6 Best Practices for Getting the Most from Your Marketing Data Analytics Strategy
Do you remember the lessons from high school history classes about the California Gold Rush? I used to enjoy learning how men collected their families and their belongings into their covered wagons and headed out west to mine for gold and, hopefully, strike it rich. The gold prospectors of the 19th century wouldn’t be lost in the great data rush of the 21st century.
For marketers, much like California’s rivers, the ever increasing availability of rushing streams of customer data present great opportunity (all without having to load your family into your horse-drawn covered wagon). Taking advantage of that opportunity, however, requires forethought and preparation. Below, I offer six best practices to consider as you embark on your data journey.
- Begin with a Question
A common challenge people run into with data is the “now what”. You go through the process of assembling a data team, identifying data sources, installing a data management platform (DMP) and being excited about having access to all of this data but… now what?
Now what is the result of not starting with a question. In a previous post, I detail how the experimental nature of the scientific process can be used to guide marketers’ campaigns. The very first phase of the scientific process is Start with a Question.
The question is key to understanding the story you want to tell and the metrics that help you convey that story. Perhaps you want to use your data to create a customer lifetime value (LTV), create customer segments or predict churn rates. Your question should be formed to support that goal and to establish a mission, of sorts, for your data analysis efforts.
- Think Big, Act Small
Experience has shown me that the natural inclination with implementing most technologies is to implement the tech, liberally apply it to all problems and expect magic. Experience has also proven that to be a recipe for failure.
It’s great to have big plans for your data efforts but realizing the goals you set is, generally, best achieved by taking baby steps. Strategically, you should absolutely consider what you ultimately hope to accomplish with your data — defining the ideal situation that allows you to realize positive ROI from the investment in the data program. Still, for myriad reasons, you should expect a long, hard slog to get there. Instead, begin with a small team with a small goal that helps familiarize your team with the technology’s capabilities and better positions you to evangelize the benefits of the technology as you reach deeper within your organization to get the data you need to achieve your long-term goals.
- Seek Acceleration Over Automation
Much of the hype over data science techniques like machine learning, deep learning and artificial intelligence has been over automation — using algorithms to remove humans from operational processes. The fact is (some may disagree with my ‘fact’), the technology just isn’t there yet to support full automation. The technologies are great at tasks like describing behaviors or using models to predict activity. Artificial intelligence, however, isn’t yet ready to make decisions unaided by human intervention.
The goal for marketers should be to use data science and analytics to accelerate decision-making by providing your team with additional tools they can use to become more efficient and productive.
- Pay Attention to People, Process and Tools
Making data work for you will require change. If your goal is truly to become a data-driven organization, the changes will reverberate through your team and beyond. You need to get ahead of it.
Governance, leadership management, training and user acceptance all fall under the People banner. Who/what team will own the your data? Who owns the data sources that feed your data? Who creates rules about adding, maintaining and removing data? What support is needed from your management to support your data efforts? How far up the chain do you need to go to achieve your goals? How will you help your manager convince her superior to buy in to your objectives? What will be your approach to training your team to use the data technologies you implement? Beyond training, what steps will you take to encourage actual adoption rather falling victim to the all-too-common try it once and return return to the old way?
Process cannot be overlooked. Regardless of what your stated primary goal is for seeking data analysis, your number one unstated goal is changing something about the way you work. That means changing the way your employees get their jobs done and, perhaps, delivering a redefined customer experience. You have to identify process inflection points where the use of data are most likely to help you achieve your goals. Process flow diagrams can benefit you by providing an illustrative representation of your processes. Moreover, buyer personas and journey maps can aid you in understanding your customers’ purchase processes and identify inflection points where data can bolster your ability to convert or retain customers.
Regardless of how you approach data analytics implementation, you will be using multiple Tools to help you get there. There is no one size fits all approach to data technology. The scope of your implementation, your data sources, the volume of your data, the types of data you need to manage, the velocity of your input and output data streams and, of course, your objectives can impact the tools you use. Moreover, it’s not likely you’ll use only a single application for analyzing all of your data. Data wrangling (I.e., converting data from a raw form into a usable one, visualizing data (I.e., using graphs to better understand data) and the actual data analysis may all rely on capabilities of multiple types of applications. It may even be the case that applications are specific to domains and/or challenges (e.g. customer acquisition, churn, attribution, content optimization, etc.)
- Trust Occam’s Razor
It would be a superb understatement to claim data analysis was easy. Even with a skilled data science team it can be challenging identifying where and how to use data science and data analytics to optimally influence your organization. The thing is, however, we shouldn’t over-complicate them either.
In an effort to perfectly model behaviors or to have highly accurate descriptions of activities, we sometimes allow analysis paralysis to take over. The principle of Occam’s Razor suggests the simplest answer is the right one. Generally, so too is it the case with data analytics. When potential complications arise, ask first if there is a simple solution for resolving your challenge. From data science techniques to metrics, start with the simplest approach first.
- Communicable Insights over KPIs
I recently came across a beautiful infographic of marketing metrics done in the form of a chemistry periodic table (found within the story). It includes many suggestions for KPIs for tracking a digital marketing campaign. It is a great look into how many options are available for measuring campaigns. Unfortunately, it leaves marketers exactly in the same position they were in before coming across that beautiful infographic — focused on metrics as the end all and be all.
Marketers don’t measure solely for the sake of gathering data. Metrics are tools used to convey ideas — to tell stories, if you will — and get your team to act. Conversion rates, social mentions and time on site, for example, can all be valuable metrics. What insights do they convey, though? What do they say relative to your strategy? What do they mean to your team? What do they mean to your superiors? How do they help you celebrate your successes or identify challenges? The metrics that matter are those that best position you to influence others to optimize your chances for achieving your goals.
The great data rush of the 21st century is changing the way we live — making us more knowledgeable and more empowered in the process. Yet, we’ve just begun to scratch the surface of what’s possible in terms of using data to make smart decisions. For marketers, getting in now still puts you in the early, but necessary, stages of applied data analytics. The best practices referenced above will increase the likelihood that your data journey is a progressive one that delivers ever increasing value to your team and your organization.
May you find many nuggets of gold in your data!
Talib Morgan is the President and founder of Analegy, a marketing technology and data analytics consultancy. Over his 20 year career, Talib has worked with big brands including well-regarded companies as AT&T, Eli Lilly, General Motors, Guardian Life, MetLife and numerous others. He is a former president of the American Marketing Association of New Jersey.