In Seattle, data science is everywhere. We have Microsoft, Amazon, Facebook, Google, and hundreds of other companies large and small, all working with data and analytics every day. The amazing, hilarious, and slightly frightening thing is that almost every one of these data science teams are approaching “data science” differently. Very similar to the vague definition of Big Data, the application of the moniker “Data Scientist” to dozens of experts in various fields and roles has created a massive number of individuals who can freely call themselves data scientists because they fit some part of the bill. For better or worse, we now have no shortage of candidates for entry to mid-level Data Scientist positions, but few of them possess the ability to consult. It’s one of those rarely encountered “rainbow unicorn” qualities that’s missing when most data scientists come from statistics or computer science backgrounds.
To illustrate the variance in what a data scientist might do on a daily basis, let’s go back to my examples from above. At Amazon, some internal data science teams spend months, or even years, solving or improving a solution for a single problem. They go way off the deep end using complex math and algorithms to improve their services even fractionally. Their job could involve fine-tuning recommendation models to get a 0.1% increase in clicks, or developing a marginally better elastic compute optimization ruleset, because either of those optimizations represent millions of dollars to the bottom line. The role of a data scientist here is clear: be really, really smart. These data scientists are wildly different from some of the ones at Microsoft, who spend most of their time engineering the application/integration of various algorithms to build additional services, or the ones at Facebook, who perform elaborate research and A/B tests on what a user sees and how that influences their behavior. The point is, data science is everywhere, in all sorts of forms, because it’s pretty much what we’re calling (advanced) analytics now.
At Neal Analytics, we often say “Management Consulting is in our DNA,” which leads to an approach that is vastly different from the myriad of other data science and analytics providers in the space. Despite the technology being cheaper and more readily available than ever before, sellers can’t just throw technology at the customer and see what sticks, particularly when working with data and analysis. Almost every company out there advertises a “silver bullet” solution which is supposed to solve all business problems and provide a miraculous ROI. Clients are rightfully skeptical, because the last thing they need is another tool to throw in their already cluttered toolbox. This is where the data scientist comes in, yes? You may hire a talented carpenter to build your solution, but if the data scientist lacks business acumen and consulting experience, odds are you’re going to build the wrong thing. No executive wants yet another IT project where the value is lost because nobody gets it, and data science is as hard to “get” as it… gets.
To ensure the solution you’re building can create actionable change in an organization, you need to start at the top, work with executive sponsorship, and tell a story with the data and analytics. One may argue that that is the true measure of a data scientist, but it is the measure on which so many who have changed their titles on LinkedIn to Data Scientist (because it’s more marketable!) fall short. Data storytelling isn’t simply generating models or visuals and explaining them adeptly, it’s the entire process of working with the client from day one to connect them with the data at every stage and lead them through the hypotheses they want to test. The data science consultant effects change by building up the client’s analytics skills, competencies, and understanding of the machinations of their business, while also delivering the insights.
The remaining key differentiator between a consulting data scientist and others is their appetite for change, variety, and chaos. Consulting firms such as Neal often have resources tagged on several projects at once, which requires data scientists with a high level of self-direction, the PM competency to juggle imposing timelines, and the drive to constantly be learning something new. It’s not for everyone, and for those that like consistency and a predictable workday, it’s sure to be a bit shocking.
The point is, we don’t need you to just be really, really smart at Neal, and our clients need more than that as well. The data science space needs enterprising individuals who are reaching to achieve more than getting really good at