Powering personalized marketing with a modern data stack

Powering personalized marketing with a modern data stack

Insights from top marketing leaders on how Customer 360 is becoming a fundamental part of business strategy in 2024.

Insights from top marketing leaders on how Customer 360 is becoming a fundamental part of business strategy in 2024.


More about the episode

An esteemed panel of industry leaders — including Snowflake CMO Denise Persson, Fivetran CMO Rachel Thornton and HubSpot’s VP of Platform Ecosystem Scott Brinker – uncover the evolving role of Customer 360 in today’s business landscape. They discuss how a holistic approach to customer data is enhancing strategic decision-making and fundamentally changing how organizations interact with their customers. 

Our experts discuss the critical need for the modern data stack and the move towards more personalized customer experiences. 

Key highlights from the conversation include:

  • How HubSpot streamlined data integration into Snowflake with Fivetran to their analytics and machine learning capabilities
  • The importance of integrating real-time data across applications and systems to enhance customer interactions
  • The shift towards hyper-personalization and its implications for privacy and data governance

Watch the episode


Kelly Kohlleffel:

Hi folks, welcome to the Fivetran Data Podcast. I'm Kelly Kohlleffel, your host.  

This week, we have a special episode — three elite marketing leaders discussing the current state of data in marketing. 

Moderated by Fivetran’s VP of Product & Portfolio Marketing, Catlyn Origitano, the panel includes Scott Brinker, VP of Platform Ecosystem at HubSpot, Denise Persson, CMO at Snowflake, and Rachel Thornton, CMO at Fivetran. 

Their conversation touches on everything from what Customer 360 looks like in 2024, to the role of the modern data stack in enhancing customer personalization and even the future of leveraging advanced analytics and GenAI in marketing. 

For the full show notes, visit us at fivetran.com/podcast. Send us any feedback or comments at podcast@fivetran.com. 

Catlyn Origitano (00:01)

Today, we're going to be talking about how that modern data stack can power personalization in marketing. I am joined by an elite group of folks today to discuss these different topics.

First, Scott Brinker, the VP of Platform Ecosystem at HubSpot. The man, the myth, the legend behind the MarTech architecture as well — one of my favorite fun facts about you, Scott.

Denise Persson, the CMO from Snowflake, and Rachel Thornton, the CMO from Fivetran. Welcome everybody. It's a privilege to have so many big names and smart people on this call today. Thank you for taking the time to join us.

I want to start with a question that might feel deceptively easy: What is Customer 360 to each of you? Leading the horse to water a bit here, in 2024, it probably looks really different. Scott, why don't we start with you: What do you think Customer 360 is, and what are the things that impact it in this day and age?

Scott Brinker (01:13)

I can't help but start with the old joke about Wanamaker’s: I know half my advertising is wasted, I just don't know which half. A lot of companies have been in a mode where they have a “customer 180,” they’re just not sure which half it is. What's been both really exciting but also challenging for companies is the explosion of data sources that marketers now have access to. It's not just the traditional marketing channels. Because so many companies have all these digital properties and digital touch points with their customers, a tremendous amount of first-party behavioral data is flowing in.

There are a lot of really exciting plays happening in the ecosystem. Now we're seeing more and more second-party and clean data room partnerships. Marketers have an embarrassment of riches with all this data, but I think the challenge for them is, when we talk about the Customer 360, it's no longer just those traditional touch points.  Did they engage in this marketing campaign? Did they get contacted by a particular salesperson? How do we make sure that we're layering in all of these different digital touch point feedbacks that we have as well?

Catlyn Origitano (02:30)

I love that. I think that's absolutely spot on. Denise, I'm curious, do you have a different definition? Anything to add? What are you seeing, and what do you think Customer 360 is these days?

Denise Persson (02:40)

We can look at it from a technology standpoint, where it's really about getting all those different interactions and data points into one place, ideally, in real-time. From a business value standpoint, it's all about getting to know the customer as a person and the specific individual needs and expectations that customer has.

We all know that marketing is about relevance. The more relevant you are to the customer, the more likely you're going to win and keep that customer. Looking ahead, Customer 360 is also the foundation for all your data maturation in a journey. If you want to take advantage of things like AI, and the next phase, Gen AI, you need to have a Customer 360 strategy and data strategy in place.

Catlyn Origitano (03:36)

I love that. One thing that stuck out to me was your mention of real-time. I don't know that we've thought of Customer 360 as a real-time metric in the past. It was something you'd map, but not in real-time. I think that's a more modern requirement for the Customer 360.

The other part that I found interesting is the idea that Customer 360 is the foundation, not the goal anymore. 

Rachel, I'd love to have your take on this: In order to get into things like AI, you do need to have everything ready. There's a lot of actual foundation and prep work to be able to do these things, and it's interesting to talk about Customer 360 as the foundation rather than the endpoint.

