This site uses cookies to enhance your browsing experience and deliver personalized content. By continuing to use this site, you consent to our use of cookies.
COOKIE POLICY
I’ve had the good fortune over my career of working with many different organizations across many different industries. Often, I’ve been in an advisory role assisting organizations with getting a handle on their data assets and finding value in the insights those assets can provide. When data is integrated, clean, governed, and understood the whole is worth far more than the sum of the parts. That’s easy to say, and hard to accomplish. In fact, finding organizations who are truly data driven is always an interesting surprise! Of course, my anecdotal data is biased- I likely wouldn’t be involved if they have already tackled that challenge!
Since I’ve pivoted into working with a team of data scientists and helping our clients to frame up data science projects, I’ve found myself thinking a lot about what guidelines organizations should follow to determine if they’re truly ready for data science. While I’m not proposing a discrete checklist, and readiness isn’t strictly binary, I thought it might be helpful to share some of my thoughts and maybe help others be introspective before deciding if and how to explore creating a data science capability within their organization. I’ve broken my thoughts into three key areas: culture, data, and skillset.
Culture and data usage have been inseparable for at least as long as data management has been around. Not all organizations view their data as a true asset, and without that shared vision it’s unlikely that exploration of those data assets to see what insights can be uncovered would be a highly valued activity. Embracing data science means embracing exploration. Oftentimes the outcomes you get were not the outcomes you expected or sought. Think about that for a second. Does your organization’s leadership celebrate that sort of grass-roots innovation, learning to fail fast and early, and being able to pivot an effort on a dime while staying focused on the value of such an exercise? That’s a tall order. While this culture isn’t the end-all enabler of data science, the presence of it will certainly aid in adoption.
You’ve heard the adage that “data scientists spend 80% of their time cleaning and integrating data”, right? That’s been fairly spot on in my experience. Data scientists are very good at fixing data for their immediate challenge, but that rarely leaves the organization’s data any better off in the long run. I’m not proposing that all your data has to necessarily live in a clean, well-structured data warehouse before data scientists can make use of it. But I am saying that if you haven’t already invested in data governance activities or built some integration, your data scientists aren’t going to reach their full potential. If your data scientists are a luxury sports car, your data is the road. If the road is full of potholes and poorly mapped out, that sportscar isn’t going to come anywhere near max speed!
Because data science is truly the nexus point of so many information disciplines, finding the right people with the right skills can be extremely challenging. Consider the following nine skills: programming, data visualization, data intuition, statistics, data wrangling, machine learning, software engineering, calculus & linear algebra. It’s unreasonable to expect any one individual to have mastered all of those, and still a challenge to build a well-rounded team that complement each other well. What skills does your organization already possess? What skills will you need to invest in?
Everything above is a suggestion based on my years of working with organizations on their data management challenges. To be contrary to my point, I am firmly against letting the great be the enemy of the good. Do I think data science activities are valuable even if data isn’t integrated and the culture isn’t fully supportive yet? Sure! The insights that might be discovered could shatter complacency, spotlight broken processes and missed opportunities, or jolt an organization awake to realize how important it is to invest in data management. Just be ready; moving forward with an initiative or operationalizing a model in a less than ideal environment can be exponentially more challenging.
I’ve met with many organizations who want desperately to get a data science capability off the ground but are stymied by so many of the above pitfalls. From data nobody trusts to poorly managed expectations and everything in between, these challenges can be pervasive and daunting; but they aren’t insurmountable. Are you interested in building a capability? Take a moment to think about your organization’s readiness. Need someone to bounce ideas off, or help explore a particular challenge more thoroughly? Our team of data scientists is ready to pull out the dry erase markers and white board some really geeky stuff with you, don’t hesitate to reach out! Together maybe we can navigate the bumpy road. Sorry, had to get one more bad analogy in.