Remember photographs? Like, pieces of paper with images you could actually hold in your hand?
They were great—until you tried to organize ‘em. Over the years, you’d collect boxes and boxes of photos and often totally forget where they came from. Was this our ski trip to Whistler or the Great Snowstorm of ‘82? Is this Timmy’s baptism or Grandpa in the tub?
Then along came digital cameras and smartphones. Google Photos and Apple Photos. Automatic organization. Meta-data.
Marketing data can sometimes feel like those old boxes of photos: a confusing assortment of numbers, trends, and consumer insights all jumbled together. And messy data is a big problem if you wanna make full use of AI marketing tools.
On this episode of Unprompted, hosts Pete Housley and Faye Campbell chat with Breedon Grauer, CEO of Fastloop, about how marketers can organize and structure their data to make sure they’re getting the most outta their smart, shiny new tools.
This data-heavy episode covers topics like:
- Why consolidating your data and making it accessible is essential to empowering your entire business
- The importance of establishing a concrete data foundation before trying to accelerate your business with AI
- How you can combine out-of-the-box AI solutions with your proprietary data to do things your competitors can’t
So buckle up, nerds, and listen to the episode to learn how numbers and robots can take your marketing to the next level.
Episode 4: AI for Data Geeks
[00:00:00] Pete: Hey, marketers! Welcome to Unprompted, a podcast about AI marketing and you. I’m Pete Housley, CMO of Unbounce. Unbounce is the landing page builder with smart features that drives superior conversion rates.
This is our fourth episode and we are starting to get deep into things. Today, we’re going to geek out on data and AI for marketers and IT folks. We’re gonna talk a little bit about where companies are on their data maturity curve, and what that even means, and hopefully anchor that back to, uh, some good thought leadership for our marketing audience today.
Today, my co-host is Faye Campbell. Faye is our Director of Revenue Operations at Unbounce and leads the Revenue Operations team. Revenue Ops is responsible for three pillars: data, marketing technology tools, and processes. Faye is the first person who will tell you that these three are intrinsically linked and your tooling data are incredibly important to the outputs. As we say, garbage in and garbage out.
Faye is one of those individuals in a company who literally has the keys to every corner of the business. Her team is thinking three steps ahead of every new program, campaign initiative, and asking: “How can we measure this? How can we optimize it? And how do we know it’s working?”
Faye, we know that marketers often need it or data teams to properly shape scope and build their capabilities and integrations. Tell us a little bit about the role that you play in bridging those two departments.
[00:02:00] Faye: Yeah, absolutely.
I think with revenue operations—whether you’re calling it revenue operations or marketing operations, I think there’s different flavors of that that you’ll see in the industry—um, I really like to think of us as a bit of a translation layer, between the business side and the tech side. And so of course, we are very lucky to be supported here at Unbounce by a team of, you know, analytics engineers, data engineers, data analysts as well.
But then on the business side, you know, I’m working with folks like you, people on your team, to help translate, you know, business requirements, priorities, vision, and initiatives into structured usable data, putting together measurement plans that are really gonna provide business value long term. And part of that, thinking three steps ahead comes down to, you know, we’re executing at this stage, but what are the questions we’re gonna be asking—three, six, twelve months from now.
[00:02:51] Pete: I think Revenue Operations is a really interesting role. And I come from a number of marketing companies where the marketing team had a general idea of what they needed from the data folks. The data folks didn’t necessarily understand the martech world, and we had to do a little bit of bridging, to get, you know, further along on that data maturity curve.
I’m super excited—today our guest, uh, his name is Breedon Grauer, and I’ll introduce him a little later—but Breedon has a very sweet spot in the industry where he helps companies bridge their technology and their marketing gaps. And I think he’s gonna be a star witness in our episode today.
But Faye, tell me a little bit about what excites you and what’s going on in martech these days.
[00:03:43] Faye: Yeah, absolutely. I think in the world of martech and marketing operations these days, looking at where we are now, data accessibility has just never been better. I think the level of data democratization, um, and accessibility that’s available to marketers today—you know, without having to write a legacy sequel, without having to unpack, you know, a data warehouse that you didn’t structure and don’t understand. I think the amount of data and the style of data available to marketers today has, you know, never been better.
[00:04:15] Pete: Yeah. I mean, that’s unbelievable. And I, I come from a world where we used to, you know, my interview questions for marketers, like, “Can you do advanced pivot tables? Can you do relational tables? Can you do VLOOKUPs? Can you, you know, all of those things.” And now so many of the tools really just use drag and drops to essentially, shape data, feed your martech stacks, create your lists, for CRM initiatives, and so on.
So, yeah, I agree with you. The world, has changed. So, the world has changed. What do you think is a marketer’s biggest challenge right now?
