BACKSTROKE: FROM SELLING HOURS TO SELLING OUTCOMES WITH RJ TALYOR

E103 BACKSTROKE_ FROM SELLING HOURS TO SELLING OUTCOMES WITH RJ TALYOR

Transcript

Dalton Anderson (00:00.942) Welcome to VentureStep podcast where we discuss entrepreneurship and suit trends and the occasional book review. What happens when AI can generate creative variations in seconds that used to take a team of five people a week to produce? The agency landscape is shifting from selling hours to selling outcomes. We're exploring the new initiatives that are allowing firms to slash internal costs while increasing revenue for their clients. Today on the show, we have RJ Tallier.

the CEO and founder of Backstroke. RJ is a 2X founder with 20 years of experience in digital marketing and the marketing tech space. He is a early leader at Extract Target, which was acquired by Salesforce, later founded Pattern 89, a creative AI platform that was successfully acquired by Shutterstock. And he's now at the helm of Backstroke using his deep industry knowledge to

help firms evolve their digital marketing efforts and eliminate the overhead that traditionally bogs down creative production. Welcome to the show, RJ.

R.J. Talyor (01:07.186) Hey, thanks, Tom, appreciate it. Glad to be here.

Dalton Anderson (01:12.014) I think we'll just jump right in on the show. The first thing that we'd like to dive in on and definitely maybe informative to the group and the listeners is just maybe an overview of a couple minutes from your perspective of how has this workflow shifted in the last, I would say like an overview of like the last five versus like the last year and a half, how it has changed.

R.J. Talyor (01:40.896) Yeah, well, mean, digital marketing agencies have, I mean, it's been like a wild, wild ride in the last 10 years, really. And, you know, in the, or like in 2015-ish, 2016, all these machine learning algorithms started emerging that really impacted the digital ad buying and bidding.

process and a lot of these digital marketing agencies made a lot of money on kind of the managing your ad spend on platforms like Facebook or I guess now Meta as well as Google and you know that's kind of bread and butter for agencies.

And then there were new solutions that came out from Google and Facebook that allowed you to automate a lot of that. like humans that were running spreadsheets no longer really had to do that. So then it became those folks started saying, all right, let's focus on creative. And instead of on optimizing for cost per click or CPM or cost per outcome.

Let's optimize the creative. These same digital marketing agencies do everything from SEO to your retention marketing to your website to positioning, to strategy, all of it. But I really think that the big shift happened with the bidding and the optimization of those ad spends.

And then, you know, in the last five years with the emergence of GPTs, know, for like GPT-3 was kind of the first one that went mainstream about what, three years ago. Copywriters started saying, wait, the machine can write the copy instead of me. Now the human creative process has interrupted. And then those LLMs started getting better and better. And now we have platforms that people can use to create.

R.J. Talyor (03:38.43) you know, replicate photo shoots. like, you know, Nano Banana or other similar platforms allow you to create content that is remarkably good. And, for the last like three to five years, all the creatives have been saying, well, Hey, the optimization people who have been interrupted by these machine learning algorithms, their jobs are in peril, but my job isn't because I'm the creative. I'm the person that creates all this content. I'm the photographer. can't replace that.

But now that's a question as these models have gone from showing six fingers, sausage fingers, weird situations to being very, very believable and on brand and beautiful in addition to performance.

Dalton Anderson (04:13.474) Hmm.

R.J. Talyor (04:23.185) It started with kind of the numbers and the data crunching and then emerged into the words that marketers use. And now it's into the creative assets. And, you know, I think that if you asked any agency at any point in that, in the last 20 years, what's your differentiation, they would say strategy. They would say, Hey, we're really strategists. We understand your brand. understand it, et cetera. But now agencies really have to deliver on that.

because the creative production, the optimization of ad spend, et cetera, is all really using, you're using the same set of tools likely that your competitors are using. And the true differentiation comes in how you apply that judiciously, taste, brand, positioning.

those types of things which are very strategic and ultimately human. So I'm not sure I did the like all of it, but that's how I see it. It's like a pretty stark adjustment in the last 10 years and that creative, like the insertion into the creative process, I think is accelerating at the fastest pace today as we talk.

Dalton Anderson (05:31.736) No, I think it's a great overview and just double clicking into the nano banana comment. I actually did a podcast, a demo of different things that you could do with the nano banana model. It's not, that's not actually a name, but that was the moniker that caught on as a joke. And then Google renamed it. think it, people were calling it nano banana on, on tick tock went viral and then Google was like, okay, fine. I'll change the name to nano banana because everyone knows this. but I was taking,

R.J. Talyor (05:43.966) Yeah.

R.J. Talyor (05:47.347) Yeah.

R.J. Talyor (05:54.451) Mm-hmm.

Dalton Anderson (06:03.061) models that had their like they had, you know, say a tan shirt and like jeans. And then I was just clipping different outfits off of Pinterest and then say, Hey, I need I need you to switch the clothes that this model is wearing to this outfit. And I want to change the background from like a tan balanced. I know with lighting, I want to be like more aggressive and and like, whatever, you know, I'm not I'm not a photographer, but

R.J. Talyor (06:08.019) Mm-hmm.

