Latest / Inside Commerce: Ecommerce Strategy, CX and Technology Podcast / EP269: Delivering One Of The Industry's Biggest SaaS Migrations With More Than 1 Million Products, with World Books Product Director
Transcript
- 0:02You're listening to the Insidecommerce podcast, hosted
- 0:05by experienced consultants James Gerd and Paul Rogers.
- 0:08Insidecommerce is your fast track to digital knowledge
- 0:10feature in interviews with leading e-commerce retailers,
- 0:13industry thought leads and the C-Suite of key tech companies to
- 0:16explore the latest in digital strategy, customer experience
- 0:18and tech. Discover what the latest trends
- 0:21really mean and get practical advice based on experience to
- 0:24help fine tune your ECOM strategy and tech stack.
- 0:27Follow us on LinkedIn and reach out to James and Paul if you'd
- 0:30like to discuss an e-commerce project.
- 0:32We hope you enjoy the latest episode.
- 0:35This episode is sponsored by Centre and Greben.
- 0:37More about them later. Another cracking episode for you
- 0:40all. Today we are talking to the
- 0:41product director at World of Books who've recently handled a
- 0:45a large scale migration, migration transformation.
- 0:48Pick the words as you like, but for those of you who don't know
- 0:52World Books, they are AP back global retailer of used books.
- 0:56I think it's over a couple of $100 million global sales.
- 1:01So lead, lead really large scale DC and they've got an extensive
- 1:05product catalogue you can expect from a a leaving retailer of
- 1:07books. And we'll get into the size of
- 1:09that later. And the reason why we always
- 1:11have a chat today with our guests, David McGee is that
- 1:15they've recently migrated across to Shopify, moving to SAS.
- 1:20So moving to SAS enterprise. This is an area that I think is
- 1:23growing the number of global enterprise brands that are
- 1:25taking SAS seriously. But doing this with such a
- 1:28massive catalogue of skews, it's very challenging.
- 1:32So before we get into the questions, warm welcome, David.
- 1:35Thanks for coming on the podcast.
- 1:37Yeah. Thanks, James and Paul.
- 1:38Great to be here. Good to good to talk.
- 1:41Before we start getting into the details of the project, do you
- 1:44want to give people a flavour for who you are, what you do?
- 1:49Yeah, we'd, we'd love to. So yeah, David McGee, the
- 1:52product director for World of Books.
- 1:54In terms of the business itself, the the clue's in the name
- 1:56really we're in the books business.
- 1:59In terms of our goal and our aspiration as a business, New
- 2:02Way need to be the world's largest and most sustainable
- 2:05dedicated online bookstore. What we're super excited about
- 2:09and in my role in particular is working out well.
- 2:11How do we use technology? How can you, I want you to find
- 2:14ways to allow our customers to have really great experiences
- 2:18when they read more and waste less to give you a little bit of
- 2:22sense of scale. Maybe just to build on a couple
- 2:24of the points that you made a moment ago, James, and both from
- 2:28an organization and platform perspective, we, you know, three
- 2:31sites in the UK, one in the US and one in Hungary.
- 2:34And then in terms of the product itself, we've got, you know,
- 2:37more than 10 million books in physical inventory.
- 2:40Most of those are individual SKU's and we have some quite
- 2:44complex ways of working in our technology around for example
- 2:48dynamic re pricing. So we do re pricing it would be
- 2:515 million products per day. So it's a pretty complex, fast
- 2:57moving e-commerce business, but an exciting, exciting place to
- 3:00be. Perfect.
- 3:02I think that's a really good intro and that takes us nicely
- 3:05on to the next. The first question.
- 3:07So obviously everything you talked about that are things
- 3:10that you maybe wouldn't naturally associate with a shop
- 3:135 based tech stack and historically you worked with I
- 3:17guess, yeah, a broad, a broader stack, different approach.
- 3:21So what led to you move into shop buy?
- 3:23What were the main challenges or like what were the main things
- 3:26that motor better? Do some needs to chop a fire or
- 3:28SAS platform with James Sir? Yeah, great.
- 3:31Quick question. I'll answer it in two parts.
- 3:33So the first part is kind of the strategic context for World of
- 3:37Books and the second part of the quite specific challenges that
- 3:40we were facing. So World of Books originated and
- 3:43as a business that sold on marketplaces.
- 3:46So while the books were selling on eBay, Amazon, Abe, Libris,
- 3:50all sorts of different third party marketplaces.
- 3:53And that's how the business really grew originally in the
- 3:55UK, but over the last number of years we've been pushing really
- 3:58hard as the director Consumer Space launched our own website
- 4:01site and started to try and take more ownership of the full end
- 4:05to end customer experience. And that really helps us from a
- 4:08brand building perspective and also reduces some of the
- 4:11commercial risk around marketplace fees, for example,
- 4:14if they if they move significantly, that had a really
- 4:17big and painful impact on our business.
- 4:19So from a from a strategic perspective, moving into direct
- 4:23to consumer is the direction we really want to go and that's
- 4:26where we're growing. We really had two main
- 4:29challenges with our legacy tech stack.
- 4:31So firstly, we had a really rapidly rising total cost of
- 4:35ownership. We were coming to the end of our
- 4:37contract with the legacy provider and we were quoted I
- 4:41think maybe a 3X cost increase in our fee and that just caught
- 4:46that kind of gave rise to the the conversation.
- 4:49Should we think about doing something else now in fairness
- 4:52to our legacy provider there, they did tell us that we were
- 4:55making I think 250 million to 1/4 of a billion API calls per
- 4:59week. So we were actually costing them
- 5:01money to support though there's some real complexity there.
- 5:04And I think you know that that that was definitely one of the
- 5:07things that kicked this this off.
- 5:09Why that was important for us is that part of our value
- 5:12proposition is that we support kind of value seeking book
- 5:15lovers and a bar cost face dramatically increases.
- 5:18It makes it a lot harder for us to support that part of the
- 5:20proposition. So that was Part 1, total cost
- 5:23of ownership. And then Part 2 really is we
- 5:26just had an inability to develop the product or to make for more
- 5:30engaging customer experiences. Most of our team was tied up in
- 5:35supporting that legacy tech stack, fixing legacy
- 5:38integrations. And we really haven't done
- 5:41anything over the last maybe two years to innovate the product,
- 5:44but she'd do anything for the customer and that might make the
- 5:48book buying experience better. And so, yeah, those are those
- 5:52are the two primary things. Obviously, moving to shop fight
- 5:55kind of gives us some limitations.