[00:04:51] Faye: I mean, honestly, I think it’s the dark side of the same coin where there’s just so much choice. There’s so many tools out there.
I see a lot of, you know, tech stacks these days, It’s an octopus of technology where you’ve got all these tentacles, all these different siloed tools. Sometimes that data is siloed, sometimes they’re not structured the same way. And so you run into, you know, a tricky problem where marketers, especially, you know, it’s 2023, the environment we’re in today, are being asked to kind of radically prioritize what their tech stack looks like and what’s in there.
You know, I, I think for marketers today when they’re thinking about implementing and selecting a tech stack, doing so with a thoughtful approach to, “What does this enable us to do, what data becomes available when we do it this way? How should these systems talk to each other? Do they have, you know, a similar data model?”
I think all of these considerations can help people unpack what can be a huge mess really, if, if not handled thoughtfully and correctly.
[00:05:52] Pete: Well, I mean, touche on that. When I was at Indochino, I think we had as many as 30 marketing software vendors, and you just get, you know, tool creep and you don’t necessarily get the synergies. And what I’d like to suggest is that marketers think about the full customer journey: you know, from pre-collect to customer acquisition, to, you know, customer lifecycle marketing, and make sure that the tools are complementary. And hopefully, and we’ll get to Breedon in a few minutes, but living off of one universal source of truth, in terms of the data, that they, rely on.
Alright, so today’s episode, we’re calling AI for Data Geeks. And we’ll be exploring that sort of crucial aspect of modern marketing where we do need to shape data and have that universal source of truth ready such that it can fuel AI technologies. And as marketers embrace AI and its potential, the quality and structure of their data clearly are gonna play a, uh, a pivotal role.
So, you know, Faye, as we think about marketing automation now, AI streamlines and automates various marketing tasks: email marketing, lead nurturing, lead scoring, campaign management, and so on. Just tell us a little bit about your wheelhouse as it pertains to automation.
[00:07:20] Faye: Yeah, absolutely. And, and so I think for us as a Revenue Operations team here at Unbounce, you know, we maintain these systems that enable our Revenue units. That’s Marketing, Sales, Customer Success. And when I’m asked about the opportunities for AI in that space—I’m not gonna lie, you know, we are still just dipping our feet kind of into the opportunities available to us through AI.
I think part of that comes from, you know, as operators, we are constantly thinking about governance, maintainability, transparency. Can our stakeholder teams trust the systems that we’re putting out into the world? And so kind of like, with those concerns and, and thoughts in mind for me, automation is, one of the most exciting frontiers for implementing AI, especially as a starting point and getting our feet wet, because, you know, you have this opportunity to have an almost semi-supervised system or a back office system that’s still being passed through some, some governance or, or checks and balances.
And as a place for us to get started, I see opportunities in, lead scoring, sorting, categorization at scale, data hygiene exercises, working with unstructured qualitative data that is historically a nightmare to try and automate off of. And so trying to find these opportunities to supercharge our tech stack without bypassing some of the, the governance gates that we like things to go through, whether that’s marketers with constraints on, you know, lists and, and logic gates.
[00:08:48] Pete: It’s really interesting—you said something like, “we’re not that far along on the AI curve in terms of our martech stack.” And I think this is probably where all marketers are right now.
We’re faced with this tsunami of AI information and we’re dabbling out there in tools—you know, ChatGPT, or Smart Copy, or Midjourney, or DALL-E or you know—but some of these automization, tools are, you know, definitely your new age. And I think the fact that we’re even having this conversation is great because we are embracing it.
And so let’s get a little more specific. Let’s talk a little bit about personalization and marketing. And of course AI enables, you know, marketing to create these personalized, one-to-one, experiences. And of course, we’re gonna need customer preferences, behavioral patterns, demographics to deliver, these highly targeted messages and offers.
[00:09:49] Faye: From my perspective, you know, personalization has been the gold standard for, for such a long time now—but from an operations perspective, it’s, it’s always been a little bit challenging or scary to try and implement because you find yourself in this position of, you know, “How do I categorize my customers into different buckets? How are we gonna define those buckets? You know, what’s the data structure that we can even collect this information and, categorize these folks?”
And so the upfront, you know, labor to sort and organize and predict—maybe that’s not the right word, to, uh—anticipate, you know, it was very manual where we get in a room and we say, “Hey, let’s name nine customer buckets,” and, and try to personalize to those. And I think the opportunities around AI and, you know, generative models is to… well, I mean, I’m curious. I guess I, I don’t know. I’m not the authority.
