Sponsored By: Hightouch | How Chick-fil-A Turns Guest Data into Caring Experiences

Summary

Chick-fil-A identified three breakdowns in its guest marketing—strategy, operations, and tooling—caused by disconnected CRM channels and siloed data, and addressed them by shifting to a composable CDP architecture built on Databricks and Hightouch.
The new architecture creates a unified decisioning layer that brings together restaurant, app, and loyalty data in Databricks, with Hightouch acting as the 'air traffic controller' that pushes trusted insights into the right marketing channels at the right time.
This video covers concrete campaign outcomes including breakfast message timing improvements driven by real-time data, personalized birthday and tier sends, a personalized end-of-year in-app experience, and a national loyalty giveaway executed at Super Bowl scale.

Chapters

FAQs

What is a composable CDP and how does Chick-fil-A use one?

A composable CDP (Customer Data Platform) separates data storage from activation, storing unified customer data in a platform like Databricks and using a tool like Hightouch to orchestrate marketing activation across channels. Chick-fil-A uses this architecture to keep guest data—covering restaurant visits, app activity, and loyalty interactions—in a single governed location while empowering marketers to run campaigns without relying on engineering handoffs.

What marketing challenges was Chick-fil-A trying to solve?

Chick-fil-A's team identified three core breakdowns: fragmented marketing strategy with no unified view of the guest journey, operational complexity from managing disconnected CRM channels, and tooling limitations that prevented real-time personalization. This video describes how the composable CDP architecture on Databricks and Hightouch addressed all three.

How did real-time data change Chick-fil-A's breakfast campaign timing?

By integrating real-time loyalty and transaction data into its decisioning layer, Chick-fil-A was able to identify the optimal moment to send breakfast-related messages based on each guest's actual visit patterns rather than fixed campaign schedules. This video presents the breakfast timing improvement as a concrete example of how real-time data changes campaign performance.

What is Chick-fil-A's AI roadmap for personalization?

Chick-fil-A's roadmap described in this video points toward moving from rules-based personalization to outcome-defined personalization, where AI determines the best action to take for each guest rather than marketers manually configuring rules. The unified data foundation on Databricks and Hightouch is the prerequisite for that shift.

