The Future of MarTech is Composable

Composable customer data platforms are bringing marketing and data teams together. Are the Channel and Platform eras of MarTech behind us?   


  • Breaking News! Marketing and data teams are happily working together—this shouldn’t be news, but most companies’ marketing and data teams operate in two different worlds. Since organisational silos exist in every business, why should marketing and data be any different? 

    The marketing and data divide is notable as both work with the same customer, product, and sales data sources and yet utilise separate technology and vendors. The data team provisions generalised data infrastructure; marketing’s data needs are met with marketing technology (MarTech) SaaS software.

    In 2021, the silos started to break down with the introduction of a new category of technology — Data Activation — which connects data warehouses with marketing’s SaaS applications. Companies like Indeed, NASCAR, and Mercari combined the customer data in their warehouse with a data activation frontend to deliver the same marketing functionality as traditional customer data platforms (CDP). 

    The MarTech industry started referring to this combination as a Composable CDP. In software design, composable describes applications that are designed in components.  These components can be connected to form larger applications like a CDP. The real power of composability comes from the ability to combine components to make other larger applications; this is exactly what’s happening now. Companies are combining their data warehouse with data activation to create CDPs and clean rooms, journey orchestration platforms, and personalised digital experiences (DX).

    Composability also applies to the organisation itself. In their annual CIO conference, Gartner predicted that The Future of Business is Composable. Nowhere else is this better illustrated than when data and marketing teams come together around a composable CDP or clean room? Data engineers utilise their skills and tools to develop a customer 360 in the data warehouse. Marketers activate and engage those customers through the same channels but without the added data integration overhead required of traditional MarTech applications. 

    We’re entering a new era of MarTech, a composable era. And while that sounds big, it merely describes what’s already happening in the marketplace. Composability provides an agile and iterative approach to MarTech by connecting existing systems and teams. At the same time, composable MarTech is a departure from how marketing technology has been delivered over the last two decades. 

    As VP of Marketing at Universal Music Group, I lived through the two earlier eras of MarTech. Initially, a channel era where we organised our tech and teams by digital channel. Another is the platform era, where we wanted to understand the customer holistically, across channels, by deploying large customer platforms.

    I want to reflect on these two earlier eras to illustrate their benefits and shortcomings to the marketer, then some recent history about how we arrived at composable solutions for CDP, clean rooms, and DX. Finally, for those interested in exploring further, I’ll identify some leading vendors providing composable warehouse-powered applications designed for the marketer.  


    MarTech Magic: I can do it all by myself

    Each digital marketing channel—email, web, advertising, ecommerce, SMS, social—brings technical challenges unique to the channel. Initially, you need the help of developers to even engage with the audiences within the channel. Over time a MarTech vendor removes the technical complexity by providing software that makes marketing to that channel easy, with no-code tools that don’t require the support of a developer or IT. For marketers, these applications seem almost magical in enabling them to reach thousands to millions of customers   all by themselves.

    Marketing Capability Curve

    As Marketers mature, they begin asking better questions and more from their martech application. Channel management and delivery are a given; marketers now want to work more efficiently, deliver better experiences, and measure against overall business objectives. In response, MarTech vendors provide more capability in each of these areas:

    • Automation: Finding ways to reduce manual steps to create marketing content, an email, or a landing page. The content authoring experience is often improved by adding a channel-specific CMS that pulls product and marketing data from external systems: DAM, ECOM, PIM, or CMS.
    • Personalisation: Use engagement and conversion signals from external channels and applications to send the right message at the right time.
    • Analytics: In addition to reporting channel metrics—opens, clicks, page views—report on the business impact of a given campaign—sales online, offline.

    It is worth noting that these capability improvements correlate to data collection from external systems. 

    Legacies of the Channel Era

    Marketing continue to operate in silos. The marketing team is disconnected from the rest of the enterprise IT and data organisation at the business level, marooned on a digital island. And within the marketing department, teams are broken down by technical delivery channels: email, web, advertising. The combination leaves everyone seeing more trees than forest. And certainly, no single customer view.


