84: Tejas Manohar: The past, present, and future of Composable CDPs

What’s up folks, today we’re extremely privileged to be joined by Tejas Manohar, Co-founder and Co-CEO at Hightouch.

Summary: Tejas envisions a new future for Customer Data Platforms (CDPs) that is more adaptable to each company’s needs. He discusses how Hightouch helps some of the world’s largest companies create “Composable CDPs” by using their own data architecture rather than purchasing a pre-packed option. He argues that this approach should help marketing and data teams align on a shared “source of truth,” not just through tech and architecture– but through operational shifts and cross-team collaboration. The conversation is an essential guide for businesses seeking to optimize their data use and enhance customer experiences.

Jump to a section

About Tejas

  • When Tejas was only 10 years old, his parents bought him a laptop for getting 100% on his Tennessee state exam. He discovered coding and started building apps for his friends.
  • At the age of 15, Tejas scored a software engineer internship at Hubspot. He built an automated interview scheduler that’s still used today.
  • He then joined Lottery.com as the first engineer leading a full stack team, building their backend, web, and mobile apps
  • At only 16, Tejas was hired as an early engineer at Segment. He would spend 3.5 years working on different product teams, including Sources, Warehouses, and finally Personas, where he managed the whole platform team
  • In June 2019, Tejas left Segment to pursue entrepreneurship.
  • Today Tejas is the co-founder and co-Ceo of Hightouch, a Data Activation platform that moves customer data from the warehouse (and other sources) to business teams’ tools. Hundreds of companies use Hightouch, including Spotify, the NBA, GameStop, TripAdvisor and GitLab.

From Corporate Travel to Reverse ETL

When asked about Hightouch’s creation, Tejas shared the interesting twists and turns of his entrepreneurial path. His initial venture after leaving Segment wasn’t directly into the customer data sphere. He founded a startup, Carry, in the corporate travel space.

However, Tejas’ departure from Segment wasn’t just fueled by an entrepreneurial itch. He had reservations about the future trajectory of the Customer Data Platform (CDP) industry. He didn’t fully believe CDPs were set to become the standard for managing customer data across industries.

Tejas then spent about nine months building Carry until the onset of COVID-19. Despite the inherent challenges of the travel industry—low margins, high human operation requirements, price-sensitive customers—Carry was growing. Yet, COVID-19 brought it to a grinding halt.

With business metrics falling to zero almost overnight, Tejas and his co-founders, Kashish and Josh, found an unexpected opportunity to reset. They decided to return to their roots in the data industry, tapping into their old ideas and experience from their Segment days. The pandemic, in all its harshness, became a catalyst for their return to the customer data space.

Tejas’ story is far from a linear narrative. The travel venture, the COVID-19 pivot, and the return to the data industry all added unique layers to his entrepreneurial journey. Looking back, Tejas feels gratitude for these unexpected turns of events, which led him back to the dynamic world of data and customer platforms.

Takeaway: An entrepreneur’s journey isn’t always a straight path. Tejas’ experiences, from his departure from Segment to his foray into the travel industry and his eventual return to customer data, highlight the unforeseen opportunities that can surface in the face of challenging times. His story underscores the importance of adaptability and leveraging past experiences to seize new opportunities in the ever-changing business landscape.

Composable CDP – The Birth and Journey of New Building Blocks

Traditional Customer Data Platforms (CDPs) are packaged solutions that collect, store and model customer data before ultimately delivering it to business tools. CDPs like Segment are all-in-one solutions that rely on separate infrastructure from the rest of a company’s data. Many large enterprises and mid-market companies have invested heavily in CDPs, hoping to enable marketers to freely explore customer data, create audiences, and tailor customer journeys across all channels. Yet, despite widespread CDP adoption, many companies were finding little value in these investments and running into persistent pain points. 

Hightouch pioneered an alternative architectural solution for customer data, which it coined the “Composable CDP.” This approach allows companies to use their own data storage and infrastructure, and instead layer on features like Hightouch’s Data Activation (Reverse ETL) to deliver that data to downstream tools. 

Many popular terms in Martech induce more head-scratching than clarity, like “marketing cloud” or “engagement hub.” Hightouch’s goal with the term “Composable CDP” wasn’t simply to coin a catchy phrase, but to instead address persistent dissatisfaction within the CDP industry and clearly define a new architectural approach. 

