Latest / Inside Commerce: Ecommerce Strategy, CX and Technology Podcast / EP341: Modernising Ecommerce Platforms Without Risking Revenue: Expert Strategies for a Seamless Transition
Transcript
- 0:03You listen to inside commerce, a weekly E Comm strategy,
- 0:05technology and CX focus podcast hosted by experienced digital
- 0:09consultants and re platform specialist James Get and Paul
- 0:11Rogers. Get insights into E Comm
- 0:12strategy leading brands. Hear from industry thought
- 0:14leaders and leading tech vendors with practical advice to help
- 0:17you improve your online CX. This episode is sponsored by
- 0:23World pay. Our topic today is or question
- 0:27we're paying on is how do you modernise your e-commerce
- 0:30platform without risking revenue?
- 0:33So we're coming at it from the is there a platform in the right
- 0:35answer? Not always, of course.
- 0:39If it is the right answer, how do you build an incremental
- 0:43capability? How do you fix problems and
- 0:44modernise your text stack and do it in a way that is, I guess
- 0:49sustainable, but also isn't going to cause you problems when
- 0:53you get to launch that could end up compromise in revenue because
- 0:56it's very easy to make change, but change doesn't always
- 1:00necessarily deliver the results you're expecting.
- 1:02And there's been there's definitely been a bit of a
- 1:05change in the market, like a a shift away from like cyclical
- 1:09multi year transformation projects towards you know,
- 1:14better, better thinking about what the right TX act go forward
- 1:16is. And this might be keeping an
- 1:18existing platform upgrading to a later version or making great
- 1:23use of APIs to bolt on capability or introducing
- 1:26middleware to, to to separate existing systems.
- 1:30So there's lots of ways that people are going through
- 1:32technological change, but not necessarily doing a massive RIP
- 1:34and replace. So with that in mind, Paul and I
- 1:38are going to riff on what we're seeing and what discussions that
- 1:41we have with people, recommendations we're making.
- 1:44Anything you want to add to that, Paul?
- 1:46No, I don't think so. Yeah.
- 1:48I think just from what you've said so far, there's obviously
- 1:51going to be lots of yeah and ways the discussion can go.
- 1:54So no, let's jump into it. Let's, let's start with the the
- 2:01question on reducing technical debt, because this comes up
- 2:03every project we want to reduce technical debt and technical
- 2:07debt exists for many reasons. It could be on like old legacy
- 2:11tech and therefore there's a lot of code inefficiencies built up
- 2:14over the years. It could be multiple development
- 2:18partners have been involved, whether it's agencies or
- 2:20internal dev teams who have different ways of coding.
- 2:23And over the years that's layered on lots of lots of like
- 2:27coding efficiencies or different ways of, of, of handling coding
- 2:31practices to lots of third parties being integrated into
- 2:37tech stack. So you get inevitable over the
- 2:40years without real scrutiny, you get technical debt, which can
- 2:44result in like code blow, inefficient themes, clunky
- 2:47integrations, hacks, workarounds, you name it.
- 2:50And also the the joy of like lack of documentation.
- 2:54So you don't actually know why things have been done or how
- 2:56they really work. So in terms of like modernizing
- 3:01without risking revenue, what, what do you advise or how would
- 3:05you try to help clients to, to understand or reduce technical
- 3:07debt so that when you go into like a new solution or improved
- 3:11solution, you're not just repeating previous issues?
- 3:14Yeah, I think, I think it's, it's really difficult this
- 3:17because I feel like I very rarely nowadays, I think 3 or 4
- 3:22years ago, I feel like I would have worked on projects where
- 3:27technical debt would have been like an explicit thing.
- 3:31And I feel like maybe over the last few years as like the
- 3:36average brand that we would typically work with, I think is
- 3:40probably simplified lots of areas of their business got
- 3:44better at documentation, moved more to SAS technologies.
- 3:49I feel like technical debt is maybe more like implied than
- 3:53explicit. And I feel like most of the time
- 3:56the goals for the projects that we're working on and more around
- 4:00like agility and cost, particularly over the last
- 4:04couple of years. And I feel like, yeah, I feel
- 4:11like there's less maybe need, I don't know if you agree with
- 4:17this, but I feel like, and maybe it's the type of client that I'm
- 4:21working with. I think from a back office, back
- 4:24office perspective, there's often lots of consideration
- 4:27around technical data. Lots of things have been built
- 4:29specifically for that business or you know, multiple systems
- 4:34that have been built in different ways, etcetera,
- 4:35etcetera. But I feel like nowadays I don't
- 4:39chat to as many businesses where there's as like technical debt
- 4:44is as much of A consideration. I think it's more about that's
- 4:51kind of like I guess like a sub goal around agility or cost does
- 4:55that. Yeah, I, I, I think so.
- 4:59I, I guess it's semantics. What I, I think it's been the
- 5:02language has been used is changing because a lot more
- 5:06people have moved to modern SAS platforms.
- 5:10People have moved, moved away from legacy implement even even
- 5:14like the more modern versions of legacy platforms like sales
- 5:17forces, Adobe's, people who've moved on to like a big commerce,
- 5:20Shopify or a centre, whichever platform because they want
- 5:25they've got, they've wanted to go away from having critical
- 5:28applications for hosting compliance, all those
- 5:31requirements that sit as a cost line with either internal
- 5:34engineer or with agencies, right.
- 5:36And therefore a thing that probably changes the the
- 5:39conversation. But where I see the conversation
- 5:40is more about let's call it theme debt where people have had
- 5:45front end themes built, they're not in a headless environment.
