AI citations are not SEO rankings: rankings could hold for years, but citations fade and get replaced by fresher links, and Profound's Citation Decay data says half of the content cited by answer engines is less than 13 weeks old. Paris Childress talks with Simon Yong of Question.Marketing about what that means for late entrants, whether AI engines rotate the sources they trust, why YouTube may be the shortest path into Google AI Overviews, Google testing checkout inside its AI experiences, the rumor that ChatGPT ads lift citations, writing for machines, and building a knowledge base from recorded meetings.
AI citations aren't SEO rankings. A ranking you protected well could hold for years; a citation fades and gets replaced by fresher links. That sounds like bad news, but it means no one is locked out: fresh, focused content gives even late entrants a real chance, and YouTube may be the fastest way in.
In this conversation, Paris Childress, founder of Hop AI and co-founder of GEOforge, talks with Simon Yong of Question.Marketing, a frequent co-host, about how AI citations behave, why freshness matters so much, and how to build content that gives AI something new.
No. In SEO, a good ranking that you protected well could hold the top position for years. Citations are not like that. Once they get stale, they fall off and get replaced by fresher links. Profound's Citation Decay data says half of the content cited by answer engines is less than 13 weeks old. The upside: no matter how late you start, fresh and focused content gives you a chance to break in.
"You always have a chance, no matter how late you get in the game, how far behind you think you are." (Paris Childress)
No. The engines differ tremendously in which brands they mention, which brands they cite, and how many citations they give. A brand could be nearly invisible in ChatGPT and well represented in Google AI Overviews, so an aggregate score hides more than it shows.
Simon asks whether an engine serves one trusted answer to everyone asking a question, or rotates a small group of trusted sources and watches how people react, swapping out the ones that underperform. Paris's view: engines optimize heavily toward freshness, so as citations decay, others get the opportunity to come in, even if the original source is still a strong answer.
Paris has seen YouTube videos make it into AI Overviews almost immediately and bump out older videos with more history and engagement. Ahrefs Brand Radar names YouTube the most cited domain in Google AI Overviews. Simon compares it to how YouTube itself works: a new video can rank at the top within minutes, gets an audition while it collects views, and is demoted if it does not satisfy viewers.
TechCrunch reported that Google is testing a Buy button on select Flipkart listings in Gemini and AI Mode in India. Tapping it opens a Flipkart-branded checkout without leaving the AI interface. (On air, Paris names Instacart; the test is with Flipkart.) Fewer people visit the retailer's site, but staying inside the AI conversation may remove enough friction to create incremental purchases.
Simon has seen posts claiming that running ChatGPT ads raises citations and mentions, and he says himself there is no proof and it would be a scandal if true. Paris sees a more plausible indirect path, the same one long debated for Google Ads and SEO: ads that send engaged visitors who convert and do not bounce can send strong signals about a brand.
Machines. People rarely read blog posts on company websites anymore, and the volume of low-effort content keeps growing. Paris frames it as pigs at a trough of AI slop versus pigs hunting truffles: information gain is the truffle. Simon's advice is a content plan built around genuinely new information, kept succinct and to the point.
Capture the knowledge your business already creates. GEOforge feeds a persistent knowledge base from sales calls, customer support transcripts, and AI voice interviews with subject matter experts. Recording everything also has strategic value: AI could surface a great idea from a meeting the CEO never attended. Simon describes a recruitment client that uploaded six months of phone calls, and the analysis showed which agents did the best work and why.
Subscribe to The GEO Show wherever you get your podcasts, and watch the full episode on YouTube: https://www.youtube.com/watch?v=9539i5_8YOc
Paris Childress: Hi, everybody, and welcome to The GEO Show, episode 51. And today, I am joined by Simon Yong from Question.Marketing, who has been my now frequent co-host. We started doing LinkedIn Lives, and now we've moved it over into The GEO Show. So let me welcome Simon to the stage. How are you doing today, Simon?
Simon Yong: Not too bad, but for some reason I strangely seem to be a little blue today. But anyway, maybe you don't see that. It's cold here in the UK.
Paris Childress: A blue, a bluish background. I don't know what's going on there.
Simon Yong: Yeah. Well, anyway, that's what it is. We've moved into autumn.