    The Quest for a Single Customer View

    Consolidating a single customer view, a customer 360, is at the heart of the Platform Era.

    This involved collecting customer data from wherever it existed in the organisation: online marketing applications and offline data like in-store purchases. From this foundation, a unified profile of the customer and history of all interactions could be built and utilised in many ways: orchestration of communication across channels, personalising the digital experience, and fine-grained targeting of the customer in advertising,

    Unfortunately, not one but many “Platforms” were introduced at about the same time. Data Management Platforms (DMPs), Marketing Automation Platforms (MAPs), Journey Orchestration Platforms (JOPs), and Customer Data Platforms (CDPs)—all were building their single customer view. 

    These platforms collected customer data from the same sources in the same technical manner—tags inserted into the site or app—yet were perceived as unique due to how the business wanted to act on the data. DMPs were for the advertising team and CDPs for the direct marketing team. Most Enterprises, like Universal, implemented both.

    Platform Implementation and Adoption Challenges

    Implementing any of these platforms is hard. Adding tags is just the beginning; you must modify the website to pass valuable metadata to the platform. Adoption by marketing teams can vary too.

    Not One Ring to Rule Them All

    Although Platforms asserted to be the one ring to rule them all, the single ring only applied to given channels. In the end, there were too many rings and too many sources of truth. 


    The Single Source of Truth is within the Enterprise

    While the marketing team implemented various Customer 360 platforms, the data team collected and organised its version of a single customer view. They gathered vast amounts of data on multiple aspects of the business, including business transactions and customer information. 

    More importantly, the data team adopted the more modern Cloud Data Warehouses (CDW) from companies like Snowflake, Google, Amazon, and Microsoft. Compared with their on-premise SQL server predecessors, CDWs are built to handle large amounts of data at internet speed and scale. 

    The SQL to CSV Dance

    Word got around about the data teams’ new war chest of information, and marketers started asking for access to feed their marketing applications. A typical workflow was to request audience data from a data analyst. The analyst ran a SQL query and sent the results as a CSV for the marketer to upload.

    Marketing and data were finally working together, but the only drawbacks were the requests’ manual and ad hoc nature.

    Enter Reverse ETL

    In 2021 Reverse ETL software started being adopted by engineering teams to automate these manual exchanges between data and marketing teams. Now an engineer could query the SQL database and directly deliver the results to wherever the marketing team needed: marketing clouds, ad networks, and CRM applications.

    Data Activation adds a Marketer Interface

    What if marketers wanted to build the audiences directly? Digital marketing is an iterative test-and-learn process. Marketers couldn’t use the SQL language required for Reverse ETL, so marketer-friendly audience segmentation and activation were added. 

    Marketers could start doing everything they wanted with their CDP, JOP, and DMP without the added time and labour of a CDP deployment. Once you add a marketer interface connected to the data warehouse, it feels like the marketing applications I have used all along.

    Composable CDP

    As I evangelised this category of MarTech, I found the current labels—Reverse ETL and Data Activation—didn’t convey to marketers its actual value. I started using Composable CDP as it resonated more. Most marketers know what a CDP is and understand from the Composable DX the benefits of separating content data from presentation.

    Composable MarTech

    Composable DX and CDP are just the beginning of this new paradigm in MarTech. Each illustrates how the foundational components of digital marketing—customer and content—are managed centrally and available as micro-services for a new breed of MarTech applications to access and utilise.  

    The Composable MarTech Landscape

    Data Activation Vendors

    Data Activation applications are the best way to transition to warehouse-powered composable MarTech. Applications in this category enable marketers to build audiences and sync those segments to their existing marketing, sales, or advertising application. All provide no-code interfaces to explore, define, and activate audiences using customer data stored in the data warehouse. In this way, the marketer experience is very similar to a Customer Data Platform. 

