Latest / Inside Commerce: Ecommerce Strategy, CX and Technology Podcast / EP309: How to Supercharge Shopify CRO: Segmentation, Personalisation & Testing, With Visually Co-Founder & Vervaunt CRO Lead
Transcript
- 0:00The fact that Shopify exists and has such a huge distribution
- 0:04allowed us to create a tool that allows brands and partners to
- 0:09execute experimentation in a fraction of the resources.
- 0:17You listen to Inside Commerce, the weekly ecom strategy,
- 0:19technology and CX focus podcast hosted by experienced digital
- 0:22consultants and re platform specialist James Go to Paul
- 0:24Rogers, get insights into ecom strategy leading brands, hear
- 0:27from industry thought leaders and leading tech vendors with
- 0:30practical advice to help you improve your online CX.
- 0:35This episode is sponsored by Shopify.
- 0:37We're looking at a really important topic today.
- 0:40It's around conversion rate optimization for Shopify brands,
- 0:45segmentation, personalisation and test them with two highly
- 0:48experienced guests who will be able to give you a massive
- 0:51amount of like practical real world insights from their
- 0:54expertise. So a warm welcome.
- 0:56Let's let's do it 1 by 1. So they confuse you both about
- 0:59talking over each other. But first, a warm welcome to
- 1:01Anya, who's visually Co founder and CEO.
- 1:04Hey, Anya. Hi.
- 1:06Hi, James. How are you?
- 1:07I'm good, thank you. Welcome.
- 1:09Tel Aviv today. So you're warmer than anyone
- 1:11listening in Europe or in our and also to Nick, who's the CRO
- 1:17lead at Vervant. Hi, Nick, How are you doing?
- 1:20Hey, very good. Thank you.
- 1:22Yeah, good. Thanks.
- 1:22Thank you for joining us. Really looking forward to this
- 1:24today. I know that you guys have worked
- 1:27on lots of different businesses, projects and got lots of
- 1:29different perspectives there. So there's not a bit of
- 1:31background in instructions. So Nick, do you want to start
- 1:33and give people a flavour for who you are, what you do, what
- 1:35sort of projects you work on? Sure.
- 1:38Yeah. So as you said, I'm CRO lead at
- 1:40Vervant. I've been here.
- 1:42In fact, it was my year anniversary yesterday at Bevon,
- 1:46but I've made it. But I've been in the world of
- 1:50kind of UX and CRO for for over 10 years, worked across loads of
- 1:54different businesses from B to B more recently.
- 1:57So like pushing into e-commerce, which I personally find the most
- 2:01interesting. I think there's lots more
- 2:02movement there. But yeah, always been agency
- 2:05side came into Bevon to kind of build the CRO service introduced
- 2:11to Anya at visually, which has been game changing, which I'm
- 2:15sure will come on to a little bit later.
- 2:17But I'm kind of working on a bit of everything.
- 2:21So I I work across the full spectrum of of CRO.
- 2:24So everything from the customer research through to ideation
- 2:28design, test development and analysis, and then obviously
- 2:31program management that sits within that and, and all the
- 2:34kind of client facing pieces too.
- 2:36So I've been doing that particular role for for probably
- 2:415 or 6 years now. And now I'm starting to just try
- 2:45and build out the team here. But yeah, really looking forward
- 2:48to getting into this subject. Very happy to have Anya here
- 2:51alongside me to do work really closely together.
- 2:53So yeah, I'm excited to get into it.
- 2:55Yeah, I'm really looking forward to align the fact you tease out
- 2:57all the different areas CRO because I think sometimes people
- 3:00think of it as is the tests that are executed don't spend so much
- 3:03time on how much insight and research analysis needs to go in
- 3:07order to drive effective experimentation.
- 3:09So looking forward to detail all of those elements.
- 3:13So Anya, so by way of instruction to people, some
- 3:16people have heard of visually, some people will know you,
- 3:19others won't. You grew up in the CRO world, so
- 3:22let's do an instruction to who you are in your background, but
- 3:24also what you've learnt from running experimentation for a
- 3:27large retailer that then LED you to building visually.
- 3:31Yeah, absolutely. Thank you.
- 3:32So I've been e-commerce for more than 16 years, most on brand
- 3:37side where I scaled the two CS from 1200 million, one of the
- 3:43largest brands I worked for for seven years and was their senior
- 3:47VP of Product and Data. We actually really doubled down
- 3:51on experimentation and we ran close to 1000 tests annually.
- 3:57So that led us to increase our per session value in over 100%
- 4:01annually and that led Meta to invite me to talk about how we
- 4:05did that. I used to give talks at Meta to
- 4:09other brands and then with time brands would come up to me
- 4:14saying, you know, nice examples. I used to kind of show
- 4:18personalizations and tests, but how do we do this?
- 4:22And for me, back in the day, it was kind of obvious.
- 4:25We had close to 100 developers and 15 people on products and
- 4:30data. So we had huge teams behind it.
- 4:35But what I realized that the brands that are on Shopify, they
- 4:39don't have these huge technical teams and sometimes they are,
- 4:45you know, 8 figure brands and they don't even have one
- 4:48developer in house, right? So I was trying to research for
- 4:52something like Clavio for onsite.
- 4:54So basically an experimentation personalization tool that will
- 4:58be platform to platform and we didn't find anything.
- 5:01And that's how I crossed sides to B to B.
- 5:05And that's how we created visually.
- 5:09Amazing. And I think this is really
- 5:10interesting because a lot of people listen on brands side.
- 5:13One of the key questions they have I guess of consultants like
- 5:18us technology founders is, is what knowledge and understands
- 5:21they have the mechanics and the challenges that brands go
- 5:24through on a day-to-day basis that impact their ability to
- 5:27execute. So it's really good that we've
- 5:28got those perspectives as well as the execution side of it.
- 5:33So let let's start. So we're focusing in on, I know
- 5:35that Vigili's focusing on the Shopify community and a Nick, I
- 5:40know that Vivant worked with a vast range of, of, of
- 5:44predominantly premium brands in that space as well.
- 5:46So we, we saw a bit of platform centric and then we get more
- 5:49broader strategic. But to Anya, I've heard other
- 5:52people in the past saying, I'm not saying it's right or wrong.
- 5:55The it's hard to do meaningful AB testing, for example, on
- 5:58Shopify due to platform constraints, due to the way that
- 6:00the technology and third parties integrate.
- 6:04How would you answer the subjection?
- 6:06How is it possible to do it meaningfully and to get good
- 6:09test executed, given the fact that you can't just do anything
- 6:12you want on a platform that's managed by a SAS spender?