- 5:57At the same time, you know, if, if you if you're on a different
- 6:00tech stack, you might have the ability to customise anything.
- 6:03And but at the same time, what we're really excited for is
- 6:06moving to Shopify with very quick access to a big ecosystem.
- 6:11Though a lot of their apps, for example, in the Shopify store,
- 6:15whether that's for loyalty or subscription or maybe integrated
- 6:19AB testing or maybe launching a mobile app.
- 6:22Those are things that we could quite quickly get off the ground
- 6:25with Shopify and understand how they did or didn't work for our
- 6:28customer and for the business. Whereas on the legacy checks
- 6:30that that kind of innovation was months or years worth
- 6:33development. So yeah, to summarize that
- 6:37reduced reducing total cost of ownership and enabling product
- 6:41innovation for the two prime May drivers.
- 6:43Greben create digital flagship stores for iconic brands of
- 6:46today and tomorrow. We are the number one Sentra
- 6:50agency in the world with more than 50 Sentra implementations
- 6:53to date. We are also e-commerce design
- 6:56specialists, platform and tech partner agnostic.
- 6:59Try our new e-commerce design subscription.
- 7:01Visit greben.com. That makes sense.
- 7:04And you mentioned or like so far we've talked about catalogue,
- 7:08the dynamic pricing side and there's obviously a number of
- 7:11other areas that are complex around business.
- 7:14But the other thing that's not unique about your business that
- 7:17I was surprised about when we first spoke was that you were
- 7:20using a single instance of Shopify and markets.
- 7:24And as part of that you've you've had access to some betas
- 7:27and everything else. But there's amongst those
- 7:30things, we'll come on to that a bit more later.
- 7:32But amongst those things, there's, you know, a lot of
- 7:34things that in theory could and should challenge Shopify.
- 7:38So can you talk us through kind of the process and kind of how
- 7:41you went through migration and how you handled some of those
- 7:44compromise areas and kind of a hurdles you faced?
- 7:50For sure. Let me talk through the process
- 7:51that we kind of went through from a migration standpoint and
- 7:55then stop me at any point if there's somewhere you want to go
- 7:57a bit deeper. But we we just started really
- 7:59with discovery. So we spent quite a lot of time
- 8:02talking to our customers about what they like about our current
- 8:05site and platform and what they didn't like and what could be
- 8:07improved. Did a bunch of tear downs of
- 8:10competitor sites and competitor user experience and just built
- 8:15some kind of best practice and some mock ups and designs for
- 8:18what we felt we could do potentially with Shopify to help
- 8:22us go on and, and, and build on that, on that solution.
- 8:26We looked at a variety of different platforms and just
- 8:29kind of tried to trade off cost, you know that which is a driver
- 8:34for us ecosystem and support and scalability and Shopify just
- 8:40came out as the candidate that we thought was the the most
- 8:42likely to work for us. The Shopify team in fairness
- 8:45them are really great. They engaged with us really,
- 8:48really well. They set us up throughout their
- 8:50organization, gave us a bunch of different points of contact,
- 8:53then really invested time in trying to understand us.
- 8:57The markets piece was a challenge as well.
- 9:00We're pretty early to market for a for a business of our size and
- 9:04scale. But one of you know if you think
- 9:07back to some of the challenges we're trying to avoid the total
- 9:10cost of ownership piece. If we'd gone with the expansion
- 9:12stores and the kind of customisation and localisation
- 9:16and then the support and maintenance of those stores was
- 9:19something we thought that try and avoid if we can.
- 9:23I also think Shopify were pretty keen to get us on markets
- 9:26because it's a great case study for them.
- 9:28So even if there were some limitations, maybe we didn't
- 9:32even realise those at the time until we actually pushed further
- 9:35into into working on the market they but from a from a migration
- 9:39phasing perspective, then we went through discovery.
- 9:41We don't kind of build out the front end though, just designing
- 9:45and building the site on Shopify.
- 9:47It'll become a bit more clear when we talk about it later.
- 9:50But one of the key things that we did there was we, we built
- 9:53the site, we designed the site in such a way that we tried to
- 9:56make it almost identical to our legacy site from a customer
- 9:59experience perspective. So that was really frustrating
- 10:03in one regard because we want to do something with better and
- 10:05more exciting and more interesting for our customers.
- 10:08But when it came to AB testing, we wanted to have the two as
- 10:11close as possible in terms of design.
- 10:14So we replicated the site in Shopify and the only major
- 10:18change was around performance. You know, if you from a core web
- 10:21vitals perspective, it was actually better, but we were
- 10:24aiming for going to design look and feel parity from a customer
- 10:28perspective on on the legacy stat.
- 10:30And then probably the largest piece of work overall was
- 10:33actually back end work. So we built custom middleware in
- 10:36house and if you think about the middle Ware, it's kind of the
- 10:39pipe work that takes all the data from our data warehouse and
- 10:42transforms that and prepares that in a way that's suitable
- 10:45for the Shopify and Shopify appropriate.
- 10:49It's all the work around product creation, product update, stock
- 10:52inventory and price updates that all flows through our middle
- 10:56work in the millions and millions of events per day.
- 11:00And that really got us to the point where we had a HEC stack
- 11:03with the front end built, we had a back end through our
- 11:06middleware. And then we moved into the, the
- 11:08piece around AB testing and then into kind of actually migrating
- 11:12the, the stores to, you know, redirects and the Google
- 11:16shopping feed as well. So that's kind of, I guess, the
- 11:18overall outline. Centre is the only commerce
- 11:21platform built exclusively for fashion brands.
- 11:23Over 300 fashion brands run on Centre including icons like Paul
- 11:26Smith, Jeff Ave. and Nudie Jeans.
- 11:28Centre solves business logic challenges for fashion brands
- 11:30are no other platform and they are on a mission to help you
- 11:32scale up rapidly globally across all online and offline channels.
- 11:36Take your fashion brand global faster, further stronger today
- 11:40question on this the the middleware that because that's
- 11:42really interesting because I think a lot of time people go
- 11:45down quite standardised routes. The middleware patchworks is
- 11:49very strong in the Shopify ecosystem.
- 11:51Then there are the enterprise stacks out there.
- 11:54But often people underestimate the cost when you start to get
- 11:58the sheer volume of events like you guys got.
