What merchandising is important?

Andre Brown Transcript:

Infinity nation and beyond, we look at the status of your digital landscape and how it affects the growth and profitability of your business. On the show we will be talking to business leads partners in the E commerce world, as well as some of our own specialists team to give you actionable insights and find solutions to help you on your growth journey today. We will consider new ideas stretch our mindsets beyond the status quo. And in the process, discover how to gain an edge on the competition and drive some great results.

AL:

Today we welcome Andre Brown from advanced commerce. Ai Andre, how you doing? Hi

AANDRE:

I’m good. Thank you. How are you?

AL:

Yeah, great. Thanks. So I think today we’re going to delve into the world of onsite merchandising and why artificial intelligence, artificial intelligence and machine learning alone won’t be your sole solution to merchandising.

ANDRE:

That sounds right.

AL:

Good. I mean, I think as you and I have spoken about many times before, I think is a lost art. You know that sort of the whole inspiring our customers when they land on the website, what might be great product that’s relevant to them. But yeah, let’s hear a bit more about your world. A bit more about advanced commerce where you’ve come from and then we’ll ticket on Sure.

ADNRE:

So very quickly, by way of background, prior to advanced commerce, I was the founder of attract and the former chief exec and that’s a business that I ran for 15 years. I really grew that from zero to becoming the leading European player in search for Merchandising, and that journey included going public in 2014. And then in 2017, acquiring Fred Hopper, which is another company that people are probably familiar with. So my background is very much about search for Merchandising, and merchandising really is about controlling the product display sequence or product curation. And as you say, it’s a bit of a lost art in many ways and the reason why it’s important when it goes to this paradox of choice. So this is the issue you face as retailers is, you know, your consumers are facing this paradox of choice. I’ve got so many options to choose from, and they’ve got a very small attention span. So you know your typical visitors and you’re going to look at the first two or three pages on your site. Therefore, they’re going to see a very small fraction of what’s available from your online catalogue. And once you know that, it suddenly makes the decisions that you take about product curation and very specifically, decisions you take about what to show on those first few pages, suddenly makes that really, really important. Much like the care and attention that retailers would pay to shelf space in a physical store. And it is basically the same attention problem, but it’s much worse than the online well because there’s no real limits to how big your product catalogue can be. Whereas at least in a physical store, you know, you’ve got physical limit of how much stuff you can put out. So that’s kind of the problem. And obviously, you know, as a business that’s the problem that we solve with our merchandising platform. But people often ask me, Well, why can’t you just automated them? Why didn’t you just apply artificial intelligence to it? And the answer to that is, well, we do use artificial intelligence and AI is really good for doing a lot of the heavy lifting. But AI alone won’t do it. And then I’m going to run through some reasons why it won’t do it. So the first reason is that AI assumes that past behaviour is a good indicator of future behaviour, which on the face of it sounds reasonable sounds like a reasonable assumption to make. But it’s certainly not true for big ticket items. So if you’ve recently bought a sofa, or you’ve recently bought, you know, set of golf clubs or a camera or something like that, you’re not going to be in the market anytime soon for another big ticket item. So certainly for similar problems, it doesn’t allow for changes in taste or size. So you know, when I was younger, I really like milk chocolate. As I’ve got older I prefer dark chocolate, you know changing tastes, or if I read out to you somebody’s clothing purchases over the last year. And let’s say off the top the 20 items they bought, you know eight of them were cut out. Six of them had velvet another six word kind of a certain colour. A I would look at that and think oh well we know what you like. So that’s what I’m going to recommend. What AI doesn’t realise is actually I’ve just described last season’s trends to you it’s not the seasons

AL:

Yes. To notice. It can’t really cope with that sort of changing in taste And then would it also be thrown out of a gifting Andre like if I bought someone a prom?

ANDRE:

