96: Natalie Miles: Building vs. Buying Martech, the Power of Generalists and Assembling a Composable CDP

What’s up everyone, today we’re joined by Natalie Miles, Head of Marketing Technology at Chime.

Summary: Don’t underestimate the role of generalists in martech; they’re your go-to for system-level thinking and breaking down data silos. Building vs buying your tech stack? It’s not black and white; successful setups usually mix both, and including engineers in the decision process is non-negotiable. Considering a CDP? Opt for a composable one to get quick value and robust data management. And if you’re venturing into personalization, it’s your team’s culture and process that’ll make or break it, not just the tools. Tune into this episode for straight-up, actionable insights that cut through the noise in the martech world.

Jump to a section

About Natalie

  • Natalie started her career as a Financial Counselor at the Consumer Credit Counseling Services of San Francisco 
  • She then took on the role of Quality Assurance Specialist at Lending Club, a fintech marketplace bank where she was eventually promoted to Operations Analyst
  • Natalie then moved over to Credit Karma, best known for pioneering free credit scores where she started as a Marketing Operations Analyst and was later promoted to Marketing Operations Manager
  • And for the last 3 years, she’s been Head of Marketing Technology at Chime, a fintech company that offers no-fee savings accounts, where she’s built and managed a holistic Martech stack supporting all channels and functions within the Marketing org

The Intersection of Financial Empathy and Marketing Operations

When asked about her transition from financial counseling to marketing operations, particularly within FinTech, Natalie illuminates how her upbringing and career have been tightly woven with mission-driven personal finance companies. Shaped by her experiences in a working-class household and graduating amid a historic economic crisis, Natalie’s focus has been on transforming legacy institutions that often operate on zero-sum models—those that profit when the customer suffers. Her goal? To align business value directly with user value. 

Natalie also emphasizes the importance of having a generalist background when working in marketing technology. She points out that her diverse experiences, including her time as a financial counselor, have enriched her understanding of system-level thinking—a key asset for any marketing technologist. It wasn’t just about marketing; it was about leveraging technology to make different teams more efficient, whether they were marketing or support teams.

In her journey through the marketing landscape, Natalie discusses the evolution from specializing in lifecycle marketing to adopting a more generalist approach once again. She believes that understanding the pain points in one marketing channel provides insights that are transferable across other channels. This is vital because while each channel has its own nuances, they also share common threads that are integral to driving growth or achieving specific business outcomes. 

Natalie underscores the concept of the “T-shaped marketer,” a term often used in marketing discussions to describe professionals who start by specializing in a specific channel but gradually broaden their scope. This broad understanding is crucial in the realm of marketing technology, where preventing data silos and powering omni-channel journeys are key.

Key Takeaway: Having a generalist background isn’t just about being a jack-of-all-trades; it’s about mastering system-level thinking. This kind of broad perspective is invaluable in marketing technology, where understanding how various components interact can significantly improve efficiency and effectiveness. By being well-versed in multiple areas, you’re better equipped to tackle complex challenges and integrate solutions that drive measurable results.

The Power of Generalism in a Specialized Marketing World

When asked about the value of a generalist background in martech, Natalie explained that many marketing organizations structure themselves around specific channels, but doing so can have its drawbacks. Specialists may be excellent at understanding the intricacies of a particular channel like Google Search but may lack a broader understanding of how to harmonize different channels for an integrated, omni-channel experience.

Natalie pointed out an often-overlooked aspect of specialized teams: they often onboard tools designed to solve specific channel needs. While this specialization can drive short-term success, it often fails to consider the bigger picture. As marketing complexity grows and companies aim for more personalized, omni-channel experiences, the need for someone who can tie all these disparate elements together becomes increasingly important.

In the startup world, this is especially significant. Startups usually kick off with generalists who can wear multiple hats and pivot as needed. As the company matures, specialists are added to the mix. Natalie highlighted the risks of over-indexing on channel-specific experts. These experts can work in silos, and this compartmentalized approach can be a roadblock when aiming for more intricate marketing strategies that require seamless coordination between channels.