- 6:17Yeah, that's a great question. I except for one area which is
- 6:24the checkout and is obviously a lot more limited in Shopify,
- 6:27although visually does integrate to checkout, extensibility is
- 6:31and one is of the only tools that does that except the
- 6:35checkout. I think it's really the other
- 6:37way around. I mean, the fact that Shopify
- 6:40exists and has such a huge distribution allowed us to
- 6:44create a tool that allows brands and partners to execute
- 6:50experimentation in a fraction of the resources.
- 6:53And I'll give a very, very specific example.
- 6:56So let's say that you're an e-commerce manager and an
- 7:00apparel brand, right? And you woke up one day wanting
- 7:04to test what happens if you emphasize your returns free
- 7:09returns policy on your cart door, right?
- 7:12So what, what do you usually need to happen?
- 7:15And I know because I used to write those JIRA tickets when I
- 7:18was a brand side. So 1, you need the developer to
- 7:23add that text right to your cart.
- 7:27Then two, you can't really show it to all carts because some
- 7:32carts, right, contain products that are final clearance and
- 7:37they're not eligible, right? So your developer needs to make
- 7:41sure that they understand the content of the cart, right and,
- 7:46and kind of customize it accordingly.
- 7:49And then three, it makes sense that it will increase your
- 7:53conversion rate in a substantial way, but what will happen to
- 7:58your returns? Maybe it will also increase your
- 8:02returns, right? And therefore your profitability
- 8:05will decrease. So now you need to find a way to
- 8:09get your returns data to your experimentation platform, right?
- 8:14And you also need, you know, maybe a developer or an analyst
- 8:18to do that afterwards, right? So that was kind of prior to
- 8:23Shopify. And visually now you can wake up
- 8:27one day as an e-commerce manager and not having any
- 8:31experimentation platform, not having any technical background
- 8:35whatsoever. And to have this test live on
- 8:39air in an hour, right? Because adding the text to your
- 8:45card drawer takes is maybe as simple as writing an e-mail,
- 8:50right? And then two, visually, thanks
- 8:54to the deep integration to Shopify, understands the content
- 8:59and the dynamic content of the cards and understand how your
- 9:02catalog is structured, so it understands if it's final
- 9:06clearance or not, right? And then lastly, we pull all the
- 9:09data right from Shopify natively.
- 9:12So we have returns, profitability, subscription rate
- 9:16all out-of-the-box with just one simple install.
- 9:20So. So as I see it, it's kind of the
- 9:22other way around. What do Mattel, Victoria,
- 9:27Beckham, Sennheiser and Glossier all have in common?
- 9:31They power their businesses with Shopify.
- 9:35Shopify is the most innovative and scaled commerce platform on
- 9:39the planet. It also happens to have the best
- 9:42converting checkout on the planet, Shopify, powering the
- 9:48future of enterprise commerce. Find out more at
- 9:51shopify.com/enterprise. So I guess, I guess this is, I
- 9:56mean, it's always worth teasing this out is, is one of the the
- 9:59challenges to brands is making sure that they have storefronts
- 10:03set up properly in data connections and integrations,
- 10:05right? So that the fair pipe platforms
- 10:07can pull this data. And I have this in discovery lot
- 10:10of time and people descope things saying, oh, it's not
- 10:12critical, we don't need that data in.
- 10:14And then you realise that later on it actually impacts your
- 10:17ability to do programs like this.
- 10:18So I think that's an important thing for people to listen is if
- 10:21you want to take experimentation seriously, you've really got to
- 10:24get that those robust integrations from other third
- 10:26parties to get data into your econ platform in the first
- 10:29place. So I think that's, thank you for
- 10:33giving a really tangible example because that was going to be my
- 10:35next question was, can you give me an example of one of the ways
- 10:37that you did? I see you've already done that,
- 10:39which is amazing. And Nick, let's talk about that.
- 10:41Let's put a commercial into this.
- 10:43You know, you advise a lot of different brands on this.
- 10:47And I know Paul and you talk about this a lot.
- 10:49I've talked about a lot of Paul and other people is the way that
- 10:52people measure CRO varies quite significantly.
- 10:55So in terms of KPIs and success criteria, where do you think
- 10:59people go wrong, How can I impact, and what for you is a
- 11:03sensible way to measure the outcomes of testing?
- 11:07Yeah. So it's a common question and I
- 11:10think it's probably very misleading down to the the the
- 11:14name of the service conversion rate optimization like it's, it
- 11:18narrows down everything that we do.
- 11:19And I think that when people think, right, we're here to
- 11:23optimize conversion rate, they think of that one metric and
- 11:26that's all they're looking at. We all know sitting on this call
- 11:29and, and plenty of people listening will know as well that
- 11:33conversion rate as a figure is a pretty flawed metric.
- 11:36It's kind of become a bit of a proxy for UX, but it's impacted
- 11:41by so many different factors that if you have a look at like
- 11:46a daily trend line of, of, of conversion rate, it's all over
- 11:50the place. So trying to use that as a
- 11:52single KPI just is never going to work.
- 11:55Like even if you, if you, if you have like a massive lift from a
- 11:59test and you look at that trend line over 12 months, you won't,
- 12:04you won't see it on there at all.
- 12:05So looking at that in isolation across 12 months, for example,
- 12:10it's just, it's, it's not going to give you the results you
- 12:12want. So what we tend to do is we'll
- 12:17look at the kind of cumulative potential of each test that we
- 12:23run. So every time we run a test, we
- 12:25produce a report and in that report is a potential monthly
- 12:29revenue contribution. You know, hopefully they're all
- 12:32uplifts and that looks great. Now a test is obviously run in a
- 12:37very short window, 2 to 4 weeks and you're looking at the
- 12:41revenue increase in that particular period.
- 12:43So you're looking at a real small slice of time, really
- 12:47specific slice of time as well. So to just go, well, my monthly
- 12:50contribution is say 10 grand, but just times about 12, surely
- 12:54that's, that's my ROI done. That would be the dream
- 12:58obviously. But seasonal impact, economic
- 13:02changes, all of these things kind of like have an impact.
- 13:05So what we tend to do is we look at that monthly figure for the
- 13:09for the test that we ran. We then extrapolate that out,
- 13:14but then we make a few adjustments.
- 13:16So we'll adjust for things like even intangible things, like the
- 13:21fact that you've added a new feature to a site.
- 13:24There's like a novelty effect, right?