[00:10:40] Pete: Well, I mean, I think in, if I just take marketing attribution circles alone, you know, we’ve often looked trailing, and we’ve looked at completed journeys, and we’ve looked at the click paths, and the fractional attribution, et cetera. But what’s more relevant is those incompleted journeys and using predictive means and AI to say these are the next two touchpoints that you would need to get this customer over the line and be a conversion.
So we’ve established the fact that, as you know, at Unbounce and in marketing, we’ve got a little bit of way to go on our martech to fully embrace AI, and we will go on that journey. But I want to flip for a moment to the other side of our house at Unbounce, which is our product, uh, and technology team.
And of course, Unbounce is a landing page, you know, builder company, and we’re the global leader in that. But for three years we’ve endeavored on an R&D project and we have a machine learning team of, I don’t know, eight or ten. And based on two billion historic conversions of landing pages that Unbounce, we’re able to shape all of that data to create an AI landing page that gives you copy, layout, blocking, CTAs, recommended length, color—all optimized for conversion. So we’re pretty excited that we’re at that place. But make no mistake, that was three years of preparing the data and, you know, developing, the AI use cases to come out, with this.
So, Faye and I have been geeking out for a few minutes here, but we’ve got even bigger geek with us today on the show. And, um, it is, my great pleasure today to introduce, you know, a longtime colleague and to be honest, a good friend, Breedon Grauer. And Breedon is one of the more sophisticated and accomplished marketers I’ve met in recent years. And Breedon is CEO of what I’m calling a “new age agency” that literally focuses on the exact topics that we’re discussing today.
So Breedon is CEO of a company called Fastloop. And Fastloop is fusing business strategy with technology solutions by literally using cutting-edge technology and operational thinking—Faye, playing rate into your operational mind. Fastloop helps build organization’s vision for the future and enables them to reach their potential.
About Breedon: He leads all aspects of operational strategy at Fastloop, including, uh, growth and execution. Prior to Fastloop, Breedon was a member of the senior leadership team at the Jim Pattison group of companies. Literally a $15 billion company, 50,000 employees, 300 companies across a hundred countries, where Breedon led a number of growth initiatives, including global marketing, intercompany sales, procurement, and technology.
Hello, Breedon, and welcome to Unprompted.
[00:13:49] Breedon: Thanks for having me, Pete.
[00:13:50] Pete: Tell us a little bit about yourself, my friend.
[00:13:52] Breedon: Well, I’m, I’m Vancouver, born and raised, passionate data geek as, as you put it. Thank you for, uh, saying that. The changes in the space that are happening right now, there’s no better time to be in business and, and to be in technology.
[00:14:06] Pete: All right, so, you know, as I mentioned, I see Fastloop really as a, a new age agency. You build strategy, you structure and shape data. You lead complete and total digital transformations, even at the enterprise level. Tell us a little bit about the opportunity that you saw, in the market and about Fastloop’s, you know, point of difference.
[00:14:28] Breedon: Yeah, so we launched Fastloop in March of 2020, so right before the pandemic. At the time, a couple of interesting things were happening. There was a mountain of data coming at businesses, and as digital transformation sped up, new systems were implemented that were creating more and more data. And what we saw was that IT teams weren’t equipped to harness that data and actually create business value.
At the same time, you had Amazon, Google, and Microsoft investing billions and billions of dollars into cloud infrastructure that can essentially take unlimited amounts of data and crunch it to do what you needed. And so as those two worlds were colliding, we thought there was a really big opportunity to help organizations large and small, really bring their data together and combine it and actually create something tangible out of. And so what I mean by “tangible” is revenue, profit, and operational excellence.
And so we take a lot of time not only focusing on the technical complexity of, how do you organize and structure and bring in your data to one location, but how do you actually use it to drive business results that your CFO would be proud of?
[00:15:38] Pete: You know, it’s interesting, Braden, we do a lot of research and we have, I dunno, 17,000 clients. But we’re always trying to find what is important to marketers, what is important in CRO land, and so on. And I can tell you that in June of 2023, revenue optimization is what is on every marketer’s mind.
And I think—you and I have talked about this a lot—really the role of marketing has really shifted to the role of growth in companies and very, very, tied to revenue. And certainly, you know, I would advocate, something called ROMI, “return on marketing investment,” so that any investment that we put out there, we are measuring that return judiciously. And, you know, it is only through data.
So, okay. So I get the idea of how you’re shaping data, driving revenue. What is your business, like, focused on? So tell me a little more specifically now on the focus here.
[00:16:39] Breedon: So most companies have data in dozens of systems. And so as standalone systems, that data is not necessarily creating that much value. However, when you combine that data together, the insights you can glean off of that combined information can be extremely powerful.
Now, how you get data out of a number of different systems can be very different and very challenging. So we do a lot of data engineering work, bringing data into cloud data lakes and cloud data warehouses.