The Composable CDP approach enables marketers to leverage the rich data sources already present in company warehouses and then activate it across various customer journey touchpoints. 

The core value of a solution should not be whether it’s bundled or unbundled, but rather, the tangible business outcomes it can drive. As companies invest in housing their data using various BI tools, from Microsoft Power BI to newer players like Looker, the potential to empower marketing teams with this wealth of data is tremendous. Composable CDPs like Hightouch offer flexibility to use this data, and don’t limit themselves to specific data collected for a rigid traditional CDP.

The term “composable” was chosen to reflect this mindset – working with existing resources, scaling with future technologies, and avoiding the rigid, off-the-shelf solutions. While the term may elicit confusion for some, the purpose behind it – empowering businesses to effectively use their data – remains clear.

Takeaway: The  “Composable CDP” emerged from the need for a novel approach to CDPs that focused on empowering marketers to use  customer data more effectively. It’s about leveraging existing data, offering infinite flexibility, and scaling with future technologies, rather than sticking to rigid, traditional solutions.

Breaking Down the Power of Composable CDPs

Probing deeper into the potential of Composable CDP, Tejas was asked to illuminate the benefits of adopting such an approach over a monolithic all-in-one packaged solution.

Tejas started by emphasizing the fundamental flaw in traditional CDPs – their reliance on a pre-defined data architecture. Brands using conventional CDPs like Segment, Oracle, or Salesforce CDP are forced to adapt their data into a format acceptable to the platform, and this restriction severely limits the platform’s capability.

In Tejas’ words, “they can only operate on data that they understand and that was built for them.” This myopic vision becomes problematic in the complex, diverse landscape of large enterprises where every business is unique and possesses an array of distinct data. 

Tejas vividly illustrated this point by citing the case of a Fortune 500 company that wanted to leverage its pet loyalty program data – a dataset unique to their business – to drive personalization and engagement. Traditional CDPs failed to handle this unique set of data due to their rigid architecture, but Hightouch’s flexible and inclusive approach allowed the company to quickly use that data for many new use cases.

The ability of Hightouch to tap into an organization’s existing data, whether it’s stored in Snowflake, Databricks, or any other system, and utilize it to deliver highly personalized experiences is at the heart of its value proposition. By contrast, the challenges of molding data to fit into a traditional CDP’s format have led to a high failure rate, Tejas pointed out, making the novel architecture of Hightouch all the more appealing.

Takeaway: The real power of a composable approach like Hightouch’s lies in its flexibility and inclusivity. It’s not restricted to pre-defined data architectures and can handle unique, diverse data sets that are crucial to large enterprises, unleashing new potentials for customer experiences.

Unpacking the Legacy Customer Data Platforms

Arpit Choudhury’s informative blog post delineates eight vital components of a legacy CDP. Choudhry’s components ranged from the basic SDK (the software to collect first-party data) to the advanced Reverse ETL (Extract, Transform, Load) process used for extracting customer data from a warehouse to other business tools. His list also included ID resolution, data quality, accuracy, consistency, and governance aspects. 

Relying on both his time at Segment and Hightouch, Tejas was asked to reflect on these elements and offer his unique perspective on the anatomy of a CDP.

data beats: https://databeats.community/p/composable-cdp-vs-packaged-cdp-components

In response, Tejas noted that this framework emphasizes comprehensiveness of features. While he and Hightouch have written extensively about differences between traditional and Composable CDPs– and developed products to address all of these feature areas- he highlighted the importance of considering the why over the what. Rather than focus on components, he suggested to focus on understandng why companies opt for CDPs in the first place. 

According to Tejas, companies pursue CDPs to leverage their customer data to personalize the customer experience and drive business outcomes. The ultimate goal isn’t necessarily about the different components or features, but about leveraging data to the fullest. Tejas emphasized that in his view, the crux of any CDP lies in three core things: a way to collect data, a way to transform it, and a way to activate the data.

Tejas pointed out that the Composable CDP approach lets companies leverage their own data warehosue– and so they can access and activate data that wasn’t originally built for a traditional CDP’s end purpose. 

From the start, Hightouch, focused on large enterprises, emphasizing compliance, privacy, and governance. Their experience with companies such as the NBA, Warner Music, PetSmart, and GameStop necessitates a high degree of data accuracy and consistency.

Tejas underscored the importance of use case-driven selection in Martech. Instead of comparing solutions on a feature basis, marketing technologists should identify the activations they need for their business and then look for the features that enable those activations.