One of the most compelling points Natalie made was around marketing organizations that prioritize outcomes over channels. An outcome-oriented approach can enable the same individual to manage paid retargeting ads while also running lifecycle campaigns, for instance. This blend of responsibilities demands a broader skill set and makes the case for generalists who can adapt to multiple marketing scenarios and strategies.

Key Takeaway: Don’t underestimate the power of generalists in martech. They bring the critical ability to weave together various marketing channels and tools, enabling a more integrated and effective marketing strategy. If your team is too specialized, you risk creating data and strategy silos that can hamper your broader marketing objectives.

Martech’s Dilemma: Engineering Constraints and the Build vs Buy Debate

When asked about Casey Winters’ article on the notion that martech is essentially for engineers, Natalie offered a nuanced perspective that extends beyond the conventional build-versus-buy debate. Casey argues that martech has evolved as a response to engineering constraints, and lifting these constraints would render third-party martech solutions redundant. Natalie, while a fan of Casey and his work, respectfully disagrees with this one-dimensional view. She highlights a reality most businesses face: the absence of unlimited engineering resources. In her experience, this constraint justifies the need for third-party solutions, especially when internal solutions often lack marketer-friendly user interfaces.

Natalie touched on the complexity of allocating engineering resources effectively, particularly in sectors like FinTech. Should a FinTech company spend its limited engineering capital on building martech products, or should it focus on actual financial products that drive consumer growth? She suggests that the more pressing question businesses should be asking isn’t whether to build or buy, but where to best align their engineering resources in line with their core competencies. This consideration often leads to a blend of in-house and third-party solutions in a company’s martech stack.

Narrowing the definition of martech to just third-party solutions is, in Natalie’s view, a limiting approach. She emphasizes that most martech stacks will inevitably be a mix of both built and bought solutions. This mix arises because even when buying a solution, substantial engineering resources are often needed to integrate it into existing data architectures. And contrary to Casey’s idealistic scenario, Natalie stresses that most companies do not operate in a utopia of endless engineering resources.

Natalie’s final point converges on the person who sits at the intersection of engineering, marketing, and data privacy. Whether that person’s title is a product manager or something in martech is irrelevant; what matters is that someone is thinking about the problem space for marketers. This individual creates the roadmap, defines the requirements, and collaborates with cross-functional teams to ensure that the chosen solutions—built or bought—are effectively implemented.

Key Takeaway: The martech ecosystem isn’t a binary choice between building and buying solutions. It’s more about resource alignment and meeting the real needs of your marketing team. Beware of simplistic viewpoints that suggest you can entirely build or buy your way to success; the reality is usually a hybrid approach tailored to your business constraints and goals.

Bridging the Gap Between Engineers and Marketers in Martech Decisions

Continuing the thread about the role engineers play in the martech sphere, Natalie weighed in on the dynamic relationship between marketers and engineers. She acknowledged that while most martech tools might cater primarily to marketers, overlooking the engineering team could lead to integration nightmares. These tools need to fit within the existing tech architecture, after all.

Natalie also brought up the common pitfall in the decision-making process for tools like Customer Data Platforms (CDP). In many organizations, marketers might go off and purchase a tool without consulting the engineering team, setting the stage for future headaches and compatibility issues. To truly get value out of martech, a unified decision-making process is critical.

However, Natalie pointed out that the effectiveness of this collaborative approach often hinges on the organizational structure. In larger, more siloed companies, the engineering team might not fully understand the challenges that the marketing team faces. In such scenarios, the need for a translator—a product manager or a specialized martech role—becomes crucial. This individual serves as a bridge, understanding both the marketing challenges and the technical framework necessary to address them.

For Natalie, the debate isn’t as binary as it seems. It’s not about whether martech is for marketers or engineers. It’s about recognizing that both have a role to play and ensuring that role is respected in the decision-making process.

