Episode #108: Revolutionising E-commerce Marketing with AI by ASK BOSCO

Speaker A: Welcome to the Milk Bottle E Commerce show brought to you by Milk Bottle Labs, one of the world's top rated accredited ah, Shopify experts. The Irish based Milk Bottle Labs team upgrade, migrate, manage and advise some of the highest performing Shopify stores worldwide. Founder Keith Matthews interviews merchants, developers, fellow agency owners and Shopify folk to share as many Shopify tips and tricks as possible. If you're planning on upgrading to Shopify 2.0 or migrating over to Shopify, reach out Milk Bottle will help. Now over to Keith.

Speaker B: Welcome back folks. Today we are joined by John Reedman, the CEO and um M the founder of Ask Posco. Ask Posco is an award winning AI marketing platform for e commerce enterprises with over 25 years scaling digital agencies and SaaS ventures. John Co founded Ask Bosco alongside Bonamy grimes from Skyscanner. Asposa recently scooped the Standout AI Solution Award at the 2025 UK Digital Excellence Awards and the Innovation Award at the UK Paid Media Awards. John is also recognized as one of the UK's top digital leaders in this year's BIMA 100 list. Beyond business, this Yorkshire native is combining his passion for cycling with charity work, undertaking an epic London to Sydney journey in segments to support Ugandan orphans through his venture ride 25 she showing that purpose drives everything he does, including Ask Posca. John, you are very, very welcome. Nice to have you.

Speaker C: Well, thanks for having me Keith. Uh, yeah, excited to sort of get into some of that. And we need to get you on the bike.

Speaker B: Oh, uh, that's where if we're in a pub we talk about that John. But we're not.

Speaker C: Yeah, well after five, after five pints you'd be cycling to Sydney.

Speaker B: I have to say I wrote that intro and I read the intro and I reread it for this recording. There's a lot in it to unpack. But look, you're very, very welcome. We, we met at uh, breakfast meeting in London there recently which sounds really, really formal and something struck me about you and the rest of your team was that you'd be a good guy to chat to and also I'm actually excited to learn more about your product. So John, at a high level, what exactly is Ask Posca?

Speaker C: Sort of my background is I've been working in digital marketing agencies for quite a long time and we used to all clients want good reporting and we used to have this challenge particularly in performance marketing agencies where there was always a disconnect between what we thought the number and what the client thought the numbers were. And there's always this confusion and if anything it's getting worse with tracking and then data between different platforms. So we ended up when I was at Search Labs and then at Summit creating, right, we're going to get some super clever people in, we're going to connect all the data up with bits of software like Funnel or Supermetrics and then we're going to build these beautiful dashboards in Tableau or Power Bi or Looker Studio. And we were creating small little industries, Keith. And the trouble is none of that was billable and clients expected it. And then also you'd have an account director or salesperson go out and say to the client, oh, do you want bespoke reporting? We can do bespoke reporting. And then you've got 200 versions of the same thing that you've then got to maintain. And basically I was actually having a conversation as all good conversations should start in a Swedish ski resort with Bonamy, my uh, my friend and now business, uh, partner investor about this problem. And I said, surely with machine learning, AI and the way the world's going, we can fix this and create a thing that's almost like a standard because there isn't a standard for cross channel report for E commerce and marketing. Everybody's got a version of the truth and I think that creates confusion and lack of clarity. So what we wanted to do was sort of connect all those dots together and give people a single source of the truth so that they can make better decisions and overlay on top of that some AI to ultimately try and help people predict where they should spend their next Euro pound, dollar, whatever. So that's what we set off on a path to do. And that's now what, what we do. So We've got about 350 different clients using Asbosco every day, a lot in the Shopify ecosystem. And it connects all your data into one place, gives you beautiful reporting which you can just talk to and ask questions of. In a generative AI way. It then benchmarks you against your peer set. I don't know if you remember Hitwise, Keith, but there used to be Hitwise was like, yes, yeah. So Hitwise died when Experian bought it. And I was like, everybody used to love Hitwise. They used to put Hitwise in their board package. And I'm like, can we not create a single metric that sort of says how good are we compared to our competitors? So the reporting side sort of says here's what's going on. And then the benchmarking says, well here's what's going on us relative to our competitors and then we have a forecasting and budget planning side which says right, what should we do next? So, so that's what we do. That's all we do. Um, and we do that main. Yeah.