it’s funny because that’s my next point is assumes that you always buy for yourself otherwise if it doesn’t make that assumption doesn’t really work right but so you know if I if I buy something from my aunt, and guess what AI is going to recommend stuff that is suitable for my aunt not for me, has been of a similar problem, which is it doesn’t allow for multiple purchases on the same device. So I don’t know what your home is like but you know, typically we’ve got an iPad in the living room. And guess what you know, my daughters tend to all use that to do the final purchase. So that then aggregates all of their behaviour into one profile. So you know again, you’re that’s why you will get odd results because there’s not necessarily your profile is matching. Sure. And then another problem is, you know, by definition, it’s retrospective. Can’t it can’t predict the future and basic users historic data to build a model and then it views the world through that model. So it’s all rare we’re looking and just to try and give you an analogy. Let’s imagine that instead of being a retailer, we were X Factor. And we’re applying AI to X Factor. We give it this kind of historic data and it says okay, we know what a boy band looks like. And our sorry to break it to you but it’s obviously not me and you are happy about it knows what a typical boy band looks like. It’s gonna be four or five of them. They’re all gonna have emo haircuts. One of them’s gonna wear the jacket, blah, blah, blah. It also knows what a female bands in a girl band or what a female singer would look like. So what happens when a young Freddie Mercury turns up? It doesn’t know what to do because it doesn’t fit into the model. And to paraphrase somebody else, you know, computer says no. And so, really, kind of the summary for all of this to me is almost a challenge to retailers which is do you want to be a trendsetter or do you want to be an order taker? So the worry with an over reliance on AI I’m not I’m not saying not to use AI because we do use AI but the sort of the worry with just trying to hand it over purely to AI. Because I think by definition, you just become an order taker. You’re competing with everybody else on the same basis. And we know that would lead to a race to the bottom on price. So you know, that’s the challenge. Do you want to be an order? taker? Or do you want to become unsettled to be a trendsetter? You’re not going to get there with AI? Because AI can’t really predict the future. It’s only working on what’s the past. So much like the analogy with with Freddie Mercury. You’ve never seen that before. But guess what, the VA an amazing trendsetter.

AL:

Yes

ANDRE:

I suppose in your experience, in my experience, and I suppose maybe it’d been different a differing areas or different levels of econ business. I think we both agree there seems to be an own unloved and lost art or just I don’t know, maybe I’m being critical but since a lot of effort is spent on where do we go and acquire new customers, what would they look like? What types of customers what products would they like to spend a lot of time and effort and money acquiring go to the site. But we don’t seem to put the same love and attention into that experience when they’re in store as we do in terms of trying to attract them. And that’s what I always have to try and understand. Why is that? And then I’d say a typical business or some business we’ve been looking at. It’s like 70% geared towards acquisition, maybe 80% geared towards acquisition and then 10%, a bit of CRO merchandising and 10% retention and I sort of try and wrap my head around that in terms of sure there’s probably a turnover growth target that someone wants to hit but is that the most profitable way and is that building the best? We took on this these podcasts call about customer lifetime value and leveraging that database asset. And this was, again, I’m always considerable considering tools that are going to help me build that greater loyalty and repeat purchase because then any money I spend up a funnel is only going to give me a greater return over that customer’s lifetime. And the value and I think it’s Yeah, I don’t know. I just I’d love to hear what you’re seeing in the marketplace. And I think we’re both in agreement that curating your range in the correct way to the type of customer landing isn’t the norm or isn’t the norm in a lot of E commerce?

AL:

I think absolutely. Right. So you know, talked about the importance of product curation or product sequencing. I’m guessing you could go to you could pick almost any site, and the chances are five different visitors to that same site will get exactly the same sequence. Even though the different democratics different demographics, some of them will be bargain hunters, some of the fashionistas, some of the, you know, loyal returning visitors. Some will be very low, certain brands, and yet everybody gets the same sequence that really doesn’t make a lot of sense to that. And I think the I think there’s probably two reasons for that. I think one reason is, you know, digital marketing is a very well known well understood discipline now. And, you know, there’s lots and lots of people doing it. I think visual merchandising thing of its proper name is still quite a niche area of expertise. And I think a lot of mid tier and smaller retailers assume that that’s an enterprise game. And that that is not big enough. And you know, when when we launched advanced commerce, my whole pitch of the launch was bringing enterprise grade merchandising to the small and mid tier retailer. So you know, we’ve designed something that’s designed to be used by non technical people. And I think that’s a problem with merchandising in the enterprise space. The technologies are very complex and technical. So they sort of appeal to big IT teams. But wouldn’t appeal to smaller retailer where maybe you’ve got a a web manager having to run everything. And that’s really the problem that we’ve we’ve tried to solve is bringing that kind of enterprise level of merchandising capability to the small and mid tier retailer so it can be done by anybody who looks after the website.

ANDRE:

And in your extensive experience in this world, in most nursing and search, what would you say is the net average benefit of of doing this over law, I’ve built a website. I add my new product in it goes into the new category comes to the category. Appreciate that’s a very simplistic view of it. But if you’re an econ manager and you are listening to this, what what might sort of grab their attention even in terms of what typical trends are using in terms of when people are using this sorts of technology.