Paris Childress: Yeah. But it still, yeah, it still looks good. So what's been on your plate at the start of the week, Simon? What's on your mind? You thinking more about answer engine optimization, GEO direction?
Simon Yong: What's, what's been on my mind is the amount of work that I've been doing through Grok and having to upgrade to the full-on big plan now. So I'm, I've eventually graduated, which I blame you for. But that said, the only, the only downside with Claude... Sorry, not with Claude, with Grok, I think sometimes it doesn't necessarily take enough tasks to work in parallel, if you know what I mean.
Simon Yong: It will want to concentrate on something, and then if you break its train of thought, it's not quite as good as Claude in that respect.
Paris Childress: Yeah. Yeah, it's pretty linear. I mean, it can work 24/7, but not as many things in parallel. I think that's what I'm seeing as well. I usually give it one thing at a time.
Simon Yong: Yeah. And it, and it doesn't seem, and it might just be me, but it doesn't seem as quick as it was when I started. No, but maybe maybe I've complicated the tasks a little bit too much. It would be really, it'd be really nice if you could see once it completes a task, this is all the steps I took, and this is how many tokens I consumed, and maybe it could then come back and say, "Well, you could have done this, this, and this to improve the workflow."
Paris Childress: Yeah. It has a huge lack of any kind of feedback about thought process, its planning, its token usage, forecasting token usage.
Simon Yong: Well, I'll see, I'll see this month whether I, whether I blow through the full $200 plan. Let's see whether that, whether I can do that this month.
Paris Childress: Yeah. Let me just see where I'm at for the week. I'm at 22%. I'm resetting in five days, so I'm probably on track to almost max it out it looks like. And I'm still mostly giving it just the podcast post-production workflow. It's a lot, it is a lot of work.
Simon Yong: But I've used... See, I've used your similar workflow for Stitch. So for the client that we're showcasing how it goes, I've used a very similar workflow, but that one actually produces the video itself. So rather than having to sit and record anything, you give it the idea, and then it goes and makes the video using Google NotebookLM. And then from that video, it then does the it doesn't do a podcast like you do, but it does everything else.
Simon Yong: But it is equally supposed to go and change the voice for, from like an American style voice to go to ElevenLabs to pick the client voice, etc. So I think the, that's the only challenge with Grok is that because you can't get it doing so many things at once like you can with Claude. I can have four or five windows open at a time with Claude. Grok, you have to give it the instruction and then wait, which that's the painful bit.
Paris Childress: Yeah. Yeah, there is a speed issue. I've noticed Claude Opus 5.5 is much faster. Have you started using that?
Simon Yong: Yeah, I've been on, I've been on that today with a few tasks, so... But again, how quickly will I blow through all of my credit usage? Who knows? But at least the See, the thing with Claude is at least it says, "Oh, you're only locked out for four hours," or whatever. Grok, you can be, well, "Upgrade now or wait five days."
Paris Childress: Yeah. So they got you. You're up on the 200 a month plan now, right?
Simon Yong: Yeah. But with the experiments that we're running, it's got to be the case. So there's gonna be so many workflows that I'm running except, you know, for the amount of social content that I wanna post, you can't really... You know, it will be building all of that sort of stuff overnight, not, you know, I don't really wanna sit with Claude and have to get it to do that and then...
Simon Yong: I've got, I've got the two talking to each other, which is useful, but then of course you're at $300 a month then, so how many more do you need?
Paris Childress: Yeah. Yeah. All right. So, what do you see happening in answer engine optimization and GEO lately? What... Who you, who have you been talking to? What are you seeing?
Simon Yong: Well, it does, it does feel like everybody's moving to GEO and answer engine optimization. There's a few... Well, let's say the, a lot of the companies in the UK and the US seem to be moving that way. Everyone who's coming at me from, like India, Pakistan, or wherever, they're all still s- firmly SEO, SEO's the way forward. But I am honestly getting offered 10, 15 different pieces of software a day, which who knows? Like, where do you start? And the- they're so...
Simon Yong: Some of them are very, very basic, things that you and I could code up very quickly using, you know, ChatGPT or Claude or Grok, and tell it to build it on Lovable, and you'd be there in a matter of few hours. You know, so there's.
