Why is it that people do not value marketing and experience and advice and knowledge? Somebody has spent a reasonable amount of money buying a piece of software to help organise their sales process. It didn't work. They speak to us. We suggest that we can take a look at what they're trying to achieve. Figure out how best to do it and then implement it. They're not prepared to spend the money, but they're also still frustrated that they can't get this piece of software that they've bought to do the thing that they wanted it to do.
If you think about larger companies who make eye watering investments in software, they factor into that years of one thousand pounds a day type consultants to help them actually figure out how to use it, get it all rolled out. If you're going to invest in software, just make sure that you've got some budget for making sure that you can leverage the software that you're going to be paying a lot of money for. So welcome back to Digital Marketing From The Coalface.
Generally been fun, isn't it?
It's now the forty third of January I think, isn't it? Forty fourth forty fourth of January. And, uh, yeah, so we're recording a first podcast of God.
So what it's like, isn't it?
First podcast of twenty twenty.
Hasn't got like yet.
Uh, no, it's still dark. It's pretty much dark all the time. What a cheery pair we are. But, um, yeah, we thought we would, um, after the huge success of the end of year podcast, he said with a smile on his face, we thought we will continue with the podcast. Um, because if nothing else, we enjoy doing it. And, you know, the handful of people who do listen to it seem to enjoy it as well. So same format, just chewing the fat, talking about the stuff that we've been doing and Julie's written loads of questions down, whingeing. We're definitely going to do some whingeing today I think. Yeah, but I think in January you're allowed to whinge aren't you?
It's the month to whinge.
Yeah. I'm gonna keep coughing as well. It's gonna be like the Fast Show. I'm basically I'm basically Bob Fleming, so just bear with me.
Oh, God. Apologies in advance. It's probably not going to get any better.
You're doing an evening with. Oh, really? Yeah. There's. I think there's one at the soon at the armadillo, which is a venue I really like down in Glasgow. I quite fancy that. And speaking of evenings with. I've got. I'm going to see an evening with Gregor Fisher.
Oh really? Rab C.
Nesbitt. Yeah. In Perth. Um. I think that's just smart.
Nothing ever happens in Perth.
I know. I've been there once before and I can't remember who I went to see. And it's a really nice venue. Oh right. So I'm going to see. Yeah. I'm going to see Gregor Fisher. I think it'd be good.
Yeah. I think he'd be.
Really liked his work. The most recent one. Did you see the most recent one which was the one with you, Gary Tank commander.
Oh, him? Um, no.
Only child or something. It's called, I think, something.
I haven't even heard of that.
Yeah, it's really good. So I've always liked his work and, you know, he's made me laugh a lot over the years and. But he's very good at pathos as well. If you think about love actually, the role he played in Love actually, there was a little bit of pathos in that. It was.
Really nice. It was a very sweet character.
I mean, you know, it was.
Oh, it was a brilliant.
Film. Well, it was no it isn't. It was, you know, he played he played a good part in it anyway. Have you got any digital marketing related questions.
Um, to start with, there's been, there's some, some lovely new buzzwords kicking about. Oh go on. And we have the great decoupling.
Okay. Isn't it about relationships.
No no no. Yeah. Maybe this is a guess what this buzzword means.
I'm all ears because I have no idea what you're talking about.
So so this is about and I've seen it in two different places. So it seems to be a thing rather than just something somebody just said. It seems to be a thing that's being talked about in various places, but it's about how you used to do SEO, which made your rankings increase, which made your website traffic increase, which made you get more leads. And it's all the great decoupling of, of the linear model of what SEO does, um, because of AI and it's now like, um, SEO, um, maybe mixture show up in an overview or it makes you show up in, uh, um, you know, Gemini or ChatGPT or whatever, or you show up in social media. You show up in lots of different places, but it doesn't actually lead to anyone clicking through to your website. It doesn't necessarily lead to traffic, but it does then lead to somebody going, oh yeah, um, who was that thing I read about in ChatGPT? Um, I'll go and look them up when I'm ready to buy and then they go, but it's not the, the obvious sort of, you know, measurable line.
Saying.
This because.
It's demonstrably not true.
Search engine journal and search engine land.
But if we look at the actual stats. Yeah. I mean, I think it's going that way.
It's going.
That way. Yeah. And I think we are seeing traffic coming to coming to us from AI. We're seeing traffic going to some clients more than others.
Yeah. I think.
Um, I think you're right.
B2C, it's um, it's happening a lot more. I think it's slower in B2B be, because it's not as.
If it was. If it was happening on any sort of massive scale, then, you know, Google would be in a bit of a predicament, seeing as it makes all its money from people looking, going to Google, typing something in, clicking an advert and going to the website, it's still sending people to other.
People's websites for like high, intense stuff. So things that people would search and put ads for. It is still working. It's more the informational stuff. So people aren't, um, you know, clicking through to read the, the early stage informational blog, posting things and then kind of following the path there. Um, they're getting that early information somewhere else. And then then maybe clicking on an ad or whatever. But the anyway.
Well, I think it rings true, doesn't it, that if you produce good, useful content, whatever that is, and you present it on your website in a well structured way, and you add in potentially add in structured data. Yeah. And you, you know, you get some traction with that content, then ultimately it can help with search, but it can also help with AI search. It will help you appear. It will help your content be cited.
It will. But it doesn't lead to immediate clicks like it used to. So you can't you can't measure it the way that you used to. Or the path isn't, isn't the same. So that's what.
It will always lead to. Clicks. Maybe you know the numbers in decline.
Yeah.
