Growth Gorilla

No B.S. Content Promotion for SaaS Companies​

Promote, Or Die! Podcast EP#03 Kristin Tynski on Creating Valuable Content, Leveraging AI for Promotion, and Winning with Data Journalism

Promote, Or Die! Podcast EP#03 Kristin Tynski on Creating Valuable Content, Leveraging AI for Promotion, and Winning with Data Journalism

posted on

Main Takeaways

  • Content promotion is about matching interesting content with the audience and channels. Content creators need to consider distribution channels and audiences from the beginning of content creation.
  • Unique value in content creation is important, and there will be a demand for unique, valuable content. It’s the beginning of the end for content that is a rehashing of existing information without adding anything new.
  • AI can help in finding the right audience, journalists, and media outlets for content promotion. It can also supplement existing PR tools to improve the understanding of potential audiences.
  • How generative AI can help in content creation.
  • Data availability benefits brands by allowing them to create content with new information and insights, which can be found using AI and large language models.
  • The agencies capable of executing more data journalism-style content are more likely to win.

Resources

Website & Social Media Links

Other AI Resources from Kristin:

Transcript

Alan: Hello. Hello everyone. It’s so nice to have you and to finally meet you, Kristin. welcome to the show.

Kristin: Thank you so much for having me.

Alan: So today we will be covering everything about content promotion as it’s specifically,relates to artificial intelligence and so the rise of the machines, right? , everybody’s talking about chat, G p T, but yeah, the first question that I like to get started with is

what’s, like what does content promotion mean to you specifically?

Kristin: Content promotion to me, really is about matching what’s interesting about a p, a certain piece of content with what the particular audience finds interesting, and also matching the the right channels. So where do those audiences consume content?

how do you, reach out to or put that content in the proper channel for the proper audience at the right time?

Alan: So there’s a ton of people that keep producing a ton of content without thinking about the promotion. So what is the right balance in your opinion, between the content creation and content promotion?

Kristin: Well, I, I think especially in the age of ai, where we’re going to see it becoming easier and easier to produce content, really content creators need to be thinking about what the distribution channels are and who the audiences are from the very start. and what unique value their content is adding that doesn’t currently exist, right?

So it’s about creating content that has new information to add to the conversation and isn’t simply a rehashing of old information. we’re be seeing a leveling up of the demand from publishers for unique, uniquely valuable content. and, and I think the, the right mix is, Really just figuring out from the very start how you’re going to promote a piece of content, who the audience is, what the channel is, and, trying to be as certain as you possibly can be, that it’s the content you’re producing is going to be well received by the audience that you’re pitching or that you’re distributing it.

Alan: Mm-hmm. . And so do you think this is the end of all of the, how-to articles, the ultimate Guides kind of thing?

Kristin: I, I, well, not immediately, no, but I, I think it’s the beginning of the end for any type of content that is, like a retelling or rehashing or, mashup of existing information that isn’t really adding anything new to the conversation. new data, new experience, some new insight. I, I think that will be the minimum. content that audiences will care about. and yeah, I think in the age of, of large language model driven search engines, you’ll have, for basic informational content, you’ll have like a winner take all sort of situation where you can’t have 50, 50 articles that can attract attention on, you know, a very basic telling of some piece of information.

You need to be the unique source for some insight.

Alan: Doesn’t mean that people won’t write as much keyword targeted content because it’s going to be mostly news, like data driven content.

Kristin: I think it’s, it’s really hard to say cuz it depends. It depends on how things work out with, with, you know, like Google’s bard and things, new search engine that’s powered by GPT three. I, I think we don’t necessarily know yet how consumers are going to interact with these large language models in a search experience, but from what I can tell, it’s going to. It’s going to steal a lot of traffic. Right? A single answer from an AI that solves your problem or answers your question is going to eliminate the need for you to go click on 10 different website to find,

um, yeah.

Alan: Yeah, definitely. Especially considering that lately the top 10 results were all similar articles and so it didn’t really make sense for most people to keep writing content on the same exact topics and try to outmatch the competition to out rank them. but yeah, specifically for short type answers content, I definitely agree, that AI is going.

Take away, like a lot of that traffic. And so you’ve written, this article on LinkedIn that, that I loved about some predictions on AI and how it’s going to change and have some impact on both the creation, but also on the content distribution kind of landscape. So what are some examples of automation for content distribution?