- 13:25When something new goes live on a site, maybe someone's never
- 13:28seen it before. That spike at the beginning when
- 13:31you run that test might be really high, but then over time
- 13:33that's going to drop off. So make an adjustment for that
- 13:36kind of novelty effect decay. There's also like the reality of
- 13:41implementation. So you might run the test and it
- 13:45looks like this, but then maybe that's an MVP version or maybe
- 13:48it didn't quite translate to to the production development.
- 13:52So there's a little bit of adjustment made there, seasonal
- 13:55and market variability, that's quite a big adjustment.
- 13:58And then there's a bit of buffer in there for any external
- 14:00factors. So it's quite long winded.
- 14:03But when you've got the kind of rough formula to work it out,
- 14:07you can quite quickly get to an ROI figure quite easily.
- 14:10And commercial leaders, all they want to see is pounds and pence,
- 14:14right? They want to see that revenue
- 14:15contribution and the impact that your testing program is having.
- 14:19And I think that's probably the easiest way to run it.
- 14:23Alongside that, we'll, we'll, we'll look at things like the
- 14:27volume of tests that we're running, our win percentage and
- 14:31what we've learnt from each test as well.
- 14:33So building a bit of a library of, of learning.
- 14:35So any of those losing tests still have delivered value to
- 14:38the program. Commercial leaders don't tend to
- 14:41really care about that stuff because they're thinking about,
- 14:44right, I'm spending X, I'm getting Y back.
- 14:46So I think that hopefully answers that commercial
- 14:50question. But yeah, we're we're
- 14:51essentially looking at revenue contribution over time.
- 14:54I like the fact you, you, you flagged the the instability
- 14:57piece because I think this is really important is not seeing
- 15:00the impact of of testing as purely short term and monitoring
- 15:05those co-ops. Because you're right, when
- 15:07somebody gets into production, it can make it differently.
- 15:09You have all these other impacts at what made the test successful
- 15:12in the 1st place might then change over time.
- 15:15And and the reality that I mean, this is what I love about
- 15:18testing is what might be right today could be wrong in six
- 15:21months. Yeah.
- 15:22And we might need a, a new version of that test to then
- 15:25find a better treatment And and actually a positive result could
- 15:28end up negative based on some some other factor.
- 15:31And you just can't, you can't predict 100% where that goes if
- 15:34you're not tracking that data long term.
- 15:36So I think that's really important.
- 15:37Yeah. And I think it's important as
- 15:39well. Like you, you touched on it.
- 15:40Then there's this. You're, you're kind of layering
- 15:42everything, right. So every time you, you get an
- 15:44uplift, that's your new baseline and you're trying laying things
- 15:47on top. But over time they sort of come
- 15:49down. So you need to keep layering up
- 15:51otherwise you're just going to end up flat again.
- 15:53And typically if you look at like a trend line, if it's if
- 15:58it's trending down and you start doing testing, you can, it's
- 16:01still going to trend down, but you might, it might trend down
- 16:04less. So it's that adding that
- 16:06compounding value as well. And yeah, responding to to
- 16:09expectations that you just set up, you know, ChatGPT probably
- 16:13being like the biggest and AI in general, like being the biggest
- 16:16kind of factor. I know we're going to come on to
- 16:17that a little bit later, but. Yeah, it's not a pod without an
- 16:20AI question. Is it these?
- 16:22Yeah, yeah. And you know, and they're a
- 16:25capability I'd like to get into a bit more of the envision is
- 16:28around like segmentation power conversation because by default
- 16:31out the box, none of the SAS vendors do strong segmentation
- 16:36personalisation. And I think a lot of the time
- 16:39tests are done to everybody. So I'd love to know like how do
- 16:43these two come together? How does visually enable this
- 16:46and could you give us some examples so people got like a
- 16:48tangible of what what is possible in terms of targets in
- 16:51specific segments or personalized assets?
- 16:55Yeah, absolutely. So first of all, I think I
- 16:59definitely agree with Nick when it comes to, you know, the
- 17:03acronym zero. I don't think that there's a
- 17:06single person that is actually practising this that likes the
- 17:10term conversion rate optimization because it is
- 17:14usually used to describe the act of digital experience
- 17:19optimization or, or management, right.
- 17:22But it just didn't catch as well as CROI guess.
- 17:26And when we think about kind of the digital experience
- 17:30optimization in a broader way, AB testing is just one tool,
- 17:35right? It's, it's a tool to, to just
- 17:37validate our hypothesis, our predictions.
- 17:40But continuous experience optimization requires something
- 17:47more broad. As you said, segmentation,
- 17:51personalization, just as you wouldn't send the same e-mail to
- 17:56someone who just entered your store for the first time, right,
- 18:01versus someone who bought five times, your site should not show
- 18:08the same content, the same navigations and so on.
- 18:11So basically and and I would just say that kind of prior to
- 18:17my Shopify TASE right, usually experimentation platforms were
- 18:22digital experience platforms and they actually did have the
- 18:27components of AB testing, audience management and
- 18:30segmentation and merchandising, right, all in one platform,
- 18:34right. And that kind of changed in
- 18:38Shopify. And there when we came to this
- 18:43space, we saw kind of disconnected apps that only do
- 18:46upsells, right, or only do product accommodations or only
- 18:49add this. And usually merchants wouldn't
- 18:52even test how these things actually impact, right.
- 18:56So if you would add upsells to your cart, right?
- 19:01And that upsell tool would actually report on huge numbers
- 19:09of attributed revenue. The the actual scenario would
- 19:14probably could be a very different.
- 19:16So for example, if I was in the cart with $100 black T-shirt,
- 19:21right and the app sell algorithm would have suggested a $50 black
- 19:30T-shirt and let's say that I would actually click on that
- 19:34because there was no AB testing, the revenue would attribute to
- 19:38the fact that I actually clicked on an outsell rather than the
- 19:41decrease that that would have created in my business.
- 19:46So that was the space kind of when we enter that.
- 19:50So what I'm saying is that these three elements of experience
- 19:56management within e-commerce are critical for a good
- 20:00experimentation. Execution and so for for some
- 20:07examples, we work with a nine figure brand Faharty from the
- 20:12states. They added 6% in conversion rate
- 20:16by personalizing their homepage based on the shoppers behavior.
- 20:21So emphasizing women's items to shoppers that only showed
- 20:25interest in women's wear as simple as that, right?
- 20:28So basically, visually, they can create audiences based on
- 20:32anonymous real time user behavior and then create
- 20:36experiences and AB test them right first that are
- 20:41personalized to these audiences. Branch an 8 or 9 figure, not
- 20:46sure furniture brand and added 9% of their conversion rate by
- 20:50showcasing specific products only for returning shoppers.
