out that initial invitation and then if they didn’t accept a reminder two days later and then a week later and then two weeks later, and then three weeks later, and then six weeks later.

Lauren Turner (33:38)
And I wanted something that all had the same theme, but that didn’t sound too repetitive and could also be as tailored as possible. And it really helped with that. As the community started to really take more shape as we have more people joining, I have to come up with new content every day. I’m usually looking to add somewhere between two and five new challenges every day. And I don’t get a whole lot of input from other folks because they’re all busy. And so this has been really helpful in getting some of those ideas of how I can really.

Lauren Turner (34:09)
Quickly create new content. And it helps me analyze where there might be some issues and what are possible solutions to help fix them. Now, as our community grows, I’m going to continue to use it for troubleshooting helping with those more tailored campaigns so we can drive reengagement for folks who might have dropped off enhanced engagement for people who are baseline active but not really doing.

Lauren Turner (34:32)
Very much and then ways to really help up level the advocacy from people who are active. And as I mentioned previously. It’s been Super Super helpful with case studies. Can you hear my dog? Can you hear molly barking in the background.

Chuck Quigley (34:46)
She’s sleeping a little bit.

Mary Green (34:50)
She’s not distracting. I don’t want to wake her up, but.

Lauren Turner (34:53)
She’s been through a lot. So. Bear with me and her. So we talk a lot about prompts. And I just wanted to highlight a couple of examples of prompts that I’ve used. In general, the more specific that you can get, and the more tailored, the better it understands what you’re asking it to do. And you can also ask it to tell you what additional information it needs.

Lauren Turner (35:17)
In order to generate the output. So when it came to naming the community, this was the prompt that I used for that. I need a name for my new customer community. The community is geared toward admins of Alice. Give me ten suggestions. Have the names short. Have them be related to gifting and use alliteration or puns.

Mary Green (35:34)
Yes. Kyle had a question, and I can’t find the thing to raise my hand as the host, which I don’t know if that just doesn’t exist or what. For things like building a business case or drafting a case study from interview notes, would you consider templatizing your gpt prompts.

Lauren Turner (36:00)
Yes. I have done that. I just have to dig into my files of where they are. One of the things that I do in my role at Alice is I send out a weekly tip Tuesday, email. And once a month I have usually one of.

Lauren Turner (36:19)
Those topics be different. Chat gbt prompts that can be used for different subject headings for gifting or different categories, or different suggestions for gifts for a specific industry or a specific type of recipient.

Lauren Turner (36:33)
So, yeah, I can’t do that right now, but I’m happy to look them up, and I can send them. When I can.

36:42 Simplified contract for better understanding and use.

Mary Green (36:42)
Cool. And just for the last example.

Lauren Turner (36:46)
I also found it really useful for contracts. So for example when we got the contract with influid of. We were looking for a standard tnc that we wanted to use, and the one that they gave us was like 15 pages long.

Lauren Turner (37:02)
And it was really good boilerplate stuff. But it was a lot of text. And it was written. In legalese. And so what was super useful was this prompt where I basically uploaded the whole thing, summarize it for me, write it to a fifth grade reading level while retaining the main points, and then ask me if there are anything else that I need to know before generating it. And then we took that rewritten tnc that was a lot easier to understand. And that’s what we put in front of our customers. So.

Lauren Turner (37:30)
That way it wouldn’t be overwhelming and scary for them to have to agree to the tnc before getting into the community.

Mary Green (37:37)
That’s a good point. I recently was working on terms and conditions for a new community this week. And I didn’t understand part of the terms and conditions that we were looking at. So I put it in chat gpt to have it explain it to me.

Lauren Turner (37:54)
Exactly. And I would love to hear. What some of those conversations with lawyers going to look like because they make a lot of money writing, obscure and difficult to understand. Copy.

Mary Green (38:09)
That’s true.

Lauren Turner (38:10)
I’m doing a lot of things when we built the one.

Kristine Kukich – The Training Sherpa ((she/her)) (38:14)
For the Thought Industries community. I gave the traditional templates of t’s and c’s to our legal team, and they went through it and generated their version of it. And then I sent it to Chat gpt and said, now make this simple.

Kristine Kukich – The Training Sherpa ((she/her)) (38:30)
And then I send it back to Legal. And Can I use this instead.