Full transcript

[00:08] All right. So, do we want to go ahead and jump in? >> So, um I think there's some legal things that we had to just show forward statement. Um we did it. Um at the end of the session, uh please complete the
[00:20] survey about our session in the app and then here you go. >> All right. So, for those of you who stayed, thank you very much for coming to our session right before lunch. Like I said, whoever did the scheduling,
[00:34] great job putting Chick-fil-A right before lunch. This is perfect timing. Um, so I'm sure most of you have heard about Chick-fil-A, and for those of you who have not, we are a restaurant company that was founded in 1946 by
[00:46] Estru Kathy, uh, when he opened the first Dwarf Grill. And then in 1967, he actually opened the first Chick-fil-A, which is what you all see have grown to like see across uh we've done like over
[01:01] 3,000 locations across the US, including Puerto Rico, Canada, and recently the UK, and Singapore. So, we're continuing to expand. But just one thing to note is that with that growth, that growth has
[01:14] been intentional. And also with that growth, we have tried to make sure that we were intentional about growing with care. And so that's a part of what we're going to talk to you all about today. Um, so the name of our session for those
[01:26] of you who are not in for the coding session, the name of our session is from channels to customers. How Chick-fil-A is reimagining technology operations and strategy to create caring customer first experiences through our omni channel
[01:39] engagement. So, I am Chararma Parker, one of our senior platform leads at uh Chick-fil-A. I have been with the company for almost
[01:52] five years, and I have the pleasure of serving our customers in our national emails and mobile channels like push notifications and SMS marketing and recently inapp messaging as well. Prior
[02:04] to my time at Chick-fil-A, I have also served in the same capacity for customers for the Home Depot, Atlanta Hawks, and State Farm Arena. No, I was not the point guard. I'm a little bit too short. Um, and I also worked at
[02:17] Carters and Ashkosh Bagosh. >> And I'm Dan Mcdanell. Um, I'm a senior product lead. um focusing on personalization across uh various channels within our business including
[02:29] Chararma's CRM channels and also uh personalization in restaurant. Um my time at Chick-fil-A has been about four years now and my background is in uh
[02:41] streaming data on through Wall Street streaming video content. Um, but coming to Chick-fil-A, we really just focused on a lot of the needs for Chararma and her team around personalization and how
[02:54] can we take some of those practices uh into the marketing space. >> I truly did not know we were being transcribed. So, that's fun to see over there uh coming up on the screen because it is not spelling my name right. It's
[03:06] fine.
[03:18] organization. >> Oh yeah sorry.
[03:32] As I mentioned before um before we started introducing ourselves. Care is a big thing with us. That is what our brand is known for. When you hear about Chick-fil-A, one of the main things that people talk about is the care that they receive within the restaurant, within
[03:45] the drive-thru. And we have really been intentional about how do we bring that care, how do we like pull on that thread across from a digital standpoint. So, this is one of the things that we start out most of our meetings with for
[03:58] everyone who works on any body of work um that has to touch the customer. So whether it is our marketing teams, whe there is our data teams, our technology teams, this is our northstar
[04:15] and a part of that is because we also know that when you know someone's story, you can care for them personally. We deeply feel that and there are opportunities that we have seen in our everyday um because you know as you all are going through like whether you're parents, whether you're just like
[04:28] working like every day there's something that comes up. We're all busy. Definitely see that. But it's really in those moments where the person on the other side of that customer interaction sees you. And I mean truly sees you. Um
[04:41] that can bring the pleasant pause to the mundane or the chaos depending on how your day is going. And one of those examples that I want to give is recently a friend of mine um and myself of course
[04:53] is the way I said that. Uh we were at one of the theme parks here in California. I'm not going to say the name but you all are we're in California. you know where I'm probably talking about. Uh we were in one of the gift shops and we were talking about how
[05:05] this was our first time at that location. We had been to their other location across the country and we were really excited because this was our first time at this particular location and one of the workers in there overheard us. We had interactions with
[05:18] that person um as they were like showing us around like the different uh store and like the things that we were looking at. But when we went to check out, they slid over these two little buttons and they said first visit. And it was a very
[05:31] small gesture, but it had a really big impact for the rest of the day because we're walking around with this button and we're feeling kind of special because we're going up and we're seeing these long lines and some of the workers were like, "Well, come on up because
[05:43] this is your first visit." And it's like, how do we take that particular type of care and like bring it to our customers at Chick-fil-A? that has been like a driving force in a lot of our
[05:55] conversations. So today, the menu, as I said, right before lunch, look at the image. Make sure you get hungry. I'm just saying. Um we're going to talk to you today
[06:07] specifically about our problems that we had to solve internally, the transformation that we had to do, how we built that, the outcomes about that, and then because we are at a data and AI
