Paul Sonneveld
Hey, everyone. It's Paul Sonneveld here. I'm your host of today's episode. And today, we're going to talk about a topic that I think is on the minds of a lot of agency owners, certainly, as I was in Accelerate, Unboxed the last few months, and I had the opportunity to speak to a lot of agency owners. I can confidently say that today's topic is actually one that most agency owners are quietly worried about, you know and the story goes something like this right AI and automation really has made the acquisition, can't get my words out today execution like faster quicker and cheaper right and clients work out that there's some cost savings there they expect fees to come down and all of a sudden the profit margins that support the agency module are getting really really squeezed right so there's a lot of fear out there it's a pretty tiny narrative.
But today's guest thinks, actually, it's the wrong hypothesis. There's something else going on there. So I'm joined today by Chris Turton. He's, I was going to say, a regular guest on MerchantSpring. We've certainly had him here a couple of times before. And one of the reasons I love inviting him back is that he is not afraid to share his no-nonsense approach or even be controversial at times, or really, share his views on how he looks at things.So I look forward to a great discussion today, Chris. For those of you who don't know him, he is the managing director at Ecommerce Intelligence in the UK. He is well known in the industry if you follow him on LinkedIn for his transparent, no-nonsense approach to growing brands on amazon And he's very willing to take a contrarian position.
If you watched the show last year where we did one on AI, he was distinctly unconvinced that AI was moving the needle at all for agencies. Now, that was eighteen months ago, so we'll give him a right of reply there, obviously. But today is all about revisiting the conversation see what's changed, and, you know where actually where agency growth is actually coming from these days. So, it's a live show that means put your questions in the LinkedIn comments or YouTube comments wherever you're watching us and Chris, and I will try and take them. So, without further ado, Chris, it's so great to have you back on the show today. Thank you.
Chris Turton
Good morning, Paul. Well, good evening for you, I suppose. Yeah, no, it's always good to be back. Thank you. Yeah, contrarian is, yeah, probably the right answer. I look into a lot of things. I have my own sort of ideas. And I think everyone has... We all have our own biases, our own POVs based on what we see, what we do, who we work with. And that's fair enough. And you're absolutely right. Yeah, eight months ago was quite a long time. And I've always had... objections in the wrong word, but I'll always try and keep things at arm's length until we can see that develop. But I've still got some ideas between what people think or what a lot has been written about AI versus what we're now seeing and how that's involving in developing into our industry. So, yeah, I'm really looking forward to our conversation today and some of the questions.
Paul Sonneveld
Well, maybe we can start off by sort of linking it back to the episode we did eighteen months ago. So last time you were on, you took a fairly sceptical line. You know, the framing was like, you know, in the context of AI, you know, game changer or shiny distraction. At the time, you leaned more towards distraction, but we're now eight months on, and that feels like a lifetime in the world of AI. Tell me, what has actually changed in your view?
Chris Turton
Well, as we've seen, the rapid nature of AI has just gone like this. It's just gone completely crazy in terms of tech. It's the fastest format of technology we've ever seen. And I think I still hold some of those views. So one of the key things that I was really- I say anti-, but one of the things that I still think is being spoken a lot of is the usage of certain elements of things within Ecom. So things like when we look at generative engine optimisation within Amazon, for instance, we keep being told about the implications of Rufus and how that works and the fact that people are asking questions over keywords and these other, and I still hold a sceptical view with how that part operates, and I do, because the data isn't fully there to support it at the moment. We're still seeing people looking by keyword or are people more specific in the questions that they ask because they're on a shopping platform? So when people are writing for keywords, again more of an opinion, I suppose here, but I will write in what I want rather than a prompt-led format.
So I still think there's some element of, is it worth it? Isn't it worth it with that? And I'll always try and use data to hold that view. So I think it doesn't harm, because if you're getting the right questions across, we can implement that to content, do better things for brands and for products in general. The thing that has changed, and I think this wasn't too much of a prevalent concept eight months ago, is this whole idea now of agency e-commerce and AI shopping. Which is where we're now seeing massive inroads into that. And it's literally down to the way that we're seeing, particularly ChatGPT and Amazon's investment in it and a few other bits and pieces, how that's evolving into a shopping model.