Once we organize and structure data for companies, we do three things. We build out analytics across the whole business. So, I think Faye, you mentioned the democratization of information. So how do you enable everyone in the organization, from the CEO down to the front office clerk, give them the information they need to make better, faster decisions on a daily basis that can directly speed up the performance of your organization. You can also use that data to drive personalization through marketing, through loyalty, through a number of different activation channels in the way you serve your customer. And you can also build data science off of that, which is AI and machine learning. And so we’re a pure play data player that covers the gambit of data in a company across engineering that data, organizing the data, and then leveraging it.
I think one of the big, opportunities we saw was a sort of mid-market type of firm who could quickly and effectively create business value from data. We’ve seen long tail heavy cost implementations that sometimes fail, but if you understand what you’re doing in the business outcomes you’re driving towards, you can really quickly and effectively create a lot of impact in the market, and that’s what we’re set out to do.
Last thing I’ll say is one of the most fun and impactful areas is marketing. And so there’s a ton happening in marketing. You mentioned all the martech stacks before. Companies are overloaded with marketing technology. You’ve got customer behaviors that are changing faster and faster every day. You’ve got diminishing brand loyalty, and you’ve got a lot of pressure on costs, where CFOs and CEOs don’t understand why CMOs are spending so much money. And so if you’re deploying that money, how do you prove—not only to your leadership, but to your customers—that you’re worthy of their business, the investments you’re making are meaningful, and that you’re truly driving the growth and success of that company?
[00:19:14] Pete: So good. And I want to pick up something you said and you talked about, you know, different systems with different pieces of data not necessarily talking to each other.
I have a question for Faye. So, at Unbounce we run two email service providers. So we have one on the acquisition path. So I can drive ad impressions, I can get, you know, an email surrender, I can do path to conversion marketing, which gets me to a customer. And then we jump over to a new email marketing system that does all the email marketing. But if you think about that complete buyer journey, or you know, customer life cycle marketing, we’re probably not linking what happened, you know, pre-purchase to what happened post-purchase to really optimize those paths.
So I just wonder if you had a, a thought on whether we should be consolidating that or whether it makes sense to have those worlds in different houses?
[00:20:05] Faye: I mean, I think it comes down to understanding your trade-offs between, you know, maybe we have multiple systems, and how do you choose the best tool for the job? When it comes down to whether it’s email tools or mass communications tools in-app messaging chat, you can get an all-in-one platform that does all those things, but is it gonna do all of those things really well? Or would you rather get the best tool for the job and then have a unified data ecosystem running in the background so that you can still get the benefits of desiloed information?
You know, I’ve been on the other side of things where someone says, “Well, hey, we’ve got Salesforce, and Salesforce is basically a data warehouse, so let’s just cram all of our information into the CRM and there we’ve solved our silo problem.” Call me in two years ‘cause I’m sure that’s gonna be a nightmare to try to untangle.
So it really comes down to, understand why you have the tool. What is it here for, what’s the value it brings, what is it best at doing? Lean into that. Let the tool be great at what it was designed to do, but then do the data democratization or the, the desiloing exercise or the, the modeling exercise somewhere else.
[00:21:11] Pete: You know, and we’re gonna talk a little bit about CDPs, later in the show, but I’m assuming that all of these disparate systems could just write an event to a CDP and you would be able to have that as a universal, you know, source of truth. Right. So we’ll get to that, in a minute.
So Breedon, you talked about, you know, multiple systems, and sewing them together. Can you give us like, a real-world example, like, clarify that for our audience in, in a, in a real-life example,
[00:21:36] Breedon: Yeah, I’ll give you a couple of fun ones we’re working on right now.
So, we work in the insurance space. We were with a large insurance provider who’s been around a long time, been very successful. And so they sell a number of insurance products: life, auto, health, farm, you name it. And so naturally, when you’re a business and you grow and you have success, you need to make technology decisions based on the need of the organization—not necessarily the need of the future of the data potential. And so what happens is you have dozens of systems that have siloed data, that have pieces of information on your customers.
And so what we did with this insurance company—and we’re sort of in the, in the end tails of it now—is we brought all of the source data together. So we had to get into their source systems, identify where customer information was stored, how do we bring that together into a data lake and data warehouse infrastructure. And what we did is we build out a customer data model where we attach all the different attributes of the customer across system to one master model so that we can really understand holistically what’s happening with the customer.
From that point on, you can really drive deep analytics of your customer base. What are they doing? How are things shifting? Where are they getting stuck? How do we drive it further? Further to that, you can then use that customer model to activate and communicate with those customers. The outcome to that is I’m actually creating more revenue per customer, and my customers are buying more insurance products. So always driving into tangible business outcomes. So that’s sort of the typical consolidation of the data.
