We are Marketbridge.

Growth consulting and marketing for modern leaders.

Play Video
Play video

Ready to reinvent growth? Let's connect.

Andy Hasselwander
Read Time: 4 Minutes

Share:

Data-driven marketing has become the go-to standard, as marketers increasingly rely on empiricism to understand customer behavior, preferences, and needs. However, promises of better results have often been missed by marketing analytics teams.

Over the past decade, hundreds of billions of dollars have been poured into marketing technology, data assets, and talent. But, the party may be ending: Gartner predicts that CFOs will slash marketing analytics teams by 60% in 2023, and a likely recession bites. While macroeconomic forces are partly to blame, marketing analytics teams’ performance over the past decade is also responsible; Gartner also found that analytics teams influence only about half (53%) of marketing decisions.

However, there are actions that marketing analytics teams can take to increase their effectiveness—and sweat the assets that they already have. Over the past decade, while working with over 50 marketing analytics teams, we have identified six common issues that hinder effectiveness. By understanding these pitfalls upfront, marketing analytics teams can operate in a more agile, scientific way.

  1. Technology Often Fails to Cure Marketing Ailments
  2. Poor Reproducibility Leading to Ongoing Maintenance Problems
  3. Marketing Evolves Quickly
  4. Marketers Are Overly Dependent on Engineers
  5. No Reporting Standards
  6. Fire Drills Trump Long-term Vision

1. Technology Often Fails to Cure Marketing Ailments

Marketing technology software has been a huge business for 20 years. CRM, marketing automation, digital asset management, and digital marketing platforms are all major investments for large enterprises. In many cases, upwards of $20M can be spent on a single martech platform migration and installation process, employing armies of consultants, and taking years.

These projects are undertaken based on promises that are too often wildly unrealistic. A marketing automation platform might promise end-to-end lead tracking, easy campaign configuration, and seamless integration with CRM—but after years, the same data problems persist, and in some cases are even worse.

Furthermore, after installation, we found 50% to 80% of the functionality of these so-called magic bullets goes unused, particularly since much of it is force-fitted in the first place.

Instead of relying on out-of-the-box solutions for analytics, use marketing software for what it was designed for: contacting customers. Build analytics and data architecture internally. By keeping your marketing data warehouse/data lake as internally built and maintained resources—and ensuring that data flows into these systems from each new marketing technology platform—data discontinuities can be avoided.

2. Poor Reproducibility Leading to Ongoing Maintenance Problems

Extraction, transformation, and loading (ETL) are the processes that move data from system to system, or from source systems to data warehouses. In the short run, it’s easier to build these flows using graphical tools or to manually create “hacked” processes. Many marketing organizations’ data pipelines are a hodge-podge of varying batch, manual, and streaming processes that are poorly documented.

To avoid this, marketing analytics departments should demand the use of text-based (SQL, Python, Apache Airflow) ETL or ELT processes, coordinated via a version control system like GitHub. All pipelines will then be transparent and traceable. If data engineering capabilities become a bottleneck, marketing analytics departments should train this capability broadly, making data pipeline creation and maintenance a core capability.

3. Marketing Evolves Quickly

Like other industries, marketing changes and progresses in the blink of an eye. Every year, new terms, systems, and strategies must be learned. Usually, these channels and technologies are designed without data structure standards.

By establishing a standard taxonomy of marketing channels, customer segments, and products—and ensuring that new marketing technology matches this technology—data continuity can be established through disruption. This is sometimes the rule of the data governance leader, but it is incumbent upon marketing analytics to advocate for metadata consistency.

4. Marketers Are Overly Dependent on Engineers

Most marketers are creative, analytical, and organized, but they tend to be avoid coding. As such, they tend to depend on others to design, develop, and integrate tech and data. This dependence on parties outside of the marketing organization can lead to bottlenecks, delays, and poorly configured hard-to-maintain systems.

Largely due to this inefficiency, we foresee that by 2030, 50% of marketing jobs will require coding-type technology skills. These individuals will also be required to understand different data structures and access procedures that marketing data uses. Think JSON, XML, and .csv files, along with batch access (FTPs) and APIs. What seems unlikely today will likely become table stakes for marketers going forward.

5. No Reporting Standards

Whereas finance departments are required to produce quarterly and annual income statements and balance sheets, marketing doesn’t have required standard reporting. CMOs and other marketing managers are left to build a patchwork of dashboards, or purchase software promising to piece together the puzzle.

To catch up to finance, marketing should focus on simplicity first. “How much, how many” reporting simply counts spending, stimulus, leads, and other key data by standard groupings (the taxonomy mentioned in three above). Once standard reports are established and rigorously QA’d, marketing analytics can move on to more advanced reporting, like multi-touch attribution.

6. Fire Drills Trump Long-term Vision

When attempting to understand marketing performance, unfortunately, short-termism dominates long-term thinking. Immediate fixes are always simpler than putting in the effort to develop solid data pipelines, universal taxonomies, and organized data frames.

To remedy this, marketing analytics teams should maintain product roadmaps. These roadmaps aren’t software roadmaps—they should outline the capabilities and use cases that will be supported on a quarter-by-quarter basis, looking out at least three years. These roadmaps should be tied to manager compensation, to ensure that they are actioned.

These Challenges Are Repairable

While these six issues are real for marketing analytics teams, they can be fixed. The fix starts with establishing a baseline of current capabilities.

What’s Next?

Marketing mix modeling example

Explore our comprehensive marketing mix modeling example to understand how various channels and factors impact business outcomes.

How does multi touch attribution work?

How does multi-touch attribution work? Discover how MTA assigns value to every customer interaction to improve marketing effectiveness.

MMM vs MTA 

Discover the key differences and benefits of MMM vs MTA in marketing attribution and how both methodologies enhance marketing effectiveness.

Subscribe
to our insights

Ready to
reinvent growth?

Skip to content