But the numbers are still enormous. Oh yeah. I mean, if you're thinking, ah, right. I won't bother, you know, creating content so that I rank in a in a normal Google search, you'd be making a mistake, wouldn't you?
Yeah, absolutely. I think particularly in the in the areas we work in. But I know that is um, that's the great decoupling. So um.
Okay.
That's what it's all.
About in search engine land and people like that. I mean, you've got to remember that people, you know, organisations like Search Engine Land, which I've got a lot of respect for, but they've got to write about something. Do you know what I mean? They've got it's just like, it's almost like, you know, like journalists in general.
Well that's it. There's got to be a story.
They want stuff to write about. They want to scare people. They want it to be sensational because guess what? They want clicks. True. You know, so, um, I don't disagree with it. I'm certainly not a Luddite. And, you know, I'm, you know, more than anyone, um, making great use of as much as anyone rather, um, making great use of AI. And I can see how, um, you rewind a little bit. Google has been trying to do this for a long time, not even with AI. You know, if you think about booking holidays and booking flights.
And yeah, it's.
Looking what the weather forecast is and they've been trying to keep people on Google.
That's it. Yeah. Featuring snippets and all that stuff. So it's the same.
This idea of like these companies which that have flourished on the back of nicking everybody else's content or at least indexing everybody else's content. They are, um, you know, they've been trying to do this for a long time, you know, basically bite the hand that feeds them.
Totally. So present the information.
What it does is it creates this feeding frenzy. It creates this kind of, um, uh, atmosphere of fear, like, oh, you know, what have we got to, we've got to do this now to keep the AI engines happy, you know, because like, it's been tedious for the last fifteen, twenty years. Oh, you know, what does Google want? We've got to do what Google wants. Google was created to index your content to make it easier for that content to be found. Creating the content was the hard bit. Yeah, it turned into like, you know, pleasing Google was the hard bit and chasing the never ending, never ever changing algorithm was the hard bit. It's tedious, isn't it?
Oh yeah.
Absolutely utterly.
Tedious. Um, you know, and you're also you're chasing something that you don't quite know the rules of. It's like, um.
It's like marketing in general.
Playing a game of football and not quite knowing what all the lines mean and not quite knowing where the goal is going to be.
When you get an analogy would be rugby, because most people don't understand rugby at all. Or cricket. Rugby, yeah. Just like rugby. Yeah. You know what's all them like? What are them lines for and how many players are there and why? What just happened? What just happened? Yeah, like.
Different numbers of points for different numbers of things. Yeah, probably just Google's basically like, like rugby. That'll do, which is good.
As we start to move into the Six Nations rugby season. True. Only yeah. Just just under just under two weeks to go till the first game this year is on a Thursday.
Thursday.
Thursday night. I don't know. I think maybe somebody forgot to book Stade de France and Ireland and France. What are we going to do then? We can't play. We'll have to play on Thursday. Oh God, I don't.
Know. Ireland and France. Maybe there's some like Saints Day on the Friday. So they and they need a day off or something and then they have the weekend off.
Yeah. Well France's secular isn't it.
France loves a religious holiday like Ireland.
Just holiday. Yeah. You realise Leslie can hear us.
She's also gonna edit it so she can always edit this out.
She's come to work today, but she's on holiday. She's doesn't tell.
Anybody. You know what, when this podcast doesn't go out because Leslie refuses to edit it because she's. She's built a barricade. Mhm.
Okay. So so what are the what do people have to think about in plain English in respect to the great uncoupling? Well.
The sort of outcome of the articles I was reading was, um, basically instead of just measuring traffic because traffic's not that great an indicator anymore of, of what's happening and, you know, measure things like the quality of your leads and how long people stay on your website and the conversion rate and stuff like the stuff, the sensible things to measure anyway, because website traffic, we've always said is only like a small part of it. And just because you've got the traffic doesn't actually mean you're going to make more business or anything. So it's measuring what happens once people get your website, because a lot of people seem to be finding that even though their traffic's gone down, their leads haven't, you know, so the people that are coming are more engaged already, more, more prepared to, to buy what you've got to sell. So it's just looking at measuring slightly more sensible things than just traffic.
Okay. I guess brand is, as always, been vital. It's always been of the utmost importance.
And that's what.
They were saying kind of more even more so now.
Was kind of part of your brand building.
It is. Yeah. Always has been. But but even more so now with with with the with the answer engines.
Yeah. So it's, it's as much about visibility as actually getting immediate traffic. It's about getting your name out there and making sure people remember you so that when they do come to buy, they, they go straight there.
Yeah. Yeah. I mean, ultimately for many things, most things, you know, the the answer engines, ChatGPT, Gemini, Claude, they can give answers. But ultimately you're like, fine, I want somebody to do it for me. Yeah. I want somebody to build it, paint it, you know, screw it together, design it, whatever. And, and so ultimately, people are still going to come to your website. And as you pointed out, what's happening or what will happen is that the people will be even better informed when they arrive than they've ever been. They've always been this idea of creating great content, and people are eighty percent of the way or whatever the figure was eighty percent of the way through the buying journey. By the time they actually engage with you, maybe it will be even more than that. Maybe we're going to get to the point where they're kind of ninety plus percent through the journey.
They're really convinced to have a conversation. Yeah. Because, um, you know, the, the answer engines have already kind of done the selling for you. Hopefully. Yeah. If you know, if you've got it right. Yeah. The other buzzword, I don't know if you've heard of it, but you know, you know, long tail keywords. We talk about long tail keywords.
What they are.
But yeah, the, the less, the less frequently used keywords that make you, um, you know, the very specific queries, which actually are going to get more important when people type longer things into AI engines, um, short head keywords.