- 20:55So furniture is kind of a long consideration.
- 20:58The beginning. It's more of a discovery phase.
- 21:00So there's no really real reason to show specific products.
- 21:04You kind of want them to explore categories more.
- 21:07So doing that new versus returning, you see a lot of that
- 21:12doing really great, great work. Another brand, 9 figure brand,
- 21:17makeup brand Merit Beauty tested adding something they call the
- 21:23Australian set to their Australian market and increased
- 21:2918% in conversion rate to that market in the in the process.
- 21:34So just a few examples, very straightforward things that if
- 21:40we think about, you know, e-mail marketing or ads are, are kind
- 21:43of the bread and butter, but are still not done as much on, on
- 21:49Psych I. Don't have interest from a
- 21:51technology point of view because in order to drive those
- 21:53personalisations you need to have information about the user,
- 21:57you need to have the product knowledge and you need to be
- 22:00able to serve the correct products.
- 22:01So basically what what in that? What other tech is being used?
- 22:04How much of that is visually doing it and taking data from
- 22:07Shopify? It'd be I think you should be
- 22:09listening to know how the the data has been analysed to send
- 22:12the content back to Shopify. Yeah.
- 22:17So basically we pull data from Shopify but we also pull data
- 22:22from Meta. So we have a native integration
- 22:26to meta, right? And we also we also take data
- 22:34from the user real time session. We do this anonymously so that
- 22:39we can also help our brands to personalize real time, right and
- 22:46of course not all shoppers are logged in, which is most of the
- 22:53data you can bring from Shopify. And you said very correctly that
- 22:58we, we definitely kind of connect the data we have about
- 23:05the user, whether logged in or not, the catalog, right, and how
- 23:12the site is actually built and how the user is interacting with
- 23:15it. And by having all the data
- 23:18together, we are able to facilitate very robust
- 23:23personalization experiences. And, and Nick, I know Paul said
- 23:29that you guys have had lots of debates around like the benefits
- 23:31of different, different types of segments like volume versus
- 23:34value, how you cut up audiences into meaningful segments to help
- 23:39drive like personalisation or testing.
- 23:42Can you talk me through what is a good approach here?
- 23:43What's a good strategy and how does it vary from from brand to
- 23:46brand? Because it will vary depending
- 23:48on the size and the type of business.
- 23:50Yeah, definitely. I think the way that I view this
- 23:55is that in if you're just getting started with with acro
- 24:00programme, I think you want to you want to aim for velocity and
- 24:04and volume. So you want to get as many tests
- 24:07out as you can to prove that the program proof programme value
- 24:11quickly. You're probably likely to find
- 24:14quite a few quick wins in the early stages that are valuable
- 24:20across the site to all users. So, you know, fixing UX
- 24:22problems, highlighting information that's maybe buried
- 24:25below folds and things like that.
- 24:27So there's, there's probably like a really nice place to
- 24:29start where you're just thrashing out lots of fairly
- 24:32simple tests. Not only are you going to get
- 24:36value from those individual tests, but you're starting to
- 24:38understand and learn what sort of the anatomy of a, of an AB
- 24:42test looks like, what resources do you need?
- 24:44What reporting is missing? What tags are broken?
- 24:48So you you're kind of doing that troubleshooting while adding
- 24:51value with the with the testing program.
- 24:54So early on, I'd say yeah, focus on on volume and then start to
- 24:58introduce kind of simple segments early.
- 25:01So, you know, it could be as basic as new versus returning
- 25:06users. And and, and you touched on a
- 25:07couple of these already, like if the returning user homepage,
- 25:12maybe they've abandoned cart, let's get them back into that
- 25:15cart journey when they, when they come back to the site.
- 25:17So there's some fairly like easy segments that you can start to,
- 25:22to leverage quite quickly. And I will say like having used
- 25:26lots of different AB testing platforms, and I'm not just
- 25:29saying this because Anya's here, but the way that visually
- 25:32handles that audience segmentation and gives you
- 25:36access to create audiences on the fly really quickly is just
- 25:42so simply designed and easy to interact with.
- 25:44So I think there's no reason why you couldn't in your first
- 25:48tranche of testing in the first six months start to explore
- 25:51segmentation. What I would say is that
- 25:55typically more mature businesses are set up to handle that
- 26:00segmentation better because they understand their audience is a
- 26:03bit better. They've maybe got more resources
- 26:05to be able to dedicate time to creating bespoke assets.
- 26:09Some, you know, like if you think you you're segmenting your
- 26:12homepage for 10 different audience segments, that's
- 26:16potentially 10 banners, 10 promotions, 10 lots of product
- 26:20recommendations, like lots of different things that that need
- 26:23to go into that. And I think if you're just
- 26:25getting started with testing, that could be quite
- 26:27overwhelming, which is why I say start with the basic stuff and
- 26:31then build up to that segmentation.
- 26:34Yeah. I was going to ask you a
- 26:35question, probably Vanya then on RESO, because I always find
- 26:39this, I work with AB to BUSB to Brand last year who were wanting
- 26:44to drive lots of personalisation and personalisation of content.
- 26:47And then the question always comes, OK, so you won 12
- 26:51segments and you want to completely personalize the home
- 26:53page and there's eight different content layouts.
- 26:56That's, that's, that's 96 bits of content instead of 12.
- 27:02Who's creating that? What is about?
- 27:04And if you're using AI, what's the, the governance process to
- 27:06ensure that it stays on? So I'd love to know what, what,
- 27:10what is that process within within visually Are you, yeah.
- 27:14Do you do people set up and create the those different
- 27:17assets for the different segments of the test?
- 27:18And visually do you then connect with like third party CMSS where
- 27:22people might already be creating different content for multiple
- 27:25segments? Like what?
- 27:26What is that connectivity there from a resource and asset point
- 27:29of view? Yeah, that's a, that's a great
- 27:31question. So currently the way that it
- 27:35works is that they actually create those experiences within
- 27:39visually. And one thing that I believe
- 27:45that you know, AI will definitely change dramatically,
- 27:49right, is our ability to cater very granular personalization,
- 27:56personalized experiences. And the reason is that instead
- 27:59of us having to actually create, right, all of those assets,
- 28:04right, AI will be able to actually execute on creating.
- 28:07And then, you know, humans will be able to decide, right, which
- 28:13variants are along with the brand, right, are approved, are
- 28:18right things strategically and then kind of go on.