And it was only really this year that those figures from Adobe started to come out, right? And people were saying, There's now a four hundred percent year on year increase in people looking for products on there. There's been a big influx of how conversion rates are working on these AI platforms. And for anyone in our industry, we've got to perk our ears about that, because even if our clients are just Amazon-centric, there's value in that. There's ideas, there's things in that. And I'm very data-led, as you know. I love margins, profitability and data to run everything we do as a business. When I think a lot of not just agencies, people in our industry are very fluffy with best practice. They think things look better this way, and they go off an opinion-based thing. I was trying to keep things at an arm length until we've got enough tangibility in that data.
So my views have changed the way of AI shopping because of what we're seeing happening. But more than this whole search engine within e-commerce framework. I still don't see that being a huge thing. I don't see the development of that. And the data that Amazon gives across is very loose. They'll say there's a fifty-four percent increase in people using Rufus. And I'll go, yeah, but from where? What's the original figure? How has that increased? How does that improve conversion? And those are the things that we do need to keep at an arm's length until there's something to say, actually, you know what, we're not seeing this.
So that's, I think, where the mindset shift has happened in some way. And I think that's healthy. We should all have POVs that allow us to see things until we get data to show us otherwise, rather than just follow the crowd. I think we all learn better that way. And that's why I'd love to see more LinkedIn is people being more, well, I see what you're saying. But from my experience, I see this and I see that. And we all learn more. But of course, everyone, especially on social media, just wants to be heard and wants to be liked and wants the same objective collective idea and if we challenge ourselves with that we all learn more we all become much much better people and much better experts for the people we serve.
Paul Sonneveld
Yeah yeah. I think it's a some some great points there. There's certainly a wave i mean i think as we see, I don't know what I want to use the word bubble or not i probably don't know enough about it whether it is or not but there's certainly a lot of momentum a lot of enthusiasm and a lot of like over the flow and there's There's not a lot of like, oh, let's, what are we actually seeing here? What's the data telling us? Let's separate fact from fiction. It's moving at a rapid pace, and it's hard to think when you're going that fast.
Chris Turton
Absolutely.
Paul Sonneveld
Now, let me talk to you about kind of from an agency point of view and more of the internals, right? Because you just sort of, I think our last webinar was much more about the external customer shopping behaviour and the like, but sort of AI and automation inside the agency. I mean, you hinted at it before, but as we've been talking over the last few months, you know, you, you've taken a pretty clear position that, to you and to your agency, a sort of AI and automation has been much more about creating client value than it is about cutting costs.
Chris Turton
Right, a hundred percent.
Paul Sonneveld
Now that sounds nice, and we all want to believe you, but make the case, you know, what does it actually look like in your business?
Chris Turton
If you identify, when you're looking for everything within a client's remit, or even for some of the sellers that are watching right now, identifying the big time areas- time is money, time is cost- that's been the biggest value piece for us. Being able to find administrative work, things that are reports-led, or things that don't give you as much visibility unless you're starting to look into it manually. For any business, not just an agency business or a service business, that takes up an exuberant amount of time.
So what we did is we specialised in, and we've got one of our team who is just focusing around this area to see where can we make a lot of those time savings? What can we do to reduce the cost infrastructure whilst allowing the expertise of our team, and we've got quite a lot of people here, Amazon Ecom experts for many, many years. How can we say in a very simplistic way, how do we reduce those stresses from those processes and get them back and get their brains into the stuff that matters? And that's providing value.
And that's where the really cornerstone of the element of AI has been so far very, very helpful. where i don't think it's been helpful so far is the idea of these quick fixes you know this automation idea of listing creativity. For instance we talk about listing optimization you know how much i love the idea of saying you know listing optimization is done it's a tick in the box and we know really with big brands They're being data-driven. An optimised listing is never fully optimised. And people are almost using AI as a crux to go, well, AI says this is fixed, and Amazon's best practice says it's fixed. I'll tick a button, it's done. And what do we find?
Everyone's in the same boat. You can see that now. You go onto Amazon, you look at typing ground coffee or a really big keyword on Amazon, and you'll see that every organic listing on that page has a perfect listing score from Amazon from somewhere like Helium10, other software providers are available, but someone like that who's using a listing optimiser score, and they're all ten out of ten, and they've all got a tick box. And people just shrug their shoulders and go, well, what's next? And that's the problem we have with it, where people are using it more as a crux to say, it's done.
