Your Customer Journey Analytics Are Broken – And Buying More AI Won't Fix It

In this CX Today interview, Rhys Fisher sits down with Ty Givens, Founder and CEO of CX Collective, and Simel Kara, CX Account Director for Digital Experience and Transformation at Effect Digital, to dig into why end-to-end journey visibility remains out of reach for most organizations — and why reaching for AI as the solution is likely making things worse.

Simel frames the core issue around what he calls a missing "conductor" — organizations tend to have reasonable visibility within individual channels, but no shared memory connecting them. The handoffs are where everything unravels.

Ty puts it in starker operational terms. Customers bouncing between a bot, a new channel, and a live agent are retelling their story every single time — because the underlying systems simply aren't integrated. And piling AI on top of that doesn't help.

"If you don't have any of those things, what's going to happen is that those issues are actually going to be amplified because AI doesn't fix that."

Both speakers land on the same closing advice: stop buying tools and start fixing the foundations. Assign clear ownership for AI oversight, document what success actually looks like, and treat journey analytics as a live operational discipline rather than a reporting exercise.

As Ty puts it, someone needs to get off the dance floor and oversee the whole room. Right now, in most organizations, nobody is. 

Transcript

SPEAKER_02 0:08

Hello and welcome to CX Today. I'm Reese Fisher, Associate Editor, and today I'm very fortunate to be joined by two guests. I'm here with Ty Givens, the founder and CEO of the CX Collective, and Simil Kara, the CX Account Director for Digital Experience and Transformation at Effect Digital. Thanks for joining me, guys. How are you doing? Very well, thank you. Yeah, doing good. Thank you for having us. No, no, absolutely. I know, you know, we've had uh we've seen a little bit off uh off-cam. We've had this in the books a while. So uh glad to finally get the chance to uh to uh get this conversation going. You know, we're going to be discussing what I think is is quite a significant pain point in the CX space right now. You know, this idea that organizations have been chasing end-to-end journey visibility for quite some time now, but many perhaps still can't quite see the full picture. And with AI now in the mix as well, you know, the cost of that blind spot is is potentially a major issue. I guess I want to start on that point, really, you know, why organizations are struggling to see the full journey. And I wanted to come to you first, Silva. You know, you work with organizations on their journey analytics strategies quite regularly. Where, in your opinion, does the end-to-end visibility problem kind of most commonly break down?

SPEAKER_01 1:26

For me, humans really should be at the helm. And where you have various touch points across a journey, there are multiple helms across those touch points. So I am a real big advocate of what I call my three C's, so communication, coordination, collaboration. Uh, but there is a hidden fourth C as well of a conductor. And uh for me, everything within that journey, there are a number of handoffs, for example, where if those handoffs aren't proceeded with or initiated, for example, um, that's where things tend to kind of drop, really, in terms of that visibility piece. Um, most organizations have a pretty good um sense of visibility across each channel. Um, so whether it be an app, for example, and we're talking omnichannel journey here, but um whether it's an app or an IVR or CRM, for example, however, the issue that most organizations have that there isn't a linking piece between those, there isn't a shared memory, for example. And that's where things tend to drop.

SPEAKER_02 2:46

Yeah, thanks for that, Syro. Makes a lot, makes a lot of sense. I like that bit about the conductor. I think that's a nice way of looking at things. Ty, I wanted to kind of come over to you now, perhaps from a contact center and operations perspective. What does it actually look like when you know a customer arrives at an agent having already passed through digital self-service AI touch points and the context perhaps hasn't traveled through with them? How how common a scenario is that still?

SPEAKER_00 3:15

Yeah, it looks like a call or a contact where the customer is spending so much time in a bot, um, explaining the problem there, not getting any answers, getting bounced to potentially a new channel, maybe voice, um, and then explaining again what happens. So every time they touch a new point, they're starting over. And the issue is not that there's no data, there's a ton of data. It's that the information is locked in different systems that are not communicating to one another. So what happens is we see that companies are funding all of these platforms, but very similar to what Simil is saying, there's nobody overseeing, like if you maybe the plumbing might be a good way to just describe it. No one's overseeing that process. And so because integration is not happening, even though you've got all the tools, the story's not there. And I think if you listen to the first 60 to 90 seconds of any contact, then you're gonna end up finding that it's re-author, reauthation, re-explanation, and that has a real impact on like AHT and repeat contacts and also your CSAT. So ultimately, it's a customer telling their story as many times as necessary to hopefully get a response from maybe someone who can help.