    Flywheel, Hightouch, and Census are the three most popular vendors in Data Activation. All connect data warehouse platforms with marketing clouds, advertising networks, and CRM applications.

    Founded in 2017 by ex-Google marketers who saw the challenges firsthand of marketing teams getting access to customer data. Flywheel’s product focuses on the marketer experience, making building audiences and launching campaigns easy.

    Flywheel released Marve, a ChatGPT-powered natural language addition to their no-code, no-SQL Audience Builder.

    Customers Include: Indeed, Mercari, Boston Red Sox, Arizona State University

    Founded in 2018 by two ex-Segment CDP engineers who saw the growth of the data warehouse within their customers. Initially a Reverse ETL service, Hightouch has added Customer Studio, an interface for marketers to build audiences and launch campaigns. 

    Customers Include: NBA, AutoTrader, Nandos, AXS

    Founded in 2018 with the first Reverse ETL application, Census has since added a no-code audience builder, Segments, enabling marketers to build and sync audiences. 

    Customers Include: Notion, Figma, MasterClass, Canva

    Other Categories of Composable MarTech

    Data Activation is just a starting point. Just as traditional MarTech has grown over the years into highly specialised applications, the same is happening for warehouse-powered applications. These are just a few of a new breed of MarTech designed for the data warehouse.

    How to get started

    Where to begin if you’re a marketer? As this transition is well underway, the good news is that your existing MarTech vendors will eventually bring these newly architected for the data warehouse to your attention. Still, I recommend feeding your curiosity and educating yourself about these new vendors in this space.

    Take Meetings

    Meet with trusted sources in your network. Your vendors, as well as your peers at other companies. 

    Meet with the new players in the Composable MarTech landscape. While traditional MarTech providers will adapt their offerings for this new landscape, the new players are already there.

    Meet with your respective colleagues across your data and marketing teams. Composability is both a technical and organisational approach. Collaboration and knowledge sharing are key.

    Iterate toward a Single Source of Truth

    While the single source of truth will be in the data warehouse, expect there to be gaps in certain marketing data types. Consider these temporary rather than long-term gaps. The Modern Data Stack provides many connectors to quickly ingest data from marketing and advertising platforms into the data warehouse.

    Reassess Digital Transformation Projects

    Consider this a proceed with caution if you are considering a traditional CDP. These are big investments in time and cost, often justified by a small data or feature gap. Innovation in the data space is moving quickly, so expect missing marketing features to be solved before you’ve completed your CDP implementation.

    Expect Agility

    Composable solutions meet organisations where they are. Data Activation quickly connects your data warehouse to your marketing applications and doesn’t disrupt your teams’ workflows and processes. 

    Composable MarTech is the Google Translator for Data and Marketing

    What I’m most excited about in the Composable Era of MarTech is how it brings together marketing and data teams without demanding more of each group. Data teams continue working with their tools and infrastructure. Marketers, likewise, build and activate audiences with their no-SQL-required applications. Companies implementing a Composable CDP often see their data teams working more closely together than ever before. And not because the marketer learned SQL or the data team learned the components of the marketing cloud’s Journey Builder. 

    Moving from Digital Marketing Departments to Digital Enterprises

    In 2022, MarTech thought leader Scott Brinker predicted this coming together of marketing with data operations very well:

    …the shift underway to move from a “MarTech stack” that is largely siloed in the marketing department to a “business stack” that spans the entire enterprise. Moving beyond just The Digital Marketing Department to The Digital Firm…. is a dramatic expansion of the scope of the scale of the work MarTech, and marketing ops professionals can engage in, working beyond the walls of classic MarTech.

    Composability in MarTech is both a technical and organisational improvement over earlier eras. Technical composability removes the redundant collection of customer data. Organisational composability enables collaboration on a common goal while letting individuals focus on their respective core strengths: marketing and data. The collaborative benefits of a Composable CDP don’t receive nearly the attention that the technical aspects do, but I believe they are core to its success. The collaboration will move us from Digital Marketing Departments to Digital Enterprises. 


    More Like This