Same with image creativity. Fantastic for launches, you know, be able to make AI images, AI video perfect. But then is the attributing value back to you improving your clicks and your conversions? And most people are just going, they just leave it. AI says it's good, let's take it off. And that's where this is becoming hugely problematic because everyone's in the same boat and they're doing the same thing. But the big brands are going, how do we use the data to get beyond that level of everyone else trying to do the same thing?
Paul Sonneveld
Yeah, when everyone turns right, you know, can we turn left or, you know, how do you zigzag and stand out from the crowd, right?
Chris Turton
Yeah, exactly. Or turn right in the first step; use it as a first step. Absolutely. You know, AI is great for speed, and speed is value. Perfect. But that next level where it really applies, especially with costings, advertising, all those, has to be done at a different level.
Paul Sonneveld
Yeah. So, in terms of, you mentioned sort of more using AI where it has been successful in terms of saving time sort of the manual reporting or investigation, or I'd love to get a little concrete here, like can you can you list some things that you've actually automated, and you know where you saw the biggest benefits?
Chris Turton
Yeah, so obviously reporting and administrative task is a big one. But if we look at it more in terms of a value piece, so we've got some great kit that we've built with Claude that looks at a very, very technical- we've got a very, very big client, multi, multi-million pound major, you would know them brand. They have a huge issue and a huge complicated issue with variations on Amazon Vendor. So we've built a structure that allows us to identify changes that have been made because the complex reasons Amazon Vendor means it's not as quick as ticking a box for content or, in this case, ingredients and information around that product. You've got to be able to do this at scale. And before, if you're just using a flat file and you have late data, then you've got to look manually through it or download everything again to see what's happened.
Utilising things like that, in Claude in a big, fast areas and say, what's changed, what's analysed, can save hundreds of hours of time and provide better value to clients. That's just one of many things. And we're working towards those kinds of tasks where they're very much low-level administrative tasks. And then we can spend that time saying, how do we reapply and redevelop that value into clients? And I think that's where a lot of people are falling down now. They're more scared of using it instead of using it, saying, let's utilise this in more of a value sense or a product development sense and actually double down on what we're excellent at doing, which should be providing value. And I think this is exposing a lot of agencies here and anyone within any service business who's not providing that value, AI is exposing that very, very quickly.
Paul Sonneveld
So can I just press you on that a little bit more because I think your position here is very much like it's not just about saving time and trying to drop some of those savings down to the bottom line and boost your margin. But it's actually about reinvesting that back into value for clients.
Chris Turton
Yeah. And that's what everything should be, right? Everything we do should be by putting value.
Paul Sonneveld
Yeah. So, I mean, that sounds great. Again, playing the sceptic here. Like if I was a client, you know, in your agency, what am I getting this year? Maybe that I wasn't getting two years ago in terms of that value.
Chris Turton
Purely depends on the type of client, but with that particular client in mind, we're talking, you've got five thousand variations here, complex issues. They've paid and invested us to resolve these issues within the vendor space. And at the time, because we are clocking on the time and what we do, that now becomes more suppressive. So we can reduce that down. By reducing that down, we can then invest in the things we can use our brains for, the things that we can build between our account managers, our performance management team, our advertising team, and really say, have you thought about this? What about this idea? Can we develop more of a strategy for Q4. Other things that we were doing before, but we're under more strained time? And they're the things that really apply value.
And of course, you can only get that information from ChatGPT if it's already publicly available you know, you can't maximise that experience piece from your team from the seventy plus different clients and two hundred plus clients that you know i've worked with and people over the years. If you're just using an ai tool and that that's where a lot of that value comes back from that human interaction piece that we weren't able to do because we were doing administrative tasks before so that's that's been one of the biggest parts of it. And i think there are there are lots of other examples but i think If the mindset has changed to saying, how do we reapply what we do to develop that and improve value for clients? That's the key thing. And that's what we're finding huge inroads with.
Paul Sonneveld
It sounds like, as you were talking there, there's really two sides of the coin. One is you're saving time on sort of repetitive, mundane tasks that's just going quicker and faster. But then it seems like some of the, I mean, you talk about, you know, vendor catalogue issues. I've got another client, you know, I think he's been going at it for about nine months trying to sort this. It's a big issue. But I think in the past, some of these projects have just been too big to digest, right? Or the price stack you've been able to, you've had to give a client was just too hard to justify. Whereas it feels like some of these harder projects have, are sort of more in reach now. They're becoming a little bit more feasible. There's still a lot of work involved, but AI is certainly making them available in terms of maybe budget and time. I mean, is that a fair, you know, is that a fair sort of statement?