Kristin: well, I think there are lots of different ways that you can, you can integrate large language models and AI into the content distribution process. when doing it well, first comes down to, as I was saying, finding the right target audience. So, at Fractal, I mainly talking about finding the right. journalists and news publishers who would be interested in the content that we’re producing.

So AI has all sorts of applications in, in surfacing the right journalists or the right media outlets or the right audiences. I mean, simply by asking them, I have this piece of content on this topic, what audiences are most appropriate for it, but it can get a lot more complex than that.

And then certainly AI tools can be integrated into like existing.

Research workflows. So existing PR tools that are meant to find journalists or find influencers or find taste makers, those can all be, supplemented by, AI technologies that can improve the individual doing it, their ability to truly understand like what the global space of potential audience is for any given topic.

But then also, of course, generative ai, like G P T three can be used for aspects of the actual outreach itself. So, helping to write pitches. Especially if you have a source of like additional data that you can inject into the prompts.

So if you have some information about the person you’re pitching, like, you know, they like dogs, you know what the beat is that they write about, you know, some additional information about, the outlet. , you can use that to help you, to help the AI write a prompt that that, matches the content that you’re pitching to what that journalist would likely be interested in, in an automatic and scalable way.

So, yeah, in terms of finding the right audiences, the right individuals, and then also in, in terms of actually writing pitches. Yeah.

Alan: So is this, in terms of the finding of the audience, does this go beyond tools like Spark, Toro, for example? what exactly are some examples of tools that already exist that can help people find audiences?

Kristin: I mean there, there are definitely, you know, a few Good Ex Spark Toro is an interesting example of, of audience insight information. if you’re talking about journalists, then you have tools like Mara and Cision and Meltwater and a few others. None of these tools have really been enabled with AI yet, which, you know, I’m sure is coming.

so yeah. And then of course there are also like influencer marketing tools that try and do something similar for social media influencers.

Alan: And so, so there’s a lot of like different platforms that have come out in the last couple years, right? So TikTok, like ets say Shopify, all these things. Obviously people that are selling products and services on these platforms, Have had to kind of adapt and start learning how to optimize their content for these platforms.

Right. So how do you think the rise and the growth of this alternative, SEO is going to affect content creation, but also content distribution on these platforms?

Kristin: I, I think we’re seeing a fragmenting of audiences across platforms and I, I think. It’s going to completely depend on the business, which channels make most sense for them, what their product or service is, and where their audiences are aggregating or congregating. some businesses it will make sense for them to operate across many or most of the social channels as well as organic search channels.

others may have one or two primary, primary places where they can really see like a positive ROI for. , the type of marketing that they’re doing on one existing channel. I think in general, audiences are, really trending towards short form content and having shorter and shorter attention spans. And, I think that’s a general trend across all forms of content, or all channels.

so yeah, I, I think we will continue to see short form video platforms like. Have an outsized influence on and really capture more and more audiences, than they’ve been able to in the past.

Alan: What are your thoughts on like, brands having to decide which platform is best for them? what’s the, best way for them to reverse engineer this from maybe some competitors?

Kristin: I think it’s really important to figure out, are there any examples? on the channel that you’re looking to market to that have done it really well. Right? And there are, on TikTok, there are many examples of, of different types of brands that are, are really leveraging the true power of, of TikTok and creating content that matches what people expect, right?

It’s not like advertorial content, it’s not sale the content. It’s short form content that is emotionally impactful and engag. and, fun, silly not taking itself too seriously. And, and then human right, like human stories are what really matter on, on social media platforms like TikTok.

So one interesting example that I’ve seen is, think it’s a company called

It’s a Norwegian. Company and they make, they make like fans and duct work and things like that, like a very boring topic. but they’ve amassed like a very large TikTok follow because they have an employee who is very interesting and they just interview this, this guy Carll who. is just like funny in, in a very quirky way, and they ask him random questions.

Most of them aren’t really even about the brand or the product. but they’ve, they’ve been able to capture an audience through his sort of like unique charisma. And then they’ve tied it back into the brand by interviewing him at. and asking him things like, you know, what do you like about working here?

What is it about this sort of setup of this sort of job that works for you? and so it’s sort of a roundabout way of introducing a brand, but, it’s done through the vehicle of, of very human interaction and

human storytelling. And I, I think that the brands that are doing well in TikTok are the ones that, are taking that sort of.

Unique approach to content creation and aren’t, aren’t thinking about it in terms of, how close to the brand, how protective am I of the brand when I’m doing a TikTok video, right?

