Episode #34
In this episode of the Digital PR Podcast, Lou and Steve are joined by Rand Fishkin, co-founder of SparkToro and long-time trailblazer in digital marketing, for a conversation about what’s next for PR in an AI-forward industry. They explore why a fixation on backlinks is no longer fit for purpose, and how a smarter, audience-first approach is key to long-term success. Rand unpacks the growing importance of brand context, trusted sources of influence and real visibility. From TikTok to Reddit, search engines to social feeds, it’s clear that the future of PR lies in where people actually spend their attention. And while AI brings its fair share of uncertainty, Rand leaves us with a sense of optimism and a roadmap for adapting with purpose.
Have a listen or read the full transcript below.
Louise:
Joining us on the podcast today all the way from the West Coast is Rand Fishkin. Rand is the co-founder and CEO of audience research tool SparkToro, was previously one of the co-founders of Moz, and is an all-round smart person on the subjects of search, digital marketing, and recently, AI. He wrote an article that caught mine and Steve’s eye about how PR is the future of marketing, and it’s that topic that we wanted to talk to him more about today. Welcome, Rand.
Rand:
Oh, thank you so much, Louise. Yeah, great to be here. And I agree, PR is absolutely the future of marketing.
Steve:
Well, you’ve preempted our first question, which is great, because you have recently become a bit of a digital PR hero recently. So talking about how you see digital PR being an important part of our new AI search future, Could you elaborate a little more on your thoughts on it? Why do you think it will be or is so key?
Rand:
Okay, just to clarify, I have heard lots of people say digital PR and mean something that is not public relations and sort of getting into press, media, sources of influence. When you say digital PR, I think you mean go find your audience’s sources of influence and be present in those places. Correct, absolutely. But I just wanna, yeah, I just wanna clarify because for some reason, like the world that I used to be part of, my previous company, Louise, was Moz, which was in the SEO space. And there’s this weird thing where people in SEO call low quality link building digital PR. And it’s very annoying because then the terms get mixed up. Anyway, I’m not talking about digital PR. I don’t think that’s the future. I think that was a crappy past. And I don’t think I ever recommended that.
So when I talk about PR, yes, I’m talking about finding audiences, sources of influence and being present in those places. And the reason I think that is absolutely the future, no matter which way this AI trend, hype trend goes, is because, you know, that has always been the most important part of marketing and all the new AI models and training data rely on, you know, I joked earlier, right, before we started the podcast that if you were to do the ChatGPT interview with Rand Fishkin, it would just be words that frequently come after other words in my vocabulary. That’s almost exactly what AI does. In fancier ways, they look for documents that seem reasonably high quality, they index all those documents, and then they build models that representationally, you know, in vector space, put words near other words, tokens, which are words and phrases, near other tokens, and then the model essentially uses a statistical means of showing words after other words. All that that means, you don’t have to understand any of that. All it really means is that if you ask ChatGPT or Gemini, which powers Google’s AI overviews, or you ask Perplexity or Claude, any of these models or systems, a question, you know, to recommend you a plumber in Seattle, it is going to look for plumbers who frequently, hopefully most frequently, come up in all the documents where Seattle and plumbing are talked about. That’s at the base, at the core, that’s all it’s doing. And PR, at its base, is let’s make sure that your plumbing business is mentioned in as many places as possible that talk about Seattle plummer.
Louise:
It’s interesting you mentioned the digital PR as two different things, as in two different meanings or confusion when it comes around there. Because I guess for your benefit around our heritage, so Propellernet, where me and Steve work, is an SEO agency, a search agency. So our digital PR was born through a kind of, I guess, an SEO need. But I think there’s such a spectrum of how people have used it. You know, digital PR can just be, I guess, at its worst, kind of guest posty, spammy, not good for your brand, but gets you a link digital PR, which is like, that’s one end. And then on the other end, it is purely just the ability to use PR for both the brand building and then also, yes, links are helpful. And sometimes it’s helpful because people click on a link and actually get to your site, but also it’s helpful for Google as well. So just for your context of where we sit on that scale, we are, yeah, I think the reason me and Steve have stuck around in it so much is because we are, it’s always been more kind of true to pure PR.
Rand:
Are you finding that clients are getting savvy about this new world and sort of they care less about links because they know links don’t matter nearly as much? And they care more about getting their brand mentioned, especially in places where people actually pay attention and where Google and AI models are likely to index that stuff.
Louise:
Well, I guess. There’s probably a range, but I guess for the most part, they’re coming to us to say, you know, you have been our go-to people around search for X number of years, sometimes 15 years, like some of our longest clients. So what do you think? And actually, it’s been an incredible part of mine and Steve’s career, because we have been in it for 10 years or so. actually, this is the biggest amount of change. So we’re reading all the things that probably, well, we’re listening to your podcast that you’ve done before, we’re reading your articles, but reading other people’s things, we’re talking to smart people within our business and stuff like that to kind of work it out. And so I guess we are the people going back to our clients and saying, well, actually, yes, there’s still a place for building links because at the moment there are still a list of blue links and that is helpful. But there is this new world opening up where we don’t have to be quite as strict on… I mean, I don’t know if you’re aware, but in the digital PR industry, you might have a situation with clients like previously where you might get on, I don’t know, CNN or something for your client. They’ve never been on CNN before. It’s incredible. It’s talking all about that. And they might go, oh, it doesn’t have a link in it.