Mary Green (38:36)
And they said, yes, they made a couple of edits, but they let it slide. Nice. Yeah, it’s super helpful. Works really well, Yeah.

Adrian Speyer (he/him) (38:46)
That’s an industry all its own making. Legalized, human.

38:49 Legalizing human augmentation; beneficial for diverse industries.

Lauren Turner (38:52)
Is a huge thing. I’ve also found it really useful for helping our csm team. So we have customers across different industries that some of them are pretty niche. And what I’ve worked on with them with Chat ept is a prompt that will basically say I’m about to have a call with name of company.

Lauren Turner (39:13)
Give me a brief summary of what the company does, how it makes money and the top three concerns. As an employee of that company, I would have.

Mary Green (39:22)
And it gives you a nice little succinct dossier.

Lauren Turner (39:26)
Of essentially who you’re talking to so you can at least have some kind of baseline level of competence in your conversation with them so that you can get through. Some of the mucky stuff and really get more to the heart of conversations with a voice that’s a lot more credible.

Mary Green (39:43)
Wow. That would be really good for job interviews.

Lauren Turner (39:48)

Cristina Levenetz (39:50)
Oh, I totally use it for Chad gbt for job interviews. I could send you the prompts, Mary. Sure. Yeah, I’d love to. I’d love to see them.

Andru Creighton (40:00)
You need to look at the video of this guy who had no background in mechanical engineering and was interviewing for a job in mechanical engineering. And plugged the whole job description into gpt and was transcribing the call in real time and plugging in the questions that the interviewer was asking him into gpt and answering them in real time with gbt. It was insane.

Andru Creighton (40:25)
I can’t remember what happened. I can’t remember what happened with the outcome of it. But it was one of those videos where it was like, there’s no way this is real. And you can see it like he’s recording himself. The screen and the computer.

Andru Creighton (40:38)
And it’s crazy that just hit the language they’re using, and he’s clearly acting as if he knows what he’s talking about. But it’s amazing. Oh, my God.

Lauren Turner (40:52)
Cardiothoracic surgeon or anything like that. But. Low danger roles. I think it would be very interesting to have someone totally unqualified.

Andru Creighton (41:03)
Relying entirely on chat gbt to do something for a day. Yeah. I think this guy was totally just doing it for internet clout, which was why it was like entertaining and funny, but it was just crazy that if he’s doing it, somebody else is doing it to really actually get the job.

Mary Green (41:17)
It would be so easy to do that too, because. It’s super easy. It would be so easy to do that.

Andru Creighton (41:27)
There you go.

Lauren Turner (41:29)
You should do a course using Chat dpt to overcome imposter syndrome. By learning how to be a real imposter.

Mary Green (41:39)
I’ll take it.

Lauren Turner (41:42)
There you go. So I just wanted to kind of dispel a couple of myths that people have about Chat gpt that you can easily pinpoint something that’s coming out of generative AI because it sounds boring and robotic.

Lauren Turner (41:56)
And I’m just here to basically give this as an example that no, that’s not true. So, yeah, talk to me entirely in Gen Z slang. Be really over the top about it, and then it spits out this, like, oh, my God, yes. Time to slate this combo with some lit Gen Z slang hunty. So spill the deep spam. What’s the four? One, one. We stand attorney chat, you know? So I’ve been basically playing with different types of arch.

42:17 Using ChatGPT to practice different archetype voices.

Lauren Turner (42:21)
Archetype voices with Chat gbt for a while, partially as a way to augment. The studies that I do for improv. So, my improv Team, one of the long form games that we do is an extended film noir. And so I can take big swaths of copy and upload it to chat GPT and then say, how would you say this if you were in 1940s gangster, in a film noir movie.

Lauren Turner (42:46)
And it will basically spit out in that voice. And then it really helps learn. How to speak like that on stage. And I’ve done that with. Mid Atlantic. I’ve done that with heavy Boston accents. New York accents.

Lauren Turner (43:03)
It’s really helpful, even with language translation, where you can paste something and say, okay, now can you spell this out phonetically so that I know how to pronounce it correctly. It’s been really helpful that way.

Lauren Turner (43:16)
What I’m really just trying to get to the point here is. The only limitation that you really can have in what type of material comes out of Chat gbt is your own imagination if you give it enough creativity in the prompt.

