Why Cloud-First CX Is No Longer Enough

CX Today’s Rhys Fisher sits down with Rupert Adair, Director of Product Management at Enghouse Interactive.

The cloud migration era has run its course, but for many enterprises, it hasn't delivered the CX transformation promised.

With Flexera's 2025 data showing over 20% of workloads being moved back from public cloud to on-prem environments, the conversation has shifted. It's no longer about whether you're in the cloud; it's about whether your architecture is built to win.

Rupert Adair explains what that looks like in practice.

Cloud got the industry moving. But moving fast in the wrong direction is still the wrong direction. Adair makes the case that the next competitive edge in CX is architectural thinking, not cloud adoption.

🔴 The cloud-first playbook has real limits: data sovereignty, uptime requirements, and legacy integration mean that simply lifting workloads doesn't automatically improve customer experience. Rupert shares a real example: a European bank that completed its cloud uplift only to hit data residency issues, latency problems, and lost functionality – making things worse, not better. 
🔴 Architecture has to come before deployment: instead of asking ‘what can we move to the cloud?’, winning organizations start with ‘what experience are we trying to deliver?’ As Rupert puts it: “Cloud first asks where should this run. Design first asks what are you trying to build.” 
🔴 Hybrid is increasingly the destination, not a detour: Gartner predicts 90% of organizations will adopt hybrid cloud strategies by 2027. Rupert argues that for many enterprises, hybrid is a deliberate architectural end state that delivers both resilience and the flexibility to innovate with AI.

For more Customer Experience tech news visit CX Today.

Transcript

SPEAKER_00 0:08

Hello and welcome to CX Today. I'm Reese Fisher, Associate Editor. And today I'm delighted to be joined by Rupert Adair, the Director of Product Management at NCOS. Rupert, thanks for joining me. How are you doing today?

SPEAKER_01 0:21

Thanks very much. Pleased to be here. Very well, thanks.

SPEAKER_00 0:24

Great. Yeah, and looking forward to this chat. You know, we're going to be we're going to be discussing TX modernization, but I guess almost through the lens of the cloud first versus on-prem versus kind of architecture debate that's uh bubbling away right now. So on that, obviously there's been a big push towards cloud first in the last two years. But what does the shift toward design first actually look like, you know, in practice, in your experience?

SPEAKER_01 0:53

Yeah, for many years we've we've been experiencing a big push towards cloud first in the CX space. I think we all would agree that. And this was really about moving everything into one model. Sometimes the impetus was urgent deployment needs, like the pandemic and working from home. Sometimes it was cost with skills challenges, a secondary reason. And, you know, often it was simply driven by the feeling that if other organizations thought it was a good idea, then so it must be. You know, cloud only vendors obviously sees the initiative to promote cloud as the only viable direction. But in practice, most contact centers have learned that the all eggs in one basket approach does break down when you hit real constraints like compliance, uptime, legacy platforms and others. So, you know, what does the new shift towards design first actually look like in practice? Well, design first flips the question. Instead of asking what can we move to the cloud, which imposes limits on your options before you even begin, you start with what experiences are we trying to deliver and what constraints do we have. And then you decide what sits in cloud, on-prem, hybrid. Start with the outcome and priorities you want, and then decide on the limitations you can live with. So in reality, you see things like Core Voice, for example, staying on a stable platform, digital, and AI layered around it, hybrid becoming a deliberate end state rather than a transition. Now, I do need to say at that point, of course, it's often an interim step too. Of course, you know, cloud is the end goal too in many cases. There's no right or wrong. But design first means that you design the customer experience, and often that's the customer AI journey, quite importantly, these days. The constraints and the operating model first, and then you decide what sits in cloud, on-prem, or hybrid. Simplest way to describe it is cloud asks, cloud first asks, where should this run? And design first asks, what are you trying to build? And only then decide where should it run.

SPEAKER_00 2:52

Yeah, I like I wanted to pick you up on you you mentioned this, like all your eggs in one basket, which I think is a is a is a good way of looking at cloud. What do you think are the key limitations of of this cloud first mindset in enterprise CX environments?