Speaker B: John, are you not like you mentioned, companies like Tableau, you know, like there's, there's businesses out there doing the same thing that Salesforce will provide publicly quoted companies. Are you not enter entering a very, very difficult space or are you pinpointing a niche?

Speaker C: I think the difference between those and every Tableau are trying to solve like a dashboard reporting solution for everybody. Right. So they're, they're saying we've got this amazing dashboarding platform. I've got a separate conversation about some conversations had with Tableau in Salesforce Tower, but they can remain for the purpose another timekeep. But basically you're absolutely right. They're trying to do a one size fits all. What we're using is my experience, my team's experience of working specifically with marketing data in Google, Facebook, TikTok and then E commerce data from platforms such as Shopify and our knowledge and experience to build something specific for that world. Right. So we're not going to build you a rapport for, I don't know, delivery. Right. Uh, of locations where your vans are going. That's not our thing. But Tableau could probably build you that. Uh, what we're doing is helping e commerce managers, marketing managers make better decisions by creating a specific tool set for the daily, weekly, monthly, quarterly reporting. So and I suppose the real niche we want to get into, and we were just chatting before we started recording, is we're about to launch a fully embedded um, built for Shopify app version of Osco. So all your channels and reporting will be inside your Shopify store, but with the generative AI capability just to ask questions. So we think most of the analytics, so if you just have the Shopify basic analytics, it's like a long list of things you can get. Then you click on it and then you get load of tables and it's just, it's got huge amounts of data in there but it's just difficult to find what you want or you use some of the other things like polar or triple. Well and they just connect your data and then take you to their app. Whereas we want to build something native within built for Shopify. But you're right, it's, we're undertaking a big thing but we think we've got years and knowledge of experience to help people on my team have been the customer. So we've got people who have been marketing managers, heads of E commerce who know this.

Speaker B: So let's go back just to your experience. So you mentioned hit wise and then you also mentioned you haven't been nominated to top digital leaders in the UK for no reason. So what's your experience, John? What's your, what's your core background?

Speaker C: Failed pilot is the start. Yeah, failed pilot. I wanted to be top gun.

Speaker B: Do you still fly?

Speaker C: I do, but uh, not enough. But yeah, it's a failed, failed pilot. Um, and got straight into computers and Internet and digital straight out of not getting into the air force. I've just had a knob obsession with uh, I suppose performance technology. So I'll do this quickly for you, Keith. I started off my career.

Speaker B: We have as much time as you need. I'm interested.

Speaker C: All right, cool. So I started off my career with a business in London called qas which was quick address systems. It's different in Ireland because like one postcode, there isn't any postcodes outside Dublin and it's like down the road on the left past the pub.

Speaker B: No, John, John, would you believe it? They've actually introduced it in the last few years and now every single location has a postcode.

Speaker C: Oh, there we go. So I used loved the fact that the post person in Ireland just knew where everybody lived and you could just put a name on it and someone and it would get there. So um, but what QAs did was they helped people type in a postcode and bring back an address. And at the time, and we're both old enough to remember, that was like magic, right? Nowadays they call that AI. Uh, it was just good old fashioned fuzzy logic database. And I got an obsession at that business helping businesses understand how they can use technology to save money. So we'd go into big call centers and go right, you could capture more orders or more addresses or send more catalogs and not get as many returns. So it was software helping people with marketing process. So that's supposed, that was like starting at the letterbox. Then I got into email marketing when it was really just mass bulk spam, uh, if I'm, if I'm brutally honest. But we started understanding how email could be really strategic in basket abandonment and those sorts of things. And then I got into the murky world of search marketing. I remember I was on a ski lift with someone trying to explain to me what SEO was and I'm like, isn't that just trying to trick Google? And they're going, uh, no no, no, no, no. And I'm like, it really sounds to me like that's trying to trick Google into showing your products over someone else's product. And they go, no, no, no, no, it's not that but basically it's trying to promote. So I got fascinated by this. I then got asked to go and help a friend of mine grow his SEO and um, PPC agency called Search Laboratory. And we grew that from sort of seven people above a Thai takeaway and leads to 200 plus people, seven plus million revenue and became one of the biggest independent that then they, I exited that business and then subsequently they've sold that business to Havas. I then had my little foray into cycling because when I exited Search Labs I wasn't allowed to work in digital for a while. So we set up our cycling business where we're cycling from London to Sydney over 25 legs and 25 years for charity. We're currently, this year we're going to cross India. So we've cycled if you, if you. Look, I don't know if you use Strava or if you're familiar with Strava, but if you looked on, if you looked on my Strava heat map, I've got a line all the way from Leeds to Baku in Azerbaijan and then we were meant to be cycling from Baku to Tehran this year. However, based on the sort of geopolitical situation, we decided that might be a bit risky to take a hundred entrepreneurs that potential hostage risks into Iran. So we're going to go to India and we'll cycle from Jaisalmer to Jaipur.