Short head keywords.
I had never heard that. And I think somebody has just made it up as the opposite of long tail.
Yeah. As in like the headline was like car insurance.
The big one would.
Be a short head as opposed to car insurance for read m g midgets. Yeah, sort of thing. Yeah.
So, um, I just just made me laugh because I have never heard that before. And I'm pretty sure somebody just made that up.
And did you, did you, did you look into it? Is that what they meant by it?
By the that is what they meant by it. Yeah. Yeah. The opposite of long tail. But I was like, I have never heard that.
So they've always been referred to as the, the head terms. The head search terms. Yeah. The head of the tail, if you like or whatever.
But short head is the opposite of long tail.
Important ones. Yeah, I don't know.
Yeah.
Well, you know, we are in an industry and, uh, that likes an acronym and likes a buzzword and.
Likes to make things up.
Likes to make shit up, basically. Yeah.
I yeah. So, um, you know, if anyone's actually come across that And can tell me that it wasn't just made up by that person. Then let me know.
It sounds like it was made up by something with nothing. Somebody with nothing in their head.
With a very short head.
Yeah, with a short head. That's right.
Um, yeah, that's all I've got in buzzwords just now. But, um, there will be more, I'm sure.
Yeah.
As um as things move.
A lot of notes though, which is encouraging. We might, we might manage to fill their fifteen minutes actually fourteen already.
So can I do my whinge now then.
Of course you can whinge away.
So right. Why is it that people do not value marketing and experience and advice and knowledge?
We're not alone.
No, marketing.
Is not alone. We live in a world where everybody thinks they're a fucking expert at everything, because they've watched a YouTube video or they've read something online. So I was speaking to, uh, Gary earlier on, who's a local painter, about some work I would like him to look at for us, you know, and if you speak to Gary, he'll just say, oh, yeah, everybody thinks they're a painter because paintings. Easy, right? Yeah. Anyone can paint. Okay, if that's the case, why do people serve a four or five year apprenticeship to become painters? It's a bit more to it, you know. People take their car to Scott down at Dinnet garage and say, I do it myself, but I haven't got time to do. All you need to do is X, Y, and Z. You know what I mean? Can you imagine the conversations they have? Oh yeah. So you are one hundred percent right. But all I'm saying is we're not alone.
Yeah. That doesn't make me feel any better. Okay.
Go on. Carry on and.
Win. So, um, somebody who will remain nameless, um, has spent a reasonable amount of money, um, buying a piece of software to help make their life easier, organise their sales process.
It's HubSpot, isn't it?
That's the one. But part of HubSpot, the sales part of HubSpot, um, so that they, they wanted to automate their processes and make their team more efficient, save lots of admin time doing stuff manually. So, um, they got it. They set up using the HubSpot AI, what they thought was going to make their life easier. It didn't work. They spoke to HubSpot, who pointed them in the right direction. It still didn't work.
So did the people at HubSpot. Were they unable to help them, or were they just suitably not very good at their job that they all they managed to do was confuse them?
Yeah, I think they just probably said, go read this thing and, um, you know, read, read the instructions, basically.
They said this, go and read these instructions because they're super helpful.
Obviously.
Obviously.
Yeah, they speak to us and we suggest that we can take a look at what they're trying to achieve, the objective, what it is.
They understand their problem.
What they want to happen. Figure out how best to do it, which clearly isn't the way that it's been done at the moment. And having had a quick look at it clearly wasn't the way it's been done at the moment. So figure that out and then implement it, but also show them how to implement it so they can do it themselves in future. Cool. I'm not unreasonable amount of money for that, just a very standard fee for the amount of work that was involved. Way too expensive.
What were they expecting? Do you think? What would have made them like. What would have quoted them? Happy?
I don't know, I think it would have probably had to be quite a lot less like fifty quid. I've no idea. But maybe just for the the actual training without the figuring it out beforehand. No idea. But they're still stuck with HubSpot that isn't working, isn't saving them any time that they can't work.
I remember you telling me it was only a few hundred quid. It wasn't actually expensive. It's not something that we particularly wanted to do. But when you see people struggling, it's in our nature to say, right, we'll see if we can help you with that. Yeah, quite a sensible amount of money.
But I'm like, okay, I'm not gonna do it for nothing, but I'm not gonna also, you know, quote, a ridiculous amount, just that, you know, it was a very fair quote. Yeah. But no, they're, um, they're not prepared to spend the money, but they're also still frustrated that they can't get this piece of software that they've bought to do the thing that they wanted it to do, and save them a lot of time and make them more efficient.
Was it mostly they didn't see the value of paying us five or six hundred quid to sort a problem out for them, or they didn't have five or six hundred quid?
Well, they said it was beyond their budget or not. Not in line with the budget or something.
And at the start, did they say what their budget was?
No they didn't.
They knew that five or six hundred quid was more than their budget, but they didn't know what their budget was. That's another conversation you can have lots of times, isn't it?
Yeah. Yeah, exactly. Yeah. What budget have you got? Oh well we don't know. And then you tell them they're like, well it's not that. So we end up having to guess. But um yeah, the value of our experience and our knowledge and the value of their time in terms of the time they would save if they got it sorted. Yeah, yeah. I've baffled flabbergasted.