SPEAKER_01 3:08

Yeah, it's it's a good question. Cloud first worked well as a catalyst, but in enterprise CX, it does have some real limitations. Um there are three, I suppose. The first is that it assumes one model fits everything, which breaks down quite quickly once you factor in things like data sovereignty, increasingly the case, by the way, uptime requirements or existing telephony estates. Second, it does tend to prioritize where things run, back to my earlier point, and you know, over what you're trying to build. So organizations end up moving workloads without improving the customer experience. Um, this is often a point of friction with agents in contact centers as a result of that. In other words, it doesn't focus on the really important part, that is what experience you're trying to deliver. And as I've said, a most common example, probably the most common example of what I mean by experience in this context, is that of the AI journey you want for your customers. And then thirdly, it often drives large, disruptive migrations rather than allowing for incremental change around what really works. So organizations move to the cloud, but not actually improve the customer experience or for that matter the agent user experience. You know, we often overlook that, but we've acquired multiple customers in every region due to hasty, ill-conceived migrations that neglected to consult the CX team, and then we work with them, obviously, to resolve that. But whether it was an issue with agent functionality, management insights, or operational challenges, the contact centers were supplied with an inadequate solution that, you know, in many cases reduced their service delivery capabilities. And that's, you know, that's a key consideration. You don't know been affecting your customer service. So, you know, there's a couple of examples I can think of. A bank uh in uh Europe came to us having seemingly completed their cloud uplift with another integrator. We weren't happy with the results. They'd lifted the core voice into the cloud, uh, but they then hit data residency issues and latency problems, as well as losing some key communications and integration functionality because of the move. So, you know, if they'd properly consulted their teams in their the project design phase, they would have realized that some of those workflows uh were actually quite important in their previous deployment. So by moving to the cloud, they'd actually made things worse. So we worked with them to redesign their solution, which ended up being a mix of cloud and premise, with lots of services running in the cloud, but the really key ones left on-premise. So in practice, most environments are more complex. So a strict cloud-first approach can force decisions that don't really fit. And I suppose that's why you're seeing the shift towards design first, in some cases, not always, um, starting with the experience and constraints, and then you know, choosing the right mix of cloud, on-prem, and hybrid. You you get a solution that might be more modular, but actually that's a real strength. It can flex with the needs of your business and without the reliance on one provider. So, you know, there's there's another example, a utilities provider that keeps stable voice routing in place in their locations in Germany, but they've layered the digital channels and all the AI around it sourced from different cloud providers. They can modify that easily and change it if they need to. So they can improve the customer AI journey quickly without disrupting what really works and needs to work without fault.

SPEAKER_00 6:26

Yeah, thanks for that, Ruth. But I think it's always really helpful to hear those kind of those specific examples of of how these migrations work in practice. And I know within that you you touched on this design first and uh the role of architecture in these migrations. I was just wondering how do you think CF leaders should be thinking about architecture before they decide you know what to what to migrate?

SPEAKER_01 6:50

Yeah, I think I'd prefer to use the phrase CX decision makers because it better reflects um the quite broad uh requirement for people to be involved at the outset here. Um but before deciding what to move or when, the the CX leaders need to step back and think in terms of architecture, not migration. So that means starting with the end state. What customer journeys do they want to support? What constraints do they operate, will they operate under? Compliance again comes up. It's in coming up increasingly with many uh, in particular European organizations and countries, uptime, existing systems they rely on, and this is the most common one to miss, and how different capabilities fit together. So think about local data sovereignty laws and regulatory compliance requirements, an increasing uh factor in in some other regions. You know, I can think of Central Europe, uh, Iberia, for example. Think about your voice, video, and core recording traffic, storage, and what can connectivity you need and how robust it needs to be. And think about the location of where your data and your third-party services are going to be. Um it's typically those. The the AI side of things typically can be overlaid on top quite easily, but they rely on the core components working properly. So, you know, if I was to break it down into key items, they would be, and I'll I'll paraphrase, I suppose, but you know, think about the experience first. Design with the real-world limits, such as security, reliability, speed. Think of the contact center as a set of building blocks. Voice, digital, AI, data, agent tools, all have to work together. Accept, and this I suppose is the big one, that a mix of cloud and on-prem, a hybrid solution, is often the right long-term answer. I think this is something people miss. There's this uh compulsion to move to the cloud that possibly blinds people from that. Focus on connecting the systems properly and then improve it in stages. You know, so delay the more risky stuff, make sure you're not affecting your customer service, and make changes where you get you know key gates and often the AI and the cloud services can come first. But don't forget about those those core platforms. Uh so you know, it there's one other thing, and that is you know, make sure that that who is going to own this is clear from the start. You know, make sure you're talking to all of the right people because that's one of the key sources of errors later down the line. People have not been consulted, and and then we find that you know we've we've we've ended up with a an inferior solution. So from there it becomes much clearer what actually works uh and what shouldn't. Yeah, I I you know just design the experience, map the constraints, break it into capabilities, and then modernize site is the way I would probably summarize it.