Another one we work in is retail. And so if you think about retail, a lot of retailers are really good at serving customers when they walk through your doors. However, customers interact with your brand across a number of digital touchpoints that ultimately aren’t stitched together the same way as a store manager when you walk in with two feet. And so when you think about how you need to combine those touchpoints together, again, to unify the profiles of your customers. You can then use that customer model or that profile unification to give them better, more relevant, and engaging messaging. And so what we’ve seen with our retail clients is we can typically produce a 2% to 4% lift in revenue.
I’ll give you one more, which is, the media space. And so we work with a number of sort of traditional and digital media companies who want to create new revenue. And so this is what we call “data monetization.” And so this organization was having trouble getting more revenue with their clients, and they’re actually losing market share to these social media companies like, you, you know, TikTok and other ones where a lot of spend is starting to go. And what we realized is that we couldn’t really attribute the success of their advertising campaigns to the media dollars that their customers were spending.
And so what we’re able to do is take the data that they had—which was CRM data, campaign data, web click data, a number of different sources—combine it with second-party data through some partnerships. When we unified that information together, what was really interesting is we could put a machine learning AI model on top and start to actually predict the actions that customers were making both offline and online. So not only can you track digital, you can also track omnichannel, which is, in, in the more terrestrial model. And so in that scenario, that customer was able to drive new revenue from net new clients, create a tire new business department, and also get more money per customer.
So each of those you’ll see is foundationally laid on data. Each of them has a, a, a sort of, some similarities in the way you model and combine data. Each of them has a different application for the way you use it, but each of them needs to tie back to how it’s gonna drive revenue and growth for the organization.
[00:25:40] Pete: And so Breedon, you just talked, you know, three examples. Insurance, retail, and media. And in those cases with these clients, where did they fit on what we talked about earlier in terms of the data maturity curve?
[00:25:55] Breedon: Look, there’s a lot of hype around data analytics and AI, particularly in marketing tech. So we can get to that a bit later where everybody is announcing what they can do and, and how they’re amazing at everything. But I think also in organizations, people get caught with buzzwords.
The reality is for businesses that are maybe just starting, they can adopt the latest technologies and tools—and if they have a simple business, it can be pretty straightforward. Most of us don’t operate in that luxury, and so businesses have been around for a long time and they’re here for a long time. And anytime you have that scenario with the change in technology we’re seeing there’s gonna become a lot of problems.
The majority of the companies we work with are pretty far away from innovation in that curve, and they need a lot of foundational work to get ready and to get prepared for it. And the problem is a lot of them delay those decisions because of whatever reason. But because the world is accelerating, the longer you wait, the further behind you are. And so, it’s a journey we have to go on with a lot of our clients.
[00:27:04] Pete: I couldn’t agree more. And so it’s helpful to me to hear that you’re going on these journeys with clients and they’re not actually that far along on the data maturity curve, but yet they’re seeing the need to get along. And then of course they can expedite that with, you know, a firm like yours versus if they try to do it just with an in-house solution.
So I’m gonna ask you the same question that I asked Faye: What excites you about marketing today?
[00:27:29] Breedon: what excites me is how you can use data to supercharge your business. What some of those things include are audiences and segmentation. So once you get your data organized, how do you understand your customers in a completely different way? How do you organize them into cohorts based on different behaviors, different locations, different ways in which they move through their day-to-day activities?
Personalization is obviously a key one. The term’s been around for a long time, but it hasn’t really been to the point where it can truly get one-to-one. And I think now you’re getting to a place where you can really get sophisticated—not only with one-to-one communications, but how you scale that through, you know, potentially millions of customers.
I’m liking a lot of stuff that’s happening with measurement going on. But foundationally what is really exciting is: the heavy lift to prepare the data is the hard work, but the fun work that our organization does, and the reason it’s fun is because of what it enables after that. Now you hear about, you know, AI and you’ve got AI built into lots of marketing technology stacks and whatnot. But really the way you can leverage AI to drive a competitive differentiation as a customer, as a client, as a business is really exciting to me. And the reality is, you actually can’t do it through the hype that you hear in the market. The old adage of “garbage in, garbage out” has never been truer.
And so, if you don’t have your data foundation laid, you cannot accelerate your business with AI. You can do some generic use cases and sort of get yourself started and play around. That’s great. If you truly want to differentiate and create something proprietary that your competitors can’t match, you need to get your data in order. And so those are some of the areas that we stay really focused on and, and that we see, emerging in the space.
[00:29:28] Pete: All right. So I mean, I think our audience and everybody realizes that raw data is insufficient. It needs to have some shape and structure to it. And in this podcast, we’re talking about making sure that that data is structured and sound, and also can feed AI.
So Breedon very, very simply: Where do marketers and their IT counterparts even begin?