Lauren Turner (43:31)
You will get really interesting stuff coming out. I’d love to hear from you if you’ve ever done something like this where you were really trying to be over the top or bring in some kind of interesting new type of voice.

Lauren Turner (43:45)

Mary Green (43:47)
I think you might remember this, but I took our first master class from Ashley Ward. And in Cast Magic. So at the Cast Magic has a community of people. Where they’re trying all these different prompts. They’re coming up with different prompts and then they share them. And then the company puts them into the platform so that you can select them and add them as a template and things like that. And somebody put in summarize this episode.

Mary Green (44:14)
Into. A Doctor Sue Style book. And make it like xyz. Long. And it was so cute. And spot on. I don’t have it now because I moved the episode into another workspace, but oh, my gosh, it was so cute. I used a little bit of it just to send her and be like, hey, I thought this was cute and you’d like to hear it.

Mary Green (44:42)
But Yeah. It’s fun to play with. If you can add that creativity. Definitely.

Lauren Turner (44:51)
I’ve had really great luck with it for various themes, for email campaigns. So I did one that was a beyonce theme, and so it would weave in beyonce song titles into the copy. Then I basically said, once it came out with a really good one, I was like, okay, cool. Now make a Taylor Swift. Okay. Now make it Michael Jackson. And it did. And so it’s great. You don’t need to be an expert on pop culture. You just have to know enough that you can ask for the right prompt. And it’ll basically.

Lauren Turner (45:22)
Plug and play that stuff right in. And it’s great. It really helps with marketing campaigns, too. Especially when you’re looking for ideas and you have to do something on a regular basis. Like My Tip Tuesday. Emails like.

Lauren Turner (45:37)
I’ve done karate Kid themed Stranger Things and basically just listing all the shows that I watch. I did a South Park one. But just ways that you can kind of tie in to the cultural zeitgeist is a nice way that you can capture people’s attention. And you don’t need to be on top of every pop culture thing going on.

Lauren Turner (45:56)
Just know enough for the reference and then it can basically do the rest for you. Anyone know the answer to this.

Mary Green (46:14)
Big Bang theory. Andrew says South Park I say Big Bang theory.

Lauren Turner (46:23)
Any other guesses? Simpsons, Simpsons.

Mary Green (46:27)
The Office right.

Lauren Turner (46:35)
It was South Park. The entire episode was about Chat GPT. How stan was trying to use it to have heart to heart conversations with Wendy. Because want to get into the emotional stuff and then. It basically spiraled from there.

Lauren Turner (46:53)
So just some ideas. We talk about your typical marketing activities just a little bit out there suggestions of ways that you can use chat gpt because it can do that. I’ve had to do rap battles before.

Lauren Turner (47:08)
Content in the style of Adrian. I know you mentioned that before with George carlin. I would love to see it do the George carlin type thing. Pop culture references. You can have it do format things. Formatted like a script. I’ve done essentially movie script for not an entire two hour movie. But if I were going to record a film short for a commercial for some kind of product that we’re doing, can you do that and it’ll actually write a script.

47:14 Create film shorts with pop culture references.

Lauren Turner (47:40)
Which can be really helpful as well. So just some overall tips and tricks that we’ve all kind of discussed before be as detailed as possible. The more you give in, the more tailored the output is going to be to what you’re looking for it to do. You can ask it to ask questions.

Lauren Turner (48:00)
Before generating. So if there’s anything that you may have left out, it can point that out to you so it can help you refine its own prompts. Thinking about tone themes, audience. Are all different ways that you can get.

Lauren Turner (48:14)
Very. Detailed and tailored output based on what you’re looking to do. Protect your IP less of an issue if you have some of these proprietary AI things where everything is kind of part of a closed circuit. But if you’re using Chat, GPT.

Lauren Turner (48:30)
Just alone. You need to check what your company policies are and kind of make the assumption that even though it’s unlikely. Your data can be seen by other people if they’re using the right prompts. And so one workaround that I mentioned previously is if I’m going to be using chat gpt to create a case study for me.

Lauren Turner (48:52)
I take the names of the company and the names of the people out and they just replace it with acme or Joe schmo and then have it write the case study and then I put the names back in later. Iterate and edit. You don’t need to redo every prompt if it gives you five paragraphs and you like four of them, tell it to rewrite the third paragraph in a different way.

