Friends Don’t Let Friends Buy a CDP
How spending the first half of his professional career at Segment drove Tejas Manohar to disrupt the 3.5 billion dollar CDP category.
September 14, 2022
In January of 2016, after being a loving customer for years, I decided to join Segment, a leading Customer Data Platform (CDP), as one of their first ten engineers. The vision made immediate sense to me: to build a central platform for organizations to collect, synthesize, and activate their customer data. Marketing teams would finally be able to activate their data freely to personalize campaigns across all their channel tools and ad networks.
The broader market at the time was just as bought in as I was. CDPs quickly rose in popularity as organizations invested in empowering their marketing teams with actionable customer data. Executives threw CDP implementations to the top of their priority lists and dedicated hundreds of thousands - if not millions - of their marketing budgets, as well as teams of engineers, to implement them. Regardless of the high costs and long implementation phases, they knew the importance of providing marketers with actionable first-party data.
It was hard to imagine that we could have been building it all wrong.
What is a CDP, and why were they exciting?
A Customer Data Platform (CDP) is an all-in-one data platform built for marketing teams. They serve as a database for all your customer data with a bundled activation layer to help you leverage your data for marketing automation.
That’s what made them special; at the time, we thought that CDPs could not only centralize all customer data into a “360 view”, but also make it easy for marketers to take action on that data.
All CDPs have at least three core components:
Data Collection: Since CDPs are a database of customer data, they need to give you a way to send data to them. To solve this, CDPs expose an API for developers to track user traits and events (e.g., the actions users take across websites and mobile applications.)
Data Transformation: CDPs usually have an out-of-the-box identity stitching functionality and tools to create custom traits on user profiles.
Data Activation: CDPs are useless unless they help you act on your data, so they have integrations enabling you to sync enriched profiles and audience segments to marketing channels.
In a legacy CDP, all of these features are bundled together in a tightly integrated all-in-one solution that allows marketing teams to centralize customer data directly inside the CDP.
What I missed at the time–what we all missed–was the rise of the cloud data warehouse and what that meant for managing and activating customer data.
Cloud data warehouses make legacy CDPs obsolete
During my time at Segment, of all the technologies in the data space, perhaps the only one growing faster than CDPs was cloud data warehouses. Thanks to the breakthroughs of market leaders like Snowflake, BigQuery, Databricks, and to some extent, Amazon (Redshift), data teams can store, model, and analyze their customer data faster and more efficiently than ever before.
Once only available to enterprises with enormous budgets and large dedicated data engineering teams, data warehouses became accessible to organizations of every size. In a few short years, they became ubiquitous.
Data teams quickly spun up enterprise-grade data infrastructure that provided entire organizations with a source of truth for their customer data. These were made even more useful by a rapidly-evolving ecosystem of tools, including dbt. These new warehouses would house ALL of an organization’s customer data from any source - rather than just their interaction events captured in their CDP. The adoption of the warehouse spread fast and the gravity of data within companies shifted to the data warehouse.
Suddenly, the “sources of truth” within organizations began to collide. Marketing’s CDP-based “truth” became increasingly incomplete and misaligned with the truth as defined by analysts in the data warehouse. Our customers began to wake up - two years and a couple hundred thousand dollars later with a CDP that didn’t have all the data they needed and a dozen meetings on their calendars titled “metrics reconciliation.”
This was not the dream we had promised.
The rise of data activation
That brings us to 2019. CDPs were inaccurate, incomplete, and expensive - but they were actionable. The warehouse was accurate and comparatively cheap - but the data was stuck inside.
That exact gap (warehouse activation) ultimately led my co-founders and I to leave Segment in 2019. We took our experience building CDPs and decided to create Hightouch - a data activation platform. We quickly saw that by leveraging a technology called Reverse ETL, Hightouch could unlock the true value of the data warehouse - enabling both marketing and data teams alike to activate data FROM their data warehouse and send it out TO their many business tools.
Doing so would lay the foundation of a new, “unbundled” CDP architecture centering on the data warehouse as the customer data store - breaking down the data silos created by now-redundant CDPs.
CDPs are dead
At the beginning of 2022, with validation from multiple Fortune 500 customers, our team here at Hightouch made a bold prediction: Customer Data Platforms are dead. Our core belief is that while the goal of customer data platforms to create actionable customer data for marketers across their tools is correct, the approach is fundamentally flawed.
Rather than sending customer data to yet another vendor's infrastructure, teams simply need a data activation platform that sits on top of their data warehouse. This activation platform offers a marketer-first UI to create and take action on customer segments (ie. Hightouch Audiences) – a platform that feels similar to a CDP; but since it sits on top of the data warehouse, offers a more centralized, more secure, and exponentially more flexible solution.
As expected, our claim brought plenty of backlash– especially from VCs and executives invested in and working for the leading CDPs. We saw tweet storms, blog posts, and Linkedin Lives slamming the idea that such a simple concept could disrupt their $15 billion category.
However, one thing we didn’t realize when we made this claim was just how soon the broader market would agree with us. And just how right we were.
We didn’t choose an off-the-shelf CDP like Segment because a substantial amount of engineering is required to get it up and running, and at the end of the day, it is just a second source of truth. Why would we buy a CDP when all of our data modeling is already being done directly in Snowflake?