- 12:01Was was cost based on the volume, the driver for customer
- 12:04or was it more about control? Yeah, this is actually an
- 12:09interesting part and this is this was one of our key failings
- 12:12and learnings throughout the migration, which was that we
- 12:16started with a third party I pass provider that that's how
- 12:20the the the project started. And we spent probably six months
- 12:24working with that provider on building an integration from our
- 12:29data warehouse through there. I passed into Shopify.
- 12:34And I mean, we got to the point where we couldn't get confidence
- 12:37in the data that was appearing in Shopify.
- 12:40So partly due to some challenges and the the data quality at our
- 12:44end and and also just the sheer volume of changes that we're
- 12:48going through. So we actually ended up
- 12:51abandoning that after six months.
- 12:53That cost us a lot of time and money and we made the decision
- 12:57that we would build our own middleware in house.
- 13:00In hindsight, it kind of, it's so obvious that we should have
- 13:02done that from the get go. We've got a big, you know, we've
- 13:05got a tech team of 80 people, front end, back end, full stack.
- 13:09We understand our own systems and our own data better than
- 13:12anybody else. We thought we would be able to
- 13:15do it faster and not distract our team by going to a third
- 13:20party. And that didn't work out for us
- 13:23though that, that, that if I was going to point to anything in
- 13:25the, in the project that was a failing, it was our initial
- 13:29choice and our, us going down the route for the third party.
- 13:32I pass. And, but then the, the major
- 13:34success was what we actually built in terms of middleware is
- 13:37super capable, very high observability.
- 13:40So we know what's going on throughout all stages of those
- 13:43transformations and, and, and really reliable.
- 13:46So it's been a massive improvement as an end result,
- 13:49but we probably wasted six, six months and easily 6 figures in
- 13:53cost originally. You know what, I think this is
- 13:57really interesting. I love the transparency on this
- 14:00because we, you know, between, that's why we add across so many
- 14:02different projects and nothing goes 100% smoothly.
- 14:06And sometimes you see a decision in hindsight as why did we do
- 14:10that? But actually in the time you're
- 14:11doing it for the right reasons. I'd love to know how did you
- 14:15get, how did you get the buy in to, to like pivot?
- 14:19Because that's a really hard, not just financial decision, but
- 14:22an emotional decision in a project because you've vested so
- 14:25much time that the the impact that can have on the people's
- 14:28motivations is so how do you get, how did you get the buy
- 14:31insurance? So people go, right, fine, let's
- 14:33just burn it and go. Yeah, it's, it's a great
- 14:36question. It was really hard and also very
- 14:39hard in every sense. Hard for the product team, the
- 14:43project teams, hard for business stakeholders, hard for the CFO.
- 14:46He wasn't very happy. But yeah, we got to the point
- 14:50really where we had a couple of things from a commercial
- 14:53perspective that were looming. So we had a deadline for getting
- 14:56off our old legacy tech stack and significant cost
- 14:59implications if we couldn't get off that on time.
- 15:01And and we also had the opportunity cost, which was that
- 15:04we believed Shopify would convert much better for us and
- 15:08deliver better customer experience.
- 15:10So when you kind of combine both of those things that the
- 15:13potential cost that was looming on the horizon if we didn't go
- 15:16live was enormous. So we gave it a couple of extra
- 15:20months. We had a, we had a, a go live
- 15:22deadline and we that past, we weren't confident in what we
- 15:27were seeing from a product data perspective.
- 15:30We gave it a couple of months of trying, fixing, collaborating.
- 15:35Can we make this work? Can we get this data to a point
- 15:38where we're really confident in it?
- 15:40And, and, and transparently part of the challenge with that was
- 15:42it because the product, the positive data volumes are so
- 15:44high, it could take you days and days and days to upload the full
- 15:48catalogue. And only then do you really
- 15:51discover that you've got problems.
- 15:52She then spent time fixing them, clearing the catalogue, re
- 15:55uploading the catalogue. So part of the nature of the
- 15:58challenge was that the system that had been in screen just
- 16:01wasn't fast enough to be able to fix problems at our scale
- 16:06quickly. But to come back to your
- 16:08question, I, I, I spoke to a bunch of people in the
- 16:10organization, my peer and engineering show, our
- 16:13engineering director, we were aligned and we just got genius
- 16:16stakeholders together and said this is not going to work.
- 16:19We don't have confidence in this.
- 16:21I'm not as product director, I'm not going to put this in front
- 16:23of our customers because I think it'll be a terrible experience.
- 16:26I don't think it'll deliver business results.
- 16:29I think we can build the middle in house.
- 16:32This is how we think we do it. And yeah, we kind of got the the
- 16:36go. So that was super painful, a
- 16:38really hard period and painful decision.
- 16:41And then, you know, we just stood up.
- 16:43We pulled in people from different product teams with the
- 16:45right skills, took people off other platforms, other projects
- 16:50they're working on, built kind of the elite team within World
- 16:54of Books from a tech and development standpoint.
- 16:57And people worked 78 hour weeks for months to build it and to
- 17:02and to get it live. So yeah, a very, very
- 17:06challenging time within the organization.
- 17:08But the end products in terms of what they delivered it pretty
- 17:11phenomenal. Right.
- 17:14Yeah, it's really interesting. I'm sure James and I can pick
- 17:16your brain on that for a long time, but maybe we'll do a
- 17:19follow up episode of Substitute one day and the next question.
- 17:23So I think the other one I found really interesting about your
- 17:25project and something that comes up a lot and that is notoriously
- 17:30difficult to do for one and is often contentious is the idea of
- 17:35testing one site versus another site.
- 17:37And you achieved this by cloud flare.
- 17:41So can you kind of talk us through how you approach that
- 17:43from a technical standpoint and then also from like a business
- 17:45standpoint in terms of yeah, process to determine the winner
- 17:49etcetera? Totally, Yeah.
- 17:51So from a, from a let me add to start with the business
- 17:54standpoint because that that helps in informant.
- 17:57But what we were really nervous about was that we would build,
- 18:01we would design, build and launch a new site and then it
- 18:04wouldn't actually convert as well as or almost in some cases
- 18:10at all. So at our scale where we've got
- 18:12hundreds of thousands of sessions per day and fluctuation
- 18:17and conversion is really meaningful for us.
- 18:19So we really wanted to find a way to understand how does this
- 18:22site perform? How does this full text stack
- 18:25perform in comparison to the legacy site in part to validate
- 18:29the Shopify proposition? Maybe We always thought that
- 18:32through Shopify with a great checkout experience and shop
- 18:35pay, you know, we would see a conversion increase that was
- 18:38kind of what we'd built into the original business case based on
- 18:41some assumptions. But we actually were really
- 18:43nervous about just turning it on and finding that we're actually
- 18:46losing money everyday in comparison to the the legacy
- 18:50site. So that was the impetus for us.