Well, I can, I can sort of match that seeing as we're in whinge mode. So we were contacted by an organisation who'd been flogged, HubSpot, and they had been sold a sales licence for twenty odd people, none of whom needed a sales licence. But HubSpot persuaded them that they did. It was the only way to achieve what they were trying to achieve. But it wasn't. But it wasn't. No. And we had some great conversations with them, and they really needed some help. And we then figured out that to achieve what they wanted to achieve, we kind of needed to develop some middleware, something that would talk to a scheduling system that they had and talk to HubSpot, get rid of the twenty odd sales licenses in HubSpot. They wouldn't need them anymore. And the people who were delivering the, the work would, uh, would simply be able to go to their Google calendar and say, right, I've got to be here at this time to do this. And it would all be really cool. And I think we said it's going to take five or six thousand quid or six thousand quid or something like that to develop the piece of software which was needed, and we couldn't see any other way of doing it. Yeah. We don't know if it was beyond their budget because they just didn't reply. They just ghosted us after. After helping them and figuring out the problem that they had, going back to them with a sensible solution, they simply didn't send us any more messages. They simply ignored us.
That's just rude.
Well, it is rude, but it does speak to the same problem that you've got. It's just this kind of dismissive, oh, that's too expensive. I'll go and ask somebody else the same question and I'll get probably get the same answer, you know, but I don't quite understand. You know, the main thing is that if you make an investment in some software, if you think about larger companies who make like investments, eye watering investments in software, I mean, yeah, my sister works at BAE systems, um, down in, down in Barrow in Furness. Um, they build nukes, nuclear submarines. They implemented, I think SAP something like that.
Right. Yeah.
And they factor into that years, years of one thousand pound a day type consultants to help them actually figure out how to use it, get it all rolled out. Now I know that, you know, that's like a big, um, you know, it's not exactly comparing, you know, like for like it's a big thing, but if you're going to invest in software, if you haven't got in-house expertise, just make sure that you've got some budget for making sure that you can leverage the software that you're going to be paying a lot of money for, for the, you know, months and years to come.
Absolutely. Because otherwise it's, it's sitting there and it's a white elephant sitting on your desk.
I mean, obviously there's another option now that you can literally talk to an AI engine and you might, you know, realistically be able to muddle your way through and figure it out.
But the problem is that Hubspot's AI engine built the, um, the workflow that wasn't working and it wasn't working because it was way, way too complicated. And Hubspot's AI engine hadn't followed Hubspot's best practice in terms of how to build the, um, the workflow.
Well, I mean, that is one aspect of AI, which I love AI, but that is one aspect of it. If you don't give it the context, if you don't actually spend the time prompting it properly and giving it context, it will produce crap. Yeah, it will just produce absolute garbage. Um, but yeah, I mean, I hear where you're coming from. It's right across the piece. I think, as I mentioned, you know, speaking to Gary about the painting, I was speaking to Scott about, you know, when he was doing, he did, he did, um, Karen hit a pothole and pop the tire and all that. And it turned out he needed the car needed servicing and MOT and, and everything. So he just dropped the car off and always have a catch up with them. And, you know, it seems to be across the piece. I mean, I think you'd be hard pushed to find a profession or an industry where people don't sort of challenge what things cost. And um.
Yeah.
It is kind of tedious, I suppose really. But the creative industries, I think, and, you know, I'll loosely align marketing with the creative industries.
Um, yeah, it sits down the middle, I think, doesn't it?
No idea. You're you're biased. Um, yeah. I mean, creative industries are just really not valued at all. I mean, everybody thinks that they're a designer. Everybody thinks. And, and the thing is now with, you know, with the with again, with the AI engines, everybody can write code. Everybody can do, do, do the what was used to be the, the sort of clever bit, if you like. Um, I don't know. It's, it's an interesting one. And uh, hopefully you enjoyed your whinge.
I feel, I feel so much better now. Thank you. Thank you for letting me do that. My pleasure. On that theme. Talk about, um, AI and coding and AI again.
Yeah. There's a workflow that we're rewriting the software we use for kind of giving us the, the, the sort of real time overview of where we are with all the work we do every month, because we do a lot of retained work and we do project work and we built some software that helps us with that, and it was time to kind of refine it and take it up to the next level. And in so doing, Phil, how our Tim geek, um, has been. Yeah, Tim has been leveraging Claude code to write it and had a Claude code frenzy over the, over the Christmas break, uh, writing, um, a new kind of like, like a base app. So if you want a web app, you always need to be able to log in and do this and do that and do the other. So he created this kind of base app that you can use, or rather another base app that you can use. And in so doing, we use that base app as the starting point for our new, um, uh, red evolution deliverables management system, as we call it. Red man. Um, he used it for that and he set it up in the workflow with, uh, with git, which we use. Um, you know, where we put all our software repos. Um, he set it up. So when I find an issue, I can just start the issue in git GitHub with at Claude, explain the issue, and then Claude goes off and fixes it and does all that work. It has been interesting if if a little frustrating at times. It has been interesting and pretty productive. Um, I also asked Claude because he wrote the software to create a test plan for us.
Oh yeah.
Which I was then going to sit and work through. And Phil said, well, why are you going to do that? Just give Claude control of your browser and tell Claude to work through the test plan, which we also did. And it was it was really smart the way it did it. However, when the humans got hold of it earlier today, the three of us sat around that desk over there and started to say, right, let's try some real world tests and real world use of the software before we roll it out to the rest of the team. And it fell on its ass almost straight away, didn't it?
It did, but it was pretty minor stuff.
It kind of was, um, minor stuff, but they were all showstoppers. It was like we couldn't continue using it and say, oh, we'll just go back to that. It was it was too big. Now, you know, it's already been fixed, apparently. So we can go back to it again.
It has. I've checked one of the things which is fine.