Head of Data & Analytics
CDPs try to do everything but aren’t masters at anything. They are also black boxes with no observability. That’s why we went with Hightouch and our data warehouse, which acts as an olive branch between data and marketing teams: letting marketers build audiences without an engineer.
Head of Engineering
Hightouch has made it really easy to sync our data in Snowflake to our various marketing platforms so we can optimize around our most successful customers. With Hightouch we can duplicate our audiences and send data to any destination of our choosing immediately.
The problem with any out-of-the-box analytics suite is that you are restricted to only doing what it supports natively. CDPs are not extensible to what you want them to do.
Strategy & Analytics
You can’t run a modern MarTech program without Hightouch.
Head of MarTech
The new era: Hightouch
By simply integrating Hightouch to their existing data warehouse, organizations are able to form what has since been coined the “Composable CDP” by the likes of cloud leaders like Snowflake and Databricks. As the name suggests, the Composable CDP offers a modular, warehouse-centric CDP alternative. Like its off-the-shelf counterparts, the Composable CDP addresses data collection, transformation, and data activation, however, any tool can be leveraged at each step to ensure that the specific needs of your organization are met in a timely (and costly) manner.
Typical composable CDPs have the following solutions as components:
- Event tracking (often Snowplow or even Segment/Rudderstack): Generate, enhance, and model rich, quality behavioral data across all platforms and channels in a common format and stream it into your data warehouse or lake.
- ETL (often Fivetran): Replicate data from your SaaS tools and databases across marketing, sales, finance/IT, product, etc. into your data warehouse.
- dbt: Once all your raw data has landed in your data warehouse, you can use SQL to clean up and transform the data into clean tables/views.
- Hightouch: Sync data from the data warehouse into the tools that business teams rely on, e.g. Salesforce, Marketo, Facebook Ads, etc.
Perhaps the most exciting aspect of the Composable CDP is that pretty much every organization with a data warehouse already has most of the components in place. That means there’s little additional implementation time or cost to embrace a warehouse-centric approach.
What are the benefits of Hightouch?
By harnessing the power of best-in-class tooling to create a Composable CDP, there are four key benefits over a legacy CDP.
360 Degree View of Customers
While legacy CDPs claim to provide a single source of truth for an organization’s customer data, the reality is they only produce a fragmented copy. With Hightouch integrated with a data warehouse, organizations can activate ALL of their customer data - not just the clickstream events captured in off-the-shelf CDPs. Now, organizations can quickly and easily leverage data from sources like POS systems, data science models and even offline data to create a true 360 degree view of customers.
Legacy CDPs lock customers into an opinionated and rigid data model. While these data models may look clean, those who have implemented CDPs can attest that in reality, businesses come in all shapes and sizes and need much more flexibility. Data warehouses have what CDPs lack — the ability to model and query arbitrary relational data. For example, representing relationships between users and entities like accounts, teams, workspaces, or devices. The warehouse provides organizations the infinite flexibility needed to support their most complicated use-cases. We at Hightouch have a core belief that your CDP should conform directly to your data - and not the other way around.
Time to Value
When buying a legacy CDP, you're forced to evaluate everything up front because you're locked into a single platform. Combined with painfully long onboarding times, it will be months before you can even plan any activation. With Hightouch, organizations are able to skip the implementation phase - simply connect to your warehouse and activate your existing customer data so that you can launch campaigns in minutes (not exaggerating), not months.
While legacy CDPs may be able to do a little bit of everything, they hold themselves back from being the best at anything. Because Hightouch is built on an organization’s existing data infrastructure, our teams have been able to focus our investment on quickly building the features that marketers want most. Supplementing our audience builder, a visual field mapper allows for teams to seamlessly customize how customer data is passed to each specific destination - while advanced features like Splits and Overlaps ensure that marketers have the tools they need to manage the most complex marketing campaigns. This is just the beginning.
Legacy CDPs store their version of your data within their own cloud environment - introducing security risks and restricting the data that can be managed in the platform. With Hightouch, your data is stored in your warehouse - allowing you to remain fully compliant (i.e GDPR HIPAA, & CCPA) while leveraging the power of a CDP to supercharge your marketing efforts.
The road ahead
As our team heads into the back end of the year, we are more determined than ever to continue to create a best-in-class activation platform that rivals the best audience management tools available in the market. Feature releases announced today, like Audience Overlaps and Audience Splits, push us further ahead - giving users deep insight as they build audiences and enabling omnichannel multivariate testing capabilities practically unseen in the CDP market today.
As I predicted in my last post, the major CDP players will begin to develop reverse ETL capabilities as they scramble to stay relevant. We fully expect all CDPs (and even some marketing platforms, like ESPs and CRMs) to eventually come around and embrace the data warehouse as their source of truth, further validating our approach from the beginning. Hightouch is and always has been purpose-built for the warehouse. Modular components being tacked onto existing CDPs and other martech may sound nice, but at the end of the day will run back into the same architectural limitations that landed them where they are today.
At the end of the day, our team–stacked with years of experience building and buying leading CDPs–has landed on one similar conclusion: Friends Don’t Let Friends Buy a CDP.
Learn more about Hightouch Audiences and the Composable CDP, or request a demo. If you’re ever interested in nerding out about the data or marketing technology space, shoot me a note at firstname.lastname@example.org.