Steve:
Yeah, it’s crushing. It’s crushing. Yeah.
Rand:
Oh, wow. Wait, how dumb do they think Google is?
Steve:
Well, that might be a follow-up podcast that we’ll have to do at another point, but it is pretty wild, the discrepancy in knowledge at the moment, because it’s still so new.
Rand:
That is not new. Okay, look…
Steve:
Take a break. We’ve riled you, Rand. We’ve riled you early on, so early on.
Rand:
I’m sorry, I’m like emotionally upset about this. I would like to grab these clients of yours who I’m sure are lovely people and I hope they’re paying well and whatever. But I want to shake them and say, surely you’ve seen, right, that there was a Department of Justice inquiry into Google’s monopoly behavior, right? And as part of that, one of the things that they do is they look at, they interviewed Google’s engineers, right, on the witness stand, right, where they can be sent to jail if they commit perjury. And so when they were asked, you know, what are the most important things in Google’s ranking systems, user and usage data has been number one for over a decade. So if you are on CNN, if your business is on CNN, and then people go and search for you in Google and click on you, or they hear about you on CNN and they search for other things and they scroll further down in the results to find you, and then when they recognise your name because of that press mention, that is far more powerful than you got a link. We haven’t been in 2007, 8, 9, you know, a link is going to change your rankings the way user and usage behaviour does for, it’s just been a really long time.
Look, I understand that information sinks in and I guess I’m probably to blame for a lot of that, right? Like I did a lot of proselytising of link building in my SEO days when I was at Moz, the world just hasn’t worked that way in such a long time. It bothers me that this is why I’m so pissed off at Google for keeping their ranking system secret. Because they are hurting not only the agencies and consultants and the SEO industry, but they’re hurting every publisher, they’re hurting every business by having this misinformation. It means you can get scammed by people who you know, are charismatic and sell you a story about what works in that field and doesn’t because it’s not transparent. It’s infuriating. I have a lot of complaints about the AI era too, right? Like I think it’s a lot more spicy autocomplete than people think it is, but at least it’s transparent, right? Like every single system, it shows you exactly how it works. Like you can go read up on, oh, how do they build these models? Oh, what’s the fine tuning like? Oh, what’s the training? You know, there’s tons of information. Any engineer can, you can literally build your own models that mimic them. So it’s just a lot more open. And I think that is a healthy thing. Security by obscurity leads to a lot of shady people getting paid $30 million to do no work.
Steve:
You see a lot of it at the moment, you know, on the likes of LinkedIn, you know, people promising the earth. And that’s the same as it ever was in lots of industries. But AI, it seems to have accelerated it. You know, people putting AI in their bio and sort of trying to look clever. Yeah, you’re not going to be able to tell me exactly how to do this perfectly and get me all this stuff. It’s fascinating.
Rand:
Last night, this was like late last night, maybe midnight, 1 a.m., Amanda, my colleague and VP marketing at SparkToro, there’s only three of us, so our titles are or whatever you want them to be. Like Amanda’s amazing. And she sent me an article that she’d written for a publication that had asked her to contribute a guest piece. And it was basically about this, you know, this idea, Steve, that putting AI in your marketing and positioning is deeply, deeply weird. It’s like saying, let’s imagine it’s 2003 and you say, “our software is made with code”. Like, yep, we kind of guessed that it might have been, right? Or, I don’t know, your accounting firm in 1996 says, we use computers. Great, amazing, I’m glad to hear it. Slide rules have been out of date for a while. It just seems deeply strange to say, we’re an AI-powered this, or I’m an AI expert,
Louise:
Yeah, I wonder, like, yeah, maybe when Excel first came out, people were absolutely mesmerised and thought it was the most incredible thing since sliced bread, and this is the similar version. But yeah, maybe we’ll, yeah, look back in probably even like a couple of years and be like, oh, that sounds a bit silly now.
Rand:
Yeah. I strongly suspect that AI in your bio is going to be one of those cringe things in a few years that just nobody does. It’ll be like saying, you know, that you were on the cyber web or whatever it is from the 90s.
Louise:
You mentioned as we were chatting around, you know, how links aren’t as important or probably maybe not important at all when it comes to LLMs and ChatGPT, things like that.
Rand:
I mean, I think my understanding of how links are treated by the LLMs is that basically You know, in the code of the document, you see that, you know, Ahref, and then equals, and the quotes, and then the link. And so the words that are inside there, you know, they’re gonna discard the code, but they’ll keep the fact that there was a link so that, for example, if you use the anchor text over here, and then you point to whatever, Rand and Steve’s plumbing supply in Seattle, the LLM will still be able to associate Rand and Steve’s plumbing supply with the sentence and the paragraph and the document. So it’s not that the link is discarded, it’s just no more important than any other word on the page.