Chris Turton
Yeah, I think everything is going to boil down to what you're seeing going on and the feedback from customers, what are they saying? where what's the value they're saying? What problems they, you know, are going to? I think people automatically will jump onto the idea of saying, " Well, do x y and z, " line by line, and the whole point we're trying to do is redevelop and squeeze that back into saying, " Well, actually this is administrative and can be resolved and done correctly because you don't always want to fall into that trap of saying, This can be automated; this can be done with an AI tool.
But then there can be a gap within that where it can go wrong like you can with the trying to optimise and squeeze your listings down i mean god knows how many people were doing that when Amazon made that change in AI titles and people just panicked knee-jerk and just changed their title links without going should we prioritise these by conversion and they just panic panic panic and did a task that ended up potentially costing them. What we try to do is build tasks that are very much regimented to asserting how do we fix this and even be able to generate things like saying, well we've been told by Amazon that we can only generate so much content with a variation, and it's going to be sent to a seller support case with every five tickets. So we can automate the seller support inquiry that goes along with that and automate that whole process end to end. And it means there's none of that good value that our team are doing is being wasted on that kind of stuff. And again, a lot of that requires experience to be able to apply that into an AI tool, which, again, I think it's a lot of people are missing.
Paul Sonneveld
Do you- I'm just curious here, do you have a way of, so when time is freed up, how do you decide where to reinvest that? And I'm particularly with, you know, I've got a particular thing in mind here, which is I know for, even for us, more on the software side, AI is very useful at continuously providing you with suggestions on what else you should be doing, right? To the point where you could burn up all that free time doing all these other things, and sometimes you have to go back and go, right, that wasn't a great use of time. It was interesting, but it wasn't a great use of time. How have you sort of, I don't know, maybe this challenge doesn't sound as familiar to you, but if you have seen this challenge internally, I'm wondering how have you put some guardrails around that or maybe some disciplines to ensure that the time harvested is not burnt on peripheral projects?
Chris Turton
Yeah, I think a lot of that comes down to planning. We build in quarterly plans with clients. We have an end goal in mind. We're doing more now because of AI for us to spend more time on building better KPIs and aligning better with client's goals. So there's always there's always things that need to be done it's always a case of trying to allocate that by the time. And we we can pivot that back into our different departments.
Is there an ads issue is there is there margins issue is there a product issue it's a case of just saying well being able to reduce that time just means that we can do more and we pivot that towards existing plans are in place i think again people are so sort of knee jerky aren't they going well i've done that now i'll look at this and you go well What's the plan behind this piece? Is it just because you feel like doing it? Is it because you've got an opinion on something? For us, it's all about that planning stage and making sure. And also, it's good for clients because they need that direction. They need to know what's going on. So that's the biggest thing for us. There's never been a case where we've gone, oh, well, people just hang around doing whatever. We've actually tried to squeeze that into doing more. That's always been the goal.
Paul Sonneveld
Yeah, that makes sense. You've got a specific strategy for each client with some specific goals.
Chris Turton
And they can be so varied. These are things we want to get to. They can be so varied. Sometimes it's revenue goals. Sometimes it can be this big club, this variations issue. Of course, this is a gigantic catalogue problem where that value is being hugely attributed. And now we're, we're reducing that down to allow them to, to get more out of us. And because of that, they're saying, what else can we do? How can we look at some of our other brands and, you know, all these kind of things.
Paul Sonneveld
Can I ask you about agency sort of client expectations and billing? So I, as I was preparing for this episode, I found, I don't, I can't say this is a real stat, but I found the stat and it said was Sixty-one percent of agency clients now raise AI during renewal conversations. And roughly twenty seven percent or about a quarter or so, of agencies have actually been asked directly for an AI discount as they're sort of renegotiating their agreements. Does it sound familiar to you? Has it occurred to you? And if so, how have you handled that question?
Chris Turton
No, we've had occasional people ask about it or have said to us, can't you just use AI for this? And then we've gone, well, in a way, but that might only do ten percent of the actual development of the task. So no, not particularly. We tend to not get people to say that they won't go through a sales course and say, how do you use AI? That's not really a common one. What we try and drive into is their their bottlenecks, their problems as a brand, and try and work backwards from there. So we're not gonna say, we're an Amazon agency, and this is what we're doing, this is what we charge, and then they're gonna go, well, how do you utilise this tool and that tool?