They’re willing to be more experimental and sort of go outside of their comfort zone in terms of the topics that they talk about, the formats

that they use and so on.

Alan: Definitely agree that this is a very good strategy to kind of gain brand awareness, maybe. , like get trust, to the brand as well. But do you think this can also translate into actual, like revenue down the line for companies or is it just a pure kind of brand play?

Kristin: I think it really, it’s on a case by case basis. Right. So like the example that I just gave, will that translate into, You know, more more sales for their, for their fans and things like that. I, I mean, it, I think it really depends, right? Like the, the TikTok algorithm is, is really a new generation of, of content recommendation system that, , we know something about, but we, we certainly don’t know everything that we would like to know about it.

from personal experience, it seems almost, psychic in terms of how well it can match, you know, my interests, even interests I’m not necessarily aware of. and so, yeah, I think it has a, has a unique value in that may, maybe those videos are actually being shown to audiences that would be interested in buying their product.

it’s it’s a little hard to say right now. but yeah, there, there’s something pretty magical about their recommendation algorithm and, I wouldn’t necessarily underestimate its ability to find the right audiences, even if the content itself doesn’t necessarily seem like it would be able to.

Alan: Yeah, it’s definitely spot on as an algorithm. It’s kind of scary sometimes, , but yeah. So you also,write in the article about how data is becoming more available, more like, more easily attainable from people. So what do you think are the benefits of this? Because the way that I see it, , it’s mainly going to create more competition, especially from like the agency standpoint,

so why do you think there are benefits in this?

Kristin: So in the article I’m talking about the availability of data being important because it allows brands to create content that really presents entirely new information and insights. and that’s, that’s the benefit that I think it. Is primarily creating, right? So if you’re a brand, there is now really near infinite number of data sets that might exist out in the world that would help you tell a story about some aspect of your product or service.

And with AI surfacing that information is becoming much easier. analyzing that data is becoming much easier. telling those stories in a scalable way is becoming much easier with the, with the help of, of large language models. So, and then also, you know, these large language models as we were discussing earlier, are, are going to necessitate content that is presenting new, newsworthy, unique information that isn’t found elsewhere.

And the best and most scalable way to do that, in my opinion, is a sort of a data journalism approach. which uses either existing data sets, scrap data sets, or custom internal company data sets and.

Alan: Definitely a benefit for companies and brands and so I get what you mean. It’s. , it’s making things more competitive for agencies, essentially, or for people that are just pitching to journalists in general, right? Because there’s more people pitching similar stories or similar kind of data.

so how do they think this is changing the landscape for, content agency, but also for, in-house content teams?

Kristin: I, I think it’s just in general, it’s going to require leveling up of the content that’s being created and, I think the agencies that are capable of. Executing a more data journalism style content are going to win. But that’s not necessarily the only type of content that I think will be successful.

Like I was saying earlier, it depends on the channel, right? So if you’re an agency that’s producing content for, you know, short form video for YouTube or Instagram, or TikTok, Then, yeah, data driven content maybe isn’t necessarily the information that you care most about. Maybe you’re, you’re really trying to tell like a very human story.

more personal one or a funny one, or, you know, an action based one or so on.

Alan: Something else that I was kind of curious to hear about from you is, digital PR and data driven, journalism type outreach. Most of the times you’re promoting content that’s been created for the purpose of doing this outreach, so, how do you think companies that have a lot of content sitting.

Just doing nothing for the business that they have, like maybe spent a ton of money creating, like what’s the best way for them to decide? Okay, so I have this bulk of content. I need to sit down and figure out the best distribution channel for this.

Kristin: it really depends on the, on the information being provided in the content, right? I mean, if you have content that’s essentially just rehashing of existing information or you know, mashups of, of different pieces of existing information, it, it’s not necessarily gonna be worth promoting, right? It’s, if it’s not new and newsworthy, you’re not going to get coverage from.

From mainstream news sites. if you have like expert opinions or things like that, like that, that sort of content, then, then share that might be worth promoting to, like industry blogs or, or industry taste makers. I think companies need to prioritize the content that they have that is truly adding some sort of unique information and focus their promotional efforts on.

Alan: So I was talking to someone else on the podcast about this and basically they were saying the only content they should promote is content that is adding some kind of value or that is newsworthy and things like that. So definitely agree. Most companies want to still try and promote lending pages or things just because they think that those pages will, will rank and bring them revenue.