Paris Childress: Trackers visibility trackers?
Simon Yong: Things like that, yeah. So things.
Paris Childress: Some prompts and then you track.
Simon Yong: There's some that look quite nice. I tried one this morning, I think it was Rise, which had, you know, you entered your website, it would sh- it would take that, then it would look at a few different questions that you might wanna come up for, then it would go off to Perplexity and Google AI Overview, etc, etc, and it'd look it'd search for you, and then give you a basic report. But then at the end of it, obviously it says, "Would you like the full report?"
Simon Yong: Which I suppose at that point you get booked into a meeting, which is, you know, it's a great piece of lead gen for them. But yeah, the whole space at the moment seems like... And I don't know whether you get the same feeling, but 'cause my LinkedIn, as will yours be, because we focus so heavily on what we do, the majority of my feed is now SEO, GEO, answer engine optimization...
Simon Yong: LLM- LLMS, etc, etc. Everything coming through. So it just, to me, I don't think I'm obviously a typical user. I am getting absolutely snowed under with every man and his dog is coming at you with the every offer under the sun. I even had a guy pitch me you know, the whole, "What are you doing? Are you coming into AI Overviews?" And I'm like, "You have looked at my profile, have you? You d- you know what I do." But clearly not.
Paris Childress: Yeah, I haven't gotten pitched too much over email, but I'm getting I'm actually getting pitched more now for podcast guests 'cause I think my the GEO Show has now started to get onto people's radar screens, though I don't think we have a huge audience yet. But it's on enough It's g- making it into enough of these agentic searches I presume, and most of these pitches are definitely written by agents, I think.
Simon Yong: Well, I'm not gonna lie, I've, I have my own agent reaching out to podcasts to book guests, you know, to book my clients on as guests. So, and it is working. We are getting people booked, so.
Paris Childress: Yeah. It's interesting how much of that now is agent to agent because I guess I'm on the other end of it right now. I'm receiving, I'm receiving pitches that are clearly agent written, but they're still good-looking prospects. I mean, they're legit looking guests. I have GrokBot I have one of my GrokBots now specifically watching out for every email that comes in as a GEO show guest pitch and to draft a reply.
Paris Childress: Well, actually, I even allow my GrokBot to make a decision whether or not we should invite the guest to an intro call, 'cause before we do the actual... Before I say yes, I say, "Look, we I wanna do an intro call, 15 minutes," we meet and explore topics and see if there's a good mutual fit here. And then I send a Calendly link with a 15 just a 15-minute slot window.
Paris Childress: And, but GrokBot has the discretion to outright reject people that look really spammy, and it has already done that four or five times. And if not it sends a, sends a nice email saying, "Okay, this is an interesting looking guest, and here's the next step. Here's the Calendly link." I've got I think I got one or two booked already that way through, just through GrokBot.
Simon Yong: Makes you wonder whether, you know, obviously being at the sharp end of using AI, whether other industries are lagging massively behind. You know, if you took, for example, a manufacturing business, I'm not saying that they're not as, you know, up to date in terms of technology as we might be, but if you took a business like that, would it mean that they might get an awful lot more appearances if well, I'm sure there must be podcasts that would be relevant to them.
Paris Childress: I think absolutely. I think almost every other industry has to be lagging further behind where we are. So th- this is a great playbook if you get a client because starting a podcast, it just gives you immediately so much content generation, and you're starting with long-form video, and then with long-form video you have so much repurpose value, so much- so many different channels to repurpose down into after that. It's great.
Paris Childress: And then the side benefits are that people will start pitching you to be on the show, and then you can, you can have these conversations, then you can turn some of those podcast guests into customers too.
Simon Yong: I believe even though I've only done, what, six, seven episodes on mine so far, I went on Google, searched "Will AI recommend your business podcast?" And it, and it finds me. So surely that's a citation for Question.Marketing that I wouldn't have had otherwise.
Paris Childress: Yeah. So i- are most of your efforts now for getting booked on other podcasts, are you trying to have people come onto yours?
Simon Yong: So I'm, I don't, haven't gone quite that far yet. So I'm... I think you sort of gathered traction after you'd done maybe 20 episodes. Am I right?
Paris Childress: Yeah. At least, yeah. Mm-hmm.
