The Unlikely Marriage of Data Warehousing & Marketing

Guli Zhu, Head of Marketing Analytics, Square

Traditionally, CIO and CMO organizations have operated separately with different mandates. One was responsible for technologies that enabled company operations, while the other was responsible for marketing and advertising to drive revenue growth and brand awareness. However, the most successful marketing practices today have shifted away from a siloed “Mad Men” approach to one that leverages technology and science to complement and improve its art. Advances in data warehousing technologies have built the foundation for quickly and scalably unifying marketing data across many disparate sources - making CIO and CMO organizations even more inextricably linked.

Data Warehouses

Seasoned marketers are often familiar with a large rolodex of industry technologies - multi-touch attribution models, ad servers, DMPs, DSPs, CDPs - but less so on the foundational role data warehouses play in making their ad spend more effective. Where does website clickstream and cookie data live? What about data around how many customers redeemed the latest promotional code? How can we consolidate cost data from all advertising partners and build a dashboard to track QTD spend against budget? How do we combine our marketing data with critical customer attributes, like lifetime value, multi-product adoption, or even fraudulent activity, to better measure the true ROI of marketing campaigns? Whether marketers realize it or not, these sets of data are being collected, stored, cleaned, and consolidated in their company’s data warehouse. This means that firms need to break down silos between marketing and IT; they’re each other’s critical stakeholders and must closely collaborate to ensure effective ad spend.

Cloud-based data warehouses like Snowflake are not only faster than many on-premise alternatives, but also much more efficient

Challenges with On-Premise Data Warehouses

Why should marketers care about something that traditionally belongs with an IT team? Historically, IT teams maintained on-premise data warehouses. If marketing wanted to store more customer data, they had to convince their IT team to order more servers, which was an expensive and time-consuming process. Marketing teams had to share a fixed amount of database compute power with the rest of the company, so someone else’s bad query could actually break reporting needed to track budgets and optimize campaigns. On top of all this, the company not only needed to employ an army of database engineers to maintain the underlying systems, but also manage large business intelligence teams to build and maintain codebases of pre-defined data cubes. Outside of the Fortune 500, companies rarely had the resources or scale needed to justify such a large undertaking.

As a result, marketers had no choice but to look for 3rd party platforms to consolidate and use their data. These platforms would provide out-of-the-box solutions that would sound too good to be true - and were. They couldn’t meet the inevitable client-specific business needs, and no client could look under the hood to ensure data was correct and being used appropriately; brands’ own data sets were essentially locked into black box platforms. There are specific reasons why a brand might want to leverage 3rd party platforms, but it should be an intentional choice with the benefits and trade-offs thoughtfully weighed - not something where their hand is forced because of operational complexity and cost.

Cloud Data Warehouses

Recent data warehousing trends, like migration to the cloud and separation of compute from storage, have transformed what marketing organizations are capable of, enabling small teams to leverage best-in-class tools without significant overhead. Cloud-based data lakes have enabled a centralized, cheap place to store any and all data needed to understand the customer journey. Cloud-based data warehouses like Snowflake are not only faster than many on-premise alternatives, but also much more efficient - they enable teams to quickly spin up large compute clusters for activities that require heavy lifting, such marketing attribution or predictive modeling, but then also quickly ramp down those same resources when not in use. Marketing & IT organizations can now pay for only what they use, and are limited only by what they are willing to pay. No longer will there be the perennial data problem of growth companies - a long and exponentially growing backlog of users and queries waiting for a very finite amount of resources that never seem to scale fast enough.

An Ecosystem to Empower Marketing Teams

In addition, there are now technology providers like Fivetran that specialize in bringing 3rd party data into a brand’s own data warehouse. All marketing teams need to do is provide the right credentials for their services (Salesforce, Facebook, Google, Marketo, SFTP, S3 buckets), and data will immediately start syncing into their data warehouse. No more APIs to call from, no more legacy infrastructure to build and maintain. This empowers all teams to leverage their core competencies to truly drive business results, rather than spend significant portions of their time and energy on data plumbing setup and maintenance. Analysts can focus on campaign optimization analytics, data science on predictive modeling, business intelligence on data visualization and reporting, and engineers on critical infrastructure that can’t (or shouldn’t) be contracted to 3rd parties.

Leveraging the Holy Grail of Data

Both cloud infrastructure and its corresponding data ecosystem are enabling technologies that have democratized what was previously only accessible by the largest brands with the most engineering resources and financial capital. It has laid the foundation for CMO and CIO organizations to work more closely together to improve automation, reduce costs, and mostly importantly, move faster on critical growth initiatives necessary for long-term company success.

Brands large and small will be now able to choose their own destiny on how and when to use their new centralized source of truth of customer data. They will be able to more quickly, easily and cheaply marry first, second and third party data across on- and offline channels to create a single, unified view of the customer with all acquisition and retention marketing touchpoints. Marketing teams will then be able to leverage this newly unified data store to achieve the holy grail of marketing - creating hyper-targeted campaigns that give the right message, to the right person, at the right time, and in the right channel.