Yeah. But yeah, I mean, there's so much chatter on TikTok around, you know, Claude code and vibe coding and everything else and what it's doing, it's, it's almost like the desktop publisher software on stilts. You know, when, when DTP software came out, everybody was suddenly a designer and producing. Yeah, truly awful stuff. But but yeah, and I think we're in a similar place with, um, with the AI generated code because it does let anybody write code. If you can talk to a machine and say, I want, I want it to do X, I want it to do Y. I want this software to do this, that or the other. It will write the software for you and come up with, um, you know, like, why don't you make it do this as well? And I could make it do this as well.
Yeah. That's the thing. It comes up with suggestions that you haven't even told it.
Yeah. So, But in amongst all that, I mean, the really clever bit about putting software together has always been getting the architecture right, making sure that you're not painting yourself into a corner, making sure that you're creating an architecture, which is going to be, um, economic with the way it uses the processing power. Um, it's going to, you know, fail safe. It's going to use the right stack to build itself. And, you know, this isn't my area of expertise necessarily. So, you know, but, but that kind of.
You know, planning and thinking.
Yeah, what, what it's done is if you can do the planning and the thinking and really put some heavy work into that heavy duty work into that, then sure, you can run the AI to write the actual code. Yeah. Because writing code to a certain extent is about understanding syntax. It's about understanding more than that. But I saw something I don't think I mentioned this in the last podcast where or maybe I did, but anyway, maybe it was just chatting to the guys.
So nobody listened to.
It. I did well, that's right. Anyway, yeah, a developer had said, you know, a seasoned like fifteen, twenty year developer had said something along the lines of he said, basically, it means I can now deliver across a whole range of, um, languages that I don't know the syntax of these languages, but I know enough, if I look at the code to make sure that it's doing it the right way, I know what to tell it to do. Um, because if you've ever done any coding, whether it's JavaScript or C or Java or whatever it is, you know, remembering the syntax is half the battle and a lot of tools like storm PHP and tools like that. They had this, you know, they've had this autocomplete function for a long time where it kind of knows, oh, you're writing a, you know, you're creating an object. So I'll put in all the bits and bobs.
And structure and.
You can fill in, fill in the custom bits that you need to do. So it's been doing that for a while. We kind of moved on from that now where you just tell it what you want and it'll just do it and it's amazing. But again, I don't know where, where it ends. Um, Because you may finish up in a situation where just like, you know, when AI, everyone started getting AI to write the content, we just basically swamped the internet with garbage content, which worryingly, maybe the AI could then use to train itself on the, on, on, on the garbage it produced. So that that feels like a death loop to me.
But yeah, absolutely. Ever decreasing circles of nothing new ever appearing.
I think it's at its most powerful with people who know what they're doing. They could write it themselves, but this just lets them get their ideas and get the prototype built much quicker. Yeah. Um, it's clearly a capable of much more than that, but it does kind of open your mind a little bit. Um.
Are we going to end up in the position where there's a lot of um, software out there that's on very shaky code and if you go and try and do anything to it, it's all just going to fall over. Or do you think it's more robust than that?
I think it's more robust than that. But but there is some truth to what you're saying, I think. I think it's, you know, if you were using, um, an AI tool to write some code for something mission critical, you know, you're probably not going to sleep at night, I think. I think there's an issue there. Yeah. But if you just need something that works to get a job done, then it's, it's brilliant at doing stuff like that. I mean, making spreadsheets sing and dance and do things which take forever when you're trying to write a spreadsheet formula. Well, just, you know, tell, you know, tell the AI what you want to achieve. It'll right. It'll write a JavaScript and ActionScript or whatever. And then you just trigger it and it'll do all this amazing stuff. So I think it's, it's kind of democratising the creation of, of some of the tools, um, software tools. And, you know, what's it going to do to, to, uh, to SaaS companies because SaaS companies are often like, they have a great idea. They then have to develop the software. And I know this because I've been in a SAF and still am in a SaaS company. So, you know, you then have to pay developers to write software to get the thing to do what it is you want it to do. What's to stop me now? Just like looking at a SaaS tool that's just come out and going, yeah, that's great. And hey, Claude, go and look at that and copy it. Right, right one. Right. My version of that. So I.
Don't make it.
Better and make it better, but it's also yours. You don't have to pay a subscription. You can just, you know, tweak it and tell it to make it do more stuff as you, as you, as you go along. Um, Christ knows. I don't mean I honestly don't know where it's going. It's it's it's mind blowing. It really is.
I know this time next year we'll be having we don't we can't even predict what sort of conversations we're going to be having.
Because this time last year you couldn't do what you can now do.
No.
That's right. It was like generating dodgy content. Yeah. And yeah, I mean, people were using it to do way more than that. People who were at the cutting edge of it. But now all this stuff we're talking about now is all mainstream.
Yeah. Yeah, exactly. And the, and also the content that it's writing is way better. It's, um, it's got a lot more sources or it's going deeper. It's the quality of what's coming out is better.
Um, I think people are learning to, to talk to it more true. You know, we've said this before, I mean, you know, like set the context, give it as much information as you can, encourage it to ask you questions. Yeah. You'll be very pleasantly surprised by the questions it asks you.
Yeah, it's quite good if you say, I want this and do this, what else do you need me to tell you? What else do you need to know? And then it'll, it'll ask you the other things that will help it give a more precise output.
Yeah. That's right. And yeah. And, and taking your time with that, there's, you know, you can very quickly just chuck something at it and it'll go, yeah, is this what you wanted? You know, whereas if you, if you take your time and spend time developing the context, and then then once it actually gets on with the job in hand. There's half a chance it'll. It'll do more. Something more productive. It'll create something that's closer to what you're actually. What you actually need.