Louise:
And if there are people whose bosses or clients are still quite wedded to, I guess, the older way of looking things around domain authorities and links and things like that, how do you think is best to approach that?
Rand:
I feel slightly embarrassed. I invented domain authority at a time when it mattered. And it was a reasonably good system for assessing one thing, which was, you know, I think we, when did Ben and I first create it? 2009 or 10 or something like that. So at the time, Louise, like it was basically this system to predict using machine learning, which is the basis of AI, right? But using machine learning, it was designed to predict the degree to which if a piece of content were put on a website, the same piece of content were put on 100 websites, which one would rank number one through 100? And domain authority 100 should be the top one and domain authority one should be the bottom one, you know, the lowest ranking one. I think that was a good idea. I think it was useful at the time, but the signals that it used were the same signals that Google sort of used at the time, although we didn’t have their user and usage data, right? So we’re using links and page rank and the content on the site and how often it shows up in rankings, you know, all that kind of stuff. Whether that is valuable today is hard to say. What I will say is domain authority, like the number that I’ve seen, and obviously many people have refined it since I left. I think Dr. Pete Myers still works on it, is my understanding. I think he still sort of owns that project, if I’m not mistaken. You know, nowadays, domain authority is mildly predictive of a website coming up more often in Google, higher in the results than lower. Generally speaking, that probably means that putting your content there is going to be more valuable. But getting a link from those places as opposed to just having your brand mentioned on the page, not so important. This is one of the reasons when I talk about how PR, like classic forms of PR, are the future, in a world where you don’t have to care about links, classic PR is so much more powerful. I can’t tell you, I’m sure, I don’t have to tell you guys, right? But like anybody who’s listening, Anytime I or you have ever tried to pitch a piece that contains a link, you know that there’s friction. You know that the editorial team is like, ugh, great, bunch of SEO spammers over here. But that is not the case when you are contributing something and you’re saying that it’s written by Rand Fishkin of SparkToro. No one blinks an eye at that. Contributing an editorial, an opinion piece, an article, That’s been standard in the publishing industry for a hundred plus years. And so, you know, it just goes over way more smoothly. It’s very easy to say, hey, you know, big publication in our space. I wrote something I think that you would like. Can I send it to you? You don’t need to link, I think your audience would really enjoy this. The odds of getting a yes from that versus, hey, can I guest post for you and put a link in? It’s not 10x, it’s 100x.
Steve:
Yeah, that removal of friction, incredibly useful for PR people. What really wanted to get your thoughts on what, because we’ve been looking at this a lot, but like, what can PR people do to be future proofing themselves, if that’s the right phrase. So like, there’s so much going on, there’s so much to learn, there’s so much to take in. And actually, some of what you’ve talked about is they’re not, they don’t really need to change what they’re doing, but how can they, is there any kind of change that you think is here or is coming where they can future-proof themselves?
Rand:
The weirdest part about this is what’s old is new again. If you were to go to an extremely successful PR and brand person in 1965, right? They’re great at messaging. They’re great at positioning. They’re great at figuring out the publications that resonate with a particular audience. Their methodology for that is different, right? It’s surveys and interviews and, you know, stats from market research firms. But what worked then is going to work better today than it worked 10 years ago. So we’ve had this sort of quarter century of weird disruption to the world of PR, and now we’re almost coming back again to a very classic form of PR. The big differences are, if you get stuck in the mindset of the 20th century world of press and media, which was essentially a small number of large publications that reach a huge audience, that mindset doesn’t work today. That’s not how anything goes. Today, we are all inside tiny filter bubbles and, you know, you and I say CNN and someone who’s 22 in Ohio is like, oh yeah, I think my parents watched that once, you know, right? It just doesn’t, but if you say name of YouTuber who only has 15,000 subscribers, they’re like, that’s the biggest person in my life. That is the most important source of media that I have.