- 5:47They've had front end themes built on Stencil for Big
- 5:50Commerce, Liquid, Shopify, etcetera and over time those
- 5:57have become bloated because they've added on customisations
- 6:01to work around theme limitations.
- 6:03They've plugged in lots of third party scripts, different
- 6:06developers have worked on it. You've got lots of code that is
- 6:10default to the theme that isn't applicable to the that store but
- 6:13hasn't been removed, might be commented out, but it's still in
- 6:16there. You've got all of this, this
- 6:18kind of noise around clean code and also coding discipline moves
- 6:23on. Let's face it, the coding
- 6:25standards, the efficiency in which you can write code.
- 6:29So that's where I see it and what I still think is a
- 6:31completely valuable conversation, because if you
- 6:32don't address that and you don't have a really professional
- 6:36approach in engineering to building that themes as
- 6:39lightweight as you can, that's when you end up introducing
- 6:42inefficiency. And I, I, I still even see some
- 6:45clients who aren't re platforming, but they're re
- 6:48theming and they might even re theme with a new agency because
- 6:51they believe that they've exhausted the engineering
- 6:52capacity of an existing partner. And whichever route you go down,
- 6:56for me, fundamentally, you don't.
- 6:58A theme is never like we built it.
- 6:59That's great. You should be auditing
- 7:01constantly because it's in flux. Yeah.
- 7:04I think that's, yeah, it's a really good point.
- 7:06And I think, yeah, it would actually be quite an interesting
- 7:10discussion around like if we had maybe different kind of more
- 7:17technical people on around kind of how you apply that without
- 7:23moving agency. Because I would say we will
- 7:26regularly start working with brands that have a really poorly
- 7:30built theme or a theme, as you say, that has been built two, 2
- 7:361/2 years ago. And then they've been operating
- 7:38a, you know, kind of MVP retangle level, etcetera, or
- 7:41like, or maybe even like not having an agent who supported
- 7:44it. And then they've kind of done
- 7:46things themselves or had a freelance work on it.
- 7:48But yeah, it's super, super common and I feel like most of
- 7:54the time what will, because I guess you've got two options in
- 7:57that scenario. If you come to a point where you
- 7:59start working with someone like us or you or any other kind of
- 8:03consultant consultancy, you've essentially got the option of
- 8:08building the scope for a project which could include kind of, you
- 8:12know, other value add areas, road map items, redesign,
- 8:15etcetera to get more value out of the prod, out of the project
- 8:18and then rebuild the theme or yeah, you invest.
- 8:23So a small amount of money. And we've had this a few times
- 8:26where and typically we probably have moved agency and there's a
- 8:30client of ours at the minute that's moving to an agency.
- 8:35And as part of that move, we've basically said to that agency,
- 8:40what is the amount of money that it is going to cost for you to
- 8:46feel 100% comfortable in the theme, to get accessibility in a
- 8:50really good players performance in a really good place, various
- 8:53kind of little bugs from an SEO perspective, etcetera.
- 8:56And also for you to be able to maintain that theme and never
- 9:01blame the previous developers ever again.
- 9:04And that might be like a 25,000 LB piece of work, but I think
- 9:09that quite, if you're moving ages without it being like, you
- 9:12know, broader project, I always think that's a really good
- 9:15exercise. And that way, yeah, you'll find
- 9:19that the developers be more efficient and, you know, you'll
- 9:22be more compliant and performant, etcetera.
- 9:25Yeah. So I think, yeah, overall, I
- 9:26agree with what you're saying. I think it's just interesting
- 9:29hearing technical debt because I feel like historically I
- 9:33probably would have associated that more with like moving away
- 9:37from SAP or a really complex Adobe or like like you say, kind
- 9:42of super complex integrations to simplify.
- 9:46But actually, yeah, you're right.
- 9:47I mean it is still technical debt.
- 9:49Ready to grow your business online?
- 9:51With World Pay E-commerce you can deliver fast, secure
- 9:55payments your customers love, backed by global reach and real
- 9:59people powered support. Scale with confidence and
- 10:03enhance your customers experience with World Pay
- 10:06E-commerce. I think one of one of the the
- 10:09biggest misconceptions out there is that you can't get technical
- 10:12debt in modern dynamics. SAS pick your pick your label
- 10:17platforms. You fundamentally can because
- 10:19it's down to engineering. Yes, the core platforms
- 10:22typically end up being a little bit more performant, a little
- 10:26bit speedier. The AP is are more modern and
- 10:29more open, although legacy ones are, you know, playing catch up
- 10:32and investing heavily there. But anyone can build a crap
- 10:35theme. Anyone can take a good platform
- 10:37and build a crap theme on it. It's it's, you know, it can be
- 10:40done. And I, I think that like
- 10:42accessibility. We were talking to each other,
- 10:43we start recording. I was having a really good
- 10:45conversation with Mikey who runs function, who they are.
- 10:52They're an engineering focused agency.
- 10:54You know them well, you work with clients them and they
- 10:57really do focus on performance accessibility in their builds.
- 11:00And they look at that code officials, they look at the
- 11:03speed and not all. I'm not criticising according to
- 11:06people, not all people have that expertise in house to get that
- 11:09right. And you'll see call web vitals
- 11:11scores some on some on re platforms and the call web
- 11:14vitals actually get worse because the discipline around
- 11:18specific errands elements of things like accessibility and
- 11:20that's and there are so many elements to it.