R.J. Talyor (06:12.915) Yeah.

R.J. Talyor (06:18.847) Right, yes.

R.J. Talyor (06:30.452) Yeah.

Dalton Anderson (06:32.287) I tried my best and the results were legit. They were more legit than I would expect. Like to do that in like 30 seconds without having, mean, people can, I mean, I would be, if I was a model, I'd be upset like, okay, like I was paid for this and like you're doing these things to me, like altering the photo. Like you didn't pay me to do that. So I would be upset about that, but just, I'm just talking about the aspect of the model and being able to use it to,

R.J. Talyor (06:50.132) Yeah.

Dalton Anderson (07:02.381) basically change or completely alter what the offering was in 20 seconds. And I'm not even an expert. So if you're an expert, you could really get a lot of power and use out of that.

R.J. Talyor (07:06.963) Yes.

R.J. Talyor (07:14.911) Yes, yeah, there's some remarkable things that you can do with that technology and there's lots of implications there. I mean, certainly the thousands and thousands of dollars that are spent on photo shoots can be replaced in some cases by doing exactly what you're talking about, like replicating a pose or a...

a shot with a different set of imagery. And you know, it's interesting because, and then you can also animate that, you can change up backgrounds, you can change out things. There's brand implications to that, there are ethical implications to it.

who's getting paid, who's not getting paid, what's in the contract, et cetera. There's also legal things to consider because in New York, and I believe one more state in June, there is now going to be a, or there's a law that you have to disclose that the image was generated if it includes a human with a disclosure there. So I'm not sure how they're going to police that or find people for it or whatever, but there's lots and lots of implications there, but on the back to like how

brands or agencies are thinking about it is we can do it more cheaply. We can do it with more inclusive model sets, including like people of different skin tones, body types, et cetera, without having to pay for those human models to be there. can show the clothes in different applications or the products in different applications in a way that we've never been able to do. you know, in some cool ways it'll allow smaller brands to compete with

larger brands who have those larger budgets and look bigger than they are, et cetera. you know, there is a human consequence in that the model is not being paid, the photographer is not being paid, the lighting person is not being paid, the person who provides all the snacks is or coordination of it all. mean, like those are jobs. So.

R.J. Talyor (09:13.597) What will those people do instead? And how will they differentiate? I think it will just become more and more competitive. And then there's this backlash. Consumers tend to say, I don't want to be lied to. And maybe they'll distrust brands that they perceive as AI. But I've actually been tracking that. Most consumers can't tell the difference between an AI-generated image and a

one that's an actual image. there's a lot of like skepticism and doubt and whatever that enters into the picture as well. So it's a it's a funny world, Alton, like just a funny like the place where no one actually has the right answer. And it's kind of looking on the sidelines, trying stuff, trying to stay out of trouble. And they don't want to be the example that everybody points to as the bad actor.

Dalton Anderson (09:44.109) Yeah, you can't tell nowadays.

R.J. Talyor (10:10.043) in space, yeah, certainly technology is cool.

Dalton Anderson (10:10.863) Getting there on the video side too, RJ, or we're getting to the video eventually, I think maybe a year from now, we might have a video that is pretty hard to tell the difference between like what was actually filmed versus like what is AI generated. And that's a whole different topic. I know that you talked about, because that could be an episode in itself, just talking about the, the ethical implications, but, and just how, does the world react? But you talked about how there's just a massive shift ongoing.

R.J. Talyor (10:25.341) Yes.

Yeah, totally.

Dalton Anderson (10:39.797) in this space. And there's a lot less manual work than there used to be. Like, you don't necessarily need a copyright team that is is churning out as many articles or content creation. You don't necessarily need as many people doing Photoshop or you don't need as many photographers or you don't need a pipeline of models with different body types and skin tones. So you get to this point, like what you'd mentioned, like these are

R.J. Talyor (10:41.406) Mm-hmm.

Dalton Anderson (11:09.301) actual jobs that people are, they're actually good at this, this niche in the industry. What, what would you suggest like as a, as a skill shift for them?

R.J. Talyor (11:21.491) Well, I think the need for that imagery is the same. Marketers want it, so become an expert in the latest tools to serve those industries. That's my, that would be my play. You know, not everybody can.

maybe jump up and become well skilled on it. But I actually would argue that he could because it's so early. If you can become the expert in how to apply nano banana, you know, filters and technology, or even as you're describing, like the video stuff is just emerging, dig in and be the person that experiments and drives the difference there. Because I do see people kind of falling into three camps. One is people,

in the industry saying, hey, this model is evolving and we're jumping or this business model evolving, we're jumping forward. We're going to be the early adopters. I see people saying, no, I've been in this industry for 20 years like me. And this is another flash in the pan. And I'm going to keep doing what I'm doing. And I'm just going to hold on.

And then I see maybe the biggest group of people saying like, I'm frozen, I'm paralyzed, I'm not sure what to do. I'm just gonna kind of watch everything and then let somebody else tell me what to do. And I think that, you know, those are all options. But what I would recommend is like dive in, play, try, experiment. Because if instead of, if I could be a photographer who also knows, you know, if I'm a photographer.

and I know how to do Photoshop, like let's rewind 10 years ago, that's a very valuable photographer.