Key Takeaway: Tejas’ encourages a shift in perspective — from a feature-based approach to a use case-driven strategy in Martech decision-making. In doing so, he positions the future of Martech not as a quest for the most comprehensive solution, but as a tailored journey to activate and leverage data for personalized customer experiences.

📫 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.

Unleashing Marketing Audiences with the Composable CDP

When asked about Hightouch’s position in the Martech ecosystem, Tejas shared the company’s journey and how it evolved to ultimately found the Composable CDP category. Initially, Hightouch launched a new technology called Reverse ETL (Extract, Transform, Load).Reverse ETL syncs data from data warehouses to business tools.. There was debate within the team about coining the term “Reverse ETL”, but the gamble paid off, allowing them to catalyze a burgeoning space.

Since then, Hightouch has continued to expand its customer data offerings. For example, audience segmentation was part of Hightouch’s vision from the beginning, and became a core part of the Composable CDP offering when Hightouch added it. With the exception of ETL and data collection – the pathways for getting data into the warehouse – Hightouch has extended its functionality to encompass virtually all aspects of a CDP.

While ETL isn’t currently on Hightouch’s menu, Tejas hinted that it may not be off the table for future consideration. Given the rising commoditization and numerous ways for companies to collect events into their data warehouses, Hightouch has not prioritized this feature. However, it’s worth noting that they are not philosophically against its inclusion. 

Despite Hightouch’s reluctance to enter the data collection arena, the company excels in delivering a wide range of services under the Composable CDP approach. It shines in its commitment to offer more than just Reverse ETL, providing marketers with an extensive product to facilitate audence segmentation and identity resolution so that they can fully use their customer data. 

Takeaway: Hightouch has emerged as a leader in the Composable CDP space, adding features beyond their initial Reverse ETL offering. By circumventing the need for extensive engineering effort and tapping into existing data sources across businesses, Hightouch enables marketers to build audiences efficiently and effectively. This approach, anchored on data warehouses, allows businesses to preserve their unique data structure and offers them the flexibility to personalize based on their distinct attributes and customers’ needs.

Are Legacy CDPs Really Dead, and Can Hightouch Replace Them?

When Tejas was asked about the perceived controversy around composable tools and their approach to marketing solutions, the conversation moved towards the role of Hightouch as a competitor or substitute for traditional Customer Data Platforms (CDPs). Critics argue that the popular sentiment surrounding reverse ETL tools only adds to the confusion, veiling the true utility and function of such tools.

Tejas, having authored several thought-provoking blog posts last year on the subject, including “CDPs are Dead” and “Friends Don’t Let Friends Buy a CDP”, was asked whether Hightouch can indeed replace a legacy CDP today. In marketing, Hightouch is often touted as a ‘fast, flexible, affordable CDP alternative’. Yet, Tejas pointed out, Hightouch doesn’t incorporate components like event tracking and ETL, crucial elements of a typical CDP.

Tejas responded to this query with an insightful admission – describing Hightouch as a CDP alternative when used in combination without other solutions like Snowplow and FiveTran would be misleading. However, he did emphasize that less than 20% of Hightouch’s enterprise clients leverage such complementary solutions, defying the notion that Hightouch is solely for data teams at technology-forward companies.

Moreover, Tejas provided examples of Hightouch’s significant enterprise customers, such as Blizzard Activision and Warner Music Group, who have transitioned from other CDPs and now consider Hightouch their CDP of choice. The primary difference between Hightouch and a conventional CDP lies in the data collection component, which Hightouch addresses through partnerships with companies like Snowplow.

Hightouch aims to facilitate data activation success for marketing teams and personalization efforts. The future of the CDP, according to Tejas, is an architecture where companies possess their own data and can activate it across different channels, allowing for flexibility and adaptability – a future that Hightouch is ambitiously pioneering.

Takeaway: Hightouch, even without the data collection component, is recognized by some of the largest companies in the world as their go-to CDP. The future of CDPs, as envisioned by Tejas, is a more flexible, adaptable data architecture that Hightouch is actively shaping. With its data activation capabilities, Hightouch is carving a new path for the evolution of CDPs.

Breaking Down the Data Truth and Bridging Martech + Data Teams

Tejas was questioned about the conflict within organizations, arising due to the diverging understanding of ‘truth’ between the Customer Data Platforms (CDPs) used by marketers and the data warehouses employed by analysts. He emphasized the potential disparities in data interpretation between different teams, leading to complications in metric reconciliation and incomplete data in CDPs.