SPEAKER_00 9:35

Yeah, thanks for that, Rupert. I think it's a really nice overview there of the of the main of the main areas. And I'm glad you mentioned hybrid there, because I was interested to hear your opinion on where hybrid does actually fit into this. I know you touched on it a little bit there, but if you wanted to go into a little bit more detail, do you do you see it as a strategic choice, I guess, or is it more of a transitional phase, which is often uh presented as?

SPEAKER_01 9:59

I I I think it could be both. Uh there's no right or wrong, of course there isn't, because every business is different. I think you know, if you if you take the approach and and our our sales teams, our pre-sales teams use the concept of mild mindful migration, um, you know, it we we see that if if organizations really stop and think about it properly, as we're recommending, they might find migration isn't necessarily the best option after all. Um the the the main thing we'd like people to realize is that for many of the people who jumped to cloud, cloud wasn't just a bandwagon, sorry, was just a bandwagon. Um, but take away that compulsion, stop and consider. Um so you know it hybrid is often described as a transitional phase, but in reality, for some enterprise CX environments, it's a strategic end state. There's nothing wrong with that. You know, it's practical, and and there are plenty of reasons why that might be the right the right way to end up with your system. You know, Gartner predicts that by 2027, 90% of organizations will adopt hybrid cloud strategies, showing that hybrid is not just the transitional approach, but likely the standard for infrastructure design. You know, that's a that's a fairly interesting stat when you think about it. Um and you know, we do see businesses backing off the cloud deployments they made, perhaps using, you know, during the pandemic, uh, because they've seen some maybe slightly better ways of doing things. So, you know, once you design around the real constraints like data sovereignty, uptime existing investments and so on, you almost always, in our experience, land on a mix of cloud, on-prem, and private infrastructure.

SPEAKER_00 11:37

Yes. Again, I was about to ask you about on-prem, which I guess is the the the other side of this hybrid uh discussion. What do you think it's fair to consider it a legacy option, which it often is, or can that be strategic as well?

SPEAKER_01 11:52

Well, firstly, I don't believe it's a legacy option. I mean, there has been that view, but uh some businesses are seeing uh on-prem as a very valid destination these days. Um, you know, there's there's there's a number of ways to describe it. Bring the cloud home, for example, uh uptime on your terms, you know, there's all kinds of uh ways of marketing uh premise, I suppose. But the real deal is that uh in in Europe in particular, we are seeing customers considering premise as a valid end state. We see in in uh our uh industry events, we see customers asking whether we can deploy AI on-prem. We have customers, I can think of uh a government customer in a northern Europe uh country who is backed off from public cloud and has taken their infrastructure in-house and has locked down inside their own private cloud. In fact, that's a hosted premise deployment. Uh and we see plenty of examples where Premise is is you know is actually being retained because of what's happening in the world, some of the geopolitical events making people perhaps a bit nervous. Uh and and as I say, some countries are becoming quite introverted in Europe, and so they are being required to be much more uh careful with where their data is and where it's sent and who it's visible to. So premise again is a valid consideration. You know, it's not a sea change at this point, I don't think, but it is a valid consideration, and it should be part of the story in any CX transition uh considerations, I think.

SPEAKER_00 13:30

Yeah, yeah, absolutely. I think those are those some good examples. Like I said, Europe is very interesting right now. I know France has its own regulations around things as well, so it is, yeah, I'd be interested to see how that kind of all plays out going forward.

SPEAKER_01 13:42

Yeah. I think I I think just just to just a comment on that, uh Reese. I think you know, cloud gives you speed, right? But hybrid gives you control and continuous change. And so, you know, you might need that premise baseline uh to be able to have the agility to do the cloud stuff. You you can't assume it all has to be cloud. Um, you know, you you shouldn't be forced to move everything at once. And actually often these hybrid deployments give you the reliability and the resilience, but then the flexibility and the agility to do uh you know quite exciting things with AI without impacting things.

SPEAKER_00 14:17

Yeah, I think that's a really nice little summary there. I think it's been it's been good there to kind of get in get into the weeds and the on the different systems that you can deploy as technology. I guess stepping out a little bit now, more generally, what does good design first CX modernization actually look like? And perhaps what are the early indicators that an organization is getting it right?