Rachel, I'm curious, what are you seeing when it comes to Customer 360? Anything you want to add on there?

Rachel Thornton (04:32)

I do think a lot of great technology now makes the Customer 360 concept more relevant and also more possible than it's been. I'm excited to be a marketer because I think there are so many great tools out there — different campaign tools, chat tools, sales tools, biz dev tools. All of those things coming together and being able to bring the customer insights across all of those different touch points, do the analyses on them, uncover insights, use them to feed the next set of customer interactions and the next set of experiences you build is super powerful.

I think the real-time thing is critical. There are so many customers who say, "I'm not just using chatbots. I'm using SMS with my customers. I'm using WhatsApp with my customers.” Every time I have that kind of interaction with them, it feeds back into what do I want to know about this customer? Especially in direct-to-consumer environments, if that customer asks for something, I want to know what they’ve bought before. What size did they buy? What can I then tell them? If they're looking for something, I want to be able to say, "Hey, we have this for you,” or, “Based on your recent history, this is what we can give you."

I think real-time becomes not only more important, but way more achievable at this point.

Scott Brinker (05:56)

I would just add, since we've got a panel of two illustrious CMOs and someone who's pretty much a MarTech nerd, one of the things that's really exciting about these architectures with the modern data stacks is that they aren't just about unifying marketing touch points with the customers. It’s really about getting this data infrastructure that is spanning the entire company. Whether it's product interactions or sales interactions or customer service interactions, having that data in one place is opening up for marketers a much richer corpus of things that they can learn from on an analytical basis or act on from a trigger basis — a very wide 360 view.

Denise Persson (06:39)

Very important point, Scott. In the retail world, for instance, demand forecasting is one of the big challenges. You want to make sure you stock up to the right amount. If you stock up too much, that's a cost. If you don't have enough, you're going to impact revenue. Customer 360 is really the foundation for demand forecasting. It’s something your finance department is going to use as well. Every industry is going to have so many different applications. This feeds into supply chain management as well. It comes back to integration between 360 platforms and all these other tools for supply chain, ERPs and demand forecasting.

Catlyn Origitano (07:29)

As Scott and Denise also brought up, it's even within your own company. This data shouldn't just be marketing. It's also all your CSMs, all the folks that interact with customers. When we talk about building or having a data stack that can take in all of those diverse sources and give you any type of intelligence, that's a pretty big ask for a data stack — something that wasn't asked of data stacks even five years ago. I appreciate that kind of power. Those are some interesting insights into what Customer 360 is looking like, especially at some really big companies.

Denise Persson (08:08)

It’s also important to talk about the fact that this is not just something marketing can own in a silo. This needs to be a governed, centralized strategy for the entire company. Your Customer 360 platform is the beating heart of the customer-centric enterprise. It’s something that everyone will interact with. In the past, it was a marketing platform, but I think we are all in agreement: this is really critical for the entire enterprise.

Rachel Thornton (08:40)

It’s nice that marketing and other teams can come together and do it. Over the years, marketing hasn't necessarily always been involved in platform decisions or larger infrastructure decisions, and I do think that data literacy and marketing now go hand-in-hand. You can sit down with your CIO, with your CDAO, and say, "If this is the platform we want to build, here's what we need from a marketing perspective,” or, “Here are the outcomes we’d like from a marketing perspective."

From an infrastructure perspective, I know that Snowflake and HubSpot both do such a great job at educating our customers so that they know, This is what I want my cloud warehouse to be. These are my choices for that. These are my choices for my CRM. This is why I would choose this option. It's exciting to see how marketers have become more data literate over the years and much more of a driving force in those data and infrastructure conversations.

Catlyn Origitano (09:41)

Absolutely. I love the idea that Customer 360 has gone from being a stagnant dashboard to the living, breathing heartbeat of the company. I think that's really inspiring. What do you all think about the shifts to personalization (some are even calling it hyper-personalization), how data plays into that and how marketers need to think about it?

Scott Brinker (10:05)

I love the fact that we're now adding “hyper” to everything. It's hyper-agile marketing. But I understand why. The pace and the granularity of what's happening here is incredibly exciting but also quite mind-boggling. I think what's exciting about hyper-personalization is that it's through direct digital channels. We have the ability to personalize particular landing pages when someone's responding to a campaign. It's how we're plugging in content, whether it's based on a campaign or the process of determining segments.

We used to have these very large buckets of segments. If not truly a segmentation of one, we are increasingly able to really get to these micro segmentations. I can think of an example in the HubSpot ecosystem where we can now identify when a customer is a customer of HubSpot and a customer of one of our partners, but they haven't integrated the two. We can do a hyper-targeted campaign to help them get those two systems connected.

This is really exciting once you have the mechanisms to be able to automate that kind of hyper-personalization, but it's all predicated on the assumption that you can actually get the data for making those decisions to the right place at the right time.