- 11:23And you can, you know, as you'd flag to me you, yes, you can run
- 11:26automated audits of that to flag key areas.
- 11:33But what you can't do is mimic the human evaluation of a site
- 11:38and look at performance and accessibility.
- 11:42So I think, I think there are multiple things that are are
- 11:45kind of intertwined in this where there is a code level
- 11:52technical debt, but then there's also a performance and
- 11:54efficiency piece within theme builds which links into all
- 11:57sorts of decisions, third party scripts being added and it's
- 12:00very easy to people to plug in apps and add scripts, but not
- 12:04necessarily look at whether those scripts are running
- 12:05efficiently. Are they adding too much load to
- 12:08the head? I've seen this one example of
- 12:10the site the other day Mike brought up and showed a client
- 12:12who's they've been really struggling with accessibility
- 12:14scores and performance. One of the big issues they had
- 12:17on CLS cumulative layout shift, which is a no go for general UX
- 12:21but it's crap for accessibility as well is there were loads and
- 12:25loads of images in the main menu that were loading ahead of the
- 12:30main banner. So a, a main banner placeholder
- 12:34was in there, but the main banner wasn't loading in until
- 12:37the images in the header in the menu and there were loads of
- 12:40them had loaded and then the main banner suddenly jumped and
- 12:42the screenshots moving around. So I, I, I think that's, so
- 12:47it's, yeah, I've got a few clients who've gone down the
- 12:49route that you've said, which is let's build better on what we've
- 12:53got. We don't, we don't have the
- 12:54business case and investment for a re platform.
- 12:56And what we can do is improve the front end and therefore
- 13:00measure objectively about what what benefits it delivers.
- 13:03Yeah, I think that's an interesting discussion because I
- 13:05mean for me, I feel like if you were doing any kind of major
- 13:09project, that stuff should be in the scope of work and statement
- 13:13work and everything else. And I think most agencies
- 13:16nowadays would be pretty focused on at point of launch or
- 13:21potentially with things like Co advisors, you might find you
- 13:24need to do some work post warranty period, etcetera,
- 13:28etcetera. But I think accessibility is a
- 13:30minimum. I feel like the majority of
- 13:32agencies nowadays that would be kind of within their baseline,
- 13:39you know, building of a theme. And I would personally expect
- 13:44most agencies and like you said, I mean, because to be fair, you
- 13:47can emulate a lot of the things that you would want to look out
- 13:52for an accessibility perspective.
- 13:55And I personally would feel that most nowadays, most agencies
- 14:00would be in quite a good position with that.
- 14:04And then I think Core Web Vitals, I mean, it's interesting
- 14:06because I guess so many agencies now and so many people like
- 14:11Barrett from our team has got like automated checkers that
- 14:15he's built with the kind of AI tools and everything else to
- 14:19check this stuff. And like it's so easy to do.
- 14:23But yeah, so many, it's so easy to like also I'd like to degrade
- 14:29in some of those areas. But I feel like, see, it won't
- 14:34be long before any platform would be having this type of
- 14:40auditing or that automated testing in these areas.
- 14:45And to be honest, like, yeah, it feels like that should be a lot
- 14:48more mainstream because if you work with bespoke technologies,
- 14:52you're a lot more likely to have like really solid automated
- 14:56testing in place. And maybe not for these areas,
- 14:59but yeah, I'm actually quite surprised that there's not more
- 15:03mainstream services or some of the mainstream technology
- 15:07partners haven't put that kind of stuff in place.
- 15:11I think there's so you're right. I mean fundamentally yes and no.
- 15:15No agency worth of salt wouldn't commit to building to like the
- 15:19the West CAG accessibility standards.
- 15:21And most, most will say that they built a AA AAA required in
- 15:26public sector, but not not in private sector.
- 15:29That is fine, but it's, there are always gaps.
- 15:34I I've not seen a single implementation that has been
- 15:37flawless for accessibility. And and that's not a criticism
- 15:41to developers because engineers aren't accessibility experts.
- 15:44And there are, there are the two bits.
- 15:45There is the the manual testing that has to sit alongside it to
- 15:48identify the little nuances when people are using like screen
- 15:52readers or specific impairments that then create blockages and
- 15:58user journeys that are not detected through scripted tests.
- 16:01So yeah, you're right, scripted testing and fundamentally
- 16:04important. But I do think that human
- 16:06oversight is needed in order to to help go from good to better.
- 16:11Yeah, Yeah, I agree with that. I think if you launch a new
- 16:14site, it should be really good in my opinion.
- 16:18But I agree that there'll always be things that could be
- 16:20stronger. And like you say, I think
- 16:22everyone works to AA. And yeah, I'm sure there's lots
- 16:25of stuff that could be better for other areas.
- 16:29Well, there's little things like the screen meet and I, I learned
- 16:32this from speaking to a couple of access accessibility
- 16:34specialists. And for example, like an add to
- 16:39basket or if you, if you don't have a visual impairment on add
- 16:43to basket, you'll get a visual cue.
- 16:45Maybe there's a state change to the button or there's a,
- 16:47something flashes up on the screen.
- 16:49Now you can have that happen, but if somebody visual
- 16:52impairment can't see that or detect it how and they're using
- 16:56a screen reader. So someone with macular
- 16:58degeneration is a great example. They won't detect the the motion
- 17:02of the movement because their their vision is is not acute
- 17:05enough. What is happening in the screen
- 17:07reading environment to say and tell the screen reader that it
- 17:10can then alert the user that that has happened that the add
- 17:13to button, the add to cart has worked.
