Marketing analytics is clearly going to be a huge business driver in the near future, but for now it’s stuck in park. The same survey of CMOs by Deloitte, Duke University’s Fuqua School of Business and the American Marketing Association that predicted a 56% increase in marketing analytics in the next three years also noted that CMOs rate the performance impact of analytics at 3.9 on a scale from 1 to 7. That’s basically a D minus.
But even more informative may be the fact that these numbers have apparently fallen--over the last two years, it was 4.1. Considering the innovations in marketing analytics and the growing volume of data over the past decade, for many this defies common sense. However, when you peel back the onion, it makes perfect sense.
First, we must take into account that data growth is a function of advances in the technological ability to generate and collect data, but it is rarely guided by marketing strategy. Consequently, data collected by the various systems a company has in place is typically incoherent, full of missing values and inconsistencies. Often the data necessary for marketing decisions resides in separate repositories--some might be in Hadoop, some might be in a data warehouse, and some might reside in a departmental silo. This makes it difficult to find the cause and effect in the data. It also makes it difficult to get your hands on the right data in the first place.
Furthermore, marketing requires current data, and this is a bit of a trick as marketing data is constantly evolving in the face of real-time events. Consider for instance how much of a monkey wrench coronavirus has put into marketing plans. No one even knew about it a few weeks ago, and now it’s shutting down sectors of the economy. Getting data that’s 3 months old isn’t really going to be helpful when it comes to making critical strategic decisions.
Making due with what’s already in place
An obvious priority for companies should be basing data collection strategies on marketing processes and goals. But how realistic is this? Think about what it takes to revamp a company's data management strategy--the term “gargantuan undertaking” doesn’t quite do it justice. Plus, it’s just not something a CMO has any real control over--suggest it all you want, but it’s not going to happen. Any realistic improvements in marketing analytics will most likely need to make do with the data infrastructure that already exists, which most likely will be highly fragmented and inconsistent.
This means that marketing needs a way to quickly find the data they need regardless of where it is and quickly figure out how to assemble it to answer the burning question at hand. But, how do you do this without an advanced understanding of data architecture, engineering and SQL, which virtually no marketing executives have? It all starts with the question. BI/Analytics is or should be about answering questions. So, any viable solution must enable users to ask a question and translate that request in a way that the machine understands. The process starts with a simple Natural Language query. For example, if the CMO needs to know the retention rate of female 18-35 customers in Florida, they must be able to pose the question in English (or their spoken language) to the data management system. The system in turn needs to identify the appropriate data across multiple systems. At that point, it can be handed over to analysts who can get straight to the analysis without having to hunt for it (where data analysts currently spend the majority of their time).
Integrating marketing and analytics
Another obvious step that needs to take place: better integration with marketing and the analytics team. But how do you do it? The ugly reality is that when a marketing executive poses a question to the analytics team, they basically have zero visibility into what happens after that. They request mission-critical information and then wait--often months--with no clue as to what’s really going on. By the time they get the answer, it’s too late--market conditions have changed and opportunities have disappeared.
This can only be solved by providing full workflow visibility, so that when an executive asks a question, they can monitor in real time who it got assigned to, and their step-by-step progress in providing an answer. Such a system would also allow the company to pinpoint where the bottlenecks are occuring, and take appropriate steps to get things back on track. Finally, marketing needs to be able to manage it all through a single platform that not only provides visibility into the workflow of a particular answer, but also provides full visibility into all analytics questions that have been posed--otherwise you’re just rebuilding the boat every time you want to take it on the water.
All of this supports what every seasoned marketer knows that their efforts need to be: agile and iterative, able to rapidly implement new ideas and make adjustments on the fly.
Making it as simple as a Google search
The benefits of effective marketing are abundant, as marketers tackle a litany of challenges, starting with:
Audience segmentation and understanding/predicting customer intent and behavior
Predicting customer churn
Analyzing markets, industry trends and cycles
Product design, iteration and pricing
Campaign optimization
Attribution modeling and understanding the customer journey
Adjusting goals and objectives
In a nutshell, marketing analytics needs to take a page from Google Search. Marketers need to be able to search for answers in the language that they speak--not SQL--and be able to gather the necessary data without any concern for where it might actually reside. When that happens, you’ll see the rating that CMOs give analytics go from 3.9 out of 7 to a 7 out of 7.
Find out how to make this happen.