Speaker B: So John, are you doing 7, 800 miles? A thousand miles on each trip?

Speaker C: Not quite that punchy. Uh, well you sound like, you sound like easily capable to do this.

Speaker B: No, I've been involved in some charity cycles where the. Well, they're all ultra cycle cycles and at that is it, are they week trips, 10 day trips?

Speaker C: Five day trips? We do 100 miles a day, fully supported mechanics, medics, masseuses, we move all, all you got to do is pedal and it's about 100 miles a day every day for five days. We tend to break it up though, so we do like 20, 25 miles. Coffee stop, 20, 25 miles, lunch. The interesting challenge for India this year, everybody's nervous about the food situation and getting Delhi belly, which you can understand. So we're actually uh, taking our own catering setup. So we're going to have our own like almost like a whole mobile event for five days which is going to Be super exciting.

Speaker B: I would have definitely said bring your own water, but you obviously have everything covered with food as well. Uh, yeah.

Speaker C: So we're taking everything to Western standards. And then. And I suppose the other thing for me, when you're organizing charity events, just because it's for charity doesn't mean it's got to be rubbish. Right. So we make it really good quality, everybody pays their own way and then raises money on top. Each year we raise about £250,000 and we support 280 orphans in Uganda. And so I wrapped that up as a bit of a project, ended up running it for three full years and we ended up doing corporate cycling events as well for people like Google, Adobe, Facebook. And that was fascinating about just connecting my network. It wasn't necessarily designed to do that, but there is, there is an interesting bond you build with somebody when you're cycling up a big hill. Yeah. Much as it is fun to have a few beers with people, but suffering on a bike for a day in the heat or the cold is, is a great way to get to know people.

Speaker B: I've done it. Not, not on the same scale that you're talking about, but not yet.

Speaker C: Over here. Not yet, not yet.

Speaker B: No, no, no, over here. And I'm sure it's the same in the UK when young men retire from football. It seems to be, uh, taking over a lot from golf.

Speaker C: I think, um, cycling is becoming the new golf. And for me it's, you can meet up early on a Saturday or Sunday morning with your mate, get out early, do some exercise, but you're still home at lunchtime so it doesn't take your whole day up and you get some proper exercise. Yeah. So that's been really, really fun. And that's actually on. That bike ride was actually where I met Bonamy, who's the founder of Skyscanner. And we just connected. I didn't know what he did, really. We just connected. Well, actually, I was cycling between Paris and Geneva and I'm a rugby lad, right. So I'm a bit heavy. I'm not a natural cyclist. And I was on my own, um, going into this headwind and he sort of flew past me and then he sort of slowed down a bit. He said, are you alright? And I, I said, I am so far from not all right. He's like, right, have this gel, dab this, drink some water. You're bonking is what's happening to you and just sit on my wheel and I will get you there. And we just bonded over my Inadequacies on a bike, basically.

Speaker B: So that's a nice segue into the fact that he is now your business partner.