- 17:15So you might find the the the unit testing goes right.
- 17:19Tabbing through works. It's got button.
- 17:21They've been able to click it grey, but they now don't know
- 17:24it's been been complete. And there's there's often those
- 17:27little endpoints of user journeys they get missed off.
- 17:31But let's move on to the next point rather than rather than
- 17:34labouring this one, which is so at the moment, let's take the
- 17:39context right now. OK, Is that actually weirdly,
- 17:41although there seems to be a bit of an uplift in interest in in
- 17:44tech investment re platform and we're seeing more projects, it's
- 17:47come at a time when the economy is still really challenging.
- 17:50You've got the the knock on impact of the Iran war adding to
- 17:53already existing pressures that are creating trading pressures
- 17:57for for e-commerce teams are adding costs up.
- 17:59They're making customers a little bit shoppers a little bit
- 18:02more nervous about what's going to happen to them and their
- 18:04disposable incomes with the potential threat of interest
- 18:06rates get up, get all this chaos.
- 18:08Yet there's a lot of discussions going on about doing re
- 18:12platform. So how, how can people justify
- 18:15re platform at ROI in that context of that?
- 18:18Actually, it's not. It's not perfect trading
- 18:20conditions. Yeah.
- 18:22So I think this, I mean both you and I will have, I'm sure being
- 18:26in situations over the last 15 years very regularly where
- 18:31essentially a business doesn't have loads of money and they
- 18:35have certain goals that they want to reach.
- 18:38And actually it would be more cost efficient for them to re
- 18:41platform to achieve those goals over a three-year period and
- 18:45that it would be to stay on and maintain their existing
- 18:49platform. So I feel like when you're
- 18:51building a business case, usually you've either got the
- 18:54fact that you're save money over a period of time.
- 18:57So you might have to like, you know, obviously assign budget to
- 19:00the build itself, let's say 250K, but equally like your
- 19:04development retainer, your, you know, hosting casts, your
- 19:08whatever other third party casts around your existing platform
- 19:13slash tech stag will come down on and over a three-year period
- 19:17or maybe even, you know, payments or whatever else.
- 19:21So therefore over a three-year period, you'll save money.
- 19:24So that would probably be the most common one that I guess I
- 19:26would say the other one would be that they would forecast uplift
- 19:31as part of the project. So yeah, either, you know, gross
- 19:36sales, whatever else forecast improvement then so making it
- 19:40more efficient to move. And then I guess the last one
- 19:44which a lot of people that I've worked with have tied to a re
- 19:49platforming is they have a budget for the year for to
- 19:54deliver a road map. And let's say that's 250K and
- 19:57then someone has turned around and being like actually we think
- 20:01we can deliver this all of this work and RE platform for that
- 20:04same amount of money. And then that's then triggered
- 20:07the project that I feel like all three of those are probably.
- 20:11Quite relevant because I agree. Like, yeah, the, the economy
- 20:16obviously isn't great, but I feel like I've also worked with
- 20:19loads of businesses that have just been in challenging
- 20:21positions and it's, it's been like those types of business
- 20:26cases that they've put forward where it either, you know,
- 20:30you're committing to a higher performance or you're saving
- 20:33money basically. Yeah, I think right now, because
- 20:36I've been going through this process with a, a client and
- 20:38it's the, we, we deliberately predicated the cost model and
- 20:44the business case on no uplift in terms of like conversion
- 20:47rate, average order value because we thought it was, it
- 20:50was quite tenuous to link it specifically to the.
- 20:54I know you can. I'm not saying you can't and I'm
- 20:56not saying you shouldn't. And we've had this debate
- 20:57before. I always start from doing, if I
- 21:00can get a business case to work without having to do that, I'm
- 21:03happier that the business case is, is stronger.
- 21:06So we've looked sunk development costs.
- 21:10So they have a lot of problems at the moment with a an
- 21:12inefficient implementation and a problematic theme that is
- 21:18causing all sorts of bugs. So whenever they do anything
- 21:20new, it creates knock on effects that has to be triaged and
- 21:24resolved and that's coming out of support and maintenance.
- 21:27It's no, not a like retainer. It's not adding anything
- 21:29proactive to the business. It's just fixing shit basically.
- 21:32And they're spending 4K around 4 KA month from there.
- 21:36And no, we know from having modelled this on on other
- 21:40platforms or or with other agencies, if you get the, the
- 21:44build right and you get the right engineering focus on it,
- 21:48then your support maintenance cost should be under half that
- 21:51just for pure like bug fix in a longer basis.
- 21:53So, so that's modelled in then it's the, it's the one of the
- 21:58big areas that comes up is the cost of staying on the current
- 22:01platform that you could remove. So it will be you need a third
- 22:04party for XY or Z, but you won't need that because the capability
- 22:10exists natively in the new platform or there isn't a an app
- 22:15that's a much lower cost that does everything that you need.
- 22:18So we've got one around like Smart Collections Logic where
- 22:20they need a third party at the moment to achieve it.
- 22:23But actually you can strip that cost down.
- 22:24That's a good few grand a year. And yeah, then I think those are
- 22:29the ones I always try and find the cost bit.
- 22:31And I think that's that's an area where it really resonate
- 22:34and you'll know as well from these discussions, right?
- 22:36CF OS want to know, OK, great. You're spending money.
- 22:39How are you giving me efficiency or saving me money?
- 22:41Where can you strip costs out to to my bottom line?
- 22:46And payments can sometimes be an area where you can actually
- 22:49strip out, not always, but and it's also possible to take away
- 22:54percentage points and give people margin B.