R.J. Talyor (12:56.479) you know, because I can do photography as well as the editing. And if I can do photography, Photoshop to edit things, and also being expert in some of these generative tools, you got to trifecta there. So it's kind of just a constant evolution. And some people are saying, no, no, no, I'm a photographer only, I'm a purist. I'm only going to be a photographer. I'm not going to do any of the production work or the after the production work.

Dalton Anderson (13:21.715) I guess, post-production, I guess.

R.J. Talyor (13:23.347) Yeah, post production. it's like, well, OK. I mean, that is a choice if you want to be the best photographer. But I would be expanding my toolkit across all these technologies to take advantage of that. So that's what I would do.

Dalton Anderson (13:42.097) That's fair. The waves coming with or without you. I had a conversation with my friend a couple of years ago. He was kind of freaking out. And he was like, I don't know what to do. Like, this is kind of, this is scary. Like this whole thing, like five years, I don't want to be a doomer, but like five years from now, like my industry is going to like, should probably completely change. And like, I don't necessarily know what, what I should do. I was like, either ride the wave or you don't like.

R.J. Talyor (13:54.91) Yeah.

R.J. Talyor (14:01.342) Mm-hmm.

R.J. Talyor (14:09.459) Yeah.

Dalton Anderson (14:09.516) It's coming. So it's going to, it's going to either, either react to it or you re you, you write it or you, you watch it happen either or speaking, speaking of waves. So what, what are the waves that backstroke is making? So can you talk about like how there, there was two really cool things that you talked about when we had met prior was like, there's this proprietary data set that you use to help drive the

R.J. Talyor (14:16.435) Yeah.

Yep.

R.J. Talyor (14:24.639) Thank

Dalton Anderson (14:38.704) of these JNI assets that is different than like if I was just doing it myself, like if I just did it myself, I would not get the same results as what you would get at Backstroom using the product.

R.J. Talyor (14:39.934) Mm-hmm.

R.J. Talyor (14:52.637) Yeah, yeah, so we built this data set. We started the business about a year and a half ago to generate marketers email campaigns.

But about four and a half years ago, I started signing up for every email marketing program I could find. like literally, I've signed up for thousands and tens of thousands at this point. I know we've automated a lot of it, but the idea was to amass a giant data set that we could understand things like what types of content are people sending, are brands sending, what types of offers are they offering at what times of day or what times of week or what times of year.

Dalton Anderson (15:09.84) Hmm.

Dalton Anderson (15:13.754) Brutal.

R.J. Talyor (15:34.214) How are they describing those either visually or in text? What colors, what fonts, what is that pattern month over month, year over year, et cetera? then we use all of that data as a trading set so that when Backstrip generates content, we generate content that is aligned with the day or the time of year or the cycle that the brand is in, not just generating content off the bat.

The other thing that we did was we started

building up our own data set of consumer preferences. And so somebody who lives in New York, who is of a certain gender, who has certain income and certain age, has lots of different preferences than someone who might look similarly but live in Indianapolis, where I live. And we've captured that through surveys and research on a quarterly basis and use that to take in combination with our data set.

generative content that is not only predictive of trends, also predictive of performance. So yeah, it's a cool data set. We're still digging into it to find new ways to do cool stuff. But that was the base for Backstroke and what sets us apart in the landscape.

Dalton Anderson (16:57.616) The, the aspect of that is, an interesting one where you were able to get a lot of times it's, it is like you have an idea, you do the idea, but I've seen it over the years. The last, I don't know, he used to call it last four, four and a half years. Like there's these companies that come out and then it gets a little bit of traction. And then there's a massive company that's like, okay, like we're going to do that. And it's, and then it's like this

R.J. Talyor (17:25.15) Yeah.

Dalton Anderson (17:26.788) this question of like, is it a feature or a product of what is being built? But if you have this extra piece of guidance in this data set and this additional data that helps curate and optimize the output of what is being generated, then that's clearly a product versus a feature. But if it was like, okay, like we are using NanoBanana just to generate this content, then anyone can really do it, but without...

R.J. Talyor (17:30.846) Yeah.

R.J. Talyor (17:43.092) Yeah.

Dalton Anderson (17:54.533) without that data, it's not going to be nearly as good. And, and the guidance that you're able to provide by like individual demographic and area, geographic area is quite interesting. How, how does, how does this get generated for like on a client side? Like if I were to contract with, with Backstreet, how does that work? I'm just kind of curious. Cause it, I think about trying to automate like my workflow to where I can

R.J. Talyor (17:57.736) Yeah.

R.J. Talyor (18:16.743) Yeah, so when you start with

Dalton Anderson (18:24.354) maybe provide you my mailing list or something like that or my target customer. And then from there, like here's my current click through late and here's like my click through to purchase. And then I need to optimize this. Like it's too low. Like how do I, how do I bring this up a couple percent or half a percent?

R.J. Talyor (18:26.707) Mm-hmm.