- 18:53We were helped and from a technical standpoint in that we
- 18:56were changing domains. So we were trading originally on
- 19:00our legacy tech stack on Web.com and we built the new site on
- 19:03worldofbooks.com. So that was I think one of the
- 19:07enablers for us that meant that we could use Cloudflare.
- 19:10We used Cloudflare workers, the split traffic 5050 between the
- 19:15two sites. We then, you know, cookie those
- 19:18visitors to try and make sure that they stayed on the site
- 19:21that they visited. And kind of coming back to the
- 19:23design in the front end piece, we wanted it.
- 19:25We, we wanted it to be as imperceptible as possible to
- 19:28customers because we knew that if customers were landing on a
- 19:31new site, a different site, that was very likely to change their
- 19:34purchase behaviour. So we thought, let's build a
- 19:36site that's as close as possible to the legacy site.
- 19:40But 1050% of traffic, let's put a whole bunch of measurement in
- 19:45in place. So we were able to measure all
- 19:46the way from not just sessions and page views, but you know,
- 19:50search usage, click through rates and add to carts, check
- 19:54out completion conversion rate and so on and just really
- 19:57understand it kind of top to bottom.
- 20:00Within the e-commerce funnel, how the legacy tech stack is
- 20:03performing in comparison to the the new shop I build.
- 20:07And then if there were issues, we'd hopefully know where to
- 20:10focus. If we had problems in, if starch
- 20:12wasn't working or if we have problems around any particular
- 20:17aspect to the funnel, we we could make changes, rerun the AB
- 20:21test and hopefully see that those had been improved.
- 20:24But ultimately what we were aiming for was to give our
- 20:28organization confidence that we could migrate all of our traffic
- 20:31to the new platform and it wasn't going to hurt us
- 20:34financially. Yeah, I love that it's, it's
- 20:39actually quite a rarity now for people to, A, be able to, but B,
- 20:44make sure they take the time to do that level of validation
- 20:47because of time pressures. And often those time pressures
- 20:50are linked into contractual terminations.
- 20:54Yeah, yeah, really interesting to hear that.
- 20:57I'm sure there's probably a lot of ECOM folk out there.
- 21:00That's where that piques their interest instantly.
- 21:03What question I've got because I, Paul and I kind of roles kind
- 21:07of they complement each other, but they vary quite
- 21:09significantly in projects. And I do a lot more of kind of
- 21:13hands on like project leadership, project management
- 21:16stuff. And I'm always interested by
- 21:17project team structures because they vary massively.
- 21:20You talked about having a large internal engineering function
- 21:22across the stack, but how else did you structure?
- 21:26Was it fully run internally apart from like a dev agency?
- 21:31Did you, did you bring in any contract resources to run a
- 21:34project management office? Yeah.
- 21:36How did you structure that and who were the key project roles?
- 21:39Yeah, it's a great, great question.
- 21:41So the, the, the running of the the running of the project was
- 21:44World of Books internal. So it was split between me as
- 21:47the product director and my peer and engineering Jolt, who's our
- 21:51engineering director. So I, I guess around there the
- 21:54the product and delivery side and Jolt Rambeck kind of
- 21:57engineering and technical side. We did bring in a third party on
- 22:02state to do front end development for us, largely
- 22:05because not because we don't have front end developers
- 22:07ourselves, but because we don't know anything about Shopify.
- 22:10So there was an opportunity for us to watch and learn and pair
- 22:14with them as they built the site and always with the goal of off
- 22:17boarding them eventually and taking, you know, overall
- 22:20control of our own site. Being in charge of our own
- 22:23destiny was always the plan. But that was the that was the
- 22:26main contract resource that we used.
- 22:28But in terms of in terms of running the, the team itself, we
- 22:32had, you know, everything you would imagine, I suppose from
- 22:35project delivery perspective around, you know, a steering
- 22:38group that we would report to, you know, on a monthly basis in
- 22:41terms of overall progress and budget and financial controls
- 22:45and whether we were on track or not.
- 22:47And then we just ran pretty standard ceremonies around, you
- 22:51know, spring planning, grooming, refining, daily stand ups and
- 22:57just stayed as close as possible to the technical team.
- 23:01I think one of the other kind of big learnings that came out of
- 23:03that was, you know, particularly for me and Schulte were and this
- 23:07is comes back a little bit to the decision around the IPASS
- 23:09piece. We certainly learnt the hard way
- 23:12that as leaders in the organization, we need to be like
- 23:16directly and intimately involved in a lot of the detail of what's
- 23:19happening. And we weren't at the start and
- 23:21that was a mistake that was a miss on our part and we need to
- 23:24own that. But I think from a from a
- 23:26product perspective, yeah, just getting really, really close to
- 23:30essentially every decision that we're making and understanding
- 23:33why and what that might mean for the business and for the
- 23:36customer was super important for us.
- 23:40Right. And And I guess the big
- 23:42question, so you talked about having 10 million skis earlier
- 23:47and I'd imagine that, yeah, I mean, I have seen some stores
- 23:50exist with that level of cataloguing past in the shop
- 23:53ecosystem, but maybe not with your other complexities and the
- 23:57volume as well. So it's typically being like car
- 24:00parts retailers and things like that.
- 24:03But I guess how did you, it sounds like you've got a, you
- 24:06know, a back office system or something else that isn't a PIM
- 24:09that's housing a lot of that data.
- 24:12But how did you approach the catalogue migration And yeah,
- 24:15how did you find like the localization side as well?
- 24:18So I think you're multi language.
- 24:21Yeah, those two stuff. Yeah, Yeah, great question.
- 24:24So we we manage all of the data in as I said, kind of our own
- 24:28proprietary back end systems. So we have what we call the data
- 24:32warehouse, where we take product data from a bunch of different
- 24:36sources and merge that into a kind of a complete set that we
- 24:40then pass through the middle Ware into Shopify.
- 24:43So in some ways, it's relatively simple because what we focus on
- 24:49is making sure that the product data that we're ingesting into
- 24:52the data warehouse is correct, and then the middle Ware will
- 24:55just pass through whatever we ask it to to Shopify.
- 24:59The only real kind of technical constraints that we ran into
- 25:02there were just around API limits.