Speaker C: Yeah, so I finished the bike ride and then I went to a business which will sort of segue into this. It called Summit Media and Summit Media in Hull, a friend of mine was running it, he was trying to sell the business and he, Redmond, I need somebody like you, just a bit commercial. I'm trying to tee the business, the agency up, can you come and have a chat with me? And ended up going, joining the business. We positioned it for a sale and it got acquired. But Summit was famous for two things. So anybody knows Summit Media would know one of these two things. They founded the business in a prison. So it's a really interesting story and there's a whole separate conversation sometime about that and to help, to help rehabilitate prisoners, teach them some skills and this, that and the other. But also they looked after Argos for 13 years. They were actually responsible I suppose over time for the Argos website, which actually in the UK is one of the. When we were looking after it was the second biggest E com website after Amazon in the uk spending a couple of million pounds a month on Google. And we built some software, Keith, at uh, Summit, to predict where to spend your Google Ad, just really to help our team and help the profit for Argos because it was so vast, it was just difficult. We had people with PhDs and degrees in maths trying to work it all out and we're like, surely there's some software we could build to do this. So that sort of started the process and then we turned that into a platform called Forecaster. And I learned a lot about building product there because every time I. So we thought we'll go sell this to everybody. This is gonna be brilliant. Unless you spent £2 million a month and had 65,000 products and 800 stores, it didn't work. So we'd got a total addressable audience for this potentially life changing ad product of about three companies. So. But what it got me in the back of my mind was what if we could build something like Forecaster for everybody? That also pulled in your Facebook data, also pulled in your TikTok data, also pulled in your Microsoft ads data. That would be wonderful. And that's really what the conversation I was thinking in the back of my mind. And I shared that with um, Bond when we were in a skiing in Sweden and he goes well that sounds like a great idea and you're going to need some money. For that. And I'm like, no, no, I've got a plan, I'm just going to slowly do it. And then he's going, no, you're going to need some proper money for that. And that's how he ended up becoming an investor. And then we've not looked back since then. We've had a few failed start. As you sort of implied at the beginning. It's hard building tech at this level with limited funds. We're taking on some massive companies. I suppose it's built by retailers for retailers or built by marketers for marketers. It's not just tech.

Speaker B: I did say that. But of course the opposing argument for that is, is that we live in a world where you can start small and if you create something which people will use, you know what, it doesn't even matter what the opposition are to going doing because they're in fact the opposition or the larger companies sometimes are highly irrelevant because they're never going to be doing what the smaller guy is going to be doing. You know, so let's jump into the product. And it is, I'm just fascinated by it, right. So I'm going to jump into the thing that I personally am more interested in than anything else and that is your AI powered forecasting. Right. So you can probably answer this one first, John. One of the biggest problems with most of these tools is that customers or merchants as Shopify refers to them, are using so many apps, so many tools and so many plugins that very often the dashboard tools or the forecasting tools or Ask Bosco can't take everything in and there's gaps. So in other words, they'll pull a dashboard from Tablio and at the bottom of the dashboard in the PowerPoint presentation would be, oh, it excludes Google Ads for a couple of months for X and X. Reason firstly for Ask Bosco, have you addressed the gaps in the existing kind of service?

Speaker C: We will be able to pull all data stitched together for the last two years as a standard is what we do. So we'll take all the Google Ad, all the Google Analytics, also all your SEO data from Search Console, plus all your Shopify data, your metadata. I'm simplifying this and my team get really annoyed with me. Do you know how long it took us to build that? And you just talk about it in like two seconds. So we've got these really super clever connectors that mean we. Because when you extract stuff from meta, uh, and this is why again some people aren't doing this. So if you did have a tableau, you'd still need a connector to get the data from somewhere into it. Part of what our secret source is the ability to take all this data from in different structures and format, line it all up together to then be able to start comparing. But nobody has got the full picture. And anybody who says they've got. There are companies out there saying they've got the full picture because of GDPR and also all the cookie consent, you can't possibly have the full picture. But we have what is available and the best picture. And then from that we can start doing some sort of analysis and modeling. And at a, uh, top level we're looking at a cost revenue model model. And what's clear to. What I want to make clear to everybody listening is we're not using generative AI to predict the future. Right. We use generative AI to help answer your questions on your data, but we use our own AI algorithms and process that we've built from the experience of looking after Argos for 13 years to then predict the future. So it's hardcore maths and machine learning. It's not just ChatGPT making up some numbers because actually some of these generative AI models aren't very good at maths.

Speaker B: What you're building is and honest source of truth and then your forecast is based on all of that.