Tejas readily acknowledged the problem’s complexity, stressing the need for a single source of truth across the board. In Hightouch’s vision, this unified truth would arise from a data set that’s not only flexible but also entirely owned by the organization, fostering more control for marketers.

However, Tejas expressed a level of realism regarding the technology’s role in resolving these disputes. He believes the idea that software vendors can eradicate any collaboration issues between teams within a company is simply overpromising. While Hightouch provides an enabling framework that facilitates resolving these conflicts, there’s an inherent human element necessary for this to work.

For any company to succeed in their data initiatives, a robust alignment and collaboration between the marketing team and the data team is crucial. Irrespective of the platform or the nature of the CDP—be it composable or otherwise—those teams must work together in harmony. The role of Hightouch, as Tejas pointed out, is to offer a conducive framework where teams can work on the same source of truth, using the same data set, and capitalizing on the data stored in their warehouses.

Tejas’ doubled down on the need for fostering mutual understanding between marketing and data professionals. In his view, communities like Arpit’s Data Beats serve an essential role, bridging the gap between these two distinctive professional arenas. By elucidating data concepts to marketers and marketing concepts to data analysts, they contribute significantly towards promoting interdisciplinary knowledge.

However, as Tejas noted, expecting marketers to understand SQL and data engineering or data analysts to build marketing campaigns is not a realistic expectation. The solution to this problem doesn’t lie in enforcing these roles to cross over fully, but in designing and utilizing software that’s user-friendly for marketing teams, while still leveraging the data and technical infrastructure provided by the data teams.

Takeaway: The reconciliation of ‘truth’ in data between marketing and data teams isn’t solely a tech or architecture problem; it requires an operational shift and closer collaboration between teams. Software solutions like Hightouch provide a solid framework to tackle this, but the human element—teamwork, alignment, and communication—remains a key determinant in solving these challenges.

The Uphill Battle for Warehouse-Native Marketing Tools 

Tejas was asked about  the progression of martech and the potential role of the warehouse-native approach. The heart of this discussion revolves around the effectiveness of martech tools that hinge solely on the data warehouse without creating superfluous data replicas.

Tejas thinks that incumbent martech providers are not far behind Hightouch in making a paradigm shift towards the data warehouse as the central data point. This move is not only predictable but also a reaffirmation of the link between martech and data. As internal data repositories like Snowflake and Databricks gain ground, the idea of siloed marketing data is rapidly becoming outdated.

Despite this, Tejas expressed reservations about a complete sweep of the martech landscape by warehouse-native marketing tools. He brought up the diversity of marketing channels, encompassing advertising, app personalization, and more. This variety makes it impractical to expect a full-scale migration onto the data warehouse. Marketers have intricate needs; while data access is paramount, they also have to deal with aspects such as IP warming, which may not be catered to by warehouse-native tools.

This isn’t to say that Tejas doubts the potential of emerging platforms like Castled.io, Vero, and MessageGears to carve a niche in the martech landscape. His perspective isn’t grounded in their inability to create good businesses, but the improbability of these tools triggering a platform shift and upending  heavyweights such as Salesforce and Adobe. Despite the undeniable advantages to marketing, data, and IT teams, the chances of a mass SaaS swap-out do not appear immediate.

Tejas highlighted the ethos of Hightouch, stating that their approach is to tackle the present problem rather than develop new platforms.

Takeaway: Tejas’s insights indicate that while the martech industry will continue to embrace a warehouse-native approach, it does not spell a complete overhaul of the martech landscape  immediately. The future likely holds more integration and convergence than a radical replacement of existing tools and platforms. Businesses need to focus on merging their data and marketing efforts seamlessly without being lured into acquiring new platforms unnecessarily.

The Potential and Limitations of AI in Marketing

When asked about the growing influence of artificial intelligence (AI) in marketing, Tejas provided an overview of the exciting opportunities and inherent challenges. The idea of AI replacing or radically altering marketing roles is a common fear among many marketers. However, Tejas maintains a more balanced perspective on the matter, acknowledging both AI’s potential and its limitations.

Tejas recognizes the transformative potential of AI in marketing, emphasizing how AI could drastically enhance the marketing data experimentation process. Current limitations often restrict marketers to a few trials or force them to rely on gut instincts due to a lack of tools to test every possible variable. However, the introduction of AI can enable marketers to create numerous audiences and launch more sophisticated and targeted campaigns.