49:20 AI can save time, but human touch matters.

Lauren Turner (49:14)
And that way you can still retain a lot of the good stuff without having to start over. And there are a couple of obvious AI tales. This may change as it evolves, but I’ve seen that every time I’ve asked it to write me an email, it starts every flipping email with. I hope this email finds you well.

Lauren Turner (49:33)
So if you happen to see that in any inbound emails that are coming to you from A. Bdr, you can probably guess that they used an AI to generate that.

Mary Green (49:45)
Ultimately, this is a really fantastic time saving tool.

Lauren Turner (49:49)
But it’s not a replacement for the humanity that you bring to the table and the work that you do. So when you think about the time that you can save with this, how are you going to use the rest of that time in a way that only you, as a human can do that is going to bring value to the company and bring value to your life. Does this mean that.

Lauren Turner (50:08)
You now have an additional hour that you can reach out one to one with some of your customers and connect with them better. Does it give you more time to research? Does it give you more time to evangelize your work across your organization? So that way.

Lauren Turner (50:21)
You can better solidify your role in the company. All these different things that you can do that are really going to help advance yourself and advance the practice. Using that extra time that Chat gpt gives you.

Lauren Turner (50:35)
So that’s basically it.

Mary Green (50:40)
From my presentation. That was awesome.

Lauren Turner (50:47)
In the conference. We basically spent the last 20 minutes. I had an assignment for each table. With a scenario like pretend that you are a healthcare company, that. Is trying to drive new business. In this particular subset of customers.

Lauren Turner (51:04)
How would you write a five email campaign in order to help drive traffic? And I would basically have a different assignment for each table. And then they would have 20 minutes to basically just play around and chat GPT iterate and really just kind of get their hands dirty and playing around with it and seeing what they come up with and then just walking around.

Lauren Turner (51:22)
The room, seeing what they were coming up with and helping where people might get stuck. So Ultimately, that’s really what it is. I don’t think anyone. Starts out the gate being an expert on this, I certainly don’t in any way claim to be it’s more just I dedicate the time to messing around with it and learning what I can and then sharing what I learn.

Chuck Quigley (51:43)
Lauren. One thing that I’ve noticed. Since I’ve written a story using Chat BT or created a story using chat gpt. And I’m going through, like, I was on the interview. The transcript, right is what I loaded in. And I’m like, oh, I don’t remember the customer saying that.

Chuck Quigley (52:03)
And even though I put in the transcript, it took what the customer said and created vastly different. Customer quotes. So then I had to go back in I’m like now I always say Use only. Direct quotes taken from the transcript from these people, whoever the customers may be. That was one thing that stood out to me.

Lauren Turner (52:25)
Is surprising. I’d say, yeah, I think so. Long as you’re aware of that, you can also use that to your advantage, because sometimes. The chat gpt output can actually be more eloquent than what your customer actually originally said. And so if.

Lauren Turner (52:41)
It makes them sound better, and they sign off on it all the better.

Chuck Quigley (52:46)
Completely. Right. Yeah.

Adrian Speyer (he/him) (52:49)
That’s how I used it when I’ve used it. Actually the quote was actually more coherent. It summarized, like twelve lines of a quote into two sentences. I was like, It’s really good. And then they signed off on it. Yeah, exactly. Take this ramble and turn it into a usable sound.

Adrian Speyer (he/him) (53:10)
But that also gets to your point where I’ve used. Is it’s also important to go back and look at the original source.

Lauren Turner (53:21)
To make sure.

Adrian Speyer (he/him) (53:22)
It doesn’t make up stories about. Things that happened or that they did. We increased our business 300%. No, they never said that.

Mary Green (53:40)
Let’s see what else. Yeah, this was really great. Lauren. Thank you for sharing.

Lauren Turner (53:47)
For having me.

Mary Green (53:49)
Anybody else have any questions they wanted to talk about. I see caitlin had some requests.

Katlin Hess (53:59)
I was just saying the same thing happened to me yesterday. And the quote it created said, We’ve been orchestrating a symphony of user reviews, and I was like he was a quirky guy, but I don’t think he said it quite like that.

Lauren Turner (54:11)
And so that’s when I was like, I need to double check this.

Katlin Hess (54:15)
And so I found if I added the word like verbatim quotes, it then pulled actual quotes from my transcript. But it made me laugh.