R.J. Talyor (18:46.355) Yeah. So we would start by connecting into your messaging platform and taking those subscribers. Let's say you have a million subscribers on your list. We'll profile them. We use a few different data services to help understand who's on that list. Are they men or women? they certain ages? Are they certain demographics? And then we'll also look at the performance of those audiences to understand what's working by

by demographic or by kind of sub-segment or.

And then we'll compare that to our data set to say, hey, here's what's different. Here's over indexing, under indexing, et cetera. And that then serves as the base model. We'll also ingest all of the creative that you've used in the past, as well as your brand standards to understand kind of what your creative vibe is. And then by generating content for each of those audiences or sub audiences, we provide predictive content generation. And the results

is that we always end up tailoring the content more to the individuals and when people get more of what they want they tend to open click and convert at higher rate so that's how we do what we do with Backstroke and it's kind of what

normal marketing team or agency would do, but the difference is that we're using AI to identify patterns that humans are incapable of finding and we're using a massive data set that hasn't existed before. So it's, know, like seven steps ahead of what you could do with a human counterpart. And, you know, it feels like we're just kind of getting started because ultimately what

R.J. Talyor (20:32.38) people want or what marketers want is to individualize messaging so all million people in your list might get a different message. That's where we're headed. But the problem with that today is trust and how do marketers feel about a million different variants of things that they haven't seen or QA'd with their own human eyes.

So that's, know, getting from here to there is what the mission is for Backstroke in the next two years.

Dalton Anderson (21:03.889) Yeah, that's a pretty interesting thing that it's a basically I would describe it. I mean, over some flying it, but a firm based model. Like it's a model based on the firm and it's an individualized firm based model that's built on like a large data set of either their stuff plus a little bit of like your guys's proprietary data create this merged firm based model view. And then it's an intakes all of

R.J. Talyor (21:13.042) Mm-hmm. Mm-hmm.

Dalton Anderson (21:33.682) their company's mailing lists and there's some data connections in the background, et cetera, et cetera. And then from there, the output is like, Hey, we, can provide better marketing campaigns than you could provide. Maybe you could still do it, but it would take you a lot longer. And also we can move much faster than you would be able to move if it was just humans. And then, then the humans would still be able to review the content before.

R.J. Talyor (21:56.732) Yes.

Dalton Anderson (22:00.977) it goes out. But it's an interesting thing that you were talking about, because like there would be like human in the loop at the end, probably right at, at this stage of like the world and technology. But then you were talking about a little bit of like in the future, there would be a want and or requirement to individualize messaging for say in this example, a million people, which one is like super cool. And I think email lists are

R.J. Talyor (22:05.576) Mm-hmm. Yeah.

R.J. Talyor (22:27.475) Mm-hmm.

Dalton Anderson (22:32.657) I don't know if they're completely slept on, but I think they're under optimized, right? And like this, this company Backstroke helps optimize this treasure trove of potential customers. think that people have email marketing lists, but like they never seem like they are customized to that person. And I'm subscribed to, and I'm a weirdo like yourself, like I'll subscribe to companies or like, okay, they've got a cool brand messaging.

R.J. Talyor (22:37.491) Yes.

Dalton Anderson (23:00.119) I'm so I mess up the data like you definitely mess up the data when you're subscribing to all those email marketing lists and she's like, okay, well, I've got all these people that I'm interested in, but you're not necessarily purchasing stuff. You just want to look at their stuff. I do the same thing, but if you're not a weirdo, you're not doing that. And so you, when you're subscribed to the company, you most likely are interested in their products. And it's just that.

R.J. Talyor (23:12.508) Right.

R.J. Talyor (23:24.755) Yes.

Dalton Anderson (23:27.513) the message isn't capturing your attention and or the call to action isn't personalized enough to you at that moment. Or you're catching me at the wrong time. Like there's a whole bunch of stuff that goes into that. Like, like when are people typically, if you get data of like winter, people typically on their phone, like within that, within that, mean, sometimes you can get a little like, like big brother, but like if you had, if you had data on

R.J. Talyor (23:37.16) Right? Yep.

R.J. Talyor (23:53.16) Yeah.

Dalton Anderson (23:55.538) Okay, this is the stuff that they'd want, but you know, currently we're sending it when they're the most busy. Like you wouldn't know if they're in a whole bunch of meetings, but like you would be able to know like when they're typically eventually I would think when, when are people most active on social media or when are they most active on their phone? And then that's when you should probably send it a personalized message, but you could also send a personalized message, but then they're just super busy and they don't, don't look at it in time.

R.J. Talyor (24:13.469) Yeah.

Yeah.

R.J. Talyor (24:23.314) Yeah, yeah. You know, yeah, it's, it's, there's a balance between personalization and being creepy. consumers tell us over and over and over again that they want personalized content, you know, like they do want it.

