I recently read an article in TechCrunch about the challenges of elusive wisdom in big data. In his article, author Ron Miller points out the difference between data collection and data interpretation.
“We want to believe getting value from big data is as simple as pouring in the data, running a program and getting insights, but in fact, it’s much more complicated than that.” Miller writes.
I found his article to be extremely informative, which is why I want to share it with you and add commentary to what he is discussing. Have questions or comments? Tweet them to Ron Miller and I.
The Challenges of Big Data
Everything from what’s going on in our bodies, what we search for using digital devices, to how we shop and purchase products is being stored in databases. There are heaps of data that, once analyzed, may tell us everything about anything and anyone. At least that’s the idea.
One of the biggest challenges to finding answers in big data is our current state of technological sophistication. We’ve made great advancements and it’s an exciting time to be a part of the industry. We’re mastering how to collect and process data, but technology still lacks interpretation capabilities. Ron Miller interviewed Pam Baker, author of Data Divination: Big Data Strategies, who points out that predictive analytics can be very accurate depending on what you’re trying to solve. Yet, there are many variables that may not be anticipated or easily measured. Machine learning eventually will reach this level of sophistication to interpret factors such as personality, work ethic, social skills and other arbitrary metrics. We’re just not there yet.
This leads to a pressing question. How much should we rely on answers from big data without human interpretation? Miller shares that, despite data accuracy, some instances of big data use may result in trust issues. Take the medical field, for example. Miller explains how analytics platforms such as IBM Watson may be used to mine data to provide options and information relevant to an individual’s health, but physicians prefer to decide how to treat the patient.
Another pain point in big data is the gap of expertise. There is high demand for data scientists and not enough experts to meet the need. Keith Rabois, an investment partner with Kholsa Ventures, shares with Miller that the worse case scenario is a “bottleneck where we go to an expert to get the answers, then wait for the results.”
Humans Necessary for Successful Data-Driven Marketing
Behind every data-driven marketing technology is a data scientist. Technology may collect and manage data, but an expert data scientist is still required to build algorithms and process information in a meaningful way.
There’s talk that marketing technology will provide all the solutions to most of our marketing challenges. The idea is that anyone can become an expert marketer if they’re using the right technology. Talent, experience, and knowledge can be overlooked when the right tools are used, according to this line of thinking.
This isn’t exactly the case.
Yes, you can teach someone to use technology. This person doesn’t necessarily have to be a data scientist to execute data-driven strategies. SaaS solutions like nectarOM provide an omnichannel execution platform that is built for non-technical marketers. The interface is intuitive and doesn’t require coding or data mining skills.
To provide this level of sophistication, a data scientist is required to build the prototype. It is the data expert who develops the applications and algorithms. In most cases, the algorithms are not rinse-and-repeat sequences for every business. Different factors matter to different brands. You need a true data scientist to implement this for you. If for any reason campaigns aren’t performing as expected with your personalized messages, you’ll also need an analyst to come in and see if your data points are being processed appropriately. These are issues a non-technical marketer can’t solve.
The Future of Big Data in Marketing
I believe we’re only in the beginning stages of developing technology that collects, processes and interprets big data. Machine learning and predictive analytics will continue to improve. It’s just a matter of time.
Scott Brinker recently released his Marketing Technology Landscape Supergraphic for 2015. An astonishing 1,876 vendors are represented across 43 categories. In 2014 only 947 marketing technology vendors were recorded. We’re seeing more of this technology include big data solutions as part of its capabilities.
We’ll see a rise in data experts and scientists, but technology developed for non-technical marketers will explode. Data scientists will continue to be an integral part of data-driven marketing. We just might not see them up-front and center as we do other marketers. Rather, they stay behind closed doors to develop technology that predicts and analyzes the data we humans cannot. I hope we continue to celebrate these experts and shed more light on their work.
Don’t want to travel the journey alone? Sign up for my email list to learn more about marketing technology, strategy, and more.