[00:29:55] Breedon: I try to keep it simple, Pete. So, the first thing is to focus on your customer. And I believe if you always focus on your customer and you treat them right, you will be a good steward of your business and what you… and the products and services that your business produces. And so the question is, what do you know about your customer today and what do you need to know to serve them better? AI can do a lot, but if you don’t truly understand your customer, you’re not gonna be able to take advantage of it.
So the first is, get a really deep understanding of your customer. And realistically, all data is—if you leverage it—is information. And so what information can you glean on your customer that you can really understand them more holistically? The second and most important is the foundation. And so how do you ingest that information? How do you combine it, so that you can actually harness it?
And so, I always think that things can become overwhelming, particularly in this space. And so if you focus on your customer and you focus on the foundational elements to simplify your complexity, that will get you started. Build a North Star vision for where you want to go. And then all you have to do is take the first two steps. Those first two steps will give you the light that drives you to the third step, to the fourth step. And it’ll advance you along—but don’t overcomplicate it.
Focus on the basics and just get moving towards and never stop looking at your North Star.
[00:31:25] Pete: All right, so we opened the show. Faye talked a lot about the different, you know, processes and martech tools that are available in a full marketing stack. We talked about getting your data ready, to be future proof so that we have a universal source of truth in one place, and that we can have the attributes, uh, about a customer so that we understand their behaviors, and their desires, and so on.
So, with all that Breedon, bridge us now a little further along in terms of areas of marketing that are driving growth or revenue, like we talked about a little earlier.
[00:32:02] Breedon: I would say, a little bit of audience segmentation, a little better measurement and attribution, channel activation and automation is becoming more important, and if you stitch those three together, you’re gonna get growth.
The key really is gaining a 360 degree view on your customer. And then using that information to really basically understand, how do you actually acquire a new customer? Like what are the basics towards a customer coming, touching your brand and any different touchpoint? And what are the triggers that get them there?
If you have those basics I talked about before, or those emerging areas tied together, you can then figure out how you acquire customers. Once you can leverage on that, you can start to figure out: How do I do it cheaper? How do I do it better? What am I willing to spend? How do you service and grow your existing customers into better customers? How do you build brand loyalty? How do you drive incentives? How do you create personalized offers? How do you eliminate and understand why customers are leaving? And so I think some of those emerging areas are great, but then you’ve gotta understand the core levers and triggers that are driving that growth and be able to continuously measure and tweak and continually adjust on the fly to make sure that it is starting to drive towards what you’re trying to accomplish.
I think the last thing I’ll say there is sort of more around attribution, which is, you know, a lot of people have CTA and CPM and all these acronyms we can use, but how do those actually lever up to effects that it is having on your customer? And how do the effects on your customer level up to the overall organizational strategy that you’re trying to drive around revenue and profit?
And so you need to be able to tie all those worlds together from a top-down approach—based on revenue to customer to metrics—and from a bottom-up approach—from metrics to customer to business results. And I think combining those two worlds, is, is is very impactful.
[00:34:10] Pete: Something that you said a few minutes ago was the “360-degree view of the customer.” Can you elaborate on what the heck a 360-degree view of the customer is?
[00:34:20] Breedon: So, in the various systems that an organization uses to manage their business—and this could be, you know, an ERP system even for business process, it could be a CRM, it could be your accounting and financial system, it could be your paid media channels, it could be your marketing stack—all of those systems have databases behind them. In those databases is pieces of information on your customer.
Again, as standalone systems, they’re not creating value—but when you tie those data sources together, you can create a 360-degree view of your customer across every single touchpoint in your business.
And that 360-degree view enables all this stuff we’re talking about and it’s, it’s an overlooked area and it’s been a struggle in the past for lots of companies to get there for a number of reasons. However, now we’re at a point where technology is where it is. You can really combine that data, crystallize a customer 360-degree model. And that model enables, essentially, everything you want to do through your martech stack, through your measurement, through your analytics, through your attribution. All the components fall into place when you have a 360-degree model that you can both use to serve your customer better, analyze your customer—and actually known customer, an unknown customer.
That’s a really interesting place to be, and that’s really where, you know, you can really get advanced from that point forward.
[00:35:53] Pete: Breedon, I get super excited about all of this, and you and I have been down this path before on data and, and this stuff excites me. I want to be a student of marketing and a student of the universe and I wanna embrace all these things that you’re talking about—and who wouldn’t if we want to be successful marketers. So, but what then challenges, you know, would inhibit marketers from advancing along the same path that we’re talking about?
[00:36:18] Breedon: The first would probably be organizational structure and design. And so what I mean by that is, because businesses operate in certain ways in order to drive process and efficiency in the company, typically, your organization in general, but your marketing organization isn’t structured correctly to adjust to the new ways of using data and using marketing technology. And so one of the challenges we see is that restructuring an organization and shifting process can be very hard.