And so as a PR, you now need to uncover your audience’s true sources of influence, which can be extremely niche. And they almost certainly exist on platforms that are owned by big tech, right? So TikTok, Reddit, YouTube, LinkedIn, Facebook, Instagram, Blue Sky threads, even a tiny bit of Twitter still if you’re sort of in, you know, tech bro rah rah world. Your job as a PR now is to go find all those sources of influence, be able to show, prove to your, you know, client or your team, your boss, the people you work with, that those sources of influence really do reach your customers and then come up with the strategy that’s going to get you into those publications. Almost certainly, two things are going on right now. One is the same old SEO link spam that we talked about from the early 2000s. That happens in LLMs too. I don’t know if you’ve watched on, you’ve probably seen some of these examples on LinkedIn and I follow a few of them on Blue Sky where people basically spam a little bit of Reddit and YouTube with videos that get no views with subreddits that like have no subscribers they get three upvotes on their post you know it’s not it’s not reach but because they use the words in the right places right, and the right words to associate things, and because the competition isn’t too stiff, and because LLMs right now are very sensitive to small fluctuations in temporal activity, meaning, let’s say that like in our plumbing example, We had, you know, Steve and Rand’s plumbing, and then it’s Tom and Louise’s plumbing. And Tom and Louise’s plumbing has been around for 25 years, and they’ve got thousands and thousands of mentions all over Seattle media. But Rand and Tom’s plumbing comes along, and we get 50 new mentions in just a couple weeks. Even though there’s way more Tom and Louise out there, because Rand and Steve are new and sort of appear to be trending, the LLM is likely to mention us and oftentimes higher up in its answers. So because of that, there’s a lot of spam and manipulation going on. So that eventually, just like Google fought spam, LLMs are going to have to fight it. They’ll, you know, they’ll get messed with and they’ll realise that they have to change and that future’s inevitable, but as a PR, man, you can move the needle so fast right now because of how these systems work if you just get that brand out there. And it doesn’t have to be big sites, so domain authority, low domain authority, doesn’t matter all that much.
Louise:
It’s interesting, like the opening up of your targeting. So I guess I’m coming at this from like a digital PR person who was thinking more about kind of SEO results. So probably just like I said, focusing on online media, which gets you great links, you know, is great stories as well, helps build your brand, but really kind of funneling in on a certain area. It is like just opening up that kind of media list to not even just being kind of traditional publication media and social media and all these other things, which, yeah, as you’ve mentioned a few times, is how a regular kind of consumer PR agency would be thinking. And actually, we were the kind of weird outliers being like, oh, well, actually, if you just focus on this, then it’s really important. You know, you can really report on this and this. And actually, I guess it was the reporting thing I just wanted to pick up on, because in terms of if you’re future-proofing yourself for PR and you’re saying that PR can be so impactful for how your brand appears in AI overviews or ChatGPT, I guess it’s showing that you do, like, you know, you kind of, you do your PR work and it has all these benefits, including in this AI search landscape. How you prove that that was your work or how you kind of connect the two. What are your thoughts on that at the moment? Is there anything people can be doing or is it a work in progress?
Rand:
It’s mostly a work in progress in terms of measuring presence in LLMs because, as you know well, you change a couple words in the prompt, the response changes. Five people ask exactly the same prompt at the same time to the same system, they get five different responses. Because it’s predictive text, right? When I said spicy autocomplete, I’m being sarcastic a little bit, but also it is selecting words that tokens, that frequently come after other tokens in the model, in the dataset. And so that statistical system means that sometimes they’re gonna show the word that only showed up in 10% of documents, and sometimes they’ll show the word that showed up in 80% of documents, and sometimes in 5% of documents. So you can get very different responses. Oftentimes in certain areas, a friend of mine was looking at, he runs an agency similar to yours in Philadelphia, and he was looking at like, hey, when you, can I find a way to ask for a search marketing agency in Philadelphia where we don’t come up? Like where we’re not anywhere on the list from ChatGBT? And the answer was generally no. Like almost every way you could possibly ask, no matter what kind of prompt, they were always in the list because they’re just in that many documents that talk about Philadelphia and marketing agencies. And that’s what you want, right? That’s what you’re trying to create for your clients as well, that their brand is so strongly associated with the problem they solve or the space that they’re in, right? That it’s just the, it’s the Dyson of vacuum cleaners. It’s the Kleenex of facial tissues. It’s the Google of search engines. That’s the thing that you’re trying to build and create. And granted, a lot of that is in the hands of the product people in addition to the marketing and PR people. But your job is essentially, you know, to amplify that.