- 25:05So firstly, when you were picking the right right API,
- 25:10right endpoint with Shopify and then just making sure that we
- 25:13had a good and a clear schedule place for product uploads, all
- 25:18of the all of the localization that we're doing is done in the
- 25:22front end. So it's it's relatively limited
- 25:27in comparison to what we could do if we had a PIM.
- 25:30It was what a PIM was one of the considerations that we had at
- 25:33the start of the the project and we spent some time working with
- 25:38the guys at Akinio, really like those guys, really like the
- 25:42product. And I think you know the time
- 25:44that's the kind of platform that could really enable us from a
- 25:48growth and scale perspective. We just couldn't, we just
- 25:52couldn't afford from a time perspective to include
- 25:55integrating with a PIM in the scope of this project given the
- 25:58schedule we were working to. So what we've ended up with is a
- 26:02bit of a Ms. match is too harsh, but we're partly managing the
- 26:07product data in our data warehouse and then partly doing
- 26:10the kind of customization localization piece on, on the
- 26:15front end. So that's definitely an area
- 26:17where over the next, you know, six or nine months we want to
- 26:21try and understand how we can improve.
- 26:23It's relatively manual at the moment.
- 26:27As I warned you both before the start of this episode, I have
- 26:30run out of battery or nearly. So I'm just quickly migrating
- 26:33somewhere that's got a plug that one sub question on that.
- 26:36So I think Algolia plays a really important role within
- 26:39your stack. And one of the big limitations
- 26:42that I thought you'd be facing would be around category
- 26:45merchandising, which is something that isn't great with
- 26:49markets at the moment. So how have you approached that
- 26:52and how have you found Algolia with such a big catalogue?
- 26:56Yeah, Yeah. Great, great question, Paul.
- 26:58So Algolia is a fundamental part of how our site operates.
- 27:02So Algolia provides our on site search experience and and it
- 27:06also provides recommendations. So and collections, do we build
- 27:11those through Algolia? It does mean that we have an
- 27:15absolutely enormous record count with Algolia.
- 27:19When you think about all of the different indexes and all of the
- 27:21different markets and languages. I can't remember what the total
- 27:25figure was, but I think it's around 60, 5 million or
- 27:27something like that when you take all of the products and you
- 27:29apply them in in a localised way.
- 27:32But you're absolutely right. Algolia sits at the heart of at
- 27:35the heart of that. We've got a whole bunch more we
- 27:38can even do with Algolia. We're we're barely kept
- 27:41scratching the surface of that products capability,
- 27:44particularly from a search perspective.
- 27:47So we've, we've set the set the search up in such a way that it
- 27:51largely replicates what we had on Legacy, which where we used
- 27:54Algarlia as well. But there's a ton more we can do
- 27:57around kind of search relevancy testing and making sure that we,
- 28:01you know, customers really find the books they're looking for
- 28:04and through that fast catalogue that, yeah, Algarlia is a
- 28:07fundamental part of our stack at the moment.
- 28:11And on the international front, they touched on it earlier about
- 28:14the, you know, you, you operate globally, you've got English
- 28:17language sites, you've got foreign language sites.
- 28:20And that's always fun to get the right hierarchy structure and
- 28:25to, to get the right setup for and managing them independently
- 28:29and have an inheritance is where you need it.
- 28:31And you went, I mean, quite early to Marcus for such a big
- 28:34global store. So what I'd love to hear is
- 28:36nothing, nothing's ever perfect in any platform.
- 28:39What have been the compromises you've had to make around
- 28:41international by going down the Market Street?
- 28:45Yeah, a few things I think. Yeah, I think that dependence or
- 28:51that link between, you know, making sure you know exactly
- 28:56what store you're updating. You've made changes as sounds so
- 28:59trivial, but actually that can cause a few more headaches when
- 29:02you have a large team working on updating something on the on the
- 29:06platform. I think a few of the technical
- 29:08challenges that we found were things that we definitely didn't
- 29:10expect, which were, and you know Paul and I have talked about
- 29:14this off blind over the last couple of months, but we found
- 29:17that Shopify actually do sometimes have a challenge
- 29:19landing visitors in the right market.
- 29:23So based on their IP address, they're not always consistent
- 29:26and reliable with that. So we there's some unexpected
- 29:30behaviour we get, we get back on sometimes a second visit,
- 29:34sometimes a customer doesn't land on the market we'd expect.
- 29:37But at the moment, we find that around 2% of our customers, for
- 29:41example, in the US on not landing on the US storefront
- 29:45when they click through. So there is some something
- 29:47within markets on the Shopify redirect dive that maybe over
- 29:51time will be optimised. We also were caught out a little
- 29:54bit by the way Shopify sets up primary markets.
- 29:58So within Shopify markets you can have lots of different
- 30:01markets, but you have to allocate 1 as your primary.
- 30:04We didn't appreciate that that market only shifts to itself.
- 30:08So the way World of Books has worked historically is that the
- 30:11UK is our largest market and the UK ships essentially to the rest
- 30:15of the world aside from our other markets.
- 30:17So, you know, we have Germany, France, Australia, the US and
- 30:24Switzerland. The UK would ship everywhere
- 30:26else, but we couldn't do that on Shopify.
- 30:28So we ended up making the US our primary market, which only ships
- 30:32them to the US. So we definitely had, you know,
- 30:34quite a lot of rework to do effort just from a
- 30:37configurations perspective relative to what we were
- 30:39expecting and shipping rates and pricing.
- 30:42And, and then the third piece, which I, I think was, you know,
- 30:45really interesting for us, but also, you know, quite
- 30:47challenging as well was the, the multi currency and multi entity
- 30:50piece. As we have two, 2 entities, 2
- 30:54legal entities and we transact in four different currencies.
- 30:57Managing that through Shopify markets in a way that's
- 31:00efficient and reduces kind of foreign exchange it cost
- 31:04exposure. I think it has historically been
- 31:07quite hard for large enterprises.
- 31:10And yeah, that, that that's an area where we had to kind of
- 31:13work through in some detail Shopify because the the solution
- 31:16that was on offer until very recently was going to be super
- 31:19painful for us. Yeah, let's come back to a
- 31:22minute because I'm really intrigued by this because I've
- 31:24not I've not worked with. I've actually seen seen that
- 31:27it's there and been looking forward to it for a while, but
- 31:30not worked on a project over there.
- 31:32But that interesting you said about markets is that I'd I just
- 31:35I mean, Paul obviously has a a better technical handle of the
- 31:39platform in detail and ID. But that, that issue with the
- 31:42the base store on markets I came across on a project recently
- 31:46where it hadn't been understood and it led to a sudden like last
- 31:50minute. Well, we need an English
- 31:52language version for an international.