[06:19] conference, we would be remiss to not also include what it looks like for our company for AI in the future. So the problem we set out to solve, I mentioned earlier that we had a desire
[06:31] to work towards showing care and that understanding. So we wanted to make sure that we were creating a a digital layer to bring this care across that could be valuable to our business but also
[06:43] valuable to our customer and also valuable to each one of us within the organization that was working on this work. So we had to first take a look at our what our current state was. How are we showing up for our customers? If a
[06:55] known customer engaged with us because they got an email from us or they opened our app or they navigated to our website, what was that experience? Was it a connected experience? Maybe, maybe not. You click the next
[07:09] one. Spoiler alert, the answer is no. Sorry. uh we were very much still taking a very siloed approach as you can see on the screen um for our communications the campaigns that we were doing and the
[07:22] priorities uh for our particular customers. So our tools were disconnected we had disjointed experiences we had a lot of gaps and we needed to uncover those gaps. So we had to get really real internally of like
[07:36] what does that actually look like not just from a customer standpoint but across our teams. Was it within our data? Was it within our tools? Was it within our technology that made up our tech stack? You all know because you all
[07:48] are working for companies where all of these different things are at play. Or was it our own internal processes that we were using to get this work done? Another spoiler alert. It was all of the above.
[08:04] So, we started to notice that our CRM channels were not as connected as we wanted them to be. This diagram is not actual architecture, but you all can see and have probably felt this painoint in your own organizations. That's why
[08:16] you're at conferences like this. Um, we had disconnected data sets, our siloed data, and we always had a need for engineers to get more data. Not saying that we don't love our engineers that we work with in within the organization,
[08:29] but we all have our day-to-days, and we don't want to add to their plates. Just saying. So as we were connecting our CRM channels, we quickly realized that the same data and connectivity patterns could also solve for like other digital
[08:42] channels that we had within the organization and even within the restaurant. So there were things that we were doing within the app, within our email space that could also help our team members within the restaurant. So we were really trying to bring that
[08:54] connectivity together. But the main thing that we saw like when we really broke this down was like to send a campaign someone had to do so much. There was manual list pulling. There was uploading it to our sending platform,
[09:08] scheduling it. That was very slow, one time and very imprecise on the timing. So understanding where we were trying to go on this care journey, we needed to be
[09:21] open to shifts because we knew that there were gaps. How do we fill those gaps? We knew that they were not small shifts. Our company could be seen as very big to a lot, but it could also be
[09:33] seen as very small and truly there's a lot of collaboration that wasn't necessarily going on on the inside um because people had their various bodies of work. Um so we narrowed this down because in true marketing fashion, we
[09:46] like to get to like three, keep it to three bullets um to our strategy, our operations, and our technologies and the three key places that those models were breaking down for us.
[09:59] So, as I said, these these were working broken down in three specific places for us. Strategically, we were working with incredibly rigid platforms um and our processes. And from a marketing
[10:11] perspective, what I will tell you is you see all of this lovely like our limited time offer promotions that come out. Just know that we are very much ahead from a marketing standpoint when we are
[10:23] briefed on these things. So, I probably know what's coming all the way in 2027. I'm not going to confirm or deny that. But that just gives you like an example of how rigid our structures are because we know what's coming so far in advance.
[10:36] And you would normally think, Sharma, that's actually a good thing. Is it though? Especially when you have a lot of priorities going on in the organization. Um, and from a technology standpoint, our teams um have their own
[10:49] predefined bodies of work. So you all are familiar with like quarterly plannings and sprints and all of these things. So you're thinking that on the marketing side, we have all of these bodies of work that we know that we have
[11:01] to promote coming up. Our technology and our data teams are like, "Yeah, but we also have not only these things that we have to promote, but there are other places of the business that we have to support. There are other things that may not be campaigns that we need to
[11:14] support." All of these things are on our road maps. and what they were doing was converging upon each other and causing a lot of swirl within our organization. Um, so yeah, there we go. Sorry, my bad.
[11:27] So once we were on the same page about those different problems across the marketing and the technology side, um, the other problem that we saw was there was a slowdown of our own processes to
[11:39] get those communications out. Um, previously we depended on heavy engineering support like I told you. Um, so we would have scenarios like you actually see on the screen. So the marketer, me, uh, would have to file a
[11:54] ticket or submit a brief depending on the channel and or the campaign. If it was our inapp work, a lot of that was like filing a ticket with our internal engineers. If it was on the email side, we did have an agency that we used to work with. We've previously brought that