Catlyn Origitano (11:30)

That's a great point. 

Denise, I'm curious, what about you? What are your thoughts on the possibility of hyper-personalization being the new personalization? How does data fit into that, and how do you balance ensuring that your data is secure and governed per the GDPR and CCPA with providing that experience?

Denise Persson (11:54)

There's nothing more important than data governance and privacy. If we look at the trends in data at the moment and all the new applications being developed, a lot of new applications are focused on data privacy and helping organizations with it. This comes back to the trust you have with your consumers. Again, trust is the foundation to the relationship.

If customers know that they're going to get a very good customer experience through data, they will accept certain things. It often comes back to where the data resides and where it's moving. Where they’ve opted in and where they haven’t. It’s more about what the organization is doing with the data. Are they giving this data to another organization? Are they sharing that data with someone who shouldn't really have it? Those are the core things to think about.

Catlyn Origitano (12:51)

I’ll be honest, this cardigan is from an Instagram targeted ad, and whoever has my personalization is hyper-personalized. It is very interesting to see that when people really pull it off, you get introduced to a new brand or a new company you've never heard of. As customers, we've all had that experience of hyper-personalization being cool and exciting. We’ve also had the experience where we're like, "Why do you think I would be interested in that?" 

I think that can often be the line between really great marketing and people who still need to do some improvement on personalization.

Denise Persson (13:29)

When you get that personalized experience from a brand you trust and you have a relationship with, you don't question it, right? You question it when you suddenly get an ad for something that makes you say, “How could they know that?” That's when you get concerned. 

Think about the time we spend scrolling. When we're searching for something and we’re served something that fits our personal expectations, it’s pretty outstanding. We have a customer, one of the largest fashion retailers on the planet, who has mastered Customer 360. They're very much beyond Customer 360 in terms of hyper-personalization and forecasting. They can forecast and predict exactly what shoe you are going to like, and that makes the world of difference when you go to the websites and everything you see is going to be served up according to your needs.

Catlyn Origitano (14:27)

Rachel, I'm curious, what are your thoughts on that?

Rachel Thornton (14:29)

The more ability we are able to deliver very hyper-personalized experiences — the more tools we have to do it, the more data we have to do it — the more important the governance and access aspects of that data become. You have all this data, there are great things you can build from it and good experiences you can deliver. It matters deeply in terms of customer trust: Who uses it? How is that data made available? Who has access to it? How are you tagging it, labeling it, prepping it and delivering it to the right people internally so that you don't have situations where a customer was clear that they wanted one type of engagement with us, but not another. We weren’t really good with that tracking, so now we've broken their trust.

I think it's amazing how far we've come with the concept of hyper-personalization, but if you get it wrong because you aren't good about governed data, regulated data, compliance, audit capabilities, and whether you’re delivering the data to the right person to build those experiences, you quickly end up with a situation where the customer is like, "No, actually I didn't give you the ability to reach out to me that way."

I think that what makes hyper-personalization work, in addition to the tools and platforms we have now, is a really strong concept of data governance, compliance, audit capabilities and the ability to put it in the right hands of internal teams to build the right infrastructure and experiences. I think that is critical. 

Catlyn Origitano (16:10)

Absolutely. I think all of this conversation, this topic and our previous one, further highlights the bridge that is being built and crossed on both sides. Marketers need to understand more about the infrastructure behind all of this because the same data you're using to do this hyper-personalization has all of these other ramifications, and your data team has to be thoughtful. If you don't know about all of this, it’s going to be more difficult for you to move faster and succeed.

With that, we're going to talk more brass tacks about that particular architecture. Scott, I'm going to turn it over to you so you can walk us through some of that. Since you are in the middle of all of this, both as a customer of Snowflake and Fivetran, and generally just in the middle of the MarTech landscape, could you walk us through how HubSpot does this? How do you think about this? What does this look like for you?

Scott Brinker (17:11)

At HubSpot, like any enterprise company, we have a huge number of different data sources for our business: back office and front office product usage data, ERP (like Anaplan), HR analytics from Workday, Greenhouse, survey data from SurveyMonkey, data on Google Drive and Google Sheets and marketing data from our own use of HubSpot. There’s a tremendous number of sources and new ones coming online constantly.

That's one of the key things you were mentioning about the evolving MarTech landscape. If there is one thing that has been constant in marketing and technology, it’s been change. Every year we keep finding these new emerging applications or new emerging channels that we want to experiment with, but it's very important that our architecture be designed in a way that allows us to plug those emerging and experimental technologies into our stack without creating a huge engineering task.

Because we can get all that data into Snowflake, our cloud data warehouse, we're able to leverage that data in a variety of destinations — everything from Looker for analytics to custom Python in our programs, ML and AI apps. 