The key thing we're seeing is where clients are finding the concept of value faster. And I'm sure every single agency has experienced this as well, where a client has given the notice in or they've gone, and they may even give you some really good feedback, and you go, okay, this is a bit weird. And then you see it again. And then when you really drill down into it, it's simply a case of value. And we've seen before where people have brought people in-house internally, and they've said that we're going to bring someone in because we want to take more control of this. We're seeing growth. We want to do this, that, and the other. And actually, it's a valid question. And I think what's happening, and again, I'm not just interested in Amazon agency or even services agency. I think this is a B2B issue overall. People are understanding a higher expectation of value because of AI, even if they don't say it. Their expectations when they want something that's deliverable to them within any format of business, that's predominantly more B2B. But service-based is more easy to quantify because Most things are kind of in the cloud. They're things that we do on computers.
But I think most of that comes down to the fact that people are not seeing or they're not understanding the value that goes into that. And sometimes you can open yourselves up with different models and how you price because people will see that and say, well, I don't see value in that, or I don't see value in paying you a tainer, or I don't see value in paying a commission on revenue, which is a conversation of the day. But it's one of those where the expectations have been raised because of AI. And that's just how it's filtered generally into the industry. That's what's happened. And therefore, the standards of the businesses has to go up because we know it's utilisable. We know that we can compare it and develop it within the means of our businesses and our processes and our structures to produce a much better product than people can't produce themselves. And we're seeing that AI, as it continues to improve and increase, we see huge value in it because it allows us to learn more and execute more. And we're developing our business in ways that allows us to see much, much more beyond Amazon and looks at the entire, not just the commerce landscape, but the retail landscape overall, because as AI shopping is growing, that has such a wide signal base of multiple retailers.
So our goal is to say, you know, Amazon is great. Amazon is still thirty five, forty percent of the marketplace. I think that's going to change slowly as we're starting to see other things creep in now with different marketplaces, AI, Ecom, all these different things. But it's still big. But we want that to be a portion of our business as well, because we want to see everything, and they all feed back into each other. So what I'm trying to say in a positive for everybody listening, everyone watching this space is if you refocus it, the actual expectations, the values, what we can do in the future is incredible. If we're able to look at something in a value-specific way now and not this whole I'm doing a listing creation. I'm doing the right kind of keywords for the title. I'm doing advertising in a certain way. It's so old school. These are things that I've been saying, you know, you've got to look beyond that since twenty eighteen, twenty nineteen. It's about value attribution and bringing that into everything you do.
Paul Sonneveld
That's really I mean, I love that point because, particularly around channel diversification, because the argument has always been The GMV isn't there. You know, this used to be the argument with Walmart in the US. The GMV isn't there, but it's the same amount of work, and it drives all this complexity in our business. And therefore, you know, it's kind of it's hard to do properly, whereas, like you're saying now, actually, with these values, that can take some of that mundane, some of that more complex work away. All of a sudden, expanding to other channels is much more accessible than it was before. And from an agency point of view, that's kind of a win there because building some expertise across these different marketplaces and all of that not just a much deeper partnership, but also like as much stronger lock-in as well.
Chris Turton
Absolutely. And people who utilise that now will future-proof their businesses as an agency perspective and future-proof those brands. That's the biggest concern now for brands is the feeling of being taken out by a challenger brand because of those ones that are utilising in a retail sense more. And as you say, GMV is getting more and more difficult to quantify, and we're seeing Amazon squeezing those fees. TikTok, we know there's a huge halo movement onto Amazon and other areas, but it's so difficult to quantify. And I know Amazon now, they launched to accelerate this idea of hooking into Walmart and TikTok, and I think that's part of the reason. But that's also because they know TikTok is a huge high-funnel driver for their sales.
So at the moment, very difficult to engage that GMV, very difficult to then get a client to invest in it. But if we look in the future, we look already, what we're saying is we're going, well, hang on, there's twenty five retailers here that are all have diverse needs and all these different marketplaces, yet they're all feeling signals into ChatGPT or Gemini or any AI model or agentic model as well that is going to create shopping results. So we have to look at that in a much wider spectrum and say, how do we actually use AI in a way we've never been able to do before, where we can actually literally open shine the floodlights and say how does a brand how does a an e-commerce business look holistically because doing that we can do much more and we can build value back into Amazon you know we've got tools now that look at pricing and listing analysis we've got a tool now that again with with the assistance of claude and we've been able to do all sorts by getting better structured forecasting models against what we've done with our existing clients.