Key Takeaway: Martech isn’t a one-size-fits-all solution catered to marketers alone. Engineers have a crucial role in ensuring the tech stack is cohesive and functional. To avoid friction and maximize the utility of any tool, include both marketers and engineers in the selection and implementation process.

Building a Composable CDP Stack: Lessons from Chime’s Journey

When asked about her experience building a composable Customer Data Platform (CDP) stack at Chime, Natalie offered invaluable insights. The central issue was the need for personalization at scale, fueled by data activation. Chime, despite being in operation for over a decade, had not yet onboarded a traditional CDP. This lack became increasingly glaring as they sought to differentiate in a commoditized space and optimize customer acquisition costs. Personalization, Natalie pointed out, requires contextual information about user behaviors, needs, and interactions with the product.

Natalie’s aha moment came from listening to a data-edge podcast discussing “reverse ETL and data activation.” It shifted her perspective on how CDPs could function without necessarily replicating data already stored in their warehouses. This led to the realization that composable CDPs, like Census and Hightouch, offered advantages such as speed to value and reduced data storage costs. These platforms could get up and running far quicker than their traditional counterparts, which could often take up to six months just to configure data ingestion.

The data security and audience portability aspects were not to be overlooked. Operating in a heavily regulated space, Chime prioritized keeping their first-party data in-house as much as possible. Traditional CDPs would necessitate sending this sensitive data out for storage, which was a significant concern.

What really sealed the deal was the alignment with engineering priorities. Going the composable CDP route meant that the engineering team didn’t have to spend valuable resources on setting up a traditional CDP. It was a win-win, resolving the needs of both the marketing and data engineering teams at Chime.

Key Takeaway: If you’re in the market for a CDP, consider your organization’s specific needs and limitations. Composable CDPs can offer speed to value, data security, and align better with modern data architectures, which could be precisely what your marketing and engineering teams have been looking for.

📫 Never miss an episode or key takeaway 💡

By subscribing to our newsletter we’ll only send you an email when we drop a new episode, usually on Tuesday mornings ☕️ and we’ll give you a summary and key takeaways.

Success! You're on the list.

The Complexity of Data Pipelines and the Ease of CDP Integration

When asked about the situation before implementing a composable Customer Data Platform (CDP), Natalie shed light on an often under-discussed aspect of data management. She recalled a complex network of point-to-point integrations built using Python scripts and airflow. The result? A six-week time frame to set up a new pipeline to a new ad destination, making the cost of exploring new channels unnecessarily high and the process far from nimble.

The obstacle wasn’t just time but also resource allocation. In Natalie’s own words, the cost of testing into new channels escalated because of the intense work needed to get the initial data import running. The situation was further complicated by the stringent need for data privacy compliance. Without a centralized framework to govern data consent, maintaining these point-to-point integrations was a daunting task, especially in a world bound by GDPR and CCPA.

However, the switch to a composable CDP was almost a night-and-day difference. These platforms are built for plug-and-play functionality. Instead of the tedious task of molding and transforming data for each marketing destination, it became a scenario of “create the model once, publish everywhere.”

But the most significant shift was perhaps in the governance of data and consent. With a centralized hub, compliance was more manageable. Whether you’re in a heavily regulated space or not, the ability to centrally store opt-in and opt-out data while ensuring a standardized definition of “customer” across tables becomes a game-changer.

Key Takeaway: A composable CDP isn’t just an upgrade; it’s a strategic move that cuts down both the time and resources required for data management while offering robust compliance and governance features. It’s time to think beyond point-to-point integrations.

The Real-Time Dilemma in Warehouse-Native Martech

When asked about the feasibility of incorporating warehouse-native customer engagement platforms that live atop a Snowflake instance, Natalie zeroed in on the critical issue of latency. In real-time personalization scenarios, traditional data warehouses often hit a speed bump. While they might be well-equipped for less time-sensitive tasks like powering BI tools, they often fall short when microseconds count.