- 28:21And and that's, I, I think that's one of the main points
- 28:25when it comes to AI and on site experience in the upcoming
- 28:32years. What Nick, one of the, for
- 28:36somebody who's obviously working with lots of different brands
- 28:38and advisory and consultative capacity as well as obviously
- 28:42doing execution. One of the things I I know a lot
- 28:45of people struggle with is, is where what data is worth it is
- 28:50it is useful to influence our testing strategy versus what's
- 28:54just noise. There's so much data that you
- 28:56can look at and how do you, how do you structure this when you
- 28:59work with a client? So things will be useful for
- 29:01brands who are trying to work out their own way of, of
- 29:03deciphering the data internally is where do you start?
- 29:06What data sets do you look at? How do you know when something's
- 29:09useful versus when something's, yeah, irrelevant?
- 29:13Yeah, you're right in that there's definitely variation
- 29:15between brands. But typically when we start a
- 29:19new engagement with with a client, we'll do like some level
- 29:22of auditing and that is a mixture of firstly, we're
- 29:27starting in GF 4 or Shopify depending on where where the
- 29:31client kind of keeps their their data.
- 29:35And really at that stage, we're just looking to kind of set the
- 29:39scene. We just want to understand what
- 29:41does the typical e-commerce funnel look like?
- 29:44What are the key landing pages that users are being driven to?
- 29:47Which pages are underperforming versus others?
- 29:50We might go into even device level.
- 29:54So you know, how, how does it look on an iPhone 14 versus in
- 29:59Safari versus a Samsung Galaxy? And in Chrome, you often find
- 30:05like little nuances there because like development across
- 30:09the board could you know, you can miss these things.
- 30:11So you might find some quick stuff there.
- 30:13But typically we're using that data to just guide us, so
- 30:18understand what the landscape looks like from a from a numbers
- 30:23point of view, and then we'll go into more qualitative research.
- 30:27So we tend to utilise, at the moment we're using Clarity
- 30:32across quite a few different clients because it's free.
- 30:35For those who don't know Clarity, most people know Hot
- 30:38Jar, so I always reference, it's a bit like Hot Jar.
- 30:40So it's doing it. It's essentially a user
- 30:44behavioural analytics tool. So you've got heat maps, session
- 30:48recordings, you can set up funnel analytics within there as
- 30:51well. And I think it's a quite a nice
- 30:53visual way of, of seeing how people are interacting with your
- 30:57sight. So if we've gone into the data
- 31:01and we found, let's say we found APDP that's really high prone,
- 31:05it's lots of people landing on that PDP.
- 31:08Let's go and understand how people are actually interacting
- 31:10with it. Let's look at some heat maps
- 31:12first. How, how far are people
- 31:14scrolling? How much of the content are they
- 31:15actually seeing? And then go a little bit
- 31:18further, which elements on that page of people interacting with
- 31:21are loads of people clicking the size guide?
- 31:23OK, that starts to tell us that there's potentially an issue
- 31:27around sizing or the messaging around sizing.
- 31:29So then we can start to formulate test ideas, start to
- 31:33build up that road map of tests. So that process from what, what
- 31:38the landscape looks like from a data point of view to how users
- 31:41are actually behaving and interacting with the site to
- 31:44then ideation and, and design of, of tests comes next.
- 31:51So that's, that's kind of like the general process we go
- 31:53through. We kind of spend a bit more time
- 31:55upfront when we first on board people to build out that road
- 31:57map of, of test ideas. But I think most of the value
- 32:02comes from the test results themselves because you're you're
- 32:06purposefully making a change and then you're seeing how users
- 32:10interact with that change. Does it, if it made a
- 32:13difference, OK, why did it make a difference?
- 32:15And then you start to learn something a bit deeper about
- 32:18your particular audience set. So as you start doing more
- 32:21testing, the test results becomes more valuable than
- 32:24anything that you're kind of gathering at GA sort of
- 32:26passively. So I feel like that that that's
- 32:29why like we create this cycle this, I mean, we want it to be
- 32:33ongoing forever. Obviously, we talked about that
- 32:35compounding value earlier as well, but that's why that cycle
- 32:38exists, because each time you run something, you learn
- 32:41something a bit more, that then tells you something else, which
- 32:43goes back into research and then you kind of continue on that
- 32:45flywheel. Do you know what?
- 32:48I'm going to come back to that because testing, documentation
- 32:51and intelligence sharing is really important.
- 32:54So I want to come back to that because that's an area I see a
- 32:58lot of a lot of gaps in, in businesses where people do cool
- 33:02tests, but then forget about it. But before I do that, I've got
- 33:04a, a question that all I always ask when I'm working with anyone
- 33:09in, in Tesla, there's brand side agency, client vendor is
- 33:13evolution from tactical to strategic tests.
- 33:15And I think a lot of things we talked about, and this is my own
- 33:18terminology. So like a more a tactical, which
- 33:21is really important. So that's things like content
- 33:23changes on a home page or it's, you know, the, the, the example
- 33:27you gave Anya about returns information being put into like
- 33:30basically check out what impact that has and conversion.
- 33:34The bit that I think is really interesting, more challenging is
- 33:36strategic testing. So a user journey for example,
- 33:39where it's not just singularised elements, it's multiple working
- 33:43together or pricing strategy bits which can have a if done
- 33:48wrong, can fundamentally screw this a bit if done well can have
- 33:51a massive material impact on revenue and margin.
- 33:54I'd love to hear like what what role does Strategic test NIV
- 33:58have here and how would you go about planning and executing
- 34:01that because it is more challenging to implement and
- 34:03manage. Yeah, it is.
- 34:07And I think it comes back to where a brand sits on that
- 34:11maturity scale of like, you know, are we just getting
- 34:14started or have we been doing this for years?
- 34:16Are we comfortable? Because in order to execute
- 34:20those strategic tests, you, you're going to have to have
- 34:23input from lots of different people, lots of different
- 34:25stakeholders. So that collaboration needs to
- 34:27exist. Everybody needs to kind of
- 34:29understand the value of experimentation.
- 34:32And without that, you know, things that could potentially go
- 34:36wrong piece of data, maybe it's missing somebody's strategic
- 34:39direction isn't right. Like you need that
- 34:40collaboration. So I think there's, the further
- 34:44you're, you are along that journey, the, the more likely it
- 34:47is that you can start to test these bigger, more strategic
- 34:53issues. And I think you mentioned like
- 34:56price testing. So Anya, I know we've talked
- 34:59about this with a mutual client of ours and they're going to run
- 35:02some tests, price testing in the US Be interesting from and I'm,
- 35:06I'm not passing the butt completely.
- 35:08I will continue to answer, but I just want to know from your side
- 35:11like how does how does that price testing kind of look like
- 35:15how, how when we get into the executions that like what
- 35:17factors do we need to to make sure that are all, all there and
- 35:21how do we report on it? Yeah.