Speaker C: So there's. Yeah, that's the easiest way is we look at the facts of what's happened. We also then look at seasonality and look at demand and then go right, where should you spend your next pound, euro, uh, dollar, uh, based on what your objectives are. So what you can put into the system and say, right, we want maximum number of sales and we want maximum revenue, which are two different things, right? All we want on the most valuable customers and then it goes right, based on the performance of your media in the past to get the most valuable customers. Here's how we think you should split your accounts and your spend at an account level across the different platforms, but also then an individual campaign level. The big vision originally was this is going to be your marketing agency in a box. I don't think we're there yet. I think everybody still wants a human to check everything. But I do think if you hear what Facebook are talking about in the media at the moment, they said, oh, you don't need an agency, just upload a photo and we'll turn it into a video ad. It's an interesting and scary time.

Speaker B: John, this is Facebook that will send you an email to tell you that your last six months data was inaccurate. Same company.

Speaker C: Yeah. And I think the where we get to, which is why the interesting sort of we want to focus on first party data in Shopify is a lot of the platform data, as you've already highlighted, Keith has gaps. So what we can do though is if we've got all the actual transactions, all the customer data, and then we can match that with all the platform data, we will have the best available view to make predictions, but there will always be gaps. And then what we try and do with clients is understand, well, what is the best attribution model. Then when we've got the attribution model, we can load that into the forecasting platform and then forecast using that as well. So you've got a lot of different options, but we're also trying to make it in a way that's super easy for anybody to use. So if you're the owner of your own store, you don't need a degree in maths to use it because I think that's the other thing with a lot of these reporting things. You need to use SQL or you need to understand how data logic works together. Uh, I just want to help someone to type in a question and go, that's the graph I want. Right. I want to see that graph every day. This is the other graph. I want to see that graph every day or I want to see that number every day.

Speaker B: So John, what's the piece that you're most excited about? I just noticed that this is highly relevant for busy Shopify stores, but it's also highly relevant for marketing agencies because they can soak in all of the data from multiple clients. You're covering a really wide base. Right, so what's the piece you're most excited about? Either a piece that you already have or a piece that you're working on.

Speaker C: The bit I'm genuinely excited about is trying to help brands ultimately save money and spend their money in the right way. I've never had. I was at Google this week, right. On Tuesday, Agency leaders conference, which I got told off for Keith, because I obviously sell software to agencies. I still am involved in a small digital marketing agency. So we still get invited to all this stuff because we built the software and tested it on our own agency.

Speaker A: Uh.

Speaker B: Ah, very good.

Speaker C: Yeah, yeah, Right. So I go to that as modo25, but with my bright yellow, I've got black on today because it's a bit hot and I don't want to get sweat marks on my yellow T shirt. Right. But I was at Google on Tuesday with my bright yellow T shirt. It's the ability agencies are under a lot of pressure at the moment to rethink how they do stuff. Absolutely rethink how they do stuff. You've got Google and Meta all saying, spend more money Google. I've never been to a meeting with Google where they said, I tell you what, we've worked out how you can spend less money and make more sense. Uh, the answer's always spend more money. Google's a for profit. What I want to try and help the merchants do is actually spend less money and make more sales across all the different advertising platforms and hopefully make more profit. Because a lot of people think E commerce is easy. E commerce is hard. It's got tight margins. You've got returns, you've got trade tariffs. If anything, you see all these people saying, oh, you've got an online business. Oh, that must be like sat there just rolling in money. It's hard work.

Speaker B: John, do you know what annoys me when I see these, uh, YouTube videos and LinkedIn posts? There's a guy posted a message yesterday saying that he brought somebody from 0 to 25 million in three months on TikTok. Anybody using TikTok shop knows the hours and hours and hours of effort. And I think it's those people that are misleading people. If we only talk about Shopify, Shopify is updating about a thousand times a year. There's two editions every year, uh, front end and account changes and also massive big tech changes at the back end as well. Even for an E commerce manager or an E commerce team to stay on top of those changes alone, let alone the 20,000 apps that are updating and that's only one platform. It's become very, very complex. So agree with you 100% on that.

Speaker C: And I think the other thing is Google and Meta are both making it harder to understand your data. They're saying let our clever A.I. uh, work it all out, uh, give us your money and we'll go spend it. That's like putting my 15 year old in charge of the sweet shop. It's absolutely madness that you set your target, they'll hit your target, but they won't beat it because they're not incentivized to beat it. They're incentivized to maximize their revenue. So, so what I'm excited about is trying to demystify and uncouple all of this because I think you certainly, uh, E commerce retail has an addiction to Google Shopping. There's an addiction to feeding the monster. And there's an element of we've all just, it's just can we look at it in a different way and can we find the data to understand? Well how should we look at it in a different way and should we actually justify spending some money on some user generated content that actually drives different type of engagement? But then nobody actually searched for our product because they already know about it and they bought it from somewhere else. So that, that's what I'm excited about is trying to help other people make more profit in their business.