Natalie acknowledged that warehouse latency creates a bottleneck, especially for customer engagement platforms aimed at real-time personalization. While there are workarounds to improve latency within the warehouse, or even to bypass it entirely through server-to-server integrations, the challenge persists. The question isn’t just about technological capability but also about actual need. Does every marketing scenario require real-time personalization, or are we, as an industry, overestimating its importance?

Importantly, Natalie challenged the universal desire for real-time personalization. There’s a cost-benefit analysis to be done here: the technical constraints might not justify the value real-time personalization adds to certain channels. Most notably, in lifecycle channels where a push notification within a minute after a user’s action might be desired, traditional data warehouses currently fall short.

Interestingly, Natalie foresaw the martech space evolving to solve this latency issue. Whether it’s a new configuration of existing technology or a groundbreaking approach to data warehousing, the lag in real-time processing could well be a thing of the past sooner rather than later.

Key Takeaway: If you’re grappling with whether to go warehouse-native for your customer engagement platforms, consider both the existing latency limitations and the actual necessity of real-time data in your specific use case. It’s not just a technology decision; it’s a strategy call.

Experimentation and Personalization – More than Just Tooling

When asked about how experimentation and personalization strategies evolve during a company’s CDP journey, Natalie didn’t immediately dive into the mechanics of the technology stack. Instead, she argued for a broader look at the issue, emphasizing that the key to effective experimentation often starts with process and culture.

Natalie strongly suggested that before even touching tooling, teams need to create a strong foundational culture around experimentation. This includes standardizing approaches with experimentation briefs, developing frameworks for prioritizing experiments, and empowering marketers to understand concepts like minimum detectable effects and statistical significance. Surprisingly, Natalie pointed out that something as elementary as forming a testable hypothesis could be a stumbling block for some teams.

Discussing the limitations of existing martech solutions, Natalie highlighted that most tools offer their own built-in experimentation features but lack a unified, company-wide view. When experimentation is happening in silos across various teams and tools, the results get muddled. As a result, you might end up with misleading outcomes because you can’t pinpoint which experience caused the observed lift or change in behavior.

Natalie identified an unmet need in the martech space: the absence of a third-party tool that centralizes all experimentation. As companies increasingly build their experimentation platforms in-house, it exposes a gap that existing martech vendors haven’t filled. They tend to operate within the confines of their specific functions rather than solving for a company-wide, unified experimentation layer.

Key Takeaway: If you’re diving into the world of personalization and experimentation, don’t fixate solely on the tools. Culture and process are the real linchpins. And while the martech industry catches up, consider an in-house solution for centralized experimentation.

Harmonizing Teams in the Intersection of Tech and People Skills

When asked about managing the complex dynamic of being tech-savvy while also handling the human aspects of team management, especially in roles that sit at the intersection of various teams, Natalie offered some pragmatic insights. She stressed the importance of adopting a curiosity mindset. In her experience, many conflicts or misalignments between teams stem from a simple disconnect in communication. Different teams often focus on different objectives, whether it’s growth outcomes for marketing or data quality for analytics teams. This difference in focus can create a “language barrier” that leads to misunderstandings.

According to Natalie, there are typically two types of misalignment. The first is when teams agree on the desired outcome but not on the method to achieve it. This is a solvable issue and usually just requires translating each team’s goals into a common language. The second, more complicated form of misalignment occurs when there is disagreement about the outcome itself. This deeper issue calls for intervention from leadership to clarify priorities.

Natalie pointed out that the martech space, in particular, is guilty of using jargon that doesn’t resonate across different teams. This contributes to the language barrier and misalignment. To address this, she advised abstracting away the jargon and focusing on what different teams actually care about. For example, marketing might talk about “data activation,” a term that might not mean much to a data engineer.

Natalie emphasized that as marketers, or anyone sitting at this complex crossroad, it’s crucial to understand the “value props” that will resonate with the diverse teams one interacts with.