When it comes to measurement, I think the thing to throw out is attribution. That is no longer a reasonable way to measure any organic marketing. People who try to do that are going to end up just throwing money at Google and Facebook, because they’re really good at attribution. And intentionally so, right? Google took away keywords so that you would spend more on paid search to get the keyword data, because then you could prove it. Facebook took away attribution from a lot of organic, and they took away organic clicks, obviously, quite dramatically. All the other platforms have followed suit. You can see that AI, is it? The links in the AI overview in Google don’t pass refer data right now. Maybe that’s a bug, but who knows? It’s hard to say. So, look, I think we’re entering an era where the way we have to invest in marketing, not to go back to the past again, and I’m not that old. I don’t want to seem like, I know I had some grey here. I don’t want to seem like I was not practicing marketing in 1965. But people who were practicing marketing in 1965, would put up a billboard in Detroit, and they put up a different billboard in Chicago, and they put up a different billboard in Milwaukee, and they put up a different billboard in Cincinnati, and then they see which one lifted same store sales in a five mile radius around the billboard over the next 90 days by the most. And then they’d take that billboard that won, and they’d put that in the other three cities, and they’d see if they got the same lift, and then if they did, they’d roll it out nationwide. And that is a reasonable way to invest in PR too, right? You can say basically, hey, we’re going to do a, whatever, a hundred day sprint, do a big PR investment, and then we’re going to see what happens over the next three months or six months. Depends on your lead cycle. It might even be a year that you really want to measure. It could only be five weeks if you have a sort of low cost consumer product. And then you want to see whether lift happens and how much, right? So you’ve got a model of seasonality and like, well, week over week over the past five years, this is what we tend to do in terms of sales during this time of year. Okay, great. And then what are the other marketing things we’re doing? All right. So we have the trend. And now how far above or below trend are we after we made this investment? Oh, we’re 5% up. That’s got to be the PR campaign. Here’s the, you know what’s great? Not to whisper into my microphone. You know what’s great? Here’s the thing, some people are gonna say, but what if something else was causing our results to go up at that time and it wasn’t really the PR? Oh no, you made more money. Yeah. So what? Sometimes you are going to accidentally or unintentionally spend some budget on marketing that didn’t work, that happened to coincide with other marketing that did work. And to that I say, that’s the game. You need to be willing to invest in hard to measure sources that you cannot absolutely 100% prove, because otherwise you will not be taking enough risk and you will not be putting enough budget and effort into marketing and you will lose out to a competitor who does. This is just how this system works. You cannot guarantee results, you can’t be constantly, you know, oh, well, we’re gonna turn on the PR for three months and we’re gonna turn it off for six months and we’re gonna see, okay, well, a year from now, even though our competitor leapfrogged us, we at least know almost exactly how much PR contributes to the bottom line. Yay? That’s not a win. Yeah, it bothers me when people think that’s how they’re supposed to do it. Anyway, rant, rant, rant, blah, blah, blah. Stop trying to attribute all your marketing. It’s okay to waste some money on marketing as long as your results are going up.
Steve:
That’s what I’m saying. Yeah. And it was such an impassioned and eloquent answer, because I’ve often thought this, like, why do we need to attribute perfectly to one or two sources? Like, there’s no need for that. Like, I get the sort of logic behind wanting to do it, because it gives you that comfort, that satisfaction. Or maybe we’ll repeat that. But ultimately, as you say, if things are going well and upwards, then keep doing more of the same. Keep taking risks. Keep trying new things. Like, yeah. So I’m 100% with you on that. And I think that’s empowering for our listeners as well.
Rand:
I think it’s a strange thing where entrepreneurs, founders, the C-suite is taught to take risks in all these other areas, right? In product and engineering, unfortunately in finance, obviously. And yet somehow taking marketing risks is anathema to them. Like they’re allergic to that one sector having any risk associated with it or any lack of clarity. It seems dumb.
Louise:
It’s an interesting time at the moment because I do think that at least everyone’s going through it broadly the same. No one has special access to this state. These changes are happening to everyone and I think something which people often mention around that is the idea of with AI overviews, you basically might get less traffic to your site because people might get your, when you’re, people searching through Google, you’re used to people finding their answer on your website, but instead Google takes the answer from your website and gives it to you in the SERP. Well, that’s a thing for brands and for our clients and people’s bosses out there. That is, I guess, what they would class as a problem. But specifically for a PR person, obviously, we rely on a thriving media industry where publications, you know, they need traffic to their site for people to read the news. And so it’s that kind of like ecosystem where if traffic is going down just across the board, we’ve already seen that lots of journalists are losing their jobs, publications have shut, things are changing a lot. PR is important and getting on these online publications as well as other places are important as well. But if there’s going to be less traffic going to them, how do you see that working out? What do you think that impact is going to be?
Rand:
The sad and frustrating news is that the big winners from that are the big tech platforms because the attention moves from, you know, I read whatever good housekeeping, I read the Guardian, I read the Wall Street Journal, I go to the YouTube channel, I go to the TikTok, I go to this subreddit, I go to these Google AI overviews. I never need to visit other websites. These 20 or 30 websites fulfil all my information consumption needs. You know, not to bring us back to this again, but remember how a few big channels and publications owned everything we read in the 20th century? That’s kind of where we’re going back to, right? That essentially these big tech companies are going to own all the attention. And so the only reasonable thing for a PR to do is to recognise that and then to say, okay, how do I get this YouTube channel to pick up my client? How do I get this subreddit to talk about my client? How do I get this whatever, you know, publication that still gets traffic from Google Discover, which is one of the few like big traffic sending sources that’s actually rising, how do I get them to mention me? How do I get into, you know, this prominent person on LinkedIn or Threads or Blue Sky or Twitter or whatever the platform is? Yeah, that’s a little bit of a different way of thinking about it versus, here’s our publication list, it’s sort of, here’s our creator slash source of influence channel on big tech platform list, and then we’re going to pitch those. But it should be similarly effective.
Steve:
I was very taken by a blog post recently, and I think you’re kind of starting to talk about this a little bit in that previous answer, I think, but you wrote a great SparkToro blog post discussing search everywhere optimisation. Can you, for anyone that hasn’t seen it, I’d recommend going to the SparkToro blog to have a read of it, but can you explain the concept of it to our listeners? And as a bonus question, Rand, if you could erase one buzzword from all PR decks and SEO articles, what would it be and why?