- 22:56So I, I think that that's an area that people should spend a
- 22:59bit more time on the internal process stuff of how much time
- 23:03we spend in doing things that are not efficient that we will
- 23:05know we will strip out. Yeah, I think the, so I agree, I
- 23:10think if you're solely migrating to another platform, you should
- 23:13never forecast uplift. But the one thing that I always
- 23:17see as an opportunity is if you are migrating to another
- 23:19platform, you are essentially even if there's you know, no
- 23:23change to the front end, you're still building everything from
- 23:26scratch in most instances. So where if I was ever to want
- 23:31to look at an uplift, I would be saying like trying to get the
- 23:36business to give me examples of things that they think could
- 23:39drive commercial impact. So let's say it's like, you
- 23:42know, in line bundle up sells on PDP or like generally things
- 23:49that can give them the ability to like better trade the site
- 23:53and then factor those into the scope of the project with no
- 23:56additional cost. And usually any business or have
- 23:59like a list of those ready to go.
- 24:01And if they were to do it on the existing platform, it's cost.
- 24:04It's like incremental cost. Whereas if they do a start of
- 24:07the rebuild, there's no cost essentially and a lot of cases
- 24:10or minimal uplift to the project.
- 24:15So I feel like that's like one of the things I would often do
- 24:19in that scenario. Yeah, For me, what one of the
- 24:22one of the really interesting processes in any technology
- 24:27change is really grilling the why are we doing this?
- 24:31What, what, what? So you always, it's either to
- 24:34add in capability, isn't it, and give the teams better tooling to
- 24:38do stuff that they know is going to help them better execute on
- 24:41their strategy. And that could be directly
- 24:42revenue generating or it could be about speeding up time to
- 24:46market where, where whatever the benefit, you can put a number to
- 24:49it because speeding up time to market is a time.
- 24:52So if you can execute something quicker, there's a time saving
- 24:54and that's a cost against the payroll.
- 24:57You're not saving on the payroll, but you're saving on
- 24:59that time element, which can be used a more value add or you're
- 25:03getting to revenue quicker. Or, or you know, you could, you
- 25:06could be saving money elsewhere on a licence, you could be
- 25:09removing a licence you no longer need.
- 25:12So that for me is one of the most fun parts is it's kind of
- 25:14grill, grill on what is driving this and what are the benefits
- 25:18and how can you time stamp or put a number to it?
- 25:21And it you should do it to every single, every single investment.
- 25:24There's always a number you can come up with.
- 25:27Yeah, yeah, great. How do you handle future
- 25:31benefits? Because I find this one a bit
- 25:33more challenging sometimes because people say, oh, look, we
- 25:35put this in, we'll get better conversion.
- 25:37And that always feels really nebulous to me.
- 25:38I know you can. You go to any vendor, they'll
- 25:40have case studies that says these guys got an uplift, but
- 25:43it's not always contextually relevant to the exact use case
- 25:46that you're putting in. How would you handle that to
- 25:48make sure modelling future benefits becomes credible?
- 25:51To be honest, we'll use other clients a lot of the time and
- 25:54obviously we primarily operate in fashion and beauty, which
- 25:59helps us. The other thing is we're just
- 26:01going to ask people. So other brands, we do that a
- 26:03lot and even if they use certain words like it was a really good
- 26:08uplift, it would have helped us to put a tiny number next to
- 26:11something even if we're really conservative.
- 26:13But essentially, because we'll usually be forced, if there is
- 26:18ACFO that doesn't want to create budget for this or an FDA or
- 26:22whatever, we'll usually be forced to do like due diligence
- 26:26on anything that we're saying will result in uplift.
- 26:30But yeah, we usually come from other brands basically I'd say
- 26:34because I agree, like we just had one recently actually and
- 26:38they were saying like this vendors told us X and the
- 26:42numbers were like ridiculous. I should expect like crazy,
- 26:46crazy positive return. But we knew that even though
- 26:51it's going to get project signed off, we knew that was never
- 26:53going to be the case. So we went and kind of got hold
- 26:56of like some real like more conservative numbers specially.
- 27:03Yeah, I, I think that that's exactly what I would do is ask
- 27:06basically if you can find a credible use case or brand that
- 27:10is comparable and therefore you believe that that within a plus
- 27:13minus like you know, 1020% it's going to be, yeah, a good
- 27:18indication that it's worth doing.
- 27:19Otherwise you say what you don't know what it's a little bit too
- 27:22nebulous to to to do it, although I do know some
- 27:25businesses who will happily put it in just to get the business
- 27:27case signed off. Let's talk about the next one
- 27:30about South come back. The question is modernising
- 27:32without risking revenue. If you're changing, there is
- 27:36there is an argument sometimes for Big Bang.
- 27:39Let's get it all in in one go. Let's benefit from all the
- 27:41feature capabilities. Let's say you're putting in a
- 27:43product discovery engine and you would you just turn on every
- 27:45single element of it, including packaged product
- 27:47recommendations, all the search piece, semantic search, all of
- 27:51the, you know, PLP merchandising, Bola versus let's
- 27:58get it into if we take product discovery engine.
- 28:01So I've done this for where I've said, let's just get where
- 28:05search is the most important thing because it's a massive
- 28:06catalog site. Search is the dominant user
- 28:08journey. It's the most complex in terms
- 28:12of getting it configured. And you go for that and you
- 28:14validate that and you and you measure the impact before you
- 28:17then turn on the other pieces. How do you work and how do you
- 28:21abide on solution simplification and MVP approach?