[12:07] work inhouse, but still the same things. we have an internal creative agency that we also have to brief on a lot of this work. Um, and then this was passed over to our engineers or our agencies to build again depending on like which
[12:22] particular channel this was actually working on. And then once all the QA and approvals were done, this work would then launch or ship. The main thing here is like once it got to the end of the process, it was locked and it was as if someone threw away the key.
[12:40] And then the second place that we saw a breakdown was in our operations. We've approached operations in various ways across our organization. So we've been balancing the needs of our business and our tech stack schools tool tech stack
[12:52] tools with the skill sets of our team members. This has evolved over time. I know when I first started within the organization, the contractor that was supporting me had a different skill set than the contractors that currently
[13:04] support me. But we've continued every year to evaluate what those needed skill sets are based on our tools and who needed to be working with them.
[13:17] Now, I will say that these workarounds meant that we had more chances of stitching things together as we were going along, which is not always a bad thing, but what this did do was it introduced a little bit more hidden costs. So, as as I mentioned, like our
[13:30] contractors and the scaling up levels, that's a hidden cost because we may have had to like remove a specific contractor because their skill sets weren't actually what we needed going into the next future and iteration of the work
[13:43] that we were doing. Um, and then also from the previous slide how I mentioned the marketing ideas are aligned on well in advance. So the bodies of work for our technology teams and our data teams,
[13:56] everything was still like very much disconnected. So we ran into different teams basically cruising along in their different fragmented silos and having to come to a screeching halt at an intersection of their work. Uh marketing
[14:08] would present the ideas, engineering teams would have to determine if the data was available and many times the needed data was in like multiple destinations. And what I will say going back to earlier when we were talking about care that's not only not caring
[14:22] like in the internal teams but it wasn't caring in the output that was getting out to the customer. >> Now the third place now the third place where it breaks down was the tooling.
[14:35] The tooling was really capping the ambition of our marketers. Uh marketers would have great ideas and our tooling just couldn't do it. whether it was our data, whether it was how our data was synced to platforms, it was just it was
[14:51] really capping what they can actually accomplish. Not just accomplish, but like dream. Um, so that was the the third area. And one of the things that we came from
[15:03] all those three was we had this theme that we all rallied around. We want to put marketing back in the marketer's hands. wasn't ready for the spirit fingers. Um, but we wanted to put marketing back in
[15:15] the marketer's hands. And there's a lot that goes into that one simple statement. If you think about it, they're asking to get rid of our engineering team.
[15:27] >> No, but if you really think about it, uh, they're they're asking for a lot. They're they're asking they need data centralized. They need connected tooling. They need tooling connected not
[15:39] just to the data but also connected to all the platforms they engage our customers with. >> Sounds like you're saying we're needy. maybe
[15:53] we work for you.
[16:07] uh for data bricks which news to us um but but we we went down the path originally said well CDP can probably solve your problem but with traditional CDPs we we ran into a lot of Well, how
[16:23] fast do you need the data? And and al although over time traditional CD SAS CDPs helped us move data faster over time, it was still saying that we needed to move a bunch of data move a bunch of data around and and really facilitate
[16:37] and and get the rest of the organization to operate in that new CDP which was a challenge. Um our team is the team that helps our marketing team but our team is not the rest of the
[16:49] organization and influencing the rest of our organization. please adopt this new CDP just for our marketing efforts. So what we chose and we landed with was a composable architecture as some of you
[17:01] may be aware of. Um but with the composable architecture we can still collect app events. We can still collect um purchase events. We can still collect events that are happening in our restaurant and those teams can be
[17:14] producing data the way they traditionally produce data. they don't have to worry about distributing it um to this new platform in this new format. Um and then we unify all that data in data bricks and we leverage uh hightouch
[17:28] as well to unify that data and build our customer profiles and then we have our marketers in that unity section activating on it. So they activated across uh various platforms whether it's
[17:40] our CRM platforms whether it's our apps or whether it's even restaurant experiences where we're providing that data to present personalized uh messages to customers. So this is kind of how we
[17:52] look at it breaking it down. Um you know data team marketing team well >> we don't like different stuff >> data team marketing team
[18:05] focus really is on uh the data and structuring the data, formatting the data, providing that golden record uh to our marketers, make sure it's reliable, make sure they know when there's outages
[18:18] uh or uh some data stream is down or some something's going on that that is unexpected. And then our marketing team is actually just using the data similar to what you saw in um customer lake. Um,
[18:31] we're doing that similar in hightouch where our marketers are just using the data that we prepped for them for their campaigns, building their journeys. And so, one of the things that's really