Simon Yong: Yeah, so I'm not, I'm not getting pitched at all myself yet. It was only last week that I properly promoted it and finished all of the graphics and the full description and then made sure that it was listed everywhere. So know, maybe in the next couple of weeks I start getting the odd... I, but if I start... If I don't get pitched then maybe I'd find out people don't like me and I become sad.
Paris Childress: Ah, no, I think that no matter what, you've gotta have thick skin doing this stuff, and it's a law of big numbers. Most, most people naturally are gonna ignore, and then people will write back. I think yeah, just gotta stick with it. I'm gonna, I'm gonna try to stick with this and try to start bringing more guests on and having good, good discussions.
Paris Childress: The first guest went really well and now already I set them up in with a free trial in GEOforge and have produced, I produced five, five content drafts just from the podcast recording the knowledge base, because that alone was great knowledge for the knowledge base.
Simon Yong: And g- and great content that, well, potential client or guest can use on his or her own socials as proof.
Paris Childress: Mm-hmm. Yeah. So what I've been seeing, and I've been looking at top stories that have been through this daily brief that I'm getting, I see so many studies that look to be really serious studies that are... There's some, there's some common themes that are emerging, and there's some other that are totally contradictory. One common theme is that you it makes no sense at all to measure your AI visibility on aggregate across all the engines, because they differ tremendously.
Paris Childress: You could have a 1% share of voice in ChatGPT, and you could have a 20, 30% in Google AI Overviews, and it's very inconsistent. So the AI engines themselves are not at all correlated in terms of which brands they mention, which brands they cite, the number of citations. Profound has been doing a lot of interesting studies, and the one that I saw today was that the average shelf life of a citation is 13 weeks. So that's just a few months, actually, four about four months.
Paris Childress: After that, it tends to fall off basically. So there's a, this concept of a citation duration, and I think people a lot of people still think about a citation as if it's a ranking, in the SEO world, rankings tended to stick. Cases, if you had a good ranking and you protected it well, you could stay.
Simon Yong: For years.
Paris Childress: Top position for years. Yeah, and then that generates business and I mean, that's the greatest success and was the greatest success in marketing for a lot of people. Citations are not at all like that. Once they get stale, if they get a little bit outdated, they fall off, they disappear, and they get replaced by fresher links.
Paris Childress: And I think what that means is that you're you always have a chance, no matter how late you get in the game how far behind you think you are, because of this hyper-freshness factor that is happening, you always have a chance to... You could just start creating a lot of content, and it's fresher and it's newer, and we've seen this with the YouTube videos that you've been creating.
Paris Childress: They make it into AI Overviews almost immediately, and they bump out other videos that are aged with history, with engagement.
Simon Yong: What I'm wondering around is, you know, let's say you gain a citation does that therefore mean that you are cited worldwide across that search term or phrase for everything? Or is there like a cohort of trusted citations that mean... You know, let's say there's... I know, a particular question gets searched 20,000 times a month. I don't know, let's make a number up, right? You are one of the answers. You're one of the trusted answers.
Simon Yong: Now, does any particular LLM serve you all the time to all 20,000? Or does it say, "We've got five people we trust here, let's serve all five. So let's, let's give them... Yeah, let's, let's give them two and a half thousand each or whatever," right? "And see what, how people react. And okay, so the one to four are doing well. This one's struggling a little bit with the interaction it's getting. Let's pop another one in."
Simon Yong: Now, surely it's gotta work like that rather than one answer all the time, one query. Okay, we've got the thing where you're saying about age or decay, like, you know, like almost like nuclear decay, like... And then, you know, we're in half-life, if you wanna call it that, right? That was the word I was looking for, half-life. So as it sort of decays, that gives the opportunity for others to come in. But there's nothing to say that you're still not the f- the best answer.
Paris Childress: Yeah. Yeah, I think that they really optimize towards freshness a lot...
Simon Yong: From what I've read...
Paris Childress: R- real.
Simon Yong: Perplexity particularly like fresh, right? But it's then getting it picked up. You know, that's the hard part, right? Is so part of the experiment that I'm running here is with the Question.Marketing brand because it's a brand new site. It only existed two, three months ago. It has relatively no presence. So if I can prove that what I'm doing with that ends up getting citations pretty quickly, that will prove that you could start anything. You know, yes, there's a lot of work gone into it.