Rand:
Okay, I’m going to think hard about the buzzword that bothers me the most, but this term that Steve’s referring to is basically the idea that rather than search engine optimisation, which referred to the few traditional search engines, Google and Bing and Yahoo, maybe DuckDuckGo or Baidu. We are now in a world where people are searching on Pinterest and on Reddit quite a bit and on TikTok and on Instagram and on AI tools. And so historically we’ve just said, well, I do SEO for Pinterest. I do SEO for Google Maps. I do SEO for fill in the blank. We help with SEO for Bing. We help with SEO for Reddit. Great. I don’t see any problem with that. I think that terminology still works fine, but maybe that means that we should be thinking of SEO as sort of… anywhere and everywhere, right? Like when you are doing the practice of, hey, I can help your brand show up in the places where your audience seeks out information. That’s my job. You don’t have to reclassify yourself or sort of come up with a new name for some new protocol. Like I do LLMEO or I do A-I-E-I-O, or you know, whatever it is.
Steve:
Hang on, I’m going to write these down just in case these are IDs.
Rand:
If you put A-I-E-I-O on your profile, I will send you $5 via PayPal. So Ashley Little from Deviation, which I think is an agency actually in the UK, was the original sort of person who named this, basically said, let’s change SEO to Search Everywhere Optimisation. And their agency, Deviation, another agency, basically calls it, hey, we help with search everywhere. Wonderful. I think problem solved. I don’t need another brand. SEO is already a well-known acronym. Most, most people who are not super, you know, whatever, plugged into the conversation in digital marketing world around what, what new acronyms are popping up. would say, oh, do you help with SEO for ChatGPT? Do you help with SEO for AI? Do you help with SEO for fill in the blank? If that’s what everyone’s calling it, it already has a name, it already is associated, the brand is far more positive than it’s ever been, you know, 10, 15, 20 years ago. I’m sure you guys remember SEO had quite a negative reputation. Today it’s just a standard part of marketing. Do we need a new term? See, it seems weird to me that we would need a new term.
Steve:
Yeah, but we, because we’re marketers, we love coming up with new acronyms and terms. I think that’s the, it’s the gift, but the curse, you know? People love that. People love that.
Rand:
Do you think there’s, is there any benefit, though, to coming up with a new term? Like, let’s say you name something. Do you actually get, like, I don’t even think, I don’t think, for example, Cory Doctorow gets a lot of benefit from having named in shitification. Like, yes, lots of people refer to him, But other than pride, I don’t think he like, no one’s sending him royalty checks for that. You know what I mean?
Steve:
No, it’s not like writing a hit Christmas song and just living off the royalties for many years to come. You’re right. Maybe we shouldn’t bother. Yeah, it’s not worth it.
Rand:
I have named a bunch of things, right? And like, I have some pride around it, but now when I see someone using it without crediting, I kind of go, oh, well, good, the terminology took off, and I don’t need to be associated. Like, what do I care? If they had mentioned me, I don’t get anything from it, other than AI tools may be getting confused, I don’t know. Oh, I’m supposed to answer the terminology one, huh? Yes. Yeah, we were going to push on the buzzword, please. AI will take your job. I think that is a bunch of baloney and it is marketing. It’s the AI companies doing marketing to try and sell fear to executives who will then believe, well shit, I’m not, you know, the executive thinks to themselves, gosh, I’m not replacing my team with AI, am I falling behind? Maybe I should pay Claude and ChatGPT and these other tools millions of dollars a year instead of paying my employees. It’s just marketing, and it’s fear-based marketing, which I generally hate, and it is marketing that threatens people’s livelihoods, which I really hate, and it is untrue, which I hate with a passion.
Louise:
You’re not a fan. I mean, I was going to say, I don’t know if you’ve seen this image go around on LinkedIn recently, because it was in London. But it’s the internet. It goes around the world. And it was a horrible set of adverts for an AI assistant company. And it basically was like, why hire a human kind of vibes. It looked like something from Black Mirror. And I did think it was going to be some sort of joke or some sort of like there was going to be a reveal that actually it was Charlie Brooker doing something fun. But I don’t think it is. I think it’s a decent company. But yeah, the whole thing was like, you know, in front of all these commuters commuting into work because they’ve been told to come back to the office and they, you know, all this kind of stuff. And it’s like, oh, great, I’m not even valuable anymore because I can be replaced by an AI. It did the job of getting the name out there, but obviously not for not for positive things.
Rand:
It’s legal, but it’s immoral and unethical. And if I were at those companies, I would I would quit if they put up a billboard like that. It’s unacceptable.
Louise:
And I think at time of recording, Duolingo a couple of weeks ago said that they were going to go AI first, and I think they rolled back on it. It’s not going down well with anyone, like the employees, the customers.