But most of the times, especially us as the agency, it’s really difficult to do and no one is going to want to link to a lending page or anything like that, that that doesn’t add any value. So do you guys have any situations in which you have clients that come to you and they’re like, okay, we want to promote these 10 lending pages because we want to just focus on making sales and we don’t care about brand awareness?

Kristin: Yeah, and that certainly has happened and, and really we have to be very candid with them in saying, You know, this, this salesy product page that you’d like to rank better isn’t something that can be, be promoted. There’s nothing newsworthy about this. There’s, there’s nothing really uniquely valuable beyond that.

It’s selling a product or service. I mean, there are, there are ways to make it more interesting or potentially valuable such that you could promote it. Typically, if it’s something that’s super close to a brand, I mean creating some sort of tool, like a calculator or, some sort of analysis tool.

especially now with, with generative AI models being opened up through APIs. There are some, it’s, it can be pretty quick to create like a really interesting and compelling tool that leverages GPT three. So, I mean it really totally depends on the company and what. with their niches. but there may be an opportunity to create some sort of like interactive piece on a salesy product page that actually is adding value to, you know, whatever that page is actually about.

Alan: Yeah. So actually, this reminds me of an example that I saw a couple years ago on, I think it was QuickBooks. They had this page on. like it was a lending page on one of their features. And they had jump links inside the page so that, main page was a lending page, but if you clicked on the jump links, the page would transform into a data-driven, research study or it would transform into a tool.

So the page could be four different subpages so that then you could use those, to link to the page, but the link juice would pass, through the main landing page. Very interesting kind of example.

So you mentioned about expert quotes and things like that, right? So for those that don’t really know and they maybe are not really into digital pr, Could you tell us more about what, reactive campaigns and news jacking campaigns are, and how do you see them changing moving forward?

Kristin: Yeah, so Reactive or Newsjacking campaigns are really just creating a piece of like uniquely valuable content, typically with some sort of data angle that, is responding to. Breaking news or something that’s currently trending in the news cycle and being talked about a lot by major publishers and a lot of journalists.

and it, it’s staying on top of what stories are trending related to specific brands or specific industries for our particular clients. And then, Being able to suggest to our clients story ideas, data-driven story ideas that would pair well with, an existing trending news story. I, I think that AI is going to help enable this, in terms of helping to create content in a rapid way, and also helping to, to possibly create data sets that that could be used. To tell late breaking stories or to add, to add new insights. then yeah, of, of course, like expert interviews and things like that are always important for when you’re pitching any sort of late breaking news story. I mean, it, it’s newsworthy in and of itself to get, to get an expert opinion on news topic.

Alan: What’s an example of analyzing a data set? Could you go to chat G p t, for example, and give it a bunch of data and be like, come up with a unique angle for this and for that, for my company. My company is, is doing this product and we talk about this, blah, blah, blah

this the kind of, like use of.

Kristin: Absolutely. I mean, I’ve written lots of different prompts that, that sort of give instructions on the type of content that I’m looking for. So data driven content, and then a theme or a topic, right? So you would give it the, the breaking news story, whatever it is, and then ask for first suggestions of data sets.

It might be, be appropriate to help tell that. and then how that data set might be used, what information that data set might contain that would be interesting and newsworthy related to your topic. so I’ve actually been playing around a lot lately with prompt chaining, which is essentially. One prompt, and then using the output from that prompt for the next prompt, and then using the output of that prompt for the next prompt so you can get like a multi-stage output.

and I actually, I wrote, I wrote a script that sort of does this for data journalism style ideas. that does exactly what I was just talking about. Takes a topic. produces the data sets or data sources that might be relevant to that topic and then comes up with data driven ideas that would leverage each of those data sets that might be newsworthy or interesting to publishers.

So I can share that with you if you wanna share it with your audience.

Um,

Alan: We’ll, we’ll put the links in the, in the chan notes.

Kristin: And then of course, AI in these large language models can also be used for the data analysis portion. So you, I mean, you can use, you can use chat c p t and basically copy and paste a small segment of, of a. Of an Excel sheet or Google Docs sheet and ask it for insights from that data.

And of course you still always need a human in the loop for these things cause they, they can and make things up. But in general, they’re really good at pulling insights out. or at least. Getting you going in the right direction of asking the right questions and finding what insights might exist inside your dataset.

The limitation currently is that, these models have a certain size window that you can prompt them with. So it’s 4,000 tokens typically. I’m not sure exactly what it is for chat g p t, it might be different than for GPT three. so you can only fit so much data inside them to have them analyze it. So you kinda have to feed it to a piece meal.

but you could also ask them in, in a general sense, like, I have a data set that contains this general information. What insights might exist in this data set?