As long as it is like personalized and not creepy and then the challenge is that brands don't have enough time to create all that content or they haven't had enough create to create that content so like Dalton you might want Over-the-ear headphones and I want air pods, you know so like now I've got to create two versions of a campaign that go out the door and You know, maybe somebody else on the list wants You know those air those earbuds that go or that the headphones that go behind your head. So if I'm like best buy

or an electronics retailer and I want to sell earphones, now I have to create three different variations of my email and I have to have three different photo shoots in order to account for that and you know now you just go down... right all of it. So you know our teams are not currently set up to execute with the current tool set all of those different content campaigns but

Dalton Anderson (25:27.025) Three different approvals, internal approvals as well.

R.J. Talyor (25:45.087) The tools have changed and that's where like Genitive allows us to do something totally different. And that's sort of the foundation of Backstroke, which is pretty fun and why we're digging into this area because the opportunity to personalize is there. The technology is now there and we are working with leading brands today to solve that problem, which is pretty fun.

Dalton Anderson (26:09.585) I find it incredibly interesting the opportunity that you are attacking with Baxter. just like the whole concept is quite cool. Like there's currently a problem being solved, but then it scales to like this thing that like you wouldn't even, you wouldn't really even be able to, you couldn't really understand like you can understand the direction, but you don't necessarily understand like how it would happen. know, like how does that, how do you, how do you have

R.J. Talyor (26:18.376) Yeah.

Dalton Anderson (26:37.713) 1 million individualized email campaigns. curious on the now is when you contract with a firm and you have this firm built model and it has say maybe there's some kind of testing phase and it's okay like it has our brand image slash language dialed or dialed in. We're good with it.

R.J. Talyor (27:04.158) Mm-hmm.

Dalton Anderson (27:07.025) Does that speed up the ability for people to iterate? And then is there from like curation to out the door is much faster given that there isn't that many moving parts. And then also I think if that stuff is dialed, like I said, then there's just less time that needs to taken to like go through the whole approval process.

R.J. Talyor (27:33.567) Yes, is the answer to that question. is once the we have a brand configuration, sometimes we call it the brand brain. But you you kind of brief in that part of the product once and as long as it knows your fonts, colors, tone.

Dalton Anderson (27:43.536) Hmm.

R.J. Talyor (27:53.829) style tastes that are all included in your brand guide, then it can create variants that are all on brand. And suddenly you as a human, like as a creative director or you as a content strategist, get to review content instead of creating the content. And

know, backstroke and these AI tools are great for lots of things, not everything. I think that human creativity is still very much required. come from a creative, I have a master's in creative writing. So I come from a very creative background and I believe in the power of creativity. But it's kind of that whole thing I was talking about earlier, like brand or agencies and brand teams would say like, oh, I'm really good at strategy. I'm really good at taste. Like I have good taste. I'm a curator of content like that type of stuff, but

we spend all of our time and actually a lot of agencies cost out like what you buy is hours. You buy hours of production time or hour or you buy.

emails created, you don't buy like in the strategy a lot of times just included because it's hard to quantify how much, how many strategies you get or how much strategy you get, you know, it's like, how much time did I think about you? You know, those are the things that are always, it's like tough to quantify, you know? Um, so instead they've gone to output and I think that the output is going to be the byproduct of that strategy and what you're to end up buying and be interested in. we as humans have to.

Dalton Anderson (29:13.266) Mmm.

R.J. Talyor (29:29.088) really be good at ideas, creativity, strategy, taste, curation, those types of things which are all like softer skills and harder to quantify in a contract. And it's also harder to describe on a resume, for example, and put on your LinkedIn page or put into a portfolio. I mean you can kind of do it, but...

Dalton Anderson (29:44.338) Mm.

R.J. Talyor (29:58.527) kind of not, so I don't know. No, I'm just off on the side down here, but it's like if we were used to charging by the hour or charging by the product and instead we need to charge by the gigawatt of thought, I'm not sure how you do that, but that's what ultimately is gonna differentiate the winners from the average.

Dalton Anderson (30:23.994) And you're not the only one thinking about that. I do know in legal, typically charged hour, debilitable hours, emails, research, all that stuff. But with, you know, recent models and updates and, you know, startups that are specializing in law, a lot of that research and verification piece that that is that the bulk of the work is

R.J. Talyor (30:29.982) Yeah.

R.J. Talyor (30:35.517) Yeah.

Dalton Anderson (30:51.184) kind of, I wouldn't say completely fleshed out, right? Like not there yet, but it's getting to the point in the coming years that it's, it's, it's built by case, not necessarily case size, not a, how many hours do I work on that case, which is a big shift. Like a lot of people probably don't want to do that because well, it's pretty profitable to just charge like, okay,

R.J. Talyor (30:56.083) You

R.J. Talyor (31:03.261) Yeah.

R.J. Talyor (31:07.72) Yeah, yeah.

Dalton Anderson (31:17.34) My minimum billing is 30 minutes. So I wrote an email, took me five minutes to write that email on billing 30. And then, you just get all this extra time that you're, billing, you're billing 60, but you're working, you know, in this example, 40 or something.

R.J. Talyor (31:27.581) Yeah.