- 35:24So it's a great question. I think that there are a few
- 35:29elements around price testing. So there is testing the actual
- 35:33prices of the products, right? And there is things like
- 35:38shipping threshold testing right now and, and sometimes it's,
- 35:44it's also a combination of both, right?
- 35:46So what will happen, right, if we'll increase this and decrease
- 35:50that. So we, we do need to make sure
- 35:52that we have a clear hypothesis around that, right?
- 35:55And when it comes to reporting, and I think that's a really
- 35:58critical aspect. We need to make sure that the
- 36:01reporting supports not only kind of the way that we are used to
- 36:05in, you know, let's say regular testing, which is kind of
- 36:09variant one versus variant 2. We also need to analyze within
- 36:13that variant that subset of products that we influenced,
- 36:17right? What actually happened to their
- 36:20conversion rate, to their AOV, right?
- 36:24So that granular reporting that belongs specifically to price AB
- 36:31tests is critical in order to actually create the right
- 36:39conclusions out of tests. So that's that's one important
- 36:45angle around that. Yeah.
- 36:47And I think that it kind of ties in again to that you, you
- 36:50started to touch on it James before about the documentation.
- 36:54I feel like if you, if you are just running, people call it
- 36:58spaghetti testing. If you're just like running
- 37:00tests left and right, you're not really learning from it, You're
- 37:02not documenting it in a robust way.
- 37:05It becomes really difficult to do anything strategic because
- 37:08you you're, you should be documenting every test result
- 37:12that you've got, learning from that, and then publishing and
- 37:15sharing that. So we use Air Table to basically
- 37:19do a lot of this and create some nice customer facing dashboards
- 37:23so that you don't have to troll through a spreadsheet to try and
- 37:26find what you tested on the homepage 12 months ago.
- 37:28Like you, you can see it really quickly and it all joins
- 37:30together. And I think that in itself
- 37:33becomes a strategic document to help you understand what
- 37:36actually moves the needle when we test it.
- 37:39So we'll, we'll even tag our tests based on like UX
- 37:43principles or something that we're trying to influence.
- 37:47So are we trying to motivate users to take action?
- 37:51Are we trying to improve usability?
- 37:53Are we trying to increase the attention on some particular
- 37:56element? And because we tag all of those
- 37:58different things, we also tag the devices, the audiences, all
- 38:01of that stuff kind of becomes like a nice matrix of OK, we can
- 38:05see of all the tests that we've run, 60% of them were designed
- 38:11to increase motivation for mobile users on the PDP.
- 38:16And when we did that, those six of those 60%, a hundred say 80%
- 38:22of them were saw a significant uplift in conversion.
- 38:26So we know when we take that step back, we can see that
- 38:30strategic landscape of what your audience responds to and what
- 38:35you should do more and what you do less on.
- 38:38And that documentation becomes like yours kind of strategy
- 38:41Bible. And it becomes a better view of
- 38:44your digital experience as a whole because you can see what
- 38:49influences work and what don't when when you change them for
- 38:52for your actual users. I I'm glad you mentioned tagging
- 38:57because that's such a critical thing.
- 38:58Like a taxonomy for for like test results to be able to
- 39:04categorise it and know. I remember a client who who
- 39:07stored loads and loads of information.
- 39:08I think there's like a SharePoint and none of it was
- 39:12categorized. So you had to trawl through
- 39:14every single test to find out what related to APDP or what was
- 39:18testing a price or you know, and engage whatever it was and it
- 39:21was just unusable. If a being upfront on how you
- 39:24structure that is really, really important question for you on
- 39:27that then Nick, because you can have these amazing repositories
- 39:30and then nobody uses them. Yeah.
- 39:32So how what is in your experience?
- 39:35How do you get client teams to like to to use the documentation
- 39:39and to learn from these repositories?
- 39:42Because a lot of effort goes into creating them, but they've
- 39:45got to use them to make them meaningful.
- 39:48Yeah, and I know I keep coming back to it, but it does it does
- 39:51kind of depend on where they sit on that scale.
- 39:53Again, you know, if, if you're just getting started and you're
- 39:56you're an econ manager, you've got 100 other things to worry
- 40:00about, you know, trading day-to-day, all of that stuff
- 40:03that you. You may not have the time about
- 40:06to be able to do it, but as, as your experimentation and
- 40:09expertise and team grows and you start to bring dedicated roles
- 40:13into that, then it's just putting ownership on those
- 40:15people. So as a, as an external
- 40:18resource, every time I'm on client calls, every time we're
- 40:22doing, you know, QBRS or anything like that, I've got the
- 40:25dashboard up, I've got the, I've got the repository up.
- 40:27So I'm constantly reminding them, this is here, this is
- 40:29here, this is here. And I think over time, where the
- 40:33resource becomes available, people should start to use it.
- 40:36I've definitely been in the boat where I've created all of this
- 40:39stuff and it's just, you know, sitting on a, on a shelf
- 40:41gathering dust. But I, I just, it's something
- 40:47I'm constantly referring back to whenever one sort of
- 40:51communicating any test result with a client, you know,
- 40:54individual test level, I'm always saying, right, you know,
- 40:57we've tagged this up with this, that and the other.
- 40:59This is where it sits in the, in the library.
- 41:00This is the learning section. Yeah, I'm just trying to, it's
- 41:05kind of repetition, I suppose, and just drumming it into people
- 41:08to, to come back to it because it should be a really useful
- 41:10resource outside of just the the optimization and econ team,
- 41:14right? It should be, it could be
- 41:15valuable for marketing teams when they're starting their new
- 41:18campaigns. Like if we did something on APDP
- 41:21that people resonated with, well, why not run that as an ad
- 41:24or something, you know? So it should, it should be
- 41:26shared and it should be collaborative, but often times
- 41:30it does kind of get stuck with that one person.
- 41:31So it's. Yeah, it's it's a really good
- 41:33point actually. They they get sanitizer to wider
- 41:36team, especially when you're doing like copy tests and
- 41:38totally have like anything that's change your tone, change
- 41:40your copy style or visual. Yeah.
- 41:44So Anya, one, one thing I really wanted to talk about as well was
- 41:47that, you know, you've obviously you've you've built the tool,
- 41:51you've integrated Shopify. And one of the great advantages
- 41:53in the SAS world is, is third party tech being pre integrated
- 41:57at a data level so that you don't have these convoluted
- 42:00technical integrations. But the, the converse that where
- 42:04a lot of people get nervous as store owners, especially the
- 42:07technical store owners, is the impact on performance, the
- 42:11impact on theme performance, the impact on page speed and how you
- 42:15can have trade-offs. So how how do you how has
- 42:19visually been architecture technically so that it doesn't
- 42:22have like a big impact on page load through onerous scripts?