Rand:
Google got asked on the earnings call, Sundar Pichai has asked, hey, are you going to replace a lot of people with AI? And his response was, Well, no, we haven’t found any efficiency gain from that, and so we’re not going to do that for a while. Maybe in a few years. It was one of those like, okay, you know what, if Google can’t figure out how to replace engineers with AI, then Yeah. This is, this is a fake trend.
Steve:
Yeah. But he’s keeping it open to sort of, uh, or keeping it open just in case something happens. Yeah. Just in case. But I did want to ask you, because actually from, from conversations that we’ve had with many people in the world of PR and marketing, there is, there is a lot of anxiety, like, and I think that’s understandable given some of the marketing we’ve seen, the fear-based marketing, some of the articles that are, that are out there.
Rand:
This is one of the few ways the press can still get clicks, right? Is by writing an article that like, you know, oh, the Anthropic CEO says he’s coming for 50% of all jobs, you know?
Steve:
Yeah, it’s true. But my question really was quite a simple one. It’s like, do you ever get anxious? You seem so eloquent, so positive, so kind of looking on the bright side of AI, but do you ever get anxious about this kind of stuff? I’m not turning this into a therapy session, Rand. I’m not bragging to your personal, but do you ever feel anxious about this kind of stuff?
Rand:
Oh my God, all the time. I feel a tremendous amount of anxiety, but I think that my anxiety as an example, there’s a, this is a very deeply specific thing, but there’s a, I was just looking at this today, basically a tax code issue which costs SparkToro six figures a year. So it used to be the case that, let’s say you do research and development in the United States at your company, you don’t have to pay taxes on the spend that you spent on R&D. That essentially is subtracted from your profits for that year. And that’s still true in every industry and every field except software development. There was this weird thing where it was included at the advent of sort of the software era by Congress in the 1990s, and then it kept getting renewed, but it was never put in with the rest of the tax code until Trump’s previous administration for unclear reasons, I’m not clear if it was like a, you know, most things with him are sort of personal vendetta stuff, but it’s not clear who he was angry at or why he would not want this in the tax code. Like he doesn’t, you know, he hasn’t historically been like, I hate software, right? He hates lots of things, but not software specifically. And so, it was confusing to figure out anyway. Now it’s, that code has expired. And so, you know, it’s costing us six figures. And also, it’s not just that. Every software company in the United States no longer has a tax incentive to do R&D. The research and development that brought you AI, for example, that OpenAI did, they got to expense all of that and basically build up tax credits for it. And so they had an incentive, and so does every other software company. And now that’s gone. And so Casey and I look at each other like, well, it’s going to be even longer before we hire another programmer. So, I think that tax code thing, I bring it up because I think that will impact employment in our field far more strongly than any adoption of AI or AI replacing jobs. That tax code piece should, if you map it out historically, was responsible for somewhere between 10 to 20% of all jobs in the field. those jobs aren’t gonna be hired for. Like people are not gonna make those investments and all the amazing technological innovations that might’ve happened won’t happen as a result. So, you know, maybe they’ll come from China, they’ll come from other places, they’ll come from the UK and Europe, right, where these types of things are getting invested. But regardless, it just, you know, that kind of stuff bothers me a lot. The tariffs, right, tariffs and trade war stuff, like those have real consequences and impact. That brings me a tremendous amount of anxiety, but it brings me anxiety because I believe that it is real, that the impact is real.
You know, I think I’m positive and optimistic about things like AI because when I look at the history of all technological innovation, they have always led to greater productivity and greater job opportunities. And generally speaking, they take away jobs that we don’t really want. The invention that took away the most jobs in the last few hundred years was the tractor, right? Like the tractor literally took away 20% of all jobs. Because you go back to 1910 and people had a lot of kids so that some of them would survive so that lots of them would pull plows and get horses to pull plows. And then suddenly, from 1910 to 1935, you didn’t need those people anymore. You could make all the food that the world needed off of far less labor. And so those people started going to cities and getting other kinds of jobs and saw sort of the rise of white collar labor as a result of this invention. Look, maybe AI takes a bunch of crappy menial jobs that are very repetitive and annoying to do and makes us all, like upgrades us all. But I don’t, it’s not going to take away jobs like the tractor did. And even if it does, that’s a positive thing. So I don’t know, I try and think in terms of like, what’s real and logical, what could really happen? And then I get anxious about things that are very damaging. And I think those are mostly political things. Like politics, as much as people don’t like to talk about politics on podcast, that’s what really impacts your job opportunities. That’s what really impacts, you know, fields and innovation and opportunity for investment and how CEOs and executives teams behave and whether they choose to hire agencies and, you know, the publishing industry getting crushed by Google. That is, in my opinion, fully a result, and not just Google, but fully a result of the United States not enforcing its antitrust laws. An antitrust case 10, 15 years ago should have said, you can’t do this. You are not allowed to own all these different things and then to start using the data from one thing to unfairly compete in another sector. You know, that’s the same thing as running the railroads and then saying, you can only ship our oil on the railroads that we own. You know, that’s, that’s what antitrust was invented to stop. So yeah, blah, blah, blah. Rand rants about politics, but it affects your life.