Simon Yong: Yes, there's a lot of content. But from that maybe I can map out what works within each engine.
Paris Childress: Yeah. Yeah. I think it's also interesting the work that you're doing with video especially. One theme that I've seen in the stories that I've been reporting on is that YouTube is anywhere from 20 to 25 or higher percent of all the citations in Google AI Overviews YouTube. And in a lot of cases, what we've even seen is if you have a video that's directly nailing a topic or the prompt, the prompt topic, then you can, you can get there immediately.
Paris Childress: And I think YouTube is maybe the shortest path to getting that visibility for your brand. It's faster and easier than getting your website, a page of your website to be cited either in Google AI Overviews or ChatGPT or something.
Simon Yong: Well, you never, you never used to be able to do that with Google. The... I've been able for years to get YouTube videos straight to the top of YouTube for, you know, some really competitive terms. What will happen is, you know, you can launch a video today and within 20 minutes it will be number one on YouTube, maybe for an hour or two. So while it gets some views, YouTube is saying this is good or it's not good.
Simon Yong: And then if you've, if you don't satisfy the answer or you're coming up for something incorrect and people are not enjoying the content or the quality's not good enough, you're suddenly demoted and you're gone forever, right?
Paris Childress: Gone, yeah. So you have a, an audition...
Simon Yong: You have an audition. Yeah, it's a good way of putting it. You've got your audition, show yourself, right? Do you, do you, right? And then aft- now that is true of real, of actual Google. And today I was searching with a client, we were in his offices, searched a couple of relevant searches to his business, and surprisingly, it didn't put me straight into an AI chat, or it did give an AI overview, but on the same page, normally you would see a whole load of links, right?
Simon Yong: It only showed one sponsored link and two organic links, and that was it. And there was no opportunity for me to scroll and choose more pages either. So it felt like they're experimenting again with more formats of search display, which they clearly are, right.
Paris Childress: You know what, one, one of the other stories that I covered today was that in India, Google is experimenting with putting... This is for e-commerce they... And then they're partnering with this big e-commerce player called Instacart. Do you know Instacart I think that they came out of India maybe. Anyway, if a user is having in Google AI, this is Google AI mode, not AI Overviews.
Paris Childress: But if a, if a user's having a conversation about, about a product, then at some point in that conversation, Google will serve up a buy button. So if the user says, "Yeah, I... You've given me all the details. I like this product. I wanna buy it." I mean, if there's clearly the intent to buy, then Google will serve the user a buy button, and when they click on the buy button, they don't go to Instacart. They open up, they open up a little...
Simon Yong: Like a pop-up.
Paris Childress: Seen a... Yeah, it's like a little, a little bubble inside of AI mode, you can complete the transaction. You can buy via Instacart inside of AI mode.
Simon Yong: But this is what's Jeff gonna do about that? Like...
Paris Childress: Amazon or...?
Simon Yong: Well, Mr. Bezos, Jeff, also.
Paris Childress: Yeah, yeah.
Simon Yong: AKA Jeff, he's going to... Something's gonna have to happen because, well, he's got Rufus now, hasn't he? So he's got his own AI, but it's within Amazon. It's not a, as, as I under-.
Paris Childress: Taking things even a step further than the agent going to the e-commerce site to buy on a user's behalf. This is actually the purchase itself is completed inside of Google AI mode, I guess it's kind of a... If I'm Instacart, I'm thinking, "Okay, well, it's sad that people aren't coming to my site to buy, but think of how many additional sales I'm gonna get," you're removing even if it's a little bit of friction, I mean, the buying experience is still pretty smooth.
Paris Childress: But if I'm if I'm inside of AI mode and I might want to stay there and I'm not quite ready yet to click away and leave and open up another website, I there's gonna be incremental purchases, and I think that AI mode, if Google's getting some kind of a commission an ad revenue from that.
Simon Yong: Yeah, it's gonna do at some point. Well, it'll just be the shopping model, I would imagine, wouldn't it? It'll be paid promotion. But did you see the... Well, and I don't know whether this might be completely fabricated, but people have been saying that if you run ChatGPT ads, your citations and mentions are higher in ChatGPT.