- 28:25Yes, I think this varies massively dependent on the
- 28:29brand, dependent on yeah, how much risk there is in different
- 28:32areas depending on the project. Because you know, we were, we've
- 28:35worked with some clients where they just want no change like
- 28:39they want like for like everything.
- 28:40And obviously that's not physically possible and you're
- 28:43moving platform, but they want to minimise change across the
- 28:46board just so that it's easier to quantify any changes
- 28:50essentially. And you know, some brands just
- 28:52want to test everything. Can you know, they'll do that
- 28:54post launch, etcetera. And then we've worked with
- 28:57others the essentially have been a complete opposite and they're
- 29:01willing to take more risk ultimately and Big Bang and then
- 29:05they just monitor everything very closely.
- 29:08I'd say ideally you probably want to be somewhere in the
- 29:10middle in most cases. And if you are changing
- 29:13something and you are 100% confident for the better, it's
- 29:17just a case of monitoring it post launch in my opinion.
- 29:21And you know really like knowing what you're changing.
- 29:24I think one of the things that we try to do quite a lot.
- 29:28So, and I know you do this as well, but essentially upfront
- 29:31when you start to map out all the risks going into a project,
- 29:36quantify the risk. So and to be fair, we would very
- 29:40rarely do it across something like search in my, if I'm
- 29:43completely honest, I think it off when we, if we were moving
- 29:46from one platform to another, in most instances, we're going to
- 29:49be moving to a new search vendor.
- 29:51I feel like what would usually happen is there would just be a
- 29:55set of things that would need to be done for someone to feel
- 29:59confident. And actually, I feel like in
- 30:03most instances, and again, maybe it's type of brand we're working
- 30:06with, but if the person feels confident in the, you know, the
- 30:09way that products are being served in the waiting and the
- 30:13kind of merchandising behind it and the results being good and
- 30:17the UIUX being good, people will be fine with that.
- 30:21And then actually that probably wouldn't be something they'd
- 30:23monitor, but something like SEO, which we've got as a separate
- 30:28thing in the notes. So I don't want to eat into it
- 30:30too much, but I think in that scenario, in any project, we
- 30:34would quantify the risk. So obviously that has to be like
- 30:39a line where you invest and has to be something that you
- 30:42essentially make sure that everyone's really investing get
- 30:45my. But I also think it's important
- 30:49to essentially know what or like at least estimate how much risk
- 30:53there is from like a non brand SEO perspective, how much
- 30:57reliance there is and be able to like forecast good, bad,
- 31:02terrible and then already have contingency budget from a paid
- 31:07search perspective and also have like reporting in place post
- 31:12launch. So I've obviously gone quite
- 31:15specific there, but I feel like, to be honest for me, I feel like
- 31:19a lot of areas we would try and keep the same.
- 31:22So for example, most of our clients would launch the exact
- 31:25same approach to category merchandising and you know, that
- 31:30would be very similar. I think the bits the need that
- 31:34we would usually I guess try and be careful with would be like
- 31:37the impact on template from templates changing
- 31:40significantly. And then in that case, it would
- 31:42just be a case of usually like monitoring numbers and having
- 31:46the right reports in place, performance across different
- 31:49channels. So some of we haven't got hit,
- 31:51I'd imagine probably talk about tracking.
- 31:54Actually a couple of times I spoke to and we've not been
- 31:57involved in these projects, but a couple of times I've had
- 31:59brands that have really claimed impacts on performance signals
- 32:03and that impacting the performance of in particular
- 32:07Meadow. It's quite an interesting one.
- 32:10So yeah, I think things like that, yeah.
- 32:13You've just got to try and, you know, be super careful with.
- 32:17Yeah, I feel like I've waffled on for a while now.
- 32:22You're muted. Actually, James, I think this is
- 32:24the first time that's ever happened.
- 32:25I've done it. I'm so favour and listen, I've
- 32:28done this classic schoolboy. My dog was barking so I muted
- 32:30myself and started talking again.
- 32:32I'm going to keep that in just to embarrass myself.
- 32:35But yeah, the ones I'm working with the furniture brand at the
- 32:37moment. And one of the things that that
- 32:40they've been doing on the current site or Magento is so
- 32:45wanted to make improvements and changes like category navigation
- 32:49and also introducing some elements like your shop by room,
- 32:52but also adding in FAQ blocks into key pages.
- 32:56Is they've added anything early on the current site to measure
- 33:00and benchmark the performance to validate before launch on new
- 33:05site. And I think that's a really
- 33:06smart way of doing it is, is even if you don't get it in, in
- 33:10the most perfectly beautiful implemented thing, just to to
- 33:12validate that it's going to have a positive vet before you change
- 33:16things like hierarchies and navigation is really, really
- 33:18important. You can use your existing site
- 33:20as a test bed for a new launch, rather than launch with
- 33:25something you think is compromised and then wait to
- 33:27implement it if you think the financial benefit will be
- 33:29significant in the first three to six months of that new
- 33:32project. Yeah, the we've actually got a
- 33:34client at the minute where I think we've got 2 clients from
- 33:38the minute we're doing. This on but because.
- 33:40The overhead to get kind of tests live has come down
- 33:45significantly and also to build something in, you know a static
- 33:51format that you can test via a testing tool.
- 33:54So we've got to this client. I'm thinking essentially they're
- 33:57going into a redesign and we are currently testing a load of
- 34:01structural changes to the PDP and all of that data will
- 34:05actually just go straight into the design process.