[18:44] important is effective and efficient collaboration. So, as you can tell, we we feed off of each other. Um so our data team and our our active our our architectural team we really focus on
[18:57] the data in data bricks and and setting up hightouch uh appropriately but then we don't just pass it off to our marketing team. We actually sit next to each other. >> Yes.
[19:09] >> We don't sit in a basement. We don't work. We physically sit together and we're constantly collaborating on these are the types of campaigns. These are the types of things we're doing. Although the tools are in the marketer's
[19:22] hands, there's still data work for us to still prep some of that data to make it easier and more efficient uh for Chararma's team. So, how did we build it? So, we we as
[19:36] you saw um with the composable CDP, we a centralized customer view paired with a single decisioning layers create consistent experiences across various channels that we connect to. It's really
[19:49] important that single decisioning layer because there's one customer and there's one decision to be made for that customer at any particular time and that that was really key to it. Whereas in the multi- channelannel space where we
[20:02] were operating that chararma said about the disconnected silos we had a lot of things that were interacting with the customer. Um our CRM team was marketing to our customer.
[20:14] our local uh CRM team, our local operators, our marketing to our customers. We even had our apps running engagements. They were all coordinated, but they were coordinated through people, not through data.
[20:27] >> Which brings back this. >> So those are the experiences that were once coordinated. >> Yeah. >> But through a lot of meetings and
[20:39] effort. So hight touch really became oh excuse me so hight touch really became our air traffic controller um from the decisioning layer and I'll get a little bit more around this um but that was
[20:53] really that unified decisioning layer making decisions for us um how Chararma's team can operate it so this is kind of where we're seeing it this is not our full architecture so this is
[21:05] kind of storytelling I see cameras coming out but this is more for storytelling um but this is kind of how we think about it. Um, so when we think about decisioning, are we decision making a decision for a customer for a
[21:17] loyalty reward and should we send them an email? Have they received too many emails? All that decisioning is being made from one centralized layer and all that customer data is sitting in our database or in our data bricks uh
[21:30] instance with our customer profiles. But that data is also looped around in real time. So our way of looking at all of this in customer engagements and Chararma will touch on one of the experiences that we built with this is
[21:43] that data should be val available in real time not your 24 hours data latency shouldn't be an issue for us marketers need marketing in their hands. So here's another way to look at it like
[21:56] data bricks is being the foundation hightouch is helping us orchestrate the journeys real-time segments identify identity resolution data activation and hightouch is then
[22:11] connecting us to our various channels um where
[22:23] customer should receive this on this channel. >> All right, >> back to you, my friend. >> Yes. So, we've talked to you about our problems that we had to solve uh within
[22:35] our organization and how we really worked towards getting everybody rallied on the same team, um around the tools, around the products, like all of that. And so, what I actually want to talk to you now about is some of the examples
[22:48] that we brought to life because of all of those collaborations. So, what you see here, it's nothing groundbreaking. Like, you can take pictures all you want. It's all it's all fun, games, all of that great stuff. But
[23:02] all we're showing here on this slide is really one of our breakfast messages and a tier message and our birthday message. Those are table stakes for what we're trying to do from a marketing standpoint across our various organizations where we have data on the customers and can
[23:15] really truly tell those stories and be very timely and relevant um into their particular lives. Um, so the beauty of this is really just like Dan said, it was putting marketing back in the marketer's hands. These types of
[23:28] communications actually used to take us a really long time to get out the door. And you're probably like, "Really, Chick-fil-A? Y'all have that data?" And it took you forever to get things out the door. 100%. Just as the person that
[23:40] is like pressing send, yes, it definitely did. Um, but now my team can activate on these these sins. Um, with everything that Dan Dan was just talking about, the work that his team has done, the work that our other technology
[23:52] partners and our data partners have unlocked for us, my team can actually unlock these things and get them in a more automated state that is easier so we can reach those customers in that right time in that right moment and
[24:04] getting towards in that right channel. >> Yeah. And one of the things just to add here is uh when we first started my team understood every campaign when they were going live. I can't tell you when the
[24:16] last campaign went live. Sharma's team handles it all. >> I'm not hiding it from him. I'm just just want to
[24:28] iterations like this. Spotify started it with their Spotify wrapped. Other companies have done these over the years and prior to we used to do these and it was just specific to our email channel. But we had another conversation because
[24:40] we do a lot of discovery and a lot of talking and a lot of collaboration internally about what this could look like if we brought this to other channels. What if the app wanted to get in on the fun? What if we created this experience specifically within the app?
[24:53] Do we have the data points? Are they real time? or they at least to the frequency that we needed them to be where we can have this within you open the app, you go to our for you section and this experience showed up for you at