As the data engineering team that worked on this has advocated multiple times before adopting Fivetran, this was an incredibly manual process. They had to extract and load, and each app had its own characteristics, so they had to get the proper modeling inside our warehouse manually.

Fivetran greatly accelerated and automated this process for them. They particularly love the dbt integration to expedite data transformation and modeling. I'm not giving due credit to all the wonderful work that they did with you, but I know they did publish an in-depth article on the Fivetran blog about how they got this all set up. For those of you who are a bit more data engineering savvy, I would definitely recommend looking that up.

Catlyn Origitano (19:24)

Thanks. That's great, Scott. I'm curious, as you've gotten more into this space and more into the data ecosystem, what's been the most interesting thing you've understood about the modern data stack that maybe you didn't know before?

Scott Brinker (19:29)

I think the big shift is that for a long time, the data that came into the warehouse was largely used for things like analytics. If we were applying them towards things like segmentation or the development of new models for how we might do Go-To-Market, it was very much a batch-oriented type process.

Because of the performance increases we’ve seen at the cloud data layer over these past couple of years, there's been this renaissance of people recognizing that there are all these opportunities to leverage this data in more direct operational feedback loops, feeding it back into frontline applications potentially in real-time. Even if it's pseudo real-time, it's tremendously powerful not to have to say, "We analyzed last month's data and this is what we recommend you do next month."

As things are happening inside our organization and that data becomes visible in our shared data layer, we then have the ability to action on that quite rapidly. I think for a lot of companies, this is very early. We're just beginning to understand the potential of this, but I feel for marketers, this is an absolute playground of new ways to think programmatically about how we engage with our prospecting customers. You can tell I'm pretty excited about what this is opening up for people.

Catlyn Origitano (21:13)

Customer 360 has been talked about as a goal for a very long time. I'm curious, especially from our CMOs, is it over-hyped? Is it possible? Is it impossible? Do you really think that this is something that marketers just make to market to each other or does it have some legs here?

Denise, I'd love to start with you first. What is your take on this?

Denise Persson (21:36) 

Well, it's definitely not over-hyped. It’s really the foundation of your data strategy that is going to enable you to take advantage of things like generative AI and general AI moving forward. We talked a lot about predicting behavior and personalization, but it’s also about marketing and taking advantage of generative AI for things like creating campaign designs. Think about how often we're trying to find a good image for a campaign. Think about when that image is automatically generated to fit the expectations of the customer, fully automated. Think about when the content is created automatically. That email that you would spend time writing is fully generated automatically for that individual customer. That's going to save us so much time on the marketing side. The foundation for that is the 360 data platform. You can't have an AI strategy without the data strategy. 

I think the biggest challenge at the moment isn’t technology. It's the know-how and the education and the data literacy, which Rachel mentioned. When I talk to other marketers and customers, they're not concerned that the technology isn't there, they feel like they don't have the capability to take advantage of it. What will solve that, of course, is that technology is going to be easier to use, but we still might not be fully there, so you need more data experience on the marketing team.

What we are seeing for those companies that are really successful is data scientists on the marketing team. We have data scientists on the marketing team here at Snowflake. Often, data scientists sit centralized somewhere on an IT or data team, and they're trying to figure out what the marketing team is trying to solve. They don't really understand the business problem. Bringing the know-how and technical data science to business challenges is going to be key. I think we're going to see more marketing teams with data scientists on them.

Catlyn Origitano (24:04)

Rachel, what about you? Do you think Customer 360 is possibly over-hyped? What's your take?

Rachel Thornton (24:09)

Definitely not over-hyped. I agree with Denise. It is something we can do now. I think that it is the foundation of any company that considers themselves customer-obsessed or customer-first. I think when you have built out your data's foundation and your data strategy, it’s then being able to apply it to increase new business, customer retention, reduce customer churn. All of those things, once you have developed your Customer 360 strategy, become infinitely easier.

I do think one of the interesting challenges on the data literacy side is making sure that marketing teams have the data analysts with them or a way to work with them closely so they can unlock insights and have data governance. You can only build the right experiences if you understand the data you have and if your data is ready and useful across teams. You have to make sure you have the right security governance compliance foundation. That’s critical.

Catlyn Origitano (25:20)

Absolutely. I think this is going to be a big year for marketing and for data, and I think we're kicking it off with this group of folks. I really appreciate it.

Expedite insights
Mentioned in the episode
Why Fivetran supports data lakes
How to build a data foundation for generative AI
How to build a data foundation for generative AI
more effective at replicating data for analytics
less data team time to deliver insights

More Episodes

Why everything doesn’t need to be Gen AI
Why everything doesn’t need to be Gen AI
Why everything doesn’t need to be Gen AI
Why everything doesn’t need to be Gen AI
Why everything doesn’t need to be Gen AI
Why everything doesn’t need to be Gen AI