Paris Childress: Have you, have you been seeing that? I haven't seen.
Simon Yong: I've seen, I've seen a load, a load of posts about that over the last week where people are saying, "Now there's no... There is no proof, is there?" Right? If there was proof, it'd be an absolute scandal. But, surely they can't be doing that because if they were, then the credibility of the platform would be at question.
Paris Childress: Yeah. Yeah. Wonder if there is an in- indirect relationship if I'm heavily advertising ChatGPT ads, when my ads show up I guess those don't really count as citations, but I would, would the You know, brand mentions actually matter for citations, but my brand is mentioned repeatedly through ChatGPT ads, would, that, would that start to indirectly trigger that, okay, there, this brand is appearing a lot more on the web? Or I don't know.
Simon Yong: Maybe, maybe that's the incentive for all of us to go and pay for a few ads.
Paris Childress: Yeah. I mean, they've, they've been saying that forever about SEO Google, Google Ads and SEO, that your Google Ads can impact your SEO.
Simon Yong: Well, Matt Cutts even came out and said there is no correlation between it, right? And just because Matt said so, Matt said so can't be the case. But I've... Well, from everything I've ever seen and all the campaigns that I've run over the years, it is true.
Paris Childress: I think there's an indirect relationship that when somebody is clicking on your ad and going to your landing page and converting, I think that those sending powerful organic signals that is a, this is a brand that not bouncing and they have a good user I mean, a good experience. They have a high conversion rate, so I do think it could indirectly impact your...
Simon Yong: To talk to the, to the not bouncing theory, the citation, i.e. The production of the content, i.e. New knowledge for the cohort to go into there, like information gain. As long as you're keeping that succinct and to the point, and people are finding it and liking it, and it answers the question, it's working. It's the people who are trying to write massive long form content, I think that's where there's a problem.
Simon Yong: You know, because they're trying to still satisfy the need of every potential buyer or person that might ever see my page, I'm trying to write for them. I think we're moving away from that. That's what it feels like anyway.
Paris Childress: Yeah, absolutely. We're moving away entirely from writing for human readers because reality is that people don't really go and read blog posts on a, on a random website anymore if it's... Most of my reading, I don't even really use Substack or these other platforms. I don't read much there. Most of my reading is either very short form in social media like LinkedIn or X, or, I mean, if it's long form, it's just more in my, in my news feeds.
Paris Childress: I can't, I can't tell you the last time I went and read a full blog post on a, on any company's website.
Simon Yong: But there must be more, more blog posts. I don't know, I'm maybe making this statistic out, up, but I would imagine there's more blog posts this year than there's ever been, just because the ability for people to publish them more quickly and easily. Although nobody's reading it.
Paris Childress: Slop. Yeah. Yeah, just a lot of slop. Yeah. I was thinking about this visual for presentation to have a bunch of pigs in a trough, eating out of a trough and to make a reference to AI slop, and then to pivot that to another picture of pigs that are hunting for truffles, and to say that basically that the difference, the difference information gain is the difference between truffles.
Paris Childress: That this is what the pigs w- really wanna find and they would eat it unless the humans weren't there to quickly snatch it up. But versus eating out of this trough, eating the slop out of a trough, which is what we're shoveling to them now.
Simon Yong: But too many people are doing that, right? There's... It just inherently the internet or the continent- content on the internet has been produced as cheaply and as quickly as possible. We're not gonna move away from that, and the quicker people understand you're gonna have to do it better and properly, that's the whole message, right? We- y- you're...
Simon Yong: I think, maybe I'll be wrong, we can revisit this video in a year's time, but I think you're still gonna find in a year's time that the vast majority of content is produced and is slop and not necessarily augmented with any particularly interesting additional information by the majority of businesses. So the opportunity is for everybody, if they are listening to this, to think, "How do I produce a content plan and stick to it that is going to be significantly able to provide information gain", right?
Simon Yong: That's way beyond what... I don't, I don't know that the majority of marketeers understand that. Do you?