- 31:54We need an English language version.
- 31:56Otherwise the the German subdirectory is not going to be
- 31:59set up properly and then it creates created lots of issues
- 32:02with language files and having to re import.
- 32:04And it's amazing how you don't know and you don't know there's
- 32:09a problem until you know there's a problem sometimes.
- 32:12It's, it's, it's totally right. And one of the challenges I
- 32:14think for people thinking about migrating to to shop fire or any
- 32:18other platform is there, there's exactly as you say, you don't
- 32:21know what you don't know. And there are things that can
- 32:24become quite important parts of your business or your ways of
- 32:28working that you just assume will be in place on a new
- 32:31platform and they're not. So how you then adapt to that,
- 32:35either in terms of changing your working practices or changing
- 32:37the customer experience is, is quite challenging.
- 32:40And there's also just no way to get an answer to all of that.
- 32:43That's one of the things we really wrestled with in the
- 32:45migration. We kind of set out some core
- 32:48fundamental requirements. This is what we want the
- 32:51platform to do. But you always miss something it
- 32:54it touches so many parts of the business and that there's no,
- 32:58there's no way to get that absolutely perfectly right every
- 33:01time. And and we definitely got caught
- 33:02out by that for sure. But I think you know net the
- 33:06overall position is, is a really good one.
- 33:08We just we you do have to make some compromises along the way.
- 33:12Yeah. And I think that's a, it's a
- 33:13really pragmatic approach to have because when you're working
- 33:17at any scale, whether it's a a single channel business or a
- 33:20large global business like yourself, you're working with
- 33:23people in technology and we're all flawed.
- 33:26And at some point you'll find something in the project that
- 33:29that is a bump that you weren't expecting.
- 33:30And the the projects that succeed are the ones that are
- 33:33proactive in handling it and don't turn it into a blame
- 33:36culture and go right, OK, that didn't go right.
- 33:39How do we fix it? So it's really nice to hear that
- 33:41because I think it gives other people a bit of reassurance that
- 33:44when when they have problems, do you know what it's not because
- 33:47you cocked up. This is nature of a complex
- 33:49project. Yeah, Toad's totally true.
- 33:52And I think the other thing that plays into that that's super
- 33:55important and it kind of comes back to the AB testing piece as
- 33:58well, is that a lot of people adopt A very project management
- 34:02approach to these kind of deliveries.
- 34:03But I think think a bit more about a product management piece
- 34:06where we build, we measure and we learn really helps with those
- 34:10with overcoming some of those challenges.
- 34:12So let's actually build something, let's test it, let's
- 34:14see whether it works or not and then adapt.
- 34:16So I think if you assume from the start that you understand
- 34:19everything fully and you kind of specified all the requirements,
- 34:23you're, you're just setting yourself up for for
- 34:25disappointment, but it just won't be that way too complex.
- 34:29There are too many moving parts. And in terms of, so let's come
- 34:36back to that multi entity and the local payer because you're
- 34:38right, the avoidance of FX fees is big and that was previously a
- 34:44big limitation on markets. How, how have you found it?
- 34:47Because it's, it's beta, it's not like it's been around for
- 34:49years. This is relatively new.
- 34:51So how have you found it? That and any, I guess any advice
- 34:54to other people who are considering going down this
- 34:56route? Definitely.
- 34:58So we ended up getting here via slightly tenuous route.
- 35:01So we had asked Shopify for this feature maybe a year ago when we
- 35:07were kind of in early talks with them and it was always on the
- 35:10road map. It was meant to come at the end
- 35:11of last year. It was meant to come early 2024,
- 35:15spring 2024 and it never landed. So what Shopify said to us was
- 35:20actually sign with Agen your, you know, massive payment
- 35:23provider and they Shopify built integration with Agen,
- 35:27recommended us to move to Agen and and say, look, you, you can
- 35:31have the multi currency and multi entity integration through
- 35:34Agen. That will that will work for
- 35:35you. We went through that long
- 35:37process. We did all the contracts stuff.
- 35:39We spent a lot of money with Agen, signed contracts and we
- 35:42actually went to push it live with Agen on a call together and
- 35:47we couldn't, we couldn't get it to work.
- 35:49We went back to Shopify and we said, you know, hey, help us out
- 35:52what's going on here? And they said we've changed our
- 35:55minds. You can't use that integration
- 35:57anymore. But we are building, we are
- 36:01building it a multi entity and multi currency on shop pay.
- 36:05It's not ready, but we'll, we'll let you have that as soon as we
- 36:08possibly can. So stupid, super difficult
- 36:11conversation there. We'd spend a lot of time and
- 36:14money and there were quite significant kind of financial
- 36:17implications for us if that didn't work out.
- 36:21So thankfully the way it then eventually worked out was was
- 36:25really good in that shop fight brought us into that product
- 36:29development from a feature perspective.
- 36:31So we got to join calls with their product manager, their
- 36:34product teams, their engineers, and they really wanted to
- 36:37understand how our business work.
- 36:39They really wanted to get feedback from us in terms of how
- 36:42their designs and how their interface was shaking up.
- 36:45And then once that was ready, they kind of on boarded us
- 36:47together. We, you know, we collaborated on
- 36:49getting on boarded to that and testing that live in anger.
- 36:55So yeah, the end result was a really good one.
- 36:57We'd actually always wanted to use shop pay because I think as
- 37:00a, as kind of a, a wallet, it's a really good customer
- 37:04experience. And that, you know, there, there
- 37:06certainly does seem to be some good data that suggests that it
- 37:09improves conversion as well. So we're on one hand, we're
- 37:13quite disappointed we couldn't use that we, we're going to
- 37:16Agen. So at least we'd have express
- 37:18checkout, but not with shop pay. The rug was pulled out from
- 37:22under us on the Agen P pull out from us and Agen and then we
- 37:27were able to kind of collaborate with with Shopify on on shop
- 37:31pay. That's certainly not the first
- 37:34time I've heard of that type of situation.
- 37:38Yeah, it's an interesting position those two are in at the
- 37:42moment. It's the next question.
- 37:44And one that I find particularly interesting is Shopify
- 37:47Enterprise. So that was announced about a
- 37:49year ago. It's never really been clear
- 37:52around what it is and, you know, where professional services sits
- 37:56within it and then what it means.