Again, I just wasn't smart enough in a spreadsheet to do that before; we've been able to do that. But also with this pricing listing model, being able to analyze pricing across the board. So, in other words, we can find out one of our clients products, and they could be, you know, give you an example with one of the largest professional hair care brands in the UK. They have a product line that we can do a one-button check. We can see all the twenty different retailers we know they're on, and we know Amazon are likely to check because they're looking for all these pricing things. And financially, you know what, you're massively, your retail price is massively under here. We've got extra squeeze for you potentially for margin. You know, we can invest more into advertising. We can do more of this. So having that landscape of retail actually will allow us to do things on Amazon we couldn't do before. And that's where the value attribution is in the future of agencies.
Paul Sonneveld
That's very refreshing to hear. It warms my heart given that we're a multi-channel analytics solution so
Chris Turton
Absolutely yeah
Paul Sonneveld
Yeah
Chris Turton
TikTok for GMV. Yeah, I mean that's just it's it's one of those it's very very expensive but the value attribution at the moment is tight and I know last night they're doing their own checkout and a d to c model directly into into websites so that might tighten things up so that there's gonna be so much more of this now . We’re seeing other marketplaces grow as well. So over in the UK, where we've had before, most of the marketplaces have gone. We've got a big one in the UK called B&Q, which is a big retailer. I suppose it's like the Lowe's of the UK. Their marketplace has grown massively, much higher than any others we've seen. And that investment is working for clients. And it's another key signal for AI, for AI e-commerce and ChatGPT. So those are the areas that we need to look at investing in as agencies.
Paul Sonneveld
Very exciting. I've got two questions left, Chris. We're sort of out of time, but I want to sort of try and see if I can squeeze in two. Sure. First one is, with all this automation, efficiency going on, redirecting, save time into more value for clients, when you look at your agency structure now in terms of whether it be pods, account managers, advertising specialists, all that, I mean, if you sort of look at that sort of thousand feet view of your agency right now, are there any structural changes that you detect versus, say, two years ago that you've made as a result of AI? Maybe, you know, realigning resources here or there, shifting, investing more in a certain type of skill, less in others, like what's changed?
Chris Turton
There is, as we kind of pivot into this new area, there's going to be much more potential for staffing within maybe more marketplace-specific and number-specific people. So we've got a team that we've got a team that I think most agencies don't have. We have one that's just solely focused on performance. So they work very heavily with MerchantSpring and looking at the P&L numbers and things to make sure that the value provided to clients is there.
And I think that's going to improve. Being numbers-driven is crucial. There are still hundreds, thousands of sellers that are not doing it. And it's very clear to see when you look at it, there's no way they're making any money off the back of that. So numbers is going to be a big player for people that are really invested in things. And yeah, AI might be able to analyse it, but can it make money? the right decisions when it looks at a much wider piece because clients have operational requirements; they have certain needs. There are certain things that you would know from an experience-led level of things so a lot of that area is going to be much much bigger.
I think than just the than the creative side the creative side is still going to be there. I personally believe because when we look at the creative side, we assume like before the creative side is done and dusted it isn't. There's going to be more use cases for real-life interaction where things like AI might produce a great video, but then certain features that are actually driving the conversions aren't listed, and they need to be there. So there's going to be all these different nuanced areas where skill sets will just change. And I think they'll become more harnessing within certain areas rather than generalised things. And that's going to be interesting for our industry because as we employ people at generic Amazon people or e-commerce managers and this, that and the other, And those skill sets are going to adapt drastically for the needs of those brands. Eventually, everyone will catch up. You know, next few years, everyone starts to catch up and will start to see roles for AI e-commerce managers or people who are marketplace-specific, but look at much wider notions.
Of course, the flip side of that is trying to get everyone, one person, to do everything, which is we're still seeing in that space. headache in itself yeah because the things you can't always automate but that's that's um that's where i see things change so from a recruitment side i still think it's strong we're you know twenty people we're going to be growing that as we develop our model and really look at not just becoming an amazon agency we are we are going to become you know a major e-commerce agency that's our goal we want to support brands holistically want to pierce into every area of their online digital shelf if you like and execute much more Brand consistency is going to be key. So making sure that's there is going to be the big area.