Speaker B: And if there was, if there was one feature of the platform that you wanted to shout about, what would it be?

Speaker C: Well the one I'm most excited about in terms of feature is not actually the forecasting is brilliant but the bit that's getting me excited. So I had a, I was sharing, they have a Shopify merch actually and I got this text uh, from a CEO of this big online shoe retailer and he goes I've just done it. I goes what do you mean you've just done it? Done what? And ah, he goes we were in a board meeting and we had a question about marketing spend in certain channels. In return the CMO didn't know the answer because it wasn't in his board papers. Normally what would have happened is someone would have taken an action, the action would have gone to the marketing team, the marketing team would have asked the agency a week later would have got the answer and would have forgotten what the question was. But I just fired up my asked Bosco and I typed in my question and I got the answer immediately to exactly what we wanted to do and we moved on and we made a decision. So having access to all your data and just being able to ask it questions and then ask follow up questions to run your business quicker is what I'm excited about. I think that will empower uh, because you see in a lot of big companies outside the C suite are data analysts and they're there just to answer the question, I need this, I need this, I need this dashboard, I need this. Well actually if you could just have something that gave you the answers and unlocked your data. Data. That's what I'm excited about.

Speaker B: That's great. Look, you've explained it really, really well. So what's the future of Ask Bosco? Is it bigger customers? Like you've got customers all over the world. That's the beauty about owning a software business. But what's the, you know, what do you see in the next Year or two.

Speaker C: So this is probably a bit bold and lofty, but we've got to go big, haven't you? You got to go big. So I think. Right, so I'll give you an example to put into context. So I stayed in a hotel on business the other week and I then texted my wife Nikki and said, oh, this hotel's should stay. What's the first thing she does? She googled it. Yeah, Googled it. Looked at the TripAdvisor reviews and said we're not staying there. Right? Yeah, yeah, yeah. But m. I'm here. It's really nice. That's because TripAdvisor has become this global platform for trust and authority. They are a globally recognized trust authority platform. I don't think there is a globally recognized trust and authority platform for E commerce, marketing and performance data right across all the channels. So what we want to become is almost like the Trip Advisor for marketing and E commerce data. So if you're an agency going out to pitch, you can go confidently pitch and say we're in our spouse, co partner. So you can trust the numbers. There's no smoke and mirrors. We're doing the best possible things we can with your data. That frees us up to work on strategy. You're not paying us to crunch spreadsheets and reports or if you're a brand, you say, we like your creative ideas, we like your thinking, but we want you to use our Asposco dashboard because we trust the numbers in that. So you're going to have to work with the Asposco dashboard and it just becomes the thing, thing for cross channel measurement and reporting. That's the sort of lofty vision and I think that's achievable because that's what people need. And nobody gets up on a Monday morning going, I'm really looking forward to crunching and copying and pasting and doing a pivot table and not knowing the answers my clients are going to ask.

Speaker B: But, but John, I think those days are gone anyway because the skill set of, you know when you're hiring an E commerce manager. Now in the old days there had to be an Excel jockey, but it's all changed. They need information. They've got information coming from so many different platforms and channels and areas and apps and services that what you're talking about makes absolute perfect sense. Look, it was, it was great talking to you. You'll give me some links for some of your clients and we put them in the show notes along with information on Ask Bosco. I wish you the best. And we will probably meet at some point again, uh, somewhere in Europe at some stage in the future.

Speaker C: We will see you on the bike soon. Keith.

Speaker B: Yes, I forgot the bike. I'll see you.

Speaker C: See you on the bike.

Speaker B: Good man. Good man. We'll talk to you soon. Thank you very much.

Speaker C: Thanks for having me on.

Speaker A: Thanks for listening. All of our episodes are available on Apple, Spotify, itunes and Google podcasts. If you're an e commerce professional with an exciting story to tell, reach out to the team on helloilkbottle labs.com. until the next time, take care.