- 37:58Things like API rate limits and everything cuttles, and I think
- 38:02it's also not completely clear as to when a merchant falls into
- 38:05it automatically versus being paid for it.
- 38:08So you've now signed up to Shopify and Shopify Enterprise.
- 38:12Can you talk us through why you wanted it and yeah, how it's
- 38:15going to work for you? Totally.
- 38:17So we, we signed up as you'd imagine to.
- 38:20Plus originally we didn't even know that enterprise existed.
- 38:25I think there was shop, I were kind of branding something
- 38:28around commerce components, I think they called it originally,
- 38:31but it didn't feel like a strong play from them.
- 38:34And we really didn't have, you know, too many conversations
- 38:37about it at all during our kind of onboarding or at the start of
- 38:40the product development piece. But Shopify actually brought it
- 38:44to us a couple of months ago. And I think this is from
- 38:47Shopify's perspective they're trying to push into the
- 38:50enterprise space like that's obvious they've built from
- 38:52they've built up from supporting, you know, small SM,
- 38:55ES and now onboarding complex enterprise businesses is the
- 39:00play they want to make. We were interested in it for a
- 39:02couple of reasons really around the increased API rate limits.
- 39:07The plus rate limits were sufficient for us on a
- 39:10day-to-day basis, but not if we were doing really large
- 39:15transactions. They slowed us down.
- 39:17So if we're kind of re uploading a whole catalogue for example,
- 39:21or very significant percentages of a catalogue that you know, we
- 39:25would run into this rate limits, we'd have to schedule those
- 39:28updates to increase. API limits were really key for
- 39:31us. They double on enterprise in
- 39:34comparison to Plus. And we also were really keen to
- 39:37get a technical account manager during our onboarding process.
- 39:40We had access to solutions architect at Shopfire who worked
- 39:44really closely with us for probably six months and because
- 39:47we don't really have any, but we don't have a background in
- 39:50Shopfire at all. We found that super powerful,
- 39:52somebody with good problem solve with us and just bring technical
- 39:57platform understanding. So having that as part of
- 40:00enterprise has been really important for us too.
- 40:03And then one of the selling points into the business.
- 40:06Particularly from like a ACFOS perspective was that with
- 40:09enterprise we got, we had a, there are Slas, those service
- 40:13level agreements and kind of credits for downtime.
- 40:16For example, do your business feels more protected than it did
- 40:21when we were just on plus? So they're right.
- 40:24There's you know whole kind of laundry list of things that are
- 40:27different on enterprise to plus, some of which are more
- 40:30applicable for Omni channel or business that have a great
- 40:33emphasis in retail, which we don't.
- 40:35But if I was going to call all the three things out that stand
- 40:38out as being really important for us, increased API rate
- 40:41limits, technical account management and then the Slas
- 40:45around uptime. I, I think we, to be honest, it
- 40:49was never made really that clear to us exactly what the
- 40:51thresholds were for saying your, your business is now eligible
- 40:55for enterprise shop. I just brought it to us and said
- 40:59we're developing an enterprise proposition.
- 41:01We have an early bird offer for you that makes it significantly
- 41:06cheaper than it might otherwise be.
- 41:08And year of the benefit though, yeah, that works well for us.
- 41:12I think I've heard 125 million USDGMV being floated around as
- 41:18as like a an entry requirement for it, but I've never actually
- 41:21seen that written down or confirmed.
- 41:26Yeah, there's a lot of different rumours around that, I would
- 41:28say. And I hear different things from
- 41:30different clients and yeah, things with Shopify.
- 41:32I think that's pretty much where they're headed.
- 41:36Yeah, I think, I think that's about right.
- 41:37But they never, they never told us that.
- 41:39They just said, you know, would you like to move to enterprise.
- 41:41And the other thing I would say is I think they, they try to
- 41:44integrate enterprise clients more closely into their senior
- 41:49organization. So they connected us with the
- 41:53enterprise CTO from a, from a tech standpoint too.
- 41:56So I think with plus you're kind of kept a little bit of arm's
- 42:00length. It feels like on support issues,
- 42:02you get to raise tick ticket, then talk to chat like everybody
- 42:06else. But on enterprise, I think they
- 42:08try to bring you in a bit closer to the overall organization.
- 42:11And I think, you know, given that our complexity that that's
- 42:15really helpful for us. Yeah.
- 42:17And, and from, I guess from a, a vendor's perspective, that level
- 42:22of connectivity, you know, it's basically this man market or, or
- 42:26people marking at each level of your business as a merchant, you
- 42:30have a counterpart in the vendor that builds longer term
- 42:34relationships and keeps you in the fold for longer.
- 42:36May not be exactly a smart play. That that's exactly right.
- 42:39And the way it worked with us for enterprise was we were let's
- 42:42say one year into our plus contract and they offered us
- 42:45another contractor essentially a renewal at the one year point
- 42:48rather than the three-year point though it works for Shopify to
- 42:51you because we're never at least one year.
- 42:53We're going to stay with them for at least one year longer
- 42:55than we were currently contracted to.
- 42:57And and it you know, also reduces our risk around maybe
- 43:01maybe that the fee fees for plus go up next year or the year
- 43:05after or whenever we were due to renew.
- 43:07So yeah, it actually works out pretty beneficially for us
- 43:10overall. Cool.
- 43:13And one final question for me is we, we, we always like to do
- 43:16the, the retrospective when you've gone through something so
- 43:19monumentous as this and you talked about the, the, the
- 43:22middleware piece, that's obviously a big, you know,
- 43:25lessons. Then what, what are the, what
- 43:27are the other key take outs or, or key things that you would
- 43:30want other people to hear about what, what could be done
- 43:32differently or, you know, what you would change if you ever had
- 43:36to go through it again. Hopefully you don't.
- 43:39Because that's the whole point of building.
- 43:41Yeah, I remember. You won't be everywhere.
- 43:43Yeah, exactly. But other people who are earlier
- 43:47on in a journey like yours. Definitely.
- 43:50So I think, I think, I think probably two things stand out
- 43:54and one would be definitely around changing the changing the
- 44:00project mindset to build, measure, learn rather than Big
- 44:04Bang. So definitely a mistake of ours
- 44:07in our early thinking, which was we're just going to build
- 44:10something brand new, completely different.
- 44:13Customers will of course love it.
- 44:14Of course it will work way better for the business.
- 44:17We're going to make loads of money and we can set off on this
- 44:20course where we're just doing something brand new.