- 42:26Yeah, great question. So having been brand side at the
- 42:32beginning of my career, most of my career, I used to suffer from
- 42:36that tremendously. And one thing that we did
- 42:42basically before we even started building visually as an
- 42:47experimentation platform, we actually solved the part of the
- 42:51architecture to make sure that we are loading as quickly as
- 42:56possible. And we did that in several ways.
- 42:581 is so basically a lot of the even very large experience or
- 43:06experimentation platforms, what they do is basically that they
- 43:10serve all the different tests or personalizations that are
- 43:14configured to all of the users upon loading.
- 43:20What we do officially is that we only serve the right experience
- 43:24at the right time to the right user and therefore we save a lot
- 43:30of of loading. So that's that's kind of one
- 43:34thing. The other thing that everything
- 43:36you build on visually on front end is served with React, which
- 43:40is currently the fastest way to load on front end.
- 43:45And I think at the end of the day, the proof is in the
- 43:51pudding, right? So, and that's a little bit on
- 43:54the commercial side. So back in the day when we used
- 43:58to integrate to a platform, we had to commit to a year or even
- 44:022 years, right? We had to actually devote around
- 44:083 months of development resources in order to integrate,
- 44:12right? And then only in a few months we
- 44:15were able to see if everything that we were told was actually
- 44:18true, right? Nowadays with visually, we don't
- 44:21have any annual commitments. We don't even have monthly
- 44:24commitments. So basically you can integrate,
- 44:27you have a free trial and then you can actually see how it
- 44:32works and if it impacts the site speed.
- 44:36And if it does, then by all means please remove us,
- 44:40uninstall and you will never hear about it again.
- 44:44But if it works, it can continue working.
- 44:46So honestly, as someone who was like has been on the brand side,
- 44:49I know that whatever I'll say won't mean as much as, you know,
- 44:55what they can actually test and see for themselves.
- 44:59Yeah. And it, I mean it's, it is one
- 45:00of the clear advantages of of preconnected ecosystems is that
- 45:05you can, you can pilot and run something very quickly at low
- 45:09overhead and validate whether or not it is right.
- 45:13The danger with that is people's install loads and loads and
- 45:16loads of things all at the same time and then don't know which
- 45:20ones having a negative impact on performance.
- 45:22So it's always like urge of caution to test, test, testing
- 45:26properly. Yeah.
- 45:28The the, the other thing that I, I would mention is that at the
- 45:32end of the day consolidation is a big part of increasing speed
- 45:37and visually definitely helps with that.
- 45:40So basically would visually consolidate AB casting,
- 45:43merchandising, personalization all in one and you know post
- 45:47purchase upsells. So, so that also kind of
- 45:50contributes to a more hygienic, you know, way of running your
- 45:54Shopify brand. Yeah.
- 45:56And I know a few, a few businesses I'm working with who
- 45:58are really keen to exploit the post purchase piece much
- 46:03smarter. You'll probably find that quite
- 46:04interesting. Do you have interest, do you
- 46:07have any like examples of some really smart post purchase tests
- 46:12and that have been run that have had positive impacts?
- 46:17Great, Great question. So one thing that we've seen, I
- 46:22don't know if it's very smart, but it's very straightforward is
- 46:25for men's apparel brands, we've seen offering kind of the same,
- 46:32you know, T-shirt or or trousers in a different color, right,
- 46:38with a slight discount really does the trick.
- 46:43One other thing we've seen working really well is kind of a
- 46:48surprise me upsell. And that actually goes a long
- 46:52way to, you know, clearing out stocks that you might actually
- 46:58want to get rid of in kind of a bulk way where you kind of
- 47:04create a bundle and you brand it in a interesting way.
- 47:10I was actually very surprised at that.
- 47:13Who caught, but that was an interesting one.
- 47:18Yeah, a lot of a lot of things that really surprised me is the
- 47:21fact that you know the same product in different variation
- 47:25work really well as as an additional upsell.
- 47:28I know. I think this is the beauty of
- 47:30testing is, is there are we all have preconceptions, we all have
- 47:35conscious and unconscious biases, and we all have
- 47:37learnings from from working with lots of different businesses,
- 47:41but learners have to be challenged And sometimes what
- 47:44works for three or four brands doesn't work the same for
- 47:48another. And and if we make those
- 47:50assumptions, it's dangerous far. So I really like this.
- 47:52I mean, I, I love it when something I know works well, I
- 47:55get proven doesn't on a site because it's like great, because
- 47:57then that site's doing something better than it would otherwise
- 48:00have done. So let's let's put our our
- 48:03crystal ball piece because I'm, I'm intrigued.
- 48:05Yeah. You guys obviously have lots of
- 48:08of knowledge in this area, like where you're seeing this
- 48:10evolving. So Nick, I'll start with you and
- 48:12then we'll come back to you, Ann.
- 48:12Yeah, like where are you seeing optimization, I'm going to call
- 48:18it rather than CRO. Where are you seeing this headed
- 48:22from client, how clients are approaching it and where are you
- 48:25seeing the role of AI coming into this and how that will
- 48:28develop in the future? Yeah, I think the on the AI
- 48:32piece, I think there's at the moment there's a couple of
- 48:35different areas where we're using it.
- 48:37So we're using it to help us with our customer feedback.
- 48:43And you know, once you've gathered a load of surveys or
- 48:47you're getting one of mine reviews or anything like that,
- 48:50you know, using AI for that to just summarise and put things
- 48:53into themes for you. It's saving loads of man hours
- 48:57sort of diving into into different reports.
- 48:59So I think from a research point of view, it's helping to
- 49:01highlight some of the key areas that we should dig into.
- 49:05And then on the testing side, I think Annie, Annie will come
- 49:11onto something in a minute. So I won't, I won't go on too
- 49:14long about it, but I do think that there's going to be, I'm
- 49:17starting to, I'm starting to use AI to like build more complex
- 49:21tests myself. So rather than me needing to
- 49:25rely on a team of developers to build a potentially complex
- 49:29feature on a site, I've built tests using AI.
- 49:33And I think that that's because because we know it's an MVP,
- 49:38it's only going to be live for a couple of weeks.
- 49:39It doesn't need to be, you know, perfectly robust code.
- 49:42It can just live temporarily and work for me and busting tool.