Furthermore, Tejas underlines his enthusiasm for providing marketers with access to data platforms like Databricks and Snowflake, which are increasingly incorporating AI capabilities. This approach aligns with his strong belief in the power of AI to augment, rather than replace, the human intelligence involved in marketing strategies.

Despite AI’s capabilities, Tejas is cautious about the notion of it replacing the marketing department. Drawing from his experience as a CEO and dealing with the intricate facets of marketing, he emphasizes the enduring importance of qualitative changes. Data-driven strategies can’t fully replicate the nuanced insights, intuition, and reasoning that human marketers bring to the table. These qualitative factors can dramatically alter program outcomes.

In Tejas’ view, while AI has the potential to radically alter the marketer’s role, the most likely scenario is that AI will augment rather than replace these roles. The real revolution lies in how AI can unlock access to crucial business data, empowering marketers to self-service data, build audiences, and understand customer cohorts. The power of AI needs to infiltrate all aspects of marketing, from brand planning and audience segmentation to personalization and experimentation.

Takeaway: While AI continues to evolve and impact various aspects of marketing, it isn’t poised to replace marketers in the immediate future. Instead, it stands to augment their capabilities, empowering them with more data-driven insights and decision-making tools. The role of the marketer is likely to undergo a transformation, one marked by the increased integration of AI but still very much driven by human intelligence and intuition.

Balancing Personal and Professional Lives

When quizzed about maintaining happiness amidst his multi-faceted roles, Tejas unraveled his approach towards work-life balance. As a co-founder and co-CEO of Hightouch, a developer, and with various hobbies and interests, he certainly has his plate full.

Tejas attributes his balanced approach to the significant investment he makes in his personal relationships. Engaging with family and friends, participating in activities outside of work, and pursuing various hobbies help him navigate the demanding nature of his professional life.

Tejas embraces the notion of being a “good amateur” in various pursuits, ranging from trying out new recipes with his girlfriend to exploring powerlifting, and occasionally playing the harmonica. These activities serve as refreshing breaks from the intensity of his professional life and provide him with joy and satisfaction. 

Though Hightouch demands his consistent and intense attention, the joy derived from personal life’s simple pleasures ensures he maintains a balance. Of course, he humorously acknowledges the possibility of AI eventually replacing him, referencing our previous discussion.

Takeaway: For Tejas, personal relationships and hobbies outside work are critical components in maintaining happiness, motivation, and achieving a balanced life. This illustrates that while professional success is important, investing in personal interests and relationships can significantly contribute to an individual’s overall well-being.

Episode Recap

Tejas offered an enlightening tour of the evolving Customer Data Platform (CDP) landscape. His insights provide us with a fresh perspective on the role of Composable CDPs in enhancing customer experiences and enabling marketers to leverage data more effectively.

We dive deeper into the practical applications of Hightouch’s composable CDP approach, we learned how this tool excels beyond reverse ETL. Hightouch eliminates the need for an extensive engineering effort and taps into existing data sources across businesses, offering marketers a more efficient way to build audiences. What’s more, Hightouch is not just recognized for its capabilities but is also acknowledged by some of the world’s largest companies as their go-to CDP.

Tejas underscored a crucial point about the evolving martech industry — that it isn’t about chasing comprehensive solutions but about adopting use case-driven strategies. Hightouch embodies this shift by focusing on data activation and personalization. The company carves a path towards a future where CDPs provide more adaptable data architectures, staying aligned with changing business needs.

A key takeaway from the episode is the importance of alignment and collaboration between marketing and data teams. Tejas explained that the ‘truth’ reconciliation in data isn’t solely a tech problem—it requires an operational shift that involves teamwork, clear communication, and alignment. Software solutions like Hightouch provide the necessary tech framework, but the human element remains instrumental in addressing these challenges.

Finally, Tejas’ insights caution us that while the industry is leaning towards a warehouse-native approach, it doesn’t equate to a complete overhaul of the martech landscape. Rather, businesses should focus on integrating their data and marketing efforts seamlessly without unnecessary diversions to new platforms. This shift should lead to more convergence and integration than radical replacements.

Listen below for a comprehensive, nuanced, and accessible journey through the world of composable CDPs and reverse ETL. 🎧

Follow Tejas and Hightouch 👇


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