Paris Childress: I think a lot of people in performance marketing do fundamentally understand it, but to be able to execute on it different thing. Yeah. It is hard. And I mean, that's really what we're trying to make as easy as possible with GEOforge, is to have a knowledge base that is persistently getting fed with either the latest sales calls customer support transcripts or logs, subject matter interviews that we can capture now through an AI interview, AI voice interview.
Paris Childress: We're trying to remove as much friction as possible and set up integrations where we can... In an automated way, we can just capture the knowledge that's already happening from what people are saying in the business over the phone and with this recording, recorded meetings and all that.
Paris Childress: And I think as the business as the business world as a whole gets more and more comfortable with recording everything, and I still think I've been recording stuff for years, but I still think that majority of businesses do not yet have meeting note takers going to all their meetings and then recording. I think that's, that is starting to change...
Simon Yong: That can change really quickly, can't it? As soon as everybody gets hold of a charm or a pocket or a app that'll do it.
Paris Childress: I think right now for privacy, people are still a little bit skittish about something that's listening and recording and... But I think that's gonna go away eventually because they're gonna start to really see the benefits of all that context if you give that context to AI, and maybe they'll see this before even the marketing use case.
Paris Childress: But if AI just had a knowledge of all of the only every word spoken in every meeting internally and with sales prospects, with customers, but also every Slack conversation, every email, imagine all that getting pooled into one central AI knowledge brain. Think of the strategic value that AI could then give to the organization. It could, it could flag issues that would otherwise never be detected, and it could surface new ideas.
Paris Childress: You know, maybe I'm the CEO and there was a great idea that just got bounced around in a meeting that I didn't attend, and I would otherwise never have gotten exposure to that idea. But AI would, would detect it in a meeting, and imagine this is an organization with thousands of people having hundreds of meetings a week somebody had a brilliant idea, it was recorded in a meeting, and AI picked it up and actually sent it straight to the CEO and said, "Hey, there is a really awesome idea here that never would've s- it never would've bubbled up to you.
Paris Childress: I found it. Here it is." That's what's possible when you record and you're willing to give of every single piece of communication happening in the company.
Simon Yong: There's, there's obviously tools out there or consultancies, etc. That would charge tens of thousands of dollars for the insight that, you know, you can now get on a daily basis from if you've set your AI, AI up correctly. Like one of, one of my clients Decoded, they're a recruitment company in Manchester, as you know, and they uploaded six months of their telephone calls, and it started telling me which agents were doing the best work and why.
Simon Yong: And, you know, it was one of, one of the guys in particular was having more conversations, longer conversations, which were then, you know, resulting in him probably closing more. So they can... You know, it's not just a training issue at that point, it's... But eventually the AI will start having those conversations, right?
Paris Childress: Yeah. And then you could take it a step further and say that this outstanding salesperson who's having longer conversations and closing deals, let's, let's turn some of what he or she is saying into some training material. Let's, let's actually try to mine training material out of his transcripts to see how we can, get some of the other sales agents to whatever techniques he's using.
Simon Yong: Well, it should, it could almost just be a prompt and pop up on a screen next to you. You know, while it could be listening to you and I chatting away, let's say I was trying to sell you, and it could be saying, "Talk about this. Talk about this." You know, like the properly prompting me to actually... Obviously the AI could, in theory, in the end, do it itself. But the ability for that to push a conversation in the right direction.
Paris Childress: Yeah. Yeah, absolutely. All right. Well, I think probably we can wrap it up here, Simon. It's... I think if you're open to it, I'd, I'd like to have you join me a little more often for maybe some shorter, more focused talks and.
Simon Yong: It's good.
Paris Childress: About this actually before we recorded. But as you know, I'm covering this daily morning brief, and it's been mostly a monologue. But think that there's another format where I can pick... I usually cover about 10 stories and usually there's one or two or three of them that are a lot bigger and more important than the rest, and then I usually try to lead with the title of the episode and the thumbnail and everything. I try to lead with the biggest story.
Paris Childress: Maybe we can start to take those outsized stories, and we could have some longer discussions and debate, maybe 15 or 20 minutes one or two of the bigger stories and bigger developments of the week.-.
Simon Yong: More digestible for the people that are following.
Paris Childress: Yeah. Yeah. I think that's where we should go next. All right. Well, well let's wrap it for today. Simon, it's great, great talking with you as always.