- 34:09But I would say that's a real rarity because this is a client
- 34:12that's in no rush, whereas every project we have, they're in a
- 34:18rush and they couldn't delay the design process to get the, you
- 34:22know, all of the data and recessed properly.
- 34:24But yeah, I completely agree. And actually on that, along that
- 34:29those lines in the past and we've been, I think this is
- 34:34quite a few of the times have been with Novos, the SEO agency.
- 34:37But actually to reduce change, they've implemented like loads
- 34:41of SEO changes that they would have wanted to make like one
- 34:45month into a six month project on the existing side just to
- 34:48reduce change on launch. Yeah, and the same principle,
- 34:52just reducing change, free launch.
- 34:55Yeah, funnily enough Novos is the the SCA partner of my client
- 35:00as well. It's a small world where you're
- 35:02working, isn't it? I do rate Sam and the team very
- 35:03highly. Let's let we've segues nicely
- 35:07onto one of the one of the big commercial worry areas around.
- 35:11Any migration, whether it is your you're upgrading on a
- 35:16current platform and relaunching the the storefront and changing
- 35:20parts of it or you are migrating from platform A to platform B
- 35:23where you will get different URL structures etcetera.
- 35:25It is avoiding SEA traffic drop after migration.
- 35:28Now it's never 100% perfect because there's always a slight
- 35:31drop. It's typically if it's run
- 35:33technically, I say it's run really smartly.
- 35:36That drop is not long term. You start to see really good
- 35:39recovery in a matter of days and long term you can often get
- 35:44gains to ranking and performance.
- 35:48But how do you when when you hear people within a business,
- 35:53you know outside the econ team who are nervous about SEA
- 35:57migration and issues and they've seen and read horror stories
- 36:00online, how do you advise people handle this and de risk?
- 36:06So I think I mean sometimes I will be significant risk and
- 36:11that does just require like a lot more planning and
- 36:14discussion. But I think and again this comes
- 36:16down to the client we typically work with that we will very
- 36:20regularly get that view. And then when when it comes to
- 36:23actually like quantifying the risk, like I say, and looking at
- 36:27like, you know, the organic keywords that are actually, you
- 36:30know, driving or at least estimated to be driving their
- 36:36kind of traffic and kind of overall visibility.
- 36:40Like a lot of the cases, you know, we 80% branded and there's
- 36:44very little risk from a brand perspective usually and not
- 36:47always, but usually. And then it's just a case of
- 36:51putting a plan in place where they know that they've got the
- 36:54contingency search budget. Yeah.
- 36:57And essentially just kind of mapping it without and building
- 37:00a model scenario based so that they know there's it can be
- 37:05counted. And I would actually say I feel
- 37:08like over the years I would have completely agreed with that in
- 37:13terms of that always be a drop. But I do feel like nowadays I
- 37:17feel like Google certainly I, I feel like you don't.
- 37:22It's been a while since I've seen a migration where there has
- 37:27been, well, I wouldn't say that we've definitely seen somewhere
- 37:30there's been a bit of a drop like you say, but I think it's
- 37:32more common, at least it's semi common now to see a migration
- 37:37where if it's managed really well for actually there not to
- 37:41be a drop. But again, I think that probably
- 37:43comes down to the mix of brand versus non brand as well because
- 37:49if you're heavily branded, it's going to be a lot less risk to
- 37:51organics. Yeah, yeah, definitely.
- 37:55Nothing's really changed in this, has it?
- 37:56Because the the search engines have got smarter at
- 38:00understanding migrations if you follow the directives that are
- 38:04set to follow. But the I, I always come at this
- 38:09from, I want a, an independent technology and agency agnostic
- 38:14tech SEO lead. And this is no, no denigration
- 38:19of agencies who have SE OS in house.
- 38:21It's just I've generally found the the independent ones who
- 38:25work across platforms have a deeper understanding of the
- 38:30migration complexity and the advice that is required.
- 38:33I'd love to be proven wrong, but that is normally the one of the
- 38:37ways that I would advise that people de risk it is it doesn't
- 38:40mean you don't listen to the agency SE OS you fundamentally
- 38:43should. It's a collaborative process and
- 38:45they really do add value and they help in discovery.
- 38:48But there are sometimes nuances in the, in the tech decisions
- 38:51and the compromises need to be made.
- 38:53And I find internal SE OS have they've got to work to the
- 38:56confines of how their developers work.
- 38:58An external person is not influenced by that, that piece.
- 39:02So I don't know, that's probably controversial and some people
- 39:05will dislike me saying that. I can, I completely agree.
- 39:08And even, yeah, we've done projects with lots of agencies
- 39:11that have full SEO teams, but we've never wanted to use them
- 39:15because even like the politics, like someone, if you're in the
- 39:18same agency environment and something's gone madly wrong,
- 39:22you're in a more way more difficult position.
- 39:25If you're and it's even though you don't want, you know, two
- 39:28parties arguing, blah, blah, blah.
- 39:29But it's it's easier to manage that I would say.
- 39:33Yeah, it is. And it's, that's why I say
- 39:35there's, there's some brilliant SE OS in an agency.
- 39:38I know I've worked with many of them and yeah, they, they're
- 39:41great. But you're right.
- 39:42The book that I always worried about is, is how much do they
- 39:45bend the need to bend recommendations to, to what the
- 39:49developers are saying can and cannot be done.