SPEAKER_02 4:44

Yeah, I think it's really interesting to you both talks. I think it resonates a lot with some research we actually did at CX today last year around kind of joint journey orchestration and it kind of defined it almost as the control layer of the CX stack, which again aligns with what both you two are speaking about. The you know, this idea that it sits above the data layer and the delivery layer. I guess in practice, how many organizations do you think are actually operating with that kind of intentional architecture, architecture rather versus stitching things together reactively uh similar? I was wondering if you might share your thoughts on that.

SPEAKER_01 5:22

Yes, I think uh a really small minority um in kind of both my experience and and generally uh in terms of industry. I think um as a as a kind of an example, uh a customer interacts with um various touch points. Uh retailer has a CVP, for example, contact center platform for delivery, but that orchestration piece is actually done by a human um stitching things together, it's not an uh a kind of an automated end-to-end journey, and that's where you've really got to put yourself in the customer, the end customer's shoes here and look at it from an EQ point of view in terms of what is that experience looking like for them and what are they taking away from it. Um, that's the genuine voice of customer peace, where if there are gaps, uh then um and if an organization isn't achieving those in filling those gaps, then that's where the the issues arise, really. Um, in terms of a lot of organizations are bolting these things on. Um, but there isn't again, um, Ty mentioned it earlier on in terms of that that connective tissue piece where someone is kind of holding these things together and and is has that ownership of the entire journey.

SPEAKER_02 6:46

Yeah, I guess on that point of kind of a lot of people potentially bolting things on inevitably part of that is a lot of the time people are just kind of throwing AI at the problem, which is, you know, uh across all aspects of the CX space now, as we all know. I was just wondering in this regard, I want to look dig into that in a little bit more. You know, Ty, coming back to you, we know that AI-powered agent assist, real-time journey guidance tools, they are increasingly common in the contact center space. In your experience, perhaps of implementing these types of tools, how much does the quality of the underlying journey data actually affect whether these tools can help agents or whether they just a little bit as Samel uh mentioned there, where they just almost add to the problem?

SPEAKER_00 7:33

Yeah, so I think one of the the biggest misconceptions I see with AI is we all think, not we, others, think that there's there's a band-aid fixed to it. Um I think the people on this call, we understand that it's not. And so it's only as smart as the information that is being provided to it. It doesn't come out of nowhere and unique hardware. And so if Assist can't see the bot transcript, it's gonna, you know, tell the uh customer more than likely something that they've already tried. And I think that this is where the human aspect comes in, where there are some things that are, you know, nice to use AI for, but then there's also that that whole engagement piece that really makes the difference. The other thing is that it only takes a couple of mistakes from assist before agents throw it out the window. Now you're now you're unreliable and I can't trust you. So you gave me two wrong answers, you're making me look bad, that's gonna affect my quality. Hard no, moving forward, I'm not touching it. And then the other thing is if it's good though, it can see what has been done, and then maybe it can provide an active next step. But again, that still requires the human in the loop. And I think that that's necessary. So we can we can use data as a um guide, if you will, on how these things are gonna work. So if your systems are set up properly, you have good tagging, you know, you have good dispositions, your integrations are solid, you know, then you're gonna have a decent outcome with AI. But if you don't have any of those things, what's going to happen is that those issues are actually going to be amplified because AI doesn't fix that. So I think at the end of the day, what it comes down to is the health of your systems. And I do understand that that is the biggest and heaviest lift that most people don't want to touch, but that's the foundation of all good things when you're deploying any AI strategy.

SPEAKER_02 9:41

Yeah, yeah, it's a really good point, Ty. And I guess moving on from what you touched on there, you know, we know that AI has been utilized in a lot of ways. And one of the ways we're seeing is kind of this shift from static rule-based journey path towards AI that can spot and design these emerging journeys on its own. Sonel, I wanted to come back to you. I guess from a measurement and VOC perspective, does that make it harder or easier to understand what customers are actually experiencing do you think?