R.J. Talyor (31:34.451) Yeah, yeah. know, I spent last week I was at this event on the West Coast that was put on. was an executive retreat with about 50 e-commerce executive leaders from brands that you would know. was put on by Commerce Next. And the number one thing that all of those executives were talking about was time efficiency.

and how much time can we save with AI and how much time can we put back onto our teams. No one was saying, let's get rid of people. Instead, they were saying, hey, our smart humans are spending so much time doing mundane tasks like renaming files or cropping photos or retouching things or pulling lists or creating a report.

And one of those executives said that their company had a certain number of hours saved. They were tracking on a dashboard that they review in their weekly exec team meeting. And they had a specific goal of X thousand hours that they wanted to save in the next year. it's, you know, it's the time saved piece is I think critical on the executives mind in these econ brands that we work with or retail brands.

And what they want to do is say, hey, you have that time that you weren't renaming photo files, go and do something human. Go do something that differentiates our brand. Think. And I was really encouraged by that because maybe the answer is great if we can get, you know, half the time that we only need half the person. But they're like, no, no, no, we want the person. We want the person thinking, not.

Dalton Anderson (33:04.786) Mm-hmm.

R.J. Talyor (33:22.91) doing mundane, repetitive work. And, you know, that's kind of encouraging to me. And so kind of going back to the question you asked earlier about like, what would you recommend to those people? It's like, well, get really good at these AI things because what your company actually wants is you to be thinking and creating and strategizing and those types of things. And we've been held back a lot by just like the TPS report that's due on Friday that now we can automate.

Dalton Anderson (33:51.782) Sorry, what does TPS stand for?

R.J. Talyor (33:56.729) I'm showing my age. That's really funny. So I'm 47 and there's a movie called The Office. Do know The Office? Yeah, like, or, but no, not that way. It's Office, The Office, right? Isn't it like the,

Dalton Anderson (34:02.607) I know.

Dalton Anderson (34:10.628) Yeah, yeah. I haven't watched the show, but the office, yeah.

Dalton Anderson (34:17.872) Yeah, the office,

R.J. Talyor (34:19.248) Office Space. No, it's Office Space is the movie. Sorry, not the office, show. It's the movie called Office Space and the guy is famous for kind of going nuts for because the boss keeps asking him for his TPS report and nobody knows what TPS means. It's just like a stupid like I got my TPS report due on Friday. So I'm only laughing because I'm old. TPS report. Office Space. It's called Office Space. Yeah.

Dalton Anderson (34:21.979) okay.

Dalton Anderson (34:42.502) I'll look it up, so it's called the office space. Yeah, I'll watch the movie.

R.J. Talyor (34:47.812) It's, Dalton, you're welcome. You're going to love it. It's really funny. that's a great, I mean, it's like a TPS reports. It's funny.

Dalton Anderson (34:57.83) So it's just a thing that you need to do like cross off, cross off your list. I thought it was some kind of marketing thing. So I was like, okay, I want to, I don't know that. And I don't think the people on the show know it's, that's why I was asking, but if it's a movie referenced and yeah, maybe, maybe I'm the only one out on this one, you know, like

R.J. Talyor (34:59.964) It's just the dumb report that's due every week that is the DPS report. I don't know what it stands, I don't think it stands for anything, it's just what they call it. No, yeah.

R.J. Talyor (35:17.72) I think it's a generational. I'm not sure how old you are, but I'm older than you. it's... Yeah, think hopefully there's another person in my age bracket who's seen this who's like...

Dalton Anderson (35:25.01) You have 27.

Dalton Anderson (35:33.654) I think there for sure is when I look at the analytics there is so TPS is worth it. Somebody's galing this whole situation inside joke.

R.J. Talyor (35:36.712) CPS report.

R.J. Talyor (35:41.008) Mmm.

Dalton Anderson (35:43.664) I think those are good points where you drop your coffee.

R.J. Talyor (35:47.262) No, I just dropped my coffee. No, not the coffee. Luckily.

Dalton Anderson (35:49.331) okay, because I would have been bad. The I think you made some good points about like time saved and then transition that time saved to higher value tasks is how I would phrase it. I think that's the essence of a lot of larger companies. Like there's just they just have antiquated systems. Things are old. There's got all this manual process, got this manual overhead. You call it tech debt or whatever it may be. But people have to do a whole bunch of manual stuff. Like typically you

R.J. Talyor (36:17.681) Mm-hmm.

Dalton Anderson (36:18.77) you wouldn't think that there's this much manual stuff, like people's, there's people's jobs where like 7 % of their jobs, 60 % of their job is just manual work. They're good employees. They're great, but they're just, they're just stuck in that, in that role until the company transitions to technology out and like transitions like the current technology out to some kind of new, new platform. But that's always kind of last on the list.

R.J. Talyor (36:34.331) Yeah.

Dalton Anderson (36:47.622) potentially because there's other initiatives that seemingly always take priority. So I think, you know, these kind of offerings and, and opportunities with, with AI in the future, you can build things that can automate tasks that are manual, that free up good talent. And of course, if you don't, if you don't have the right talent, your company, then you can go a different direction. But if you already have good talent, and then you want to keep your good talent, good talent, hard to hard to find.

R.J. Talyor (37:06.183) Yes.