So we tend to see two metrics being positively affected. One is average order value. And the other is the is the monthly volume of orders, which is probably a product of conversion rate. So so it’s quite typical to see a client may be improved their ARV by 10 12% and maybe increase their volume of orders say by 5%, something like that. But then the combination of those two can be quite significant can be in the 20s in terms of percentage points.

AL:

And that’s assuming no extra traffic. So that’s from the same traffic net net benefit. Okay, I think so just highlights how key is in and does your solution also then feed into email tools? Is it purely so you can you can kind of connect it with an email tool

ANDRE:

So typically what people would do is they often do the customer segmentation within the email system within the CRM system. So we have a function called Persona. So I can create a persona within my console. And I can link an email link to that persona. So I could say for you, our let’s say we thought you were a bargain hunter. And I’m now sending out an email, or I’m targeting Facebook ad or something like that you then as long as we have that connection to our persona. When you arrive at the at the site. That persona will then trigger a merchandising strategy. That’s aimed at bargain hunters. Typically, whatever you do, whether you’re browsing or searching, your product sequence will be based on what’s on sale, or possibly what’s cheap, with good stock availability. Whereas if we think you’re a fashionista as a maybe a different persona, then that would trigger a strategy that says I’m going to show you the latest stuff, but maybe maybe I’m going to show you descending price order. So I’ll show you the most expensive things first, or you could do something like you could create your own sort of custom range for price. So you might identify a group of people who are really mid range price. And so the best strategy would be d power the top of the price curve D power the the bottom end, so you just focus on the mid range price. So it’s about giving people a much more appropriate product sequence. So that’s why you improve your ARV because you’re not you’re not showing stuff that’s on sale to somebody is a fashionista. You’re showing them the newest stuff. Yeah. And they’re less price sensitive. So that’s why that would boost the ARV and Germany your volume of orders goes up then the more appropriate things that you show to people and that’s really a function of conversion rate.

And with the customer, the end user in mind do you get to track or measure I mean obviously AV and conversion rate but any other metrics that you can measure in terms of how it gives them a better experience and how they then enjoy that brand. Buying from that business further. I mean is it is there any correlation in terms of Trustpilot stars?

AL:

Sure, there aren’t I can’t point to a particular case study. We’d also integrate with Google optimise so I know a lot of people use that to try and get that kind of information on search.

ANDRE:

It’s trying to see that mug that 360 view of how it’s helping the retailer in the business sell more appropriate products in front of the appropriate eyes and then is was there a Yeah, I was just intrigued if there was a correlation the other way around. Okay, any other points in terms of the merchandising the AI strike ml, that alone won’t help you achieve your ultimate we’re sort of saying here’s an improvement in conversion rates you said improvement in customer

ANDRE:

I think I’ll just go back to what I said at the beginning, which is once you realise that your typical visitors only going to look at the first few pages. It suddenly makes the decisions you take about what to show first incredibly important. And yet most sites will show exactly the same sequence every single visitor and that’s you know, that’s where there’s real value, to be had. disfiguring does figuring that out. What should I show our versus what’s your show, Andre? What should it show Lily and its ability to to make a change quickly sit immediately in effect, and start to measure whether it’s made any effect or not. And if it does great if it doesn’t roll back, try something else. It’s just that ability to be agile.

And typically with a solution like yours in play. What sort of timeframe does it take to get this up running and generating positive results? I mean, there must be so it depends on what frame you’re on. If you’re on Shopify, we could get you live same day. Assuming you have all your ducks lined up in a row. I mean, I think typically we quote we tend to quote like four to six weeks for implementation

AL:

cool, any other bits you want to pearls of wisdom?

ANDRE:

Not that I can think of I’m sure as soon as I finished the call. I think of them and I’ll have to

AL:

Well, I think that was really useful say it’s agree with you. I think there’s an area where more attention to be put in and where a solution like advanced commerce could just help. All econ businesses just be that little bit savvy or that little bit smarter and just be you know, drive more revenue from the traffic they’re already sending to the site. So I think for me, it’s definitely on the bucket list when looking for growth strategies. And if people want to find out a bit more about advanced commerce, what’s

ANDRE:

the best way they can they go see look at our website, which is advanced commerce.io They can find me on LinkedIn or they can email me so I’m Andre at Advanced commerce.io. We’d love to chat to you.

AL:

Cool. Andre, thank you so much for your time. Great to clenches I’m sure we’ll speak soon. Take care.

ANDRE:

Bye