SPEAKER_01 10:11

I think in the in the current climate, definitely harder. There is a or has been um a level of predictability with with journeys, but the introduction of AI, and Ty mentioned it earlier, we've talked about the word guardrails. I think uh currently guardrails is the biggest word we could use with artificial intelligence. Um, who knows in the future? I think societally and certainly in an organization uh or organizations, uh, that word is key at this stage because uh it introduces a huge level of risk, number one, um, within any any kind of journey, but um there is a high level of influencing a customer for their voice of customer piece to be negative uh based on that uh utilization of artificial intelligence. So traditional means in terms of surveys and and kind of uh uh uh pieces where you are capturer feedback, the good old tried and trusted way is sometimes the best way. Uh no doubt things will improve as they have over kind of obviously in terms of tech over the last couple of couple of years and and and beyond, but I think uh the more automation and uh and an artificial intelligence that you that you introduce into a into a piece again. I go back to that EQ um uh mindset, you are taking that emotional um uh connection away because the customer sees it as artificial intelligence, even though they're not meant they might not be familiar with it, they do understand the society that they are living in. So, yeah, I I I think sometimes it's good to have a level of pre-mapping um and a level of kind of old old school uh um methods as well, as well as the facilitation of artificial intelligence, but having those guardrails in place.

SPEAKER_02 12:20

Yeah, Ty, I was just wondering to come back to you on that, you know, talk of guardrails, talk of obviously the potential risk of AI within within these journeys. In your experience, who in a CX organization typically owns that problem, you know, and how perhaps well equipped are most teams to manage that?

SPEAKER_00 12:39

Yeah, I smile because usually nobody owns it, and that's that is the problem. Um I always say that support leaders and people who operate support teams have to be in the past, present, and future and work across people, process, and technology simultaneously. Um, and I don't think that if you if you don't sit in the seat, you don't fully understand it. So I think there's a gap, right? So there's, you know, the vendor who might have set it up, there's the internal IT team, and in most cases in my history, um, the IT team doesn't really want to touch a lot of the support tools and the things that are needed there. And then that will leave that on a support ops leader. But most of the time, the person running the team is not necessarily someone who's focused on operations as a whole because managing people is its own beast in itself. Um and then we also end up finding that, you know, once the agents start complaining or CSTAT starts dipping, we've already hit a place where we're seeing the results of that instead of being proactive. And unfortunately, many organizations have to kind of see that dip as a signal because being proactive sometimes just doesn't feel right. It always kind of feels like that's not gonna be our problem. So ultimately, I would say that ops, right? Either an ops team, ops leader, ops person would be the right owner for this. Maybe the QA function, but again, I have I could go on a diatribe about that. I'm not gonna do that. But um, ideally, you wouldn't want to have someone who's um who understands the journey end to end. They can identify where there are opportunities to improve and also has the skill set to actually implement change. Now, you ask the question, how equipped are teams to handle this? And I'm gonna tell you that most are not equipped to handle this. Um, and and that's not I we focus on smaller teams most of the time. We do big teams too, but in the smaller teams, there's literally no one because the leader is wearing so many hats that they're just trying to get through what happened last night, not even today, get through last night in order to get the team ready to function and focus on the day. And so, because of that, there's no one overseeing. Someone gave me an analogy once. Um, I I don't remember, I think it was an old boss, but he talked about um being on a dance floor, like in a ballroom, you're on a dance floor. And he's like, if you're on the dance floor with everyone and everyone's dancing, you don't realize that there's a cliff at the end and people are falling off. So somebody has to go to the top at the balcony and oversee the dance floor and tell everyone, don't go over there, come back. And I think that that's what teams are missing. There's nobody overseeing the dance floor.