R.J. Talyor (37:17.404) Yes.

Dalton Anderson (37:17.68) And then the other other counterpoint is if you're just starting something, you need less people to get started. So like you can be a little bit more lean. Whereas before, like if you needed if you needed to do something and it's either we have got to do this vendor thing that's going to cost 50 plus grand a year or I've got to hire a whole bunch of people to like do this or do some outsourcing piece because I can't afford to like hire all these people full time. Well, you could.

R.J. Talyor (37:23.964) Yeah.

Dalton Anderson (37:45.116) vibe code some kind of contraption to help you out. And maybe you need a third or a half of what resources you'd need prior.

R.J. Talyor (37:52.402) Yes, totally. Yeah, I mean, we think about that at Backstroke. It's how do we build with a smaller team? And you have to think AI first, which is tricky. Again, like I'm...

of a generation I've built, you know, software companies my whole life. And it's not been that way. You don't think, or I've had to like catch myself and say, no, no, no, we need to do that AI first, which means can we code it with cursor instead of asking a software engineer to do it? And the software engineer becomes a QA person or how do we build a QA agent to do the QA and then have a software engineer who understands the whole system oversee it? Or how do we have a security expert or somebody who's familiar with SOC 2 and security data privacy?

oversee it rather than do. And so it's like this same shift and like, how do you like, not how do do less, but how do you Yeah, it works, Mario. It's like how do you actually

Dalton Anderson (38:48.339) How you work smarter. I mean, that's saying is always that saying is always like, work smarter, not harder. But now it's like legitimately you need to work smarter. Like you need to you need to use the tools that you have available. And I think that starts with having a culture at your company that is accepted, like it's accepted to use AI to solve problems. Whereas like other companies from either experience or from anecdotal information from other people, it's like, OK, well.

R.J. Talyor (38:55.207) Yeah. Yeah.

Yes.

R.J. Talyor (39:08.21) Yes.

Dalton Anderson (39:17.459) We're going to want you using AI because you need to be doing the work. We don't want you to be using some other thing to do work. You need to do it.

R.J. Talyor (39:21.447) Yeah.

R.J. Talyor (39:25.349) Yeah, I mean, it's like, you know, we don't need another analogy, but here's one. It's like someone saying, Hey, go build this house and you can either use a hammer and like a manual saw or we'll give you this auto nail gun and, you know, a chainsaw.

And you're like, well, I'll use the nail gun and the chainsaw. And you're like, well, but like the nail gun might explode. We're not sure. And like the chainsaw might cut your arm off because it might have power failure and it might cut your arm off too. So like, just be careful because we're not sure. That's kind of what we're saying right now with these AI tools. like, here, like use this. But like, we're not sure of all the impact or the consequence yet, but like you should use it. And so it's like...

It's really, there's risk in it. And it'll probably build the house, you know, probably no one's gonna get their arm cut off, whatever, but like, when something happens, everyone's like, see, I told you. So it's like, we're in this like, transitionary space where everybody knows that they should work smarter, but they don't want to be the one who has the chains that cut their arm off. Yeah.

Dalton Anderson (40:25.779) Yeah.

Dalton Anderson (40:35.527) Yeah, it's like the person pointing the finger. I told you so. the way that I the way that I mean, I do a smaller scale, but the way that I lead the team and just the way I approach it with people is, hey, like if AI is doing something, you got no idea what it's doing, then you shouldn't. You shouldn't be using it. But if it's helping you get to 70 percent faster than you would prior, then you saved you've saved 70 percent of the time on the project. So that's a plus. And I think

R.J. Talyor (41:03.037) Correct. Yes. Yes. Yeah.

Dalton Anderson (41:05.767) Like getting to the first wrong answer is faster than doing it yourself. And so you just do that. Then you optimize whatever the output was prior. And then maybe you, if you're consistently doing that, then you build some kind of agent for yourself and do it that way. But I think that's most optimal.

R.J. Talyor (41:10.107) Yeah, yeah.

R.J. Talyor (41:20.167) Yeah, yeah, yeah, yeah, yeah, yeah, agreed.

Dalton Anderson (41:26.675) Question about, question about like founder lessons. So you've done this a couple of times. And so do you have any key takeaways from people that are listening to this show and they're like, Hey, I want to be a founder one day. RJ is super cool. Like you did the pattern 89 and it was acquired and then you were an early leader at the name is eluding me. Yeah. And now you're doing this like

R.J. Talyor (41:32.88) Mm-hmm.

R.J. Talyor (41:44.092) Yeah.

R.J. Talyor (41:47.982) Exact Target, yeah. Yeah. Yeah, I mean, I wish I had some great lessons, to be honest. I like it's

You know, like, I don't, I mean, I don't like, would just say the things I've learned. I don't know that they're like advice. I think like I have children. People ask that too. Like, what is your advice? I don't know. Like everyone's kind of just making it up. You know, like I can tell you what's worked for me and not work for me and what's whatever.

Dalton Anderson (42:16.659) Yeah.

Dalton Anderson (42:22.845) Maybe not advice, but what about your thought processor approach on different things? I think that's more material, I think you could describe that.