- 49:46So from a technical standpoint, I do see it kind of helping
- 49:52maybe users who don't have additional development
- 49:55resources, helping them to be able to do more with the tools
- 49:58that they've got available. And I think in terms of
- 50:03optimization in general and where it's going, I, I think
- 50:06people are at such different stages.
- 50:09And I'm, I'm always surprised when I speak to, you know, I
- 50:12think, and you mentioned it at the beginning at podcasts, like
- 50:15some people who have these massive brands who you would
- 50:19just assume have got these huge teams behind them running all of
- 50:21this stuff. Like they really don't.
- 50:24And I think there's still this level of everybody kind of
- 50:28getting on board with testing 1st and it becoming part of the
- 50:32ecom slash marketing team standard set of people that you
- 50:37hire, you know, everybody knows you, you're going to have a paid
- 50:39person, an SEO person, a trading person, etcetera.
- 50:43But the optimization thing still feels auxiliary.
- 50:47It feels like something that you might do dip in and out of and
- 50:50it's not, I don't think it's especially in e-commerce.
- 50:54I don't think it's a standard practice across these teams.
- 50:58So I'm hoping that that will continue to to build traction
- 51:02and develop. And then I think for those who
- 51:05are already doing it and are, you know, ahead of the game, I
- 51:08think that they'll just start to explore more personalisation
- 51:15because of AI making that kind of available to us.
- 51:18I think understanding how users are interacting in real time is
- 51:23going to change as well. So being so at the moment, you
- 51:26know, and you touched on it earlier, we know if somebody's
- 51:28shown an affinity to to women's where we can start to tailor
- 51:31content in that way that requires input at the moment
- 51:34from a human. I think that there's potential
- 51:37to for tools to build in something that's predictive.
- 51:41So I understand how my user journey is working and start to
- 51:45build and tailor my user experience around that, but test
- 51:49it. So I think there's, yeah,
- 51:51there's a lot, a few different things, but I, I think it's kind
- 51:54of carrying on as we are, but introducing AI to, to, to make
- 51:58things a little bit smarter and easier to manage.
- 52:01Yeah, that makes a lot of sense. And Anya, from the, the
- 52:04technologies, the vendor perspective, where, where do you
- 52:08see this headed from a a platform capability point of
- 52:11view or an integration board where yeah, what, what can
- 52:15people expect coming in terms of features?
- 52:18Yeah. I think from kind of an
- 52:21overarching perspective, I think AI will be kind of the one that
- 52:25is executing meaning things like, you know, building the the
- 52:30variants, analyzing right building segments as as Nika
- 52:36mentioned, that will require much less human involvement than
- 52:43than before. And that will actually leave
- 52:47much more time and room for humans to own the strategy to
- 52:52get more and more creative to understand not only kind of what
- 52:57is happening, but the why that Nick is always kind of trying to
- 53:02to get to, but it takes time, right?
- 53:05And also to prioritize from a strategic and a business
- 53:10standpoint. At the end of the day,
- 53:12everything we're we're doing in experimentation is about growing
- 53:16the business, is about understanding the customers and
- 53:20what leads us is basically the business priorities, right?
- 53:24If the business is telling us, OK, we need to decrease return
- 53:30or we need to increase sales, but in this new category, right,
- 53:34that's what we optimize and that's what we try to grow.
- 53:38So I think that it's, it's, it's an interesting time and and as I
- 53:48said earlier, I think personalization will become more
- 53:51and more granular as our ability to execute in a larger scale
- 53:57will evolve. And we are very, very excited to
- 54:03be part of of this movement and launching our first identic
- 54:11layer in just a few weeks, specifically around building.
- 54:18A lot of the times what we've seen is that, you know, the
- 54:22issue is, is not around what to test.
- 54:26People have lots of ideas, but how to test, right do not get
- 54:30stuck in this these development cycles.
- 54:34And it's it's really amazing to see how how that technology
- 54:40works. Yeah, the agentic side I'm
- 54:43generally interested in because I think there's a lot of hype
- 54:46around it. But we on the hype.
- 54:47There are some really practical use cases where enabling people
- 54:51to interact with tools to query, question and build like prompts
- 54:55that help them. Dr. automated actions is going
- 54:58to make life a lot simpler for people to execute.
- 55:03I could, you know, I could carry on talking and on this topic for
- 55:06how I love it. I find it fascinating and you
- 55:09both have a wealth of knowledge, but I'm not going to use up your
- 55:11entire day. But I guess if people want to
- 55:14learn more about what each of you do, there will be contact
- 55:17details on the landing page on our website if anyone listening.
- 55:20But let's argue Anya, what's the best way for someone to reach
- 55:23out if if they're on the Shopify, their interest is
- 55:25piqued and thinking, Oh, I want to, I want to probe more on how
- 55:28visually would work for us. Yeah, so they can reach out on
- 55:34our our site and of course I'm also on lake.
- 55:38Then they want to see more on how visually works.
- 55:42We have a lot of content there, love to to show them around.
- 55:48Fantastic. And Nick, what's the best way to
- 55:51reach you? I know people can go to the
- 55:52website in general, but if they wanted to drill into CRO and and
- 55:57yeah specific challenges for their business.
- 56:00Yeah, same as Anya, if you, you can catch me on LinkedIn, I'm
- 56:04trying to to post a little bit more frequently and try and add
- 56:08some value. So, so feel free to to follow me
- 56:10there. Definitely follow Anya
- 56:12everybody. She posts some real thing stuff.
- 56:14All right, or e-mail me nick@bavant.com and let's have a
- 56:20chat. Fantastic look both, thanks so
- 56:22much. Really enjoyed this and I look
- 56:26forward to, to, to seeing the story evolved with both visually
- 56:29and obviously the, the CRO function that you've built out,
- 56:31Nick. Cheers James, been a pleasure.
- 56:34Thank you so much. Thanks for having me.
- 56:38You're listening to the Inside Commerce podcast, hosted by
- 56:40experienced consultants James Gert and Paul Rogers.
- 56:43Inside Commerce is your fast track to digital knowledge,
- 56:46featuring interviews with leading e-commerce retailers,
- 56:49industry thought leads and the C-Suite, key tech companies to
- 56:51explore the latest digital strategy, customer experience
- 56:54and tech. Discover what the latest trends
- 56:56really mean and get practical advice based on the experience
- 56:59to help fine tune your e-commerce strategy and tech
- 57:02stack. Follow us on LinkedIn and reach
- 57:04out to James and Paul if you'd like to discuss an e-commerce
- 57:07project. We hope you enjoy the latest
- 57:09episode.