SPEAKER_02 15:37

Yeah, that's that and that's a really great analogy. And I think it's also a good point you raised there around those smaller teams in particular. I think, but particularly as a day or a lot of places perhaps are guilty of uh speaking a lot about these bigger companies, but like you said, these smaller teams, I guess that's an even an even bigger issue as you outlined there. That's uh yeah, it's a really good point, Ty. Thank you. I guess just to close things here then, guys, this is a question I had for both of you. Welcome to you first as well. Um if you could fix one thing about how organizations uh approach journey analytics right now, uh what would it be?

SPEAKER_01 16:17

Good question. I would say that organizations and individuals to a certain extent, they have got to stop treating journey analytics as a uh simply a reporting exercise, for example. It's not a tick box, it's something that is a live operational um matter, and there is a tendency for organizations to build uh with every software design software available, the most incredible, beautiful journeys, but not actually physically map out what those mean, a for internal customers, but also external customers as well. And I use the term where um customer experience is always kind of looked through a lens of the end customer, but really it's experience management where you're looking at it from the lens of a customer, but also your internal um uh teams teams as well. So I think again, Ty mentioned uh fixing the plumbing earlier on, in terms of kind of you know making sure that that is in a good place before you add all the data and you add all the intelligence onto that as well, and all the kind of good things that artificial intelligence brings. Um, the other thing I would add as well is organizations need to get to speed with where the speed of AI is going. Um, it is at an exponential rate, and uh no doubt by this time next year it will have moved on 10 times more than what it is right now, and it's scary right now. So um there is a hearts and minds piece there as well with with people. Some people are obviously uh fearful of AI, um, and uh a number of organizations that that that I know and and and I'm aware of are putting in those measures from a hearts and minds piece as well, in terms of yes, this is change management, but also this can help us to be more efficient, to be more commercially um uh successful, for example. So I think before adding more AI on top, um just take a step back and just look at um ultimately what the objective for your customer is.

SPEAKER_02 18:48

Yeah, yeah, thanks very much, Cimel. Uh Ty, um, same question to you.

SPEAKER_00 18:52

Yeah, I I second that for sure, Simil. You're we're on the same page. I think that we need to fix those handoffs before buying any more intelligence. Don't buy another tool until you understand exactly what you need that tool to do for you. And I and don't be sold by someone telling you that you need something you're not sure you need. Take the time to sit back, document, outline, whatever you have to do to figure out like what does success look like before you do anything else. So um, that's one thing. The other thing is someone has to be responsible for this. And it's worth it. The ROI is there. You need to have a person or a team or a group or someone who is responsible for overseeing the way that you guys are using and leveraging AI. Um, I swear on a daily basis, I feel like every day there's something new that happened that wasn't there yesterday that I'm like, gosh, like I can't, I can't keep up. Like, oh, we can do that now. Like we couldn't do that last night. It happens that fast. And then you add to it, like, I have the benefit, Simo has the benefit of knowing whether or not what AI is suggesting or recommending is the right thing to do because we've been doing this work for so many years. So when you get a suggestion, you're like, where did that come from? That doesn't make any sense. But imagine a novice leader or someone who's just starting up or recently promoted getting bad information and actioning it because they just don't know any better. So I will say that although the roles that we're seeing and support and the expectation of those roles are changing with AI, I still think that there's so much value in like really true experience, hands-on experience, having done this before, having been there, done that, seen that. I think that that is so important in order for us to leverage AI to make it make us all more efficient and effective. I personally use it more so as like my admin to kind of do some of the busy work that just kind of is. But when it comes to the thinking work, I need to do that because I have the expertise.

SPEAKER_02 21:02

Yeah, absolutely. I think that's some really, really great advice from both of you there. I want to, yeah, thank you very much for joining me today. Like I said at the jump, it's uh I think it's a really, really interesting topic right now, but I also think it's quite a technical topic, or at least from the outside looking in, it can be. So I think speaking to people like yourselves who have the expertise and the experience in this field is uh is a massive help. And I'm sure the audience are gonna get a great deal from this. So thank you very much for your time. Thank you, Raven. Thank you. Great. I did also just want to quickly thank our audience as well for tuning in. If you enjoyed this, and I'm sure you did, please do uh remember to like and subscribe to the channel and head on over to cxthoday.com for more stories like these. Until next time.