SPEAKER_01 14:39

Yeah, I I think from what we we see with our our customers and our partners, actually, um good design first CX modernization tends to look quite different from traditional programs. You see a clear target architecture defined up front, typically these days based on customer journeys and specifically customer AI journeys and their can and any constraints and and and how those capabilities fit together. As opposed to big technology-led migrations where you're literally forklifting the whole solution into the cloud and that's that's it. So, you know, execution is usually in phases, it's incremental. Umganizations tend to want to improve specific areas like digital channels, AI, or agent tooling first, because they can, right? That's the nature of a lot of these systems. They are agile and they can be overlaid. And you know, at the same time, then that allows organizations to keep the stable systems in place until there's a clear reason to change them. So the only indications of a of a good CX migration, a good, ideally a design first migration, is that um you know the new capabilities are faster to deploy. Um so AI again could be deployed with fewer disruptions during change and better alignment between the IT and the CX teams. Come back to that point, I suppose, but uh that that communication is key. And you also see more consistent outcomes. So projects lending value quickly rather than slipping or being reworked. So in short, if they're delivering improvements steadily without major disruption and decisions are being driven by business impact rather than platform choice, that's really important, they're probably getting it right. Um in contrast, if it feels like a technology program instead of an experience-led one, it's probably drifting away from a true design first approach and it's likely to have problems later down the line.

SPEAKER_00 16:37

Yeah, just leading on from that, I guess, and and you covered this a little bit, but I was wondering what are the kind of what are the things that most organizations overlook when they start in a CX modernization program?

SPEAKER_01 16:50

Yeah, that's a good question, isn't it? I think uh probably I would say it's it's about uh organizations not fixing how they work before they start changing the technology. So, you know, understanding what it is that that in the business works, the workflows for customer journeys, uh how the agents work, um, you know, all of that stuff that affects the the customer experience is probably the top of the tree rather than you know the costs of of what you got you've got in your infrastructure. So, you know, where where it fails, businesses are focused on the tools, the platform and the cloud, all of that, but they never really agree who owns the customer experience, who makes the decisions, or how IT and the business are supposed to work together. And that's quite a big deal. We see that quite often. So what happens, of course, is it's quite predictable. Things then take longer than expected. People pull in different directions and you end up redoing things. So companies that get this right don't start with the technology. They start by getting alignment on ownership, priorities, and what good actually looks like. And I think we can say with hand on heart from an inch house perspective, that the most successful CX deployments we've seen are the ones where the CX and the IT are fully aligned from the outset. They've agreed the priorities between them, the project plans, and you know, everything all the way up to the customer acceptance acceptance is part of the process. You don't do that first, it almost doesn't matter what platform you pick because it's likely to struggle. So, you know, I was talking to our pre-sales uh leader in uh Western Europe, and she said to me, you know, we try to be involved as early in the process as possible to help whoever we're working with to think about the design from the outset. Often we start with the CCAS solution supported by, say, all of the components that make up the AI customer journey. But in our kit bag, you know, we've got the ability to offer hybrid or even premise in a phased and component approach. So we will typically engage them early on, get them to talk to each other and consult and understand all those building blocks. Um, because you know we would meet much more resistance if we assumed that simply going wholesale to cloud was right for everybody. We know that's not the case. Often it is, but not always. And we have to make sure we get that right.

SPEAKER_00 19:08

Yeah, yeah, I think that's uh that's a great place to leave things. Thanks, Rupert. I really enjoyed this chat. I think when it comes to modernization, it it's inevitably it's that cloud on-prem kind of debate. But I think it's really helpful to have these discussions where we dig into know the pros, cons, and differences of all these solutions, and like you said, how they can work together.

SPEAKER_01 19:28

That's right. Uh you know, cloud first got the industry moving, Greece, uh, but it also pushed people towards a one-size-fits-all. What we're seeing now is a shift of design first, starting with the customer experience and the constraints and the whole environment actually and how the whole environment actually works. You know, Flexera's 2025 report finds that over 20% of workloads and data have been moved back from public cloud to private or on-prem environments. That was a result of organizations realizing that cloud hadn't quite delivered, and then effectively going back to the drawing board. So design first and then action.

SPEAKER_00 20:02

Perfect. Like I said, that's uh that's a great way to end things. Um, I did also want to just quickly thank our audience for viewing as well. Uh, if you enjoyed this, please do remember to like and subscribe to the channel, and also remember to head on over to cfstoday.com for more stories like this. Until next time, thanks for watching.