The second is the sort of talent acquisition and retention, right? Like who are the people you have on board? How are you training and upskilling them? And how are you ensuring they’re growing and thriving in their role?
I think the third would be investment. this stuff can’t happen overnight—and I mean more in a time and a resource perspective. So are you willing to put in the time and effort required to invest in the change that’s gonna need to happen over time in order to get you from where you are today to where you want to be?
And then I would say the last one is there’s lots of data—but no vision on how to use it, on when to use it, on why to use it, on what’s real, what’s not. And so people get then overwhelmed and don’t start with the foundation. They start with a shiny tool that they think can get them there, and you just end up back at square one—because now you’re paying all these licensing fees and SAS fees that are never ending going up, your operating costs are going up, and you still can’t leverage your foundation.
[00:37:49] Pete: Why don’t you give us, like, a concrete example of, like, some cool martech solutions that you might be installing for your client—and very specifically, you know, if they actually have AI capabilities, because that’s what we’re here to talk about today.
[00:38:04] Breedon: One of the most maybe hyped-up marketing technologies and misunderstood marketing technologies as a customer data platform. And so, a CDP is one of my favorites. Our team has done a lot of work—not only implementing them, understanding them, building partnerships with them, and, and it’s a very powerful tool. It’s a very powerful platform.
And there’s a, there’s another opportunity we see, which is what we call a “composable CDP,” which is sort of a set of tools, open source tools you will build on your own to create the same outcome. But essentially what a CDP aims to do is develop a single source of customer truth for your business, help you model that information—like we talked about on a customer 360 model, very similar concept—and then actually leverage that information to activate across all your digital channels.
[00:38:58] Pete: Amazing. And I think so many companies are going through this, including Unbounce, right? We’ve done a big data transformation. We have a new CDP. Faye, are you excited about CDP Land?
[00:39:09] Faye: Yeah, I mean, we’re Segment customers. I’ve definitely drunk the Kool-Aid on, on the CDP because, I think when you’re coming from an environment where you have a rat’s nest of, you know, marketing and CRM and, and finance technologies, you waste so much time trying to integrate those individually to each other. And so when you can crack open the flow of data and really start to centralize everything something like a CDP, for a team like ours where—you know, we are not engineers—and so how can we quickly and efficiently and accurately kind of enable our teams and, you know, bring all these disparate data sources together in a single place and then repackage it and send it back out either for reporting purposes or to operationalize.
And so for me, you know, understanding the, the difference between all those different use cases, ‘cause it’s one thing to, you know, structure data to power, you know, a BI platform and get it into your Tableau or your looker or, or what have you. But then it’s quite another to say, okay, we need to notify this person and we need to trigger this campaign and we need to set this flag on this account. That requires, you know, a different mindset for how you’re gonna structure your data.
You know, Breedon, I’m super interested on kind of like your perspective on like the build versus buy when it comes to CDP. What do you unlock versus being able to buy a platform that specializes in integrating with your existing tech stack to kind of limit the amount of code or in-house development that you have to do?
[00:40:41] Breedon: So, um, buying a platform—like you talked about Segment, the leader in, in customer data platform, they’re a fantastic company market leader—they’ve built full platforms to do a number of things.
The first thing is to allow you to ingest multiple marketing data sources into their platform. The second is to help you create a marketing 360-degree profile by unifying those different data sources together.
One of the really great aspects about the platform side is identity resolution. So what a CDP platform will do is it will track all of your unknown profiles. So for your acquisition marketing, it’ll track people coming to your site, into app, wherever they are where we don’t know who they are—they haven’t logged in, they haven’t purchased on ecomm, something like that. It’ll track all that information about ’em and keep a repository of it. As soon as that customer buys or identifies themself, it unifies that profile under the customer 360-degree model. So now, even though your customers bought once, you have all this rich information on what’s happened previously. So that’s to me, in my mind, one of the most powerful components of a CDP.
Once you have that, it does cross-channel activation, so it will actually drive your—fancy term—“customer journey orchestration” and actually push out to the emails. Push out your app notifications, push audiences to your paid media platforms, and then because you bring everything together into one holistic platform, it has really deep and simple measurement capabilities. So the dashboards it produces are pretty strong for marketing teams.
If you’re building your own platform, well, you need to ingest data from your source systems, right? So you need to do that via data engineering. You need to get information under your CRMs, outta your accounting system, paid media sources. Well, you would do that through ETL or ELT.