Yeah, definitely.
Where I'm finding it particularly useful is, you know, I'm one of those people who, who can get, I don't know, almost a bit down staring at a screen, trying to start something. I kind of know where I'm going with it, but I just struggle to start it. And that's where it's really helped me. It's just kind of like you can spill the beans to it, give it the big picture. And it might say, well, you could look at this. You could look at that. How about this? This is an approach that's. And suddenly you're like, oh, great, I've got direction. You know that. But I only know that what it's telling me is any good because I know what I'm doing. If you know what I mean. If you know, I could ask it to create me. Um, I don't know. I could say, you know, design me a piece of software that does something or other, like works out whether a bridge is going to fall down or not. And it could go, there you go. There's, there's a there's a piece of software because I don't know how to design a bridge so it doesn't fall down.
Yeah. You have no idea if it's any good or not.
Um, yeah. I've no got no idea if it's any good or not. But I think when you use it to help you and you are a subject matter expert, you are, you know, you're good at what you do, then as an assistant, it's amazing. I know that the, the Claude core work, everyone's talking about that on TikTok just at the moment. I really need to dig into that and see what that's all about, because that's using multiple agents to do things, you know, so you've got basically when you walk away from your desk, it's off busy doing stuff for you, that kind of thing. Again, it's, it's, it's mind blowing. I mean, it's undoubtedly putting people out of work, you know, it must be. It absolutely must be. Yeah. It must be. Um, if you're a writer, you know, freelance writer, for example. Um, that's, it's got to be a challenge.
Yeah. I mean, unless you're very technical and it's, um, you know, a very specific area if you're just sort of churning out blog content for, you know, SEO purposes. Um, there's not a job there anymore.
No, it's the same with, you know, with, um, you know, some of the sort of low level software development, like creating website templates and things like that.
Yeah.
Here's a design. Turn it into a WordPress template and it'll do it now. Chances are it'll still be a bit ropey, but.
But yeah, it's a starting point. Yeah.
We live in very interesting times. We really do.
Yeah. It's um.
Yeah, it just the last twelve months have been mind blowing in, in the, where we are now compared to this time last year. Yeah. And like you say, crazy. Where the hell are we this time next year?
Yeah. We have no idea. You can't predict at all. Um, talking of, um, AI and having it, knowing what.
You're talking about, AI. Well, yeah, it's just a fact.
Yeah. It won't last.
It won't.
Last. Um.
Like the internet.
Google ads and AI Max. Not quite there yet. Maybe it's fine for, you know, high sell em cheap widgets and anything slightly more complicated. It's not it's not getting it yet. And you know, we've tried it and it didn't really get it. And I've read a few other things saying and it doesn't work with low volumes of data. It needs a lot of data to feed the machine. So it can can work out what it's doing. Otherwise it can very easily go off it completely and get the completely the wrong end of the stick or just be so.
Desperate end of the stick you're mixing your metaphors.
Go off and get the wrong end of the stick, like go fetch the wrong end of the stick or something. I don't know what I'm talking about.
Stop talking.
Stop talking. No, if it doesn't have a lot of data, it just desperately goes and tries to find something and, you know, can get very off the beaten track, which is yet again, the wrong metaphor. But it's Monday and it's January, and I'll stop talking now.
Mind you, a bird in the hand is worth ten in the bush. Um, so in conversation with Google on this subject, because. Because we do Google ads, pay per click, paid search and all the rest of it, you know, we get a chance to speak to the reps at Google. Lucky us. Yeah. To be honest, I have engaged with them more over the last couple of months and got more value out of them. But what's the where I'm going with this is like even they concede that because of what we do, and it's like low volume type search terms because we predominantly work in B2B and predominantly that B2B is industrial engineering, that kind of thing, the volume isn't there. No. And that's where the AI stuff really struggles. It doesn't have a large language model to, to fall back on. It just seems to be. Yeah, we've seen it firsthand. And you immediately think, well, I do because of my imposter syndrome. I mean, I immediately think I'm doing something wrong. I must be feeding it the wrong information or not giving it the right guidance. But fundamentally, I mean, what Google the Google guys. You know, I've spoken to said was the volumes not there? Yeah. It's the same with like, if you're trying to feed conversion data back in because we do, you know, we're doing increasingly offline conversion tracking. So we can actually say to Google, this was a good enquiry. Give us more of them.
Yeah.
Even then, if you're only getting two or three a week or, or five or six a.
Month, it needs lots of.
Them. Can't do anything.
I think if you've got lots of them, it can go all right. I can see, I can see the ones you want and give you more of them. But it's just trying to like, you know, use the budget and, and find something. Then it's just, you know, casting its net.
When I was doing some research on this towards the end of last week around like the, the automated conversion tracking and all the rest of it, you know, the offline conversion tracking alongside the, what do they call it, enhanced or advanced conversion.
Enhanced conversion.
Yeah. Yeah. And basically, you know, it was what, where we were getting to with it was fundamentally like with the low volume, high value type keywords that we're working with, you know, manual PPC, manual.
Cost per click to.
Just be, just be constantly in there reviewing it. I mean, you've done a lot of that with some clients recently, you know, where we let the AI do its thing and you like religiously went in like picking out the.
Good in every single day.
Making the negative keyword. And it just didn't get it, did it?
It was no even what we were paying.
For the clicks, some of the clicks we were paying for were, I mean, Google should be giving us a refund. It's ridiculous.
Yeah, they were way off. They were way off. Um, and it wasn't, it wasn't learning. It wasn't learning that that kind of ones weren't right. And that kind of ones were right. It wasn't getting it at all. Yeah. So um, yeah, we completely revamped.