Paul Sonneveld
Great. I've got one question left. And I've just noticed that this is a live episode, by the way. And Ahmad is giving us a few questions. I'll try and see if we can get one of them. You know what? I might pick one. I'll go to Ahmad first because he asked too.
Chris Turton
Sure. These rules are great. These never stunned me a little bit.
Paul Sonneveld
I never even read them beforehand, but let me read it out loud. Ahmad, thank you for raising the question, actually. It's good because now I can prove to the rest of the audience that it's in fact a live show. What is the distinction for agencies when every agency is using AI for automation? You know so for some of those core tasks, like report generation, listing pricing analysis you know i guess this question is: how do you differentiate in commodity, commoditised AI automation?
Chris Turton
Well, it all ties back to the value isn't it? If you can build something that creates a report that's that's perfectly aligned, that you understand, that provides value to the client; if you can produce listings that are producing a higher level conversion rate than your average. Again, people don't realise they can do that in category insights to see how their products actually perform rather than just guess with AI. And when you say price analysis, again, it depends on how and what you're using. If you're using price analysis as an option just on Amazon or in marketplaces, that might not be the full spectrum. The question there is a bit of a vague one because it really depends on what Ahmad said there with the example he's given. It depends on what those values actually represent holistically and whether or not that ties back to the brand.
If you're running an AI Amazon agency, I think you'll have the same problem as if you're running just an Amazon PPC agency. You're only looking at part of the issue, part of the actual development. Amazon PPC agency is going to look at the conversion of the product listings itself, otherwise, you spend too much money on advertising. It's exactly the same with this concept of AI. If you're using a button press for everything, you're only going to get so much value from that because it depends on the level of what you're looking at and the value back the client's seeing. That's the problem. And tools that are doing that, they're probably doing, you know, twenty percent of the work. You know, what we're doing now with AI e-commerce, there are AI tools out there, loads of them you can use to look at your AI SEO, right? There's a hundred of them.
Type in your website and it will give you where your prompts are, what your visibility is. But that's about thirty percent of what we're doing now because the e-commerce side is much more regimented and structured. And that's only from our experience that we've developed. So really, I think that the difference in there is really the value you provide and what those businesses are doing. We see there is such a contrast of agencies. We have some amazing agencies in this country. We have some ones that will tell a business that they've got to just run ads for twelve months at a loss, and then twelve months later they'll be profitable. I don't know where that logic comes from. Those are the kind of guys I would think are just going, I'll just press AI for everything and get the same old result as everybody else. Sorry it's not a clear answer there Ahmad, but it's very difficult for it to be, it really depends on what that value looks like. If you've got product A, B and C in AI, what's the actual value attribution it's giving to the client? And that's where the differentiation is going to be.
Paul Sonneveld
Yeah, really good point just doing the things with ai isn't isn't enough it's about where you exactly and how you enjoy it to your clients so hundred percent um thanks for your question amount really appreciate that. Okay, so last one for me then i thought we'd finish on a confession. So at the beginning of this episode we talk about like how we can all run away with ai and and you know the hype and all that course, I wouldn't say that you've done that but is there something that you know maybe you try to automate or you try to get the team to focus on you thought it was a great idea at the time but you know it was just it was actually not such a great idea it was a mistake or it just didn't work you had to pull back you know anything that you'd like encourage other agency owners you know not to make the same mistake
Chris Turton
Yeah what i always say with anything you build within any ai tool is have the end in mind the amount of people that start and have that thing we've done we've done that numerous times where we've said we've started with an idea of saying well how do we look at forecasting and then how do we look at our average advertising data, and you build all these little bits and pieces and these tools into play, and then you present it in front of a client, and then you go, I actually can't quantify the value or I can't see the actual value endpoint.
We've done it quite a few times. I've built numerous things, and then, so the team, we've tried something, built it, and actually it's probably the best idea right now, or it still needs some work. And it's always an ongoing process, but you have to kind of do that. But if, like I said, if you start with the end in mind, I think you'll save yourself a lot of time. I mean, Deloitte, the big UK finance accountancy firm, they put out a thing about two weeks ago that said it's like, Everybody who uses AI on their survey has found no improvement in their productivity, in their work productivity.
