A cure for marketing’s ‘Frankenstack’ syndrome

For marketing leaders to design the optimal technology stack, they must start to focus on customer profiles and not channels and applications.

sales-marketing-planning-0.jpg

 A Cure for Marketing's 'Frankenstack' Syndrome? A 5 step approach. 

The latest research from the 2021 Connected Benchmark report found that the average organization runs 843 individual applications. The report also found that only 29% of applications on average are integrated. With the average lifetime of an application being just four years, organizations must evolve away from being hierarchical and hardwired toward being flexible and open to change -- a composable enterprise. 

Marketing departments represent a line-of-business with many business applications that are mostly not integrated. The marketing technology landscape has exploded in the past decade. Just look at the trend of Scott Brinker's Marketing Technology Landscape logo map over the past ten years:

martech-landscape-2020-martech5000-slide.jpg

Marketing Technology Landscape Supergraphic 2020 - Scott Brinker

State of Marketing research from Salesforce shows that marketing leaders top challenges are customer engagements in real time, innovation, creating a cohesive customer journey across all channels, unifying customer data sources and sharing a unified view of customer data across business units. None of these challenges can be successfully met if marketing leaders are unable to define the optimal technology stack for the marketing jobs to be done. 

To better understand how marketing leaders can manage their modern technology stack, I asked one of the smartest digital marketing professionals that I know to share his thoughts. Martin Kihn, senior vice president of Product Strategy at Salesforce, and co-author of Customer Data Platforms, has studied digital marketing organizations and technology landscapes for many years. Here are Kihn's thoughts on how marketing leaders can design the optimal technology stack:

dv-kin-verticle-thumbnail.jpg

 Martin Kihn, senior vice president of Product Strategy at Salesforce, and co-author of Customer Data Platforms

Choice is not a hardship, but this rampant vendoritis reflects a real problem with most modern marketing tech stacks: in the words of Avril Lavigne, they're complicated. Forces of inner fragmentation and outer proliferation, combined with the pressure on marketing to deliver faster returns, result in that now-common phenomenon of the "Frankenstack."

Frankenstacks come in all shapes and sizes -- that's what makes them Frankenstacks. They have in common an ad hoc design-in-flight structure, where a need to connect and reconnect systems that were not designed to work with one another struggles to serve the ever-changing needs of impatient business users. The result is a hodge-podge of technologies, each with its own particular purpose, the sum total of which can obscure the actual customer it's trying to serve. User-friendly, it's not.

How did this happen? Well, like most accidents years in the making, it wasn't planned but burst from a build-up of real-time reactions to events outside the marketers' control. Let's call these these the Two C's of Marketing Chaos.

These are Chaotic Channels and Chaotic Communication

Chaotic Channels: In 2007, smartphones appeared on the scene with the launch of Apple's iPhone, and within three years about 50% of the U.S. population had an iPhone or other smartphone. Today that number is closer to 80%.

The growth of social networks was even more dramatic. When Facebook began making connections off-campus in 2005, about one in twenty U.S. adults used social media. Today, 72% of U.S. adults use an increasingly diverse array of networks, from Facebook  (69%), YouTube (81%), Instagram (40%) to Twitter (23%), LinkedIn (28%) and TikTok (21%). As the leading network, Facebook (which also owns Instagram), reaches approximately one in three people (2.8 billion MAU) on the planet today.

Chaotic Communication: As each channel appeared, a team was spun up to support it; and over time, these teams did what teams do -- focused on itself. Cross-channel internal communication was strained and often broken. Volatile consumer-driven developments -- We need an iPhone app! We need an Influencer strategy! -- put unprecedented strain on already stressed marketing teams, forcing them to adapt to new modes of communication and consumer preferences, even as they tried to reboot their data collection, analytics and creative management functions.