[25:05] the end of the year. Um so we tested that out. It went beautifully. Um and you'll see the stats here of like this was number two for like placements um clicks for 2025. The only thing that
[25:18] beat it was when the cows took over. If you have been a fan of Chick-fil-A for a couple of years, the cows take over during the summer. Um, and so this was the only thing that actually beat this particular experience. Um, we had an
[25:31] increase in incremental revenue that we saw from this. That is one of our success metrics that we're kind of looking through. Um, and then this was also I can't see the other one, but yeah, great stats.
[25:43] >> Number two for campaigns for orders. >> There we go. There we go. Um you see the stats here on the screen but what is not showing is the collaboration that happened internally for this. Um we
[25:55] actually had we found out this from a marketing standpoint. Our engineers get an opportunity to bid on work that they do with our marketing teams and this was one of the top bidded um bodies of work
[26:08] uh that our engineers bid on because they were so excited to bring this to life. um we were able to do it within the for you section of our app. Email was also still a part of it but not this personalized and push also came into
[26:22] play um based on whether you had engaged with the email or not. But the main thing here was the data and the tooling that everything that Dan talked about that this unlocked for this particular experience for us.
[26:39] And so because we had so much fun working together at the end of the year last year, we said why not run it back? Um so in February also if you have been paying attention to our brand and you don't have to, it's fine. Um in February we did the first national giveaway that
[26:53] our company has done in I believe seven years. It's been some time. I wasn't at the organization at the time when it when we did it. Um, but everything that Dan talked about, all of that tooling, all of that work that we had to like
[27:06] strip down and rebuild on the back end helped us be able to do um, one of our national giveaways, which was wildly su successful. Um, it went across channels. Again, what you're seeing there is
[27:19] actually one of our in restaurant placements, but you'll also see the email and the push notification that went out as a part of that. So, we've had great success uh when we kind of pulled all of this stuff together. >> And one thing on this campaign that uh
[27:32] we didn't have before with the unified data was access to loyalty data in real time. >> Yep. >> We we would batch loyalty data before in the past and make it available to Chararma's team. This is probably prior
[27:46] to my time too, but it would be batched and it would be available in 24 hours. So, knowing how many customers actually used the reward, didn't use the reward, and then actually we had a follow-up campaign to the actual giveaway that was
[28:01] um what would you guys call it exactly? >> It was an extension. >> Second chance extension. So, we had a second chance extension that actually went out to those uh signature and >> um red members. >> Yeah. So, those are our top two
[28:14] >> our top two tier members that um but we needed that real-time data to tell us who are going to get those new rewards um and who should we remind that the reward is expiring. So, we needed the
[28:26] real-time data to constantly be looping around and central centralized. Before you move from that, um, another thing that I do want to point out that was so key here, this was a breakfast giveaway. And our breakfast time window is between
[28:39] 6:00 and 10:30 a.m. So, if you if you're like familiar with how fast data moves and how sins move. The work that Dan talked about that we unlocked helped us
[28:51] to be able to make sure that the sends went through within that time window. prior to we would still be making a send like that throughout the day and we would have to be okay with the fact that like somebody may have gotten a
[29:03] breakfast email at like 10 o'clock at night and oops you can't go ahead and order your breakfast. So this was like it was really good to unlock a lot of that work for us.
[29:15] So being at a data and AI conference I would be remiss if I did not talk about what we have on our road map. One of the things that we have planned coming up is just like, okay, you've moved from multi- channelannel to omni channel. How
[29:27] are you guys going to incorporate AI? So, I will tell you before I get fully into this slide, we recognize, because this is a disclaimer, we recognize that
[29:39] a lot of our stuff has not been perfected. You could still see disconnected experiences from us. So, I don't want anyone to come up to me later and just like call that out. We know we're very self-aware. just telling you
[29:51] um but this is the journey in which we're trying to go on as we kind of get towards more um utilizing of AI. It is like in a in a like pulling a push
[30:03] pushing a rock up a hill but it's not that painful. I will tell you just just just want to stress that um we are doing different changes within our structure because the way in which our creative
[30:16] was currently doing is like our creative team would fully design these emails and as I mentioned before very rigid structures. We knew like months in advance what needed to come out. So they would design them just fully for that.
[30:28] Now what we're doing is working with them on like give me the pieces. I'm a huge fan of Lego. I like to get my Lego pieces and like build the things together and that's what we've worked with them on. Just give us that in a modular template type of style and we
[30:42] can work with that. Give me the give me the creative, give me the copy. We work with that from a content standpoint. Same situation with offers. We know that everybody um needs a different offer. That's not something that we can kind of