R.J. Talyor (42:27.215) Yeah. Yeah. know, like, yes. Yeah. And I'll, yeah, I will answer your questions, like lessons, like one, people matter more than you think. And, you know, with building Backstroke, I've, gotten to work with people that I've worked with before. And that's been awesome because we kind of get over the part where you get to know you and your quirks and things you're good at and bad at.

Dalton Anderson (42:54.035) Mm-hmm.

R.J. Talyor (42:54.301) whatever and it's just like, let's just get down to work and know that RJ is not good at that part. Okay, so they know that I'm not good at that part. So there's no like discovery of that. You kind of like already have the shorthand, which is really good. So finding people you love to work with and that you have worked with before. I actually frankly don't know how some of these founding teams who like meet each other through founding matchmaking make it work.

Dalton Anderson (43:04.594) Yeah, I know.

R.J. Talyor (43:17.307) Like, you know, because you're effectively getting married to that person in a work marriage of sorts. And it's like, you just met them and you like thought they were cool over a, like a hack, hackathon weekend or something. I mean, I don't know. Like that just seems, that seems kind of nuts to me, to me. Yeah. And I've gotten burned by that, you know, type of thing in the past. So one, people really, really, really matter. And then the second is like,

Dalton Anderson (43:27.759) Over read it. Yeah.

Dalton Anderson (43:35.853) Yeah, yeah, yeah, I agree.

R.J. Talyor (43:46.812) Just that it's cliche, like just listen to the customer because like the market will say, this is a funny window. The market will say AI is taking over the world. AI is eating everyone's lunch. AI is, you know, whatever, but like the customers we talk with, that's not the case. They're experimenting. They're trying. They're trying to learn.

And so kind of what the market says or what you see on social media is very different than what's happening in the trenches. And the case studies or the examples that you see are the first of many, but the first of examples of success and all of the failures along the way are not displayed.

in those case studies or on stages or in social media and all the customers we're working with are finding good successes, but they're also finding a lot of failures. So working really closely with customers, we have two core values that we think about every day. is accountability and empathy and, you know, an accountability side. It's that's kind of clear. Like we are accountable to numbers. We're accountable to metrics and outcomes, but there's also empathy because we're changing someone's livelihood.

changing someone's job. It's scary, it's overwhelming. They're dealing with other things that are outside of work. I would say really, really listening to customers is the way to win. And then the other one, maybe like the other thing I've learned is, or,

I've actually wished at some point in my career that I was less emotional, which sounds like a weird, like I feel very high, very low, and I feel like I'm like, get off a customer call or a call like this. And I'm like, so excited about what we're doing. And then I go to another customer call and I'm like, shoot, we're doing the wrong thing, or we have Mark. And then I got another one, I'm like, oh, awesome. And it's so high and low and everybody describes.

R.J. Talyor (45:45.722) startups in that way and it's just true and I feel like I react to those highs and lows very you know like

I just react to them. And so then it kind of colors my view. so, you know, with backstroke, I know that I do that. So I'm aware of it. I'm a little bit more like, okay, this is the time where I get a little freaked out. Or this is time I get overexcited about something. So I can recognize those patterns about myself. So with Pattern 89, my previous startup, we successfully exited to Shutterstock, it was great. But I had those highs and lows and I thought, oh, you one day I thought we're going get acquired by whatever

for billions of dollars and the next day I thought we're doomed, we're never gonna make it. And those feelings still come in with backstroke, but I think, okay, I know, I understand what those are. understanding kind of your response to things is really important. Or that's my learning for myself. Yeah, yeah, I'm just like, okay, no, I'm allowed to have that emotional response, but like, know that like...

Dalton Anderson (46:47.401) to help pull you back to baseline, I guess.

R.J. Talyor (46:54.749) the target that we're on is still legitimate. You don't have to throw everything out. Just know that there's highs and lows every day. So those are things I've learned.

Dalton Anderson (47:02.676) I think this is great. I this is great. I resonate with things you're saying. I think a lot of people just sign up to like, it's just highs and lows, highs and lows. And like, you just got to deal with it and it's going to affect your day or whatever. I think your perspective about like there is a highs and lows that happen, but also recognizing that it's going to have an effect on you, but also recognizing that that effect is temporary. that doesn't, that shouldn't deter you from your long-term goal and what you signed up with, like signed up for and who you signed up with.

R.J. Talyor (47:08.317) Mm-hmm.

R.J. Talyor (47:14.843) Yeah.

R.J. Talyor (47:24.902) Yeah.

R.J. Talyor (47:30.717) Yeah. Yep. Yep.

Dalton Anderson (47:34.814) curious how do people if they want to learn more about yourself or backstroke should should contact you or get involved or how should they look into backstroke?

R.J. Talyor (47:42.919) Sure. Yeah, so we're just backstroke.com. Just like the swimming stroke, backstroke.com and I'm RJ at backstroke.com. Check us out. Message me, love to chat with more.

Dalton Anderson (47:54.965) course. Wherever you are in this world, good afternoon, good evening, good morning. Thanks for listening and hope you listen in next week. I'll stop the record.