2023 Predictions: MarTech has a “Data Problem”

When MarTech wakes up on January 1 of 2023, it will find itself with a particularly bad hangover. Yet unlike previous years where the effects dissipate relatively quickly, this year MarTech is in for a real headache that isn’t going away anytime soon.

The good times are over. Budgets are slashed, the macro is uncertain, and consumer behaviors are changing.?

Gone are the days of seemingly free money to be used for “growth at all costs”.

Gone are the days when new customer acquisition budgets were seemingly only limited by Facebook’s ability to scale audiences.

Gone are the days of the freely spending customers with their pent-up COVID demand with lots of excess cash.

Underlying this is a set of three developments that have been underway for the past several years - and all of which are now alliding in a way that will cause material disruption for the category. These developments span economic, technological, and ecosystem changes - each of which on its own would be hugely disruptive.?

First and to no surprise, the economy is finally in a bad place after a 15-year bull market. This will put pressure across the entire category - and technologies that are actually driving value will persist. While this is true for all categories of spend - what’s unique about MarTech is the tremendous confusion that exists within the category. Over the past 15 years of a bull market, the category has exploded to over 10,000 vendors - and as organizations look to reconcile their overly complex MarTech architectures, they’ll undoubtedly find a high degree of redundancy.?

Second, the cloud has enabled a next-generation of data platforms that are finally emerging as a massively disruptive force. In fact, the greatest MarTech redundancies are rooted in the closed loop and insular data strategies that define so many MarTech tools and applications today. Led by Snowflake, the cloud data platform today has emerged as the most complete and accurate view of the customer across the enterprise. And for many MarTech tools, this shift represents an architectural shift in their core data capabilities.

Finally, the deprecation of the cookie has had a series of cascading effects. In end marketing use cases, advertising has become more expensive - which compounds with macro-economic factors to put pressure around downstream opportunities to optimize conversion rates and LTVs. These new use cases further propel first-party data strategies and cloud data warehouse adoption.?

With this as a backdrop, let’s get on to the predictions.

Prediction #1. The cloud data warehouse will emerge as the backend for a new set of MarTech tools.?

As I detailed earlier this year, Snowflake is on a path to redefine the future of the data application - and marketing seems to be a core area of focus.?

These new architectures unlock a new world of data - powered by cloud-enabled enterprise data strategies that provide incredibly accurate and complete views of the customer. They also bring other key benefits across governance, data science, and more.?

The category will start to fracture across those who align with this strategy and those who do not. The latter will find limited success within a dwindling population of enterprises that haven't progressed against their cloud data strategies. Incumbent marketing clouds, including Salesforce, will push their “Genie” data offerings as they conflate differences between CDP and enterprise data strategies. With this, they’ll attempt to compete for head to head with Snowflake & BigQuery, and they’ll lose.

Successful MarTech adopters of a warehouse-centric strategy will be fully integrated with the underlying data infrastructure in a way that serves three main benefits: (1) complete parity with the underlying data - whatever is in Snowflake is in the application, (2) data backends that are fully transparent and extensible vis a vis underlying infrastructure capabilities, and (3) fully integrated governance.

I’ll get deeper into #3 in a subsequent prediction.

Prediction #2. Reverse ETL will transform from a standalone category to a check-box feature that MarTech tools tout on their websites.

As the cloud data warehouse proliferates with SQL as the key interface, MarTech apps will quickly realize that they need to change their ingestion patterns. We've started to see this already across multiple MarTech categories starting with Braze and now also Segment and mParticle.

While these moves will be disruptive to emerging vendors within the reverse ETL category - we’ll also start to see the effects of incumbent vendors attempting to ingest massively complex & large datasets that previously were siloed within the warehouse. This will put massive pressure on the core data capabilities (or lack thereof) of these systems - and create a next wave of challenges that certainly won’t be solved in 2023.?

Prediction #3. "Let's talk about governance", says the modern CMO

Data privacy and regulation laws are not new - GDPR has been around since 2016. Still, with more individual countries creating their own rules and newly forming regulations such as Schrems II, it will become more complicated for marketers with global audiences to maintain compliance while leveraging data for effective engagement.

From a data perspective, these privacy regulations are a subset of a broader set of concerns around data governance that includes security, compliance, privacy, and data quality. Today governance is poorly understood or talked about outside of IT circles - but this will change in 2023.

For example, reverse ETL is terrifying from a governance perspective. The idea of copying out your customer data to all your MarTech applications is insecure (a hack to any one of them will expose data), has compliance challenges (supporting GDPR erasure across multiple applications), and results in quality issues (with each application having different data schemas and contracts).?

Governance is complicated, but that doesn't take away from how critical it is. Major changes won't happen until 2024, but next year will be a foundational awareness year.

Prediction #4. Layoffs will continue next year, and the data scientist won’t be excluded this time around

To be successful in business requires a clear connection to end value. Having a Ph.D., working with petabyte-scale data, or developing complex multi-layer neural networks - none of these grant immunity to this simple truth.?

The problem with so many data scientists and data science teams is that they just don’t get it!

But for those who do - they will continue to collaborate with marketing counterparts in an environment that will require a much finer-grained approach to optimization and efficiency gains. Optimizing customer lifetime value requires a deep understanding of customer behaviors and the ability to model it. It’s not just about predicting customer churn - but working cross-functionally with marketing teams on how to use these predictions to change customer behaviors and avoid negative outcomes.

Doing so requires a degree of business acumen - but it also requires the right tooling. MarTech stacks that enable seamless connectivity from data and data science environments will be critical. This plays directly back into prediction #1 - and the ability to seamlessly work across data & marketing environments is the table stakes requirement.

This prediction will ultimately lead to changes in the fundamental role of the data scientist. Many of the “generalist” data science roles we see today will absorb into business functions to focus on marketing insights or product analytics. Data science specialists who are working on well-understood business-critical problems (for example, training neural networks for Google’s search algorithms - or ChatGPT) - will continue to thrive.?

Prediction #5. Every e-commerce business will also become a media business?

COVID brought all retailers online, and online competition is now more intense than ever. As consumer spending continues to soften, retailers will look closely at their conversion costs and look into opportunities of monetizing their traffic through brands that are willing to pay for traffic.

This trend is already well underway with Netflix launching an ad-supported tier and will further accelerate first-party data adoption as retailers scramble to understand both the advertising and conversion dynamics of their customers.?

Amazon has been doing this for years and has a dominant presence today. Expect many more brands to follow - and for some, like Netflix, to emerge as primary ad destinations.?

Prediction #6. The convergence of AdTech and MarTech

This one really sits at the center of many trends. First, the cloud data warehouse will have similar effects on AdTech as it will have on MarTech. Among other things, this will enable consistent data views across the entire customer lifecycle - along with the key benefits described in my first prediction.

Second, pressures around general business efficiencies will force brands to optimize CAC to LTV ratios by looking holistically across the customer journey from ad bidding to conversion optimization to core retention. The “handoff” between acquisition and retention teams will necessarily blur - and in pursuit of optimizing this ratio, teams will need to rethink how they work together and collaborate. This collaboration will be a massive forcing function to create more consistent tooling across what’s historically been a disjointed set of AdTech & MarTech applications.

Finally, cloud enabled data will also provide more seamless sharing from third-party acquisition contexts to first-party engagement. Cleanrooms have risen in popularity in the past couple of years - platforms like Snowflake permit seamless sharing of cloud-enabled data across first, second, and third-party contexts.