- 39:50And how do you unpick that? If something's fundamentally
- 39:53wrong, it's far harder internally to have that
- 39:56conversation and to get by. And then it is if an externally
- 39:59contracted person is saying no, this is a blocker and This is
- 40:02why. Yeah, interesting debate.
- 40:06That's a, that's a bigger debate for another time as well.
- 40:09But we've got broadly time for one more question.
- 40:11The one that I wanted to get onto is.
- 40:13So again, coming back to that modernizing without risk in
- 40:15revenue is what one of the areas I think is really fascinating is
- 40:19looking at how technology change can make e-commerce teams less
- 40:23dependent on developers. Now I'm not saying developers
- 40:26aren't still critical, they fundamentally are.
- 40:28Senior engineers are absolutely essential in getting good
- 40:31solutions delivered and maintained.
- 40:33I'm talking about how you can take away like low level tasks
- 40:36from developers to free their time to add to value add bigger
- 40:40projects, you know, really do focus on solution architecture
- 40:43and that intelligence piece, not just having to fix issues.
- 40:47But so it's, it's like how I'm how I'm seeing and hearing teams
- 40:51doing things like AI prototyping.
- 40:54I just spoke to Tom and Co. We had by the time everyone's
- 40:58listening to this Tom and Co episode will be following the
- 41:01week after Topps tiles did their internal team did prototyping
- 41:07for new click and collect check out flow, which enabled them to
- 41:10hone down what they wanted, how they wanted to work in Canada
- 41:12over to the development partner who could then build a proper
- 41:16engineered production ready system.
- 41:18So that all the I think this is a really interesting how you can
- 41:22actually factor in thinking about the de risking of this
- 41:25because it's about what you can do more and control more
- 41:27yourself or having to pay somebody and maybe speed up
- 41:30their ability to execute for you.
- 41:32Yeah. So I mean, to be honest, that's
- 41:34a massive trend across the industry.
- 41:35So I mean, we do it all the time.
- 41:38We'll build prototypes of functionality and give it to
- 41:42agencies nowadays just via the the AI tools out there and AI
- 41:48extensions for different tools. And by actually I was working on
- 41:52a project last year and it was with a brand that actually one
- 41:57of the founders, he was like to in and throwing loads around the
- 42:01design or something. And he actually just went and
- 42:03like spent a night building like a really solid prototype using
- 42:08lovable and actually like, it was like it was.
- 42:12Yeah, just I reckon that probably saved a number of weeks
- 42:17on the project just because that was a really complex part of the
- 42:20business, really complex journey.
- 42:22And for him to just be able to literally throw up exactly how
- 42:27he wanted it to work. Yeah, it was pretty game
- 42:30changing. But yeah, I mean, I think
- 42:32overall to, to your question, I feel like that's the like, yeah,
- 42:36probably cost for me, the biggest goals at the moment,
- 42:40that driving projects cost is probably still #1 #2 is agility
- 42:45and relying less on developers and time to value from a
- 42:50features perspective, which are kind of the same thing.
- 42:53So yeah, I can, I completely agree with that.
- 42:55But I think that prototype and thing is really interesting
- 42:59because I definitely can see the impact that's already having on
- 43:03like retainer relationships and like the amount of like
- 43:07discovery that's needed for like, like new features.
- 43:11And yeah, I personally like love that kind of stuff as well.
- 43:15And I think it's yes and Changer.
- 43:18Well, I, I think there's, there's two things is it gives
- 43:20greater control internally and, and I guess boards and C-Suite
- 43:23want to know that people are controlling and in control of
- 43:27change because it has a big commercial impact if you do it
- 43:30right. But the second bit is if you can
- 43:34take time away from really high, highly experienced qualified
- 43:39engineers to not have to get involved in those early phases
- 43:42and you leave them to apply their time on.
- 43:46Think about whether the solution can be improved.
- 43:48What's the best way to get that into the the tech stack and the
- 43:51overall solution that added value engineering that will
- 43:54deliver a more robust, smooth and performance solution that is
- 43:58a way better use of their time. Otherwise in coming back to your
- 44:01point about managing like the budgets in in builds as you
- 44:04quickly get through time, if you don't have a really mature well
- 44:09thought through brief from concept because you haven't yet
- 44:13thought through it enough and therefore you need to use more
- 44:15of their time to help you nail that down.
- 44:17So I think there's good ways that you can get better time to
- 44:20value out of developers. Yeah, that kind of brings us in.
- 44:24I know we could spend a lot more time to write lots of topics,
- 44:26but people would fall asleep if everyone listening.
- 44:30We'd love to know what out of this was the most interesting.
- 44:32What questions should we have asked that we haven't that you
- 44:35would like to hear us answer? Because that will help us to
- 44:37look at maybe potential follow-ups in this episode.
- 44:40But thanks as always for listening.
- 44:42We do hope you've enjoyed it and do reach out and let us know any
- 44:45comments. You're listening to the Inside
- 44:48Comments podcast, hosted by experienced consultants James
- 44:50Gert and Paul Rogers. Inside Commerce is your fast
- 44:53tracked digital knowledge, featuring interviews with
- 44:56leading e-commerce retailers, industry thought leads and the
- 44:59C-Suite key tech companies to explore the latest digital
- 45:01strategy, customer experience and tech.
- 45:04Discover what the latest trends really mean and get practical
- 45:07advice based on the experience to help fine tune your E Comm
- 45:09strategy and text. Follow us on LinkedIn and reach
- 45:13out to James and Paul if you'd like to discuss an ecommerce
- 45:15project. We hope you enjoy the latest
- 45:17episode.