We shouldn't lose sight of what Being Google. LinkedIn for that matter. All of the pay per click models just want you to spend money. Yeah. You know, if they think, oh, well, you know, it might be right. Charge them.
Charge them. There's a big connection there somewhere. Then, um, yeah, we'll stick it out there. But yeah, because it if, if you keep it very precise, it's not going to spend the whole budget. And they obviously go, well there's this, there's like fifty quid available and you're only spending twenty. I'm going to try and spend more of the fifty.
That's right. It's it borderlines on the um unethical. Yeah. In my opinion. Yeah. What they do. I mean they would always argue, oh well you know, you told it this and it thought that and blah, blah, blah, blah. And you know, that key word is, is related to, to the subject.
Yeah. I mean, broad match the. Yes, it's broad, but there's the word match in there as well. So it's not just broad. It actually has to kind of match. And I think it was, um, it was very, a very loose interpretation of that, um, that phrase, what we were doing.
If you imagine like, you know, we look at it and we go and we sort of maybe we'll speak to Google about it, or we'll write, add that to the negative keyword list. So we don't pay for any more clicks that, you know, we've spent like forty quid on, on clicks that were never going to be of any value to us. And you sort of go, well, you know, okay, add a negative keyword, keep monitoring it. And if you think like hundreds of thousands of businesses all over the world, millions of businesses are all doing the same.
They're all forty quid times. Hundreds of thousands.
Yeah. And so you can see why it's just a money making, not scam, but it's a money making scheme. Yeah. And it's kind of not in their interests to really narrow it. Right.
Absolutely.
And when you do with it with paid search, when you do narrow it right down, you know, you just see like, right, we've had, you know, five clicks this week sort of thing because you bring it right down because as soon as you give it a little bit of leeway, it starts, starts showing, showing your adverts and making you pay for absolute pathetic, ridiculous clicks. Yeah, you tighten it right down.
You get nothing.
To try and stop it. And you get and you get nothing. Yeah, it's just not.
And clients don't like seeing small numbers even though like, yeah, these are these are small numbers because we're not paying for the nonsense, but they get nervous when the numbers are small as well. Mhm.
That's right. Um, I am a fan of paid search, but, you know, it sometimes feels like so difficult to actually get meaningful, you know, meaningful clicks. Yeah. You know, it really does because again, because of the niche that we're in, I mean, when, you know, when cam was here and his background was, was B2C, you know, so yeah, we put this together, took it out there, it'll get a million clicks over the next, you know, week or something. And that'll give us loads of data and then we can.
Refine it.
And figure it out. And it all just works because the, the AI is on overdrive because it's getting all this data coming in. But as soon as you move into some of the, you've got like.
Half a dozen clicks. It can't learn from that small numbers. However, in other fun news, um, ChatGPT is going to be introducing ads.
Isn't that interesting? Yeah.
But, um.
Surely others are as well. I mean, surely they will.
They will once, um, if one's doing it, they'll all do it. It's just gonna be um like display sort of advertising paper.
Banners and.
Things. Um well paper impression rather than pay per click.
Oh of course it is. So because we've just really, you know, we've just been talking earlier on that you've got it.
Yeah.
So that you can get loads of information and you're not clicking through to people's businesses and people's websites yet.
Yeah.
So yeah. So they're thinking, oh yeah, people are not going to click through. So we'll just pay for the number of times we show. I mean that is that has got like problem written all over it. Yeah, it really is because pay per impression again, if brand awareness etc. you know, it can be powerful, but I can see where that's going. You know, it's like, yeah, we've shown your advert a thousand times.
Yeah.
Who to? We're not really sure. What were they searching for? Well, we can't really tell you. Yeah, okay. Here's a thousand quid or whatever it is. You know what I mean? It's it's it's great. I mean, it is the model. I mean, the internet runs on advertising, doesn't it?
Yeah, exactly.
Go to newspaper websites to see that.
And oh.
God. They've been put together.
By so hard.
Chimpanzees. I mean.
Just.
Mind blowingly bad.
It's so difficult to actually read anything online that, um, that has advertising. I mean, but I mean, ChatGPT has a paid model, but obviously it's not enough to actually, you know, make it profitable for them as yet.
So you won't see the advertising if you're paying for ChatGPT. Is that how it works?
I don't know. I think there's maybe like an in between. If you pay less like Netflix, you pay less if you get the ads. Yeah. Yeah. Why are you why are you not getting ads on Prime when you're paying for Prime?
That's why are you getting them?
Yeah, because you've gotta pay more to not get the ads.
So it'll.
Be.
Something like that. I started, I paid for Paramount because I wanted to watch mob Land and land, man. I think, uh, I also watched Tulsa King, which was an absolute scream. It's really, it's really yeah, it's, it's really ridiculous, but very watchable with Sly Stallone as a, as a mobster who comes out of prison after a twenty five year stretch. Okay. It's very I enjoyed it. Karen didn't like it, didn't watch it, in fact, didn't see any of it. I don't think she just decided it wasn't for her. But I find it very odd that like, I'm paying them so I can stream these things and yet they're throwing adverts at me. Although the adverts are all for other things on Paramount, I think predominantly.
Okay, the um.
Maybe.
The prime ads are. Yeah. For all sorts of things and then.
Yeah.
It, yeah. The first time that happened, I'm like, wait a minute, what, what have I done wrong? Like why? And then it just, you know, yeah, there are some, some programs that have ads.
TV do it. I think it might, okay. Might do Apple doesn't.
The Apple.
TV doesn't.