For example, take a luxury automotive brand I worked with for some years. In the 2000's, they were doing sophisticated CRM programs using email channels in a batch-and-blast form, and employing a relational database and SQL queries to pull lists from a store of customer profiles. A separate team was created to manage and run the brand website. When smartphone use grew, in the 2010's, a separate team was created to manage the "mobile strategy," and this team acquired mobile analytics software. Then social networks became important, and a separate "social team" was created, which in turn acquired its own social management and listening tools.

The result was a natural process of accumulation, both of tools and teams, and the brand ended up with what its Chief Digital Officer described as "a hot mess." There were at least four uncoordinated teams, and three different agencies (creative, media planning and CRM), each aligned around a different channel. They did not share information because their channels were fiefdoms. Even more critically, each team had its own martech and ad-tech systems, each of which had its own customer profile.

cdp-book-cover.png

Customer Data Platforms: Use People Data to Transform the Future of Marketing Engagement, by Kihn and O'Hara


So by the time the brand did a data audit, they discovered that a single customer could have data sitting in up to twelve different databases, which were not connected.

The result? Obviously, the marketing processes and tech budget were not optimized or efficient. Tasks were duplicated, and simple updates -- such as changing a customer's last name when they got married, or updating their address when they moved -- took too long. But the worst business impact was on consumer experience. To put it kindly, the customer's experience was not coherent. They might enter information on a website and have to repeat it on the phone; take an action on a mobile app that was not reflected when they logged into the website; follow one of the brands on social media and be sent an invitation to follow that same brand via email.

Bear in mind that this brand was successful; its digital marketing had won prizes. All of its team members were competent professionals, and its agencies were highly respected. It was not lack of intelligence that led to the "mess" they were in. It was simply the rapid rate of change -- "Internet time" -- they endured, and the nature of ad hoc adaptation that led to creation of a Frankenstack.

There is a way out of the Frankenstein syndrome

In recent years, more advanced teams have begun to rethink their customer data architectures in fundamental ways, leading to the current interest in Customer Data Platforms (CDP). Often in concert with broader enterprise business transformation efforts, marketing technologists and their IT colleagues have put time into conceptual exercises or North Star marketechtures pointing the way toward a more efficient future.

picture1.png

Personalized conversations - analyze, know, personalize, and then engage


What would an idealized marketing architecture look like? It would differ by industry and requirements, of course, but its high-level components would probably look something like this. It would require:

  • Unified User Profile: A unified customer profile that included data from all the various systems you've built up over the years, matched by ID and harmonized for use
  • Smart Segments: The ability to do segmentation, machine learning and AI on this data
  • Plan and React: A campaign management and decision capability, where the marketer could set up pre-planned journeys for customers and prospects, as well as develop rules (including machine learning-driven decisions) to handle in-bound events such as an unknown customer arriving on an website or mobile app
  • Engage: The ability either to directly engage with the customer (e.g., by sending an email or SMS message) or interact with engagement systems that reach the customer on their channel of choice, often via API's
  • Optimize: Finally, the ability to close the loop by capturing conversion and other signals, reporting results, and enabling or even recommending ways to improve outcomes

The good news is that digital marketing teams have become more adept at cross-channel coordination. The reactive departments that built the Frankenstack have evolved into strategic planners focused on long-term solutions. They know the answer won't come quickly or via rip-and-replace, but rather through a focus on foundational elements of the stack such as the all-important 'single source of truth.'

So what's the cure for the marketing Frankenstack syndrome? Focus on customer profiles and not channels and applications. And when a marketer has a stable, complete view of their customer data, she can finally say, "It's alive!"


This article was co-authored by Martin Kihn, senior vice president of Product Strategy at Salesforce and co-author of Customer Data Platforms. Previously Kihn led a research and advisory agenda as Research VP at Gartner around programmatic advertising and media, digital marketing analytics and the application of data science to marketing and advertising. Before joining Gartner, Kihn worked for ten years at digital marketing and ad agencies in NYC and Minneapolis. Kihn is also author of three memoirs, including "House of Lies: How Management Consultants Steal Your Watch and Then Tell You the Time," which is the basis of a Showtime series of the same name.