You know, extract the data, load the data, transform the data. And so there’s, there’s tools out there best to breed, like a Fivetran or there’s open source technologies. Some of our favorites are like Apache NiFi. This helps you get data out of your system and into your storage, which is typically a cloud data warehouse. So, AWS has Redshift, Google Cloud has BigQuery. Databricks is another leading, uh, lakehouse that typically sits on top of Microsoft Azure and there’s Snowflake. So you need to get the data out and you need to build it into your data warehouse, data lake infrastructure, right? And that is the location where you build your customer model.
A lot of organizations want their customer information secured, privacy compliant in that infrastructure. And so when you need to do that, you can do a lot of really strong, robust customer modeling, in the tools we just talked about.
Identity resolution becomes really hard when you have a pre-built tool. So that’s one tricky area where—you know, customers are changing, profiles are updating, things are adapting, and so that’s a, a, a longer conversation you need to figure out.
So depending on the sort of data strategy of the company, the sophistication, and maybe where the company’s made investments to dates, I think there’s pluses and minuses to each. And then there’s also hybrid models, which I don’t think we have time to get into, but you might need a piece of one and some of the other, and you might need to find your secret sauce, So, there’s a number of ways to, to skin the cat, for instance, and it’s a, it’s a really exciting space. It’s moving fast.
[00:44:11] Pete: Okay. So we’ve talked a lot about shaping data, the universal source of truth, right to CDPs. How about a couple of changes you’re seeing in AI marketing that you think are powerful?
[00:44:26] Breedon: I mean, go back to CDPs or really any type of marketing technology. So particularly with the growth of generative AI, of large language models, of the big investment happening, virtually every marketing technology tool a company is leveraging today is gonna be implementing or already has some type of AI. And that’s only gonna accelerate.
So there’s a lot of benefits you’re gonna be getting from the tools you’ve invested in, right? And so things like recommendation models on top of a stack to supercharge what you’re doing today where you can kind of turn it on and it starts advancing further and further—so your performance is getting better, your customer metrics are getting stronger, without necessarily as much heavy lifting on the human side. That’s gonna accelerate, significantly.
What, what we’re sort of focused on and where we see a lot of opportunity, Pete, is the open source tools that are available to build custom AI solutions that are gonna be proprietary to the customers. You can leverage those out-of-the-box solutions that are coming to all your marketing tech, but that’s only gonna get you so far. Okay? Everybody else can leverage those too. If you combine your unique marketing tech stack, your unique combination of data infrastructure, with robust machine learning and AI tools, you can do something that your competitors can’t.
And so things like generating audiences, building out thousands and thousands or hundreds of thousands of recommendations simultaneously, providing different prices for individual customers, one-to-one at scale. Then the traditional models like churn prediction and customer lifetime value, accelerating those way beyond what you’re able to do today is really amazing.
And so we spend a lot of time as partners with Microsoft, Google, Databricks—they’re all building their own, or they have their own large language models and their own different AI models. But what they’re doing is they’re allowing you to integrate them with their open source technology where you can combine it with your own data and then really own your own future. And that to me is where it really gets exciting. And I think you’re gonna see the big winners and a lot of companies try to struggle.
[00:46:41] Pete: All right. If there’s one key message that you’d like to give to marketers based on our last 45 minutes of chat, what would that be? And Faye, maybe I’ll go to you first.
[00:46:54] Faye: Know your customers, know your strategy, know what you’re, you’re trying to accomplish, and structure your data strategy around that. It influences how you architect your website. It influences what you’re doing in Google Tag Manager or Telium or Adobe or what have you. At the point of measurement, you should already have an opinion about how you’re going to use this information later, either for the benefit of the business or for the benefit of the customer experience.
Try to think a few steps ahead. Ask yourself, you know, how are we gonna measure this? What are the questions I’m gonna be asked six months from now? Um, and how am I gonna be able to show the value in the revenue that we’ve been able to drive?
[00:47:31] Breedon: I would echo that and say, do all of that on a strong foundation of data and fundamentals. And if you do that and stay focused on what Faye mentioned—a strong foundation, a bulletproof strategy, consistency in execution—it’s gonna be amazing what you’re gonna be able to accomplish in the next one, two, three, five years.
The last thing I would just say is, you know, be careful about all the hype. Everybody is selling generative AI. All the systems and platforms and apps are talking about what they’re doing. Stick to what Faye mentioned with your strategy. Build it on top of a strong foundation and it’s gonna be amazing what you’re gonna do in the next few years.
[00:48:22] Pete: Amazing. All right, listening audience: I think the moral of the story today for me is if you’re not a geek, partner with a geek. Seek out these people who are data first and understand the marketing technology because you will be a more effective marketer, and a business person for it.
So, that was Unprompted, a podcast about AI marketing and you, and I want to really thank, my co-host Faye, and my very special guest and friend Breedon. Have a good day.