The cheaper Netflix has ads. I'm on Netflix without ads.
Yeah, I think I am as well. Yeah.
So yeah, there's a, there seems to be the, you know, free with lots of ads and pay a little bit amount and get some ads and pay more and not get ads. I don't know if that's where ChatGPT is going, but I'm guessing so.
But you know, if you think back, like when advertising started appearing online, you know, the big kind of savior was that it was, it was contextual. Yeah. It would it was advertising tailored to you. We know who you are, what you're interested in, and we're going to show you adverts for things that you might be interested in. And we say to advertisers, you know, we understand like your demographic and we can target them precisely. So like people like me, they're going to show me, you know, adverts for golf clubs or adverts for motorbikes or adverts for cars or um on adverts for um dolls houses. No, let's not go there. Um you know, adverts for things I'm interested in. So you know, it's like that was the big promise wasn't it. That seems to have gone the adverts I see when I'm on YouTube are ridiculous. I mean things that I have no interest to me whatsoever. Really? Yeah. If I'm watching, if I'm doing some research or I'm just watching, you know, for leisure, something on YouTube, like a review, you know, a car review or a motorbike review or something like that. I'll get adverts for um, panty liners, for example, while I'm watching, while I'm a bloke watching an advert about cars.
Are you using the red evolution account or your own account?
I think you know, logged on, logged in as me.
Okay. Unless you've got a continence problem, then that's bizarre.
Very odd. But yeah, the adverts don't seem to be relevant to me.
I get the YouTube, I get a lot of ads for like Monday dot com, which is which we actually use. And then, um, you know, marketing kind of stuff. It seems to pretty much know who I am.
Okay. I don't know, maybe I'm doing something wrong.
Um, you obviously watched something or somebody else watch something on your account. You watch something that's really confused it. I don't know.
It's really confusing me now as well.
Yeah, I think it it's quite easy to set it off on a tangent and um, you just have to watch one thing and then it goes, oh, I'm going to show you lots of that. You're like.
Well, no, that's tick tock. I mean, I'm a fan of tick tock, but if you watch, if you dwell on something on tick tock, I mean, I'll sometimes, you know, I like, I just like video content. Yeah. And I'll sometimes like have tick tock on while I'm brushing my teeth because I'm listening to the news or something. And I'll just put some Tik Tok videos on. I'm just brushing my teeth and, and if I'm in the middle of flossing or something, so I don't get a chance to pass past the video.
So you actually watch something that you don't really want to watch suddenly.
Oh no, I'm gonna get loads of videos like that now. Yeah. So it's.
The tick tock change.
Very quickly.
Tick tock algorithms really sensitive. I think you have to watch a few things and then it goes off on that direction. But yeah.
I still love it. I still, I mean, the educational content on Tik Tok is, I think is great. I think the short form video, you know, it's got so it's got legs. It's still.
Got.
You know it's so powerful. I think it is. I mean, I, I'm just.
I'm, I use Instagram but you know, it's the same. It's reels. Reels are what people watch. You know it's there. It's well over half of everything on that is watched on, on video is.
Usually when you're running videos that you do for TikTok, you stick them on Insta as well.
No, I don't the reason I put them on TikTok.
Oh that's right.
Is because, um, pretty much I didn't, nobody on TikTok knew me and I didn't, I didn't really use TikTok. And so I put running videos on TikTok so that nobody I knew would see them. And then people I knew started seeing them, which was like, that wasn't actually supposed to happen. So, um, no, I don't put them on Instagram because all my friends are on Instagram and I don't want to just be like, oh, I went for a run. So I put it on TikTok just to, you know, for my own accountability really. And as it was, it seemed more anonymous, but I was in the village the other day, um, in a shop, and someone went, ooh, I've seen you on TikTok. I'm like, oh, I'm famous now.
Famous.
Famous.
It's a start.
Yeah.
All you need is some sponsors now. Yeah. Free running shoes and things.
Yeah, I'm working on that. I'm not sure that, um, I'm too far down that road yet.
Okay. What else you got? We've been talking for fifty minutes, by the way.
Um, I'm out of list.
Very good list. Did you just create that list through the week as it went on, or did you sit down today and write a big list?
Oh, no. Well, we were supposed to record this last Thursday, but all the roads were flooded and we couldn't get to the office. So, um, but I put together like a newsletter for us at the end of every month based on like, what's been going on. So, um, stuff I found while I was doing the newsletter. So by next week I have nothing. So it's your turn.
You'll start extending your list, I'm sure. Or I might start putting stuff in my list as well. Again.
Good luck.
I'll try. Okay. Hopefully it won't just be about AI all the time, but obviously AI is what everyone's talking about at the moment and what everyone's using.
I know, and it's hard. It's hard to find any news that isn't AI related, to be honest.
Yeah. Well, AI or Trump, we could talk about Trump.
We could, but I'm going to America in a couple of weeks and I'd actually quite like to get into the country.
Yeah, that's right. You are aren't you? Yeah. Looking forward to it.
I am looking forward to it.
Trade mission to Florida. Yeah. In February.
I know. Great time to go to Florida. And then New York, which is at the moment under about ten feet of snow. And people have been told not to leave their houses. So I'm kind of hoping that will not be the same when I'm there.
Right. You've been listening to whatever it was, uh, but it's called Digital Marketing From The Coalface with me and Julie. And, um, hopefully you've enjoyed it and we'll keep doing it. And if the numbers stay around about where they are, we'll keep doing it. If the numbers tail off, we'll just say it and not do it anymore.
We'll keep going until someone tells us to stop.
Stop. Thank you.