- 44:22But I, but I think bringing more of that kind of, I suppose it's
- 44:25an agile mindset where you say, well, actually we're going to
- 44:27build something, we're going to measure, we're going to measure
- 44:29it, we're going to test how it's working.
- 44:30We're going to learn and implement that feedback loop.
- 44:33I think trying to do that at any point possible possible in the
- 44:37migration is really helpful. So you actually can use real
- 44:41data to make decisions on rather than just gut or intuition.
- 44:45I think it would look, I think it would be great that the site
- 44:48looked like this. I think it would be great if
- 44:50this was what customer experience looked like.
- 44:53What we really tried to do was actually just build something
- 44:55and then see how it worked and try to minimize that cycle time
- 44:59and just learn as fast as we possibly could.
- 45:01So that would be 1 build measure, learn where possible
- 45:04and use data. The second piece is definitely
- 45:07coming down to the piece around partner selection and in
- 45:13particular for us it was around the iPad and just truly
- 45:17challenging ourselves as to whether build or buy was the
- 45:20rest the best thing to do and whether the partners we had
- 45:23selected really had the the technical chops to integrate
- 45:29with our with our business. And that's said not at all as a
- 45:32criticism of those partners who I think do amazing things for
- 45:36other merchant around the globe. But war of the books, it's just
- 45:41a uniquely difficult and complex business to work with.
- 45:44So I think partner selection is is definitely a key part for us.
- 45:49And then the final bit, which is something I don't think we ever
- 45:51really cracked. It's it's a hard nut to crack.
- 45:53But deciding at what point to pull the trigger from a
- 45:57migration perspective is really difficult.
- 46:00And this is not necessarily around the, the, the, the third
- 46:04party piece and replacing them and building them or in
- 46:06middleware. This is actually around when do
- 46:08you launch the site, when is when have you done enough to
- 46:13that customers won't have a bad experience.
- 46:16Maybe it's not, maybe it's not the experience you want in an
- 46:18end state. Maybe it's not a perfect
- 46:20experience, but deciding at what point you have enough in place
- 46:24that the customer experience is good enough and the business
- 46:27results will also be good enough too is really hard and we
- 46:31definitely didn't get that right all the time.
- 46:33So for example, during the AB test, there were features on the
- 46:37new site that just were not available to our existing
- 46:42customers that they would find on the old site.
- 46:44So we had challenges around account activation, you know,
- 46:47bringing like the accounts over and activating those reliably
- 46:51when you're when you're AB testing 1 storefront over
- 46:54another. Wish lists are a big part of our
- 46:56business. Wish lists with different
- 46:58customers didn't necessarily like that.
- 46:59So I think just, you know, listening to customer feedback,
- 47:02taking that on board and then making the working out how to
- 47:06make the right decision at the right time to say, yes, we're
- 47:09going to migrate. It's imperfect, but we at this
- 47:12point we just have to pull the trigger and then we'll fix
- 47:14forward. That was something that we, we
- 47:16probably had a few hiccups with and maybe didn't always get the
- 47:19right call. We maybe aired sometimes on the
- 47:21side of speed and getting things across the line and probably
- 47:25sacrifice quality to some extent, sacrifice customer
- 47:27experience maybe on occasion. And I think we, we, we probably
- 47:30should have done better on, on some of those choices.
- 47:34It's it's such a hard decision path because it's the classic
- 47:37triangle of cost, speed and like performance quality.
- 47:42And whichever 1 you focus on the most creates a compromise in at
- 47:45least one of the others. And there's no perfect answer to
- 47:49this, is there? Because I have, I have clients
- 47:50who are it's, it's about a day and therefore it's velocity.
- 47:55But you, you then you realize that you've cut a few corners
- 48:00subconsciously like in QA or UAT and then you get the these knock
- 48:05on effects and it's, it's very hard to know which which, which
- 48:09one to go for. It's so, it's so true.
- 48:11And James, one of the only things that we found helpful in
- 48:14that was just deciding what good looked like, deciding what
- 48:17success looked like and kind of trying to unite the whole team
- 48:21around that. So for example, in the AB test,
- 48:24we obsessed over conversion. We didn't say all we care about
- 48:28is conversion rates, but kind of all we cared about was
- 48:31conversion rate. And and if conversion rate is
- 48:35equal to or better than the legacy site, we press go.
- 48:39And I think that kind of clarity, it can be massively
- 48:42helpful sometimes. And a lot of the time you've got
- 48:45different competing opinions. Finance want this to happen.
- 48:48Engineering what obviously the best quality possible product is
- 48:52thinking about the customer and the customer experience, getting
- 48:55agreement around what good looks like.
- 48:56And ideally that being 1 core measurable metric is massively
- 49:01powerful. Yeah, I think, you know, I think
- 49:04that's a really good ending point because I could, yes, I
- 49:06know Paul could, I could ask you hundreds more questions.
- 49:08I'm fascinated. I think it's like you've you've
- 49:11also got a huge amount of experience and it's a really
- 49:13interesting project. So I definitely would would love
- 49:16to follow up at some point and find out more.
- 49:18But David, really appreciate you coming on.
- 49:20It's been a, it's been a really interesting episode, so thank
- 49:22you. No, it's been great.
- 49:24It's been great. It's been almost like therapy,
- 49:26talking some big things through. So yeah, James and Paul, thank
- 49:30you both very much for having me.
- 49:31It's been great. We need to start charging
- 49:33because we're therapists now. Yeah.
- 49:35So there's there might be a minimum hourly rate, but but
- 49:38everyone, everyone's listened. I hope you've really enjoyed it
- 49:40as well because it's, it's a, it's a cracking topic.
- 49:43Do reach out. We'll, we'll put David's link on
- 49:46the landing page if you want to reach out and connect with him
- 49:48on LinkedIn. But keep your ears peeled for
- 49:50next week's episode. Take care, everybody.
- 49:57You're listening to the Inside Commerce podcast.
- 49:59Hosted by experienced consultants James Gerd and Paul
- 50:02Rogers, Inside Commerce is your fast track to digital knowledge,
- 50:05featuring interviews with leading e-commerce retailers,
- 50:08industry thought leads and the C-Suite key tech companies to
- 50:11explore the latest digital strategy, customer experience
- 50:13and tech. Discover what the latest trends
- 50:16really mean and get practical advice based on experience to
- 50:19help fine tune your ECOM strategy and tech.
- 50:22Follow us on LinkedIn and reach out to James and Paul if you'd
- 50:25like to discuss an e-commerce project.
- 50:27We hope you enjoy the latest episode.