And then use that as a basis for doing your investigation.

Alan: So do you think this is going to be difficult time for data analysts in particular

Kristin: I, I think it’s a renaissance for data analysts. I, I think that these. Generative AI tools are going to make their job so much easier. And I think we’re gonna see lots and lots of new companies crop up that, enable the workflows of, of data scientists, data journalists, data people in general to make their jobs easier, to explore the data more thoroughly, using large language models to supplement data sets and extend them. doing, doing automatic, sorts of informational parsing that wasn’t really possible before. So like doing, doing embeddings and clusterings and then having a large language model actually label clusters in like a natural language way. That’s a really interesting, way that it can be used. They’re really almost an infinite number of new applications and, and most of them haven’t even been thought of yet.

So to me it’s a really exciting time for anyone who’s interested in telling stories of.

Alan: So can you share, an example of a successful campaign, maybe where you used, these tools,

Kristin: I mean, certainly on my LinkedIn there’s a bunch of recent case studies of, of content that we’ve created and, and then of course on fractals website there are some additional case studies and examples of work that we’ve done. but we’re reusing AI in, in multiple different aspects of the campaign process and the campaign promotion process as well.

So, primarily in like the ideation phase and then the data analysis phase. so it’s hard to point to, it’s not like the AI is doing the whole campaign for you, right. But it is helping you to find all of the interesting and potentially newsworthy aspects of a data set. And so, you know, it, it’s hard to say exactly how much better it’s made us at that, but, but certainly it’s made us faster. and I think more accurate and more thorough.

Alan: We’re heading towards the end of the interview. So what are some of your favorite tools and resources that marketers can use to, give AI a try or try to improve their processes and systems using AI for content creation and distribution as well.

Kristin: Yes. So, Google Collab is probably one of the tools that I use most, is really just like a, a Jupyter Notebook inside of the Google environment, where you can share, your code with other people and then also have access to, some powerful computers that can run modern AI stuff. So that, and then of course, The APIs for these AI technologies that are coming out.

So GPT three has an API chat. G P T will have an API soon, but you can, you can use

their their web interface of course. I’m also currently pretty obsessed with, with other generative AI technologies and models, like mid journeys especially, amazing to me. I think the, the quality of images coming from there, using the.

Human reinforcement learning stuff on top of these large transformer models is yielding incredibly impressive stuff. and then I think there’s, there’s text video coming soon. there’s a new technology called, I think it’s called 11 Labs or something like that, that does, voice clothing in, in. A dangerously impressive way.

Right? Like I tried it the other day and it actually clone my voice to the extent that I couldn’t

tell the difference.

Yeah. So I

mean, uh, every

every week there’s a new technology.

Alan: yeah, so I was using, the script this morning. Uh,I don’t know if you know the tool, but it’s basically a transcription slash video editing kind of tool. And so it can do the voice cloning as well. So if you want to add some extra snippets of, like, like audio with your own voice, you can actually do it and it’s really close.

Kristin: yeah. Amazing new technologies coming out, and I think in the next year it’ll only get crazier.

Alan: So the last question to end the interview is if you could have any eighties movie character promote your content, who would it be and why?

Kristin: yeah, I think, I think Doc from, from back to the future, I, I like how he’s kind of a, a zany genius, who is really kind of fearless and yeah, I think that’d be a great representative for a

brand.

Alan: Any final thoughts that you’d like to leave the audience with and where can people stay in touch with you and find you?

Kristin: you can email me Kristen frack.tl, or if I’m on LinkedIn, I’m posting quite a bit on LinkedIn lately. free AI-based scripts and things for automating aspects of content creation and content promotion. and yeah, I’m, I’m, I’m super excited for the, the next couple years in, in content marketing, seo, pr.

I think that these new AI models are going to fundamentally change the way that we work, the way that people search, and the way that content is created. And I think those that stay on top of the most recent advancements will. A huge leg up, in terms of being able to create next level content that will be capable of getting the attention that I think brands are looking for.

So it’s sort

of a renaissance, I think.

Alan: Yeah, so I’m definitely excited as well. And so yeah, everyone, let’s go follow Kristen, and thanks for your time, Kristen. It’s been great to connect and I’ll see you soon.

Kristin: Yeah. Thanks so much for having me, Owen. I’ll see you later.

Alan: Bye-bye.