Louise:
Yeah and also I feel like in a simplistic way like you know that that thing you mentioned about like the tax codes within the software industry you could envisage someone looking at figures like in five years time which shows that like employment is down on this and just think well obviously that’s AI and actually like if you delve a little bit deeper into it then you get a bit more kind of human choices that have led to it but yeah you can just easily see that AI is going to be used as quite a oh it’s AI in a positive way but also in a negative way like and yeah there’s actually a lot more things behind it.
Rand:
Okay, we have one more thing we have to rant about. Do you folks let your PRs write their content, their article submissions with AI?
Steve:
No, no, we don’t. We actually have quite clear guidelines on it. We allow, you know, like research and stuff like that, copy checks, and maybe just to sort of, what’s the word? I guess you call it like jazzing up. I was gonna say zhuzh. Zhuzh up, like something that actually is proving quite difficult and taxing and just gives you the idea, but the way we use it is like, It gives you the starting point and then you have the human element. That’s how we want to use it. That’s how we think it should be used ethically and best for us.
Rand:
I think that is so, so, so smart. I think this will give you a true competitive advantage because right now there’s a ton of AI produced content that goes out there. I know that none of the AI tools say they can, all the analyses have basically been, oh, they can’t detect real human writing from AI generated content, whatever, I don’t care. What I do care about is you can feel, whether it was technically written by a human or technically written by an AI, in paragraph one and two of every article you read, you can tell if it’s a person or a person that sounds like AI, whatever, AI speak, which could be human or could be AI. And I think strong opinions, creative novel thoughts, weird ways of writing, unusual takes on subjects, strange paragraph formatting, all the kinds of things that like a human writer would do in literature and would do when they, I don’t know, you know, write a romance novel or like that stuff is so much more valuable than ever. Yeah, I have found myself doing this like because I consume some of this content. I find myself writing more like a like a weirdo. Do you know what I mean, I find myself wanting to convey opinions that are less professional less polished less normal
Louise:
Yeah. On LinkedIn, sometimes if I see a spelling mistake, I’m like, well, at least it was written by a human. So I’m like, great, that’s a good sign. Whereas before, you’d be like, oh, a bit unprofessional, spelling mistake.
Steve:
And you said at the start of this podcast, I find it really interesting, and maybe to let the listeners behind the curtains, but we said, you know, if you say anything that, you know, if we make a mistake or anything, then we can edit it out. But you said, no, the mistakes are actually the best part. I agree with you, because actually, that is what makes us human. It shouldn’t all be polished and perfect and everything neatly in a row. The conversation, the independent thought, the free thought, the interesting ideas are what makes us. So yeah, that’s made me feel incredibly positive about the world in general, actually. So thank you.
Rand:
Yes, because this is why I think, look, if AI was going to take jobs and get us all gigs in the arts world and we could all be creative and we could all express ourselves, that would be a beautiful future. I am super into that. Let’s all go to the theater more and see a bunch of actors on stage. Let’s create more art. And so, you know, when that comes to our field, I do a lot of writing, you all do a lot of writing and submit a lot of writing. creating stuff now, I think requires more humanity than ever, which is the exact inverse of what all the sort of ecosystem is telling us, right? They’re telling us to adopt AI. They’re telling us that everything’s going to be AI driven. They’re telling us that all the big tech systems are driven by machine learning and AI. And it’s not just that I find myself wanting it, I find that stuff resonates more. It’s more effective because it stands out from the crowd. And standing out from the crowd has always been a way to do better marketing.
Steve:
True. Very, very true. And a lovely kind of full stop, I think, on the episode. Yeah, a great finisher. Thank you so, so much, Rand, for your time, for your insights, for your positivity, honesty, and the mistakes in there as well, you know, that make us human. Thank you very much. We would heartily recommend, if you don’t already, reading the SparkToro blog. We really enjoy it. And people can find you on LinkedIn. Is there any other ways people can contact you if they want to get hold of you, work with you, you know, hear from you that you’d want to publish on this platform right now?
Rand:
Yeah, I’m actually most active on Blue Sky where I’m at Rand Fish. And then I also use LinkedIn quite a bit. So LinkedIn is a good place to follow me. If you want to subscribe to the SparkToro blog, that’s great. And obviously, folks are welcome to try SparkToro out for free. And my other company, SnackBar Studio, I think we just closed our job positions, but we’ll probably be hiring again. in a little while, so if you are in the video game industry or you’re interested in playing a chef in 1960s Italy who fights boars and collects pancetta for making her carbonara, you should come check out the SnackBar at the End of the World, which is probably about 18 months away from launch, but you can get it on the email list at snackbarstudio.com.
Steve:
Wonderful. Do check that out. Thank you again, Rand. Thanks, everyone, for listening. We’ll see you next time on the Digital PR Podcast.
Rand:
Thanks for having me.
Louise:
Bye!