So they've used it in work, and half of the things that are being generated are not actually improving that productivity level. And I think that's because people don't have a task, a problem in mind. And say, how do I fix this first? What's the reverse way of fixing this first? And that's where I think things can become valuable. They just think, I can automate this now. I don't have to do this. And they'll try it, and they'll go, oh, this looks a bit crap. And I've got to do more myself to try and fix it. I think that's the biggest problem.
So make those mistakes. Try different things. Utilise MCPs like MerchantSpring, you know get that information in you can you can do some fantastic things with looking at those those data points and cross-referencing them, um, and make like make sure you've got the end in mind you know, don't joe i wouldn't just do something like install you know we'll talk about you know the MerchantSpring MCP integration and go oh i'll plug it in and i'll just say tell me all these numbers and you'll get all that data fantastic but really the logic should be okay now i've got the power of all this information what's my end goal what do i want to see do i want to see average net profitability by my clients do i want to see average tackles by by client do i want to see whether and how do i utilize that within what i'm trying to do rather spend hours trying to finally figure out something and realize it's not worth it that's that's the thing.
Paul Sonneveld
Back in thirty years ago or twenty-five years ago, when I was a consultant, we used to have this great technique which is, before we do any analysis, we do anything, we had no AI at the time, you literally get a, you know, get an A-four piece of paper, you get a pencil, and you, and you, and you almost like create the slide with like, there's no data, but you sort of like, this is the end result that I'm looking for. What is the analysis I now have to go and do to actually get that result? So, you know, the good news is you can get AI to the second step these days. But, you know, maybe start with a piece of paper.
Chris Turton
Very true. Yeah. But tying it back, I was going to say, it's very sweet. It's great. It is great to be controlling things. And as I said, there are different things that obviously we spoke about in this kind of thing. Yeah. But the time you've got to you've got to be able to make that mark. And in eighteen months time we've had another conversation and go, things look very different again now. You know, we're saying this is where we're thinking things are, this is the data we have, and actually that might change again. So I always want to encourage people to have that and also just, you know, just prove that everything you say isn't always right. I'm still sceptical about certain things, and I think that's a healthy way to be. And as long as you've got information to back those views, we want that. We want to see those, and we want those challenges. I love that when people do that on LinkedIn. Helps us all more become better people, I think.
Paul Sonneveld
Yeah, that's a good encouragement to all of us. Thank you so much, Chris. It's been a wonderful episode. It's been great to be the beneficiary of your knowledge, of your experience, of the journey that you've been on as you built out, you know, one of the great agencies in the UK. For anyone that's tuning into this episode live or maybe afterwards and wants to just pick up the conversation with you and, you know, what's the best way to get hold of you?
Chris Turton
Yeah, I'm really active on LinkedIn. I'm sure a few of your viewers like me on LinkedIn, so I'm there always ranting daily. I'm also doing a lot more short-form stuff now on TikTok and Instagram. So on TikTok, chris_turton and instagram i'm the chris turton so i've just launched those there can be loads of new short-form videos, helper guides videos the whole works and um yeah just just getting that feedback let me know your experiences, what differs what have i said that you think is utter nonsense tell me you know i'd love to see it from other people's angles and what they do with their clients and businesses.
Paul Sonneveld
Fantastic. Well, thank you so much, Chris.
Chris Turton
No, thank you.
Paul Sonneveld
Looking forward to talking about our next episode together. It's always a pleasure.
Chris Turton
Absolutely.Fantastic. Thank you. Bye.
Paul Sonneveld
All right, everyone. That's a wrap for today's live episode of Marketplace Masters. If you've registered for this, the recording will be with you by tomorrow. And this is certainly one worth sending to whoever owns ai automation, you know, even kind of client expectation management in your business. I think there's some really useful lessons in today's episode of course, our full on-demand library where you can see all of our historical episodes i think we're well over a hundred now you is at merchantspring.io just head there look for the resources section you'll find them all listed there. If you want more from the agency best practice series as well,
And of course if you're an agency trying to get reporting and interpretation you know, across your client portfolio, across a whole range of marketplaces, including like the MCP that Chris mentioned. That's exactly what we do here at MerchantSpring. So feel free to drop me a note, and we can have a follow-up conversation as well. But thank you so much for joining us. And I look forward to seeing you in our next episode next week. Take care.