Mary Green (54:27)
That’s a good point. Lauren. Do you have any cheat sheets or anything like Christine? Posted with the personas and context and tasks. Do you have any cheat sheets? That kind of. Summarize. Maybe some of the tones or.

Mary Green (54:44)
Anything like that you would think of.

54:47 Maintain friendly and consistent tone for customers.

Lauren Turner (54:49)
I don’t have any off. Offhand. I tend in a lot of my customer facing stuff. I want to keep the tone friendly and casual. And so that tends to be. More consistent with what I’m asking it to give out. I obviously would be asking for a different type of tone if I were presenting to a community of attorneys or surgeons or morticians, I’m sure they probably don’t want fun and happy and friendly for that.

Lauren Turner (55:18)
But. A lot of it is also what I’ll do for just kind of theater stuff to learn. Okay. If I were trying to master a cajun accent for whatever reason. Give me this block of text and then spell it out phonetically so that I can learn to mimic what that sounds like.

Lauren Turner (55:38)
I’ve done that a few times, but as far as having anything in a kind of neat little like here’s your cheat sheet for how to speak like a Cajun or. How to use the Mid Atlantic fancy accent. I don’t have that. Maybe that’s something over my vacation that I can put together.

Mary Green (55:56)
No, you don’t have to work over your vacation. Not for me. I’ll let it slide this time. I had something before, but. I don’t think I still have it. I tend to you might all think this is funny, but I tend to still have some psychological.

Mary Green (56:16)
Issues with Chad gpt. So one, if I go in and I create content and I keep telling it how to edit it, I feel like I’m wasting content. And then two. My most common. Prompts are things like, how do I say this and make it sound more professional.

Mary Green (56:37)
Or what’s? Like what are alternatives for this work? Give me a longer list of suggestions. So often that I start to think, how stupid are you. This is what you use chat tbt for.

Lauren Turner (56:55)
I actually used it for my review. So part of our performance review is you have to write a self assessment.

Mary Green (57:03)

Lauren Turner (57:05)
Those are hard. It was like, okay. So I wrote what I was going to say, and then I uploaded it. I said, okay. Make this sound as professional as possible. Strip out any of the emotion to it, and find that balance between.

Lauren Turner (57:18)
Matter of factly listing the accomplishments and not sounding like and not sounding braggy and still providing opportunity for growth. And it did. And it was Super, super helpful that way.

57:32 Custom instructions feature enhances ChatGPT’s Premium version.

Adrian Speyer (he/him) (57:33)
I was just going to say. If you’re on the premium version of Chat, you can. I don’t know if it’s in the free one, so I’m sorry if it’s not, but there’s a function where you can put custom instructions in the settings, which you can tell it how you want it. So instead of you always having to tell it what you need.

Lauren Turner (57:48)
You can. Say.

Adrian Speyer (he/him) (57:53)
Talk. Or you are a and you’ve confined. I mean, there’s a whole bunch of people you can find online YouTube videos and things so you can set it up already to. Be which I know it’s scary. Maybe for some content writers, but you could say.

Adrian Speyer (he/him) (58:07)
I am a content writer who likes the Oxford. Comma. And you can kind of build out the whole thing, right? And use data and use statistics. And you kind of give it all the instruction. Then what would ever prompt you do it will use that custom instruction.

Lauren Turner (58:24)
But you can turn it on or off from the settings and chat ept if you so decide.

Mary Green (58:29)
Yeah. Somebody’s going to come up with a chat gpt that whenever you do multiple prompts like that, it’s going to be like, how dumb are you? Do you want to go back to kindergarten? And people will love it.

Mary Green (58:42)
Because. That’s just the society we have today. Yeah. I mean, I bet you it’s already out there. If you scratch a bit, I’m sure someone’s ready. Yeah. I called my husband at one point. I’m like, we should write it like, GPS, like, create a GPS app that works. But when you screw up, it’s like, Is this why your parents don’t like you.

Adrian Speyer (he/him) (59:03)
Why it was actually kind of joining my wife, because I’m trying to lose weight. So it’s like, do you really need that? Donut. Adrian.


You may also like

CMAweekly is a community for Customer Marketing, Advocacy, and Community pros.

Subscribe for the weekly newsletter

     2024 All Right Reserved