[30:54] go from a blanket standpoint. we really want to go from a one to all type of thing to really getting closer to that one to one or hey even if you give me like one to few right now I'm I'm
[31:06] totally fine with that. Um, and then with the channels we have, as I mentioned before, email, push, SMS, we have our inapp. Like there are different channels that we can activate, but we want to really activate those things
[31:19] together in that omni channel way that we were talking about. But not only that, we want to make sure that we're getting it to you in the right channel because if we see we're sending all of these emails to you and you're not engaging with them, but we see that you're actually engaging with push,
[31:32] let's figure out how to make sure that that's coming to you in that right channel. And then from a timing and frequency standpoint, we've been testing things um like our time zones, best time
[31:44] to send, like all of these various things, but really how are we seeing that these these times and the frequency in which they're getting them are impacting whether they're coming and making that sale for us. Um and then our
[31:56] organizational structure, I've talked about it a lot. We have changed the way my team is structured. Dan has like worked on like his team structure because we're really trying to make sure that we're working together in a cohesive way to bring a lot of this work
[32:09] to life and I would also say the same thing within our ops model. Um so yeah it's it's a lot of transformations. >> Yeah >> a lot of change. >> Multi channel brought us some omni channel transformation but AI brings a
[32:22] separate transformation as a whole. >> Um once once teams get comfortable with omni channel which is still a little bit uncomfortable in our space. um once they
[32:34] get comfortable with omni channel here comes AI out of nowhere and all the whole blows up a little bit of the operations from omni channel. So we're working through the transformations and how AI will help us uh assist Chararma
[32:49] in in building those onetoone experiences. >> Yeah. And we also like to say that we um we're climbing with care and confidence. And I will say that it's something that we truly do believe within our organization because if we do this
[33:02] incorrectly, y'all have been like talking about AI and you've seen what it does for other companies. There is a way where you can use too much AI and it becomes a sea of sameness where it's telling you the same triggers each and
[33:14] every time. So everybody's getting like a blanketed message that is just not right for them or it's way too onetoone and you're missing out on your fringe customers. Um because if you're setting up things, you're trying to get that
[33:26] next purchase based off of purchasers. What are you giving to the people that haven't purchased? What are you giving to the people that haven't engaged with your stuff? So we are really trying to make sure that we are climbing with care and confidence and only in the unique
[33:38] way that we can do that from a Chick-fil-A standpoint. And so here what you're seeing is like the traditional marketing that we've talked about like you can segment your customers. You craft those experience
[33:51] with the rules and journeys. You analyze the outcome. It's like rinse and repeat each and every time. And what we're trying to work towards is getting to a place where instead of us crafting those experiment experiment rules and
[34:03] journeys, we actually want to flip that on its head and define those outcomes. We know from our brand strategy team like what our particular goals are for our campaigns. But how can we drill that down for our particular campaigns? How
[34:15] can we actually drill that down at a channel level? How can we drill that down at a local level to where we know what those goals are upfront and then are allowing the tooling to optimize for what that next action needs to be, what
[34:28] that next message needs to be for the customer to achieve that. And I will also say I'd be remiss to say like we're still going to have humans included in that. I have to say that because if our legal team gets a hold of that, I need
[34:41] to make sure I continue to say that humans will still be involved. I am not going to have an agent that I'm just going to name and it's going to be a part of my team. I will also be there as well. So >> yeah. Well, one thing I will add is that
[34:54] uh because we unified data through data bricks and unified our decisioning through hightouch I if if we were in this AI space and we were still operating in a multi- channelannel way it'd be really hard to
[35:07] add a bunch of agents because all those agents and those different business units would be going after the customer potentially not aware of other uh agents that are going after the customer leading to multiple touch points.
[35:21] customer frustration. Um, so what it really has done, not just from a technology standpoint, it's got teams to think about the data first, uh, rather than the experience first and has
[35:33] teams working closer and closer together around what is the experience we want to craft, not just in the individual channel that I'm building, but how does my individual channel affect your channel? By unifying the data, it
[35:47] actually brought collaboration unintentionally. I mean, you know, it's fun. So, that is it. We have gone through our problems, what we had to do internally
[35:59] to kind of shift with those, the things that we've done, the the things that Dan's team has unlocked, um, the things that like examples that I've shown. And so, we really just want to tell y'all like thank you for sitting in on our
[36:11] session, not leaving because you were part of the coding session. And it was our great pleasure to present to you all today. >>

Learn more about the Databricks Data and AI platform.

The information provided herein is for general informational purposes only and may not reflect the most current product capabilities or configurations.