Key Takeaway: The secret sauce to team alignment isn’t some complex framework but rather effective communication. Be the translation layer between different teams by understanding their goals and using language that resonates with them. Speak less like a marketer and more like a human who understands what other humans care about.

The Elusive Art of Finding Balance in a Multifaceted Life

When asked about the holy grail of work-life balance amidst her roles as a martech leader, a mom, a wife, and a film enthusiast, Natalie quickly dispelled the myth that anyone has it figured out. Instead, she likened her daily experience to walking a tightrope while juggling. For Natalie, it’s about acknowledging that she’s not going to be able to “do it all” and being intentional about which ball she’s willing to drop at any given time. The peace and empowerment come from this level of intentionality.

Natalie stressed that there are times when one aspect of her life will receive less attention. Maybe it’s a day with excessive screen time for her child or a day where work only gets 50% of her energy. These trade-offs are inevitable, and the real skill lies in prioritizing. Parenthood, according to her, is a crash course in this form of rapid-fire decision-making. Parents become masters at determining what needs immediate attention and what can wait.

Another aspect Natalie brought into focus was the concept of a personal “North Star.” For her, this North Star is her child. Everything she does, including her career, is a means to provide opportunities or set examples for him. Having a clear focus allows her to make choices that align with her core values, offering peace of mind in the chaos that is daily life.

Key Takeaway: You can’t be everything to everyone, and that’s not just okay, it’s real life. Make peace with the fact that you’ll have to drop the ball sometimes and focus on what truly matters to you. That’s your North Star, and it’ll guide you when you have to make those tough calls on where to invest your time and energy.

Episode Recap

Generalists are your secret weapon in martech. Forget the myth of the jack-of-all-trades being a master of none. The real magic happens when someone can connect the dots between multiple areas. They’re the folks who can break down the silos that choke your data and muddle your strategy. They understand that every piece of your tech stack interacts in complex ways. Bottom line: you want these generalists on your team. They help you get more from your martech investments by integrating solutions that actually drive measurable outcomes.

Now, let’s talk building vs buying in the martech space. It’s not an either-or situation; it’s a ‘best-fit’ scenario. Forget the misleading viewpoints that suggest you should fully build or buy your stack. Reality check: the most effective setups are usually a blend, customized to your specific needs and constraints. And don’t just bring marketers into this discussion. Engineers are key. They ensure your tools not only work together but work well. That’s how you get a tech stack that’s cohesive and functional.

When diving into Customer Data Platforms (CDPs), be strategic, not reactive. Composable CDPs can be a game-changer—they’re fast, secure, and they mesh well with modern data architectures. Think of it not just as an upgrade, but as a high-impact strategy move. You save time, cut resource costs, and get top-notch data management features, all in one. But don’t rush into it. Consider your unique needs. Need real-time data for customer engagement? Weigh the existing latency limitations. Think about the genuine necessity of real-time information for your specific case.

Last but not least, if you’re wading into personalization and experimentation, remember: your tech stack is only as good as the culture and processes behind it. Tools are just that—tools. Your team’s mindset and workflow are what will make or break your personalization efforts. While the martech industry is still evolving, if you need specialized solutions for experimentation, don’t hesitate to build them in-house.

So why should you tune into this episode? Because you’ll walk away with no-nonsense, actionable insights that help you not just navigate but thrive in the martech space. From leveraging the power of generalists to making savvy choices in CDPs, we dig deep into what really matters. So don’t miss out—listen to the episode and arm yourself with the insights to stay ahead 🎧👇

Follow Natalie and Chime 👇


Intro music by Wowa via Unminus
Cover art created with Midjourney

See all episodes

📫 Never miss an episode or key takeaway 💡

By subscribing to our newsletter we’ll only send you an email when we drop a new episode, usually on Tuesday mornings ☕️ and we’ll give you a summary and key takeaways.

Success! You're on the list.

Future-proofing the humans behind the tech

Leave a Reply