Twitter Chat: Data Science 101, Video Edition from Content Marketing World

I had the pleasure and privilege of being the guest subject matter expert for Content Marketing World’s Twitter chat on data science 101, and the questions were so good, they deserved longer answers than just a handful of tweets. If you want to read the chat as it happened, you’ll find the original tweets here.

Let’s browse through the questions, then watch the answers. These are the questions asked in the Twitter chat, edited for clarity; click/tap on a question to skip to the answer:

1. What is data science and why does it matter in content marketing?
2. What kind of analytics can data science uncover?
3. How can marketers adopt a data science mindset? Outside of hard, analytical skills, what soft skills should marketers possess?
4. What tools are useful to help marketers dig deep into their organization’s data?
5. What steps make up a data science lifecycle? Where do you begin?
6. How can predictive analytics make your content more effective?
7. Not every marketer can (or wants to) be the data scientist for their org. What should we look for when hiring an FTE or partner?
8. Where do our current analytics tools lack, and how could a data scientist help?

Now, let’s tackle the answers.

What is data science and why does it matter in content marketing?

Data science is the practice and professional of extracting meaningful insights from data using the scientific method. The science part is critical and what sets apart data science from analytics, engineering, and AI.

What kind of analytics can data science uncover?

This is an interestingly worded question. Analytics are fundamentally about explaining what happened, the what in your data. Very often, as stakeholders we also want to know what’s relevant (data overload) and then why (insights). That’s where data science can help, especially with understanding what’s relevant. Methods like regression, clustering, classification, and dimension reduction can greatly assist us in finding out what really matters.

You’ll never create something out of thin air – always derived from your initial data. That’s why domain expertise matters – to know what else is available.

How can marketers adopt a data science mindset? Outside of hard, analytical skills, what soft skills should marketers possess?

Data science is exactly what it sounds like: performing science with data. The soft skills which make for a great scientist thus transfer to a data scientist and any marketer who wants to adopt a perspective of using the scientific method to improve their marketing. The seven data science soft skills are:

• Curious
• Humble
• Open
• Patient
• Persistent
• Passionate
• Resilient

It’s so important to note that if your workplace lacks or actively opposes these qualities, your ability to grow will be seriously hindered.

What tools are useful to help marketers dig deep into their organization’s data?

The answer to this question depends on the level of skill a marketer has in data science, specifically the technical and statistical skillsets. I’d put the available tools in categories of beginner, intermediate, and advanced. Beginner tools help marketers extract and report on the data itself. Intermediate tools help marketers start to understand patterns and relationships in the data. Advanced tools help marketers manipulate, transform, and distill the data.

• Beginner: Spreadsheets, Google Data Studio, the various data sources
• Intermediate: IBM Watson Studio, Tableau Software, IBM Cognos
• Advanced: R, Python, SQL, Scala, Spark, Neo4J

What steps make up a data science lifecycle? Where do you begin?

The short answer to this question is to define the problem and hypothesis, prepare your data, explore your data, test your hypothesis, build a model, validate the model, and then deploy and observe. Each stage is composed of multiple sub-steps.

How can predictive analytics make your content more effective?

Predictive analytics comes in two flavors – understanding and building a predictive model of what makes something work, and time-series forecasting to predict when something will happen. Both techniques are invaluable for boosting your content marketing power.

Not every marketer can (or wants to) be the data scientist for their organization. What should we look for when hiring an FTE or partner?

This is a critical question because there’s a significant shortage of trained data scientists. Those who exist and are qualified are “reassuringly expensive”. Thus, be on the lookout to evaluate the 6 skill areas that a data science individual or agency must have: coding, stats & math, data engineering, domain expertise, business expertise, and science expertise. If hiring, you may need to hire a team rather than a single individual.

Beware of “crash course data scientists”! These folks generally have only one of the six skill sets and limited or no practical experience. Remember that expertise is all about knowing what’s going to go wrong – anyone can do things well when everything is perfect.

Where do our current analytics tools lack, and how could a data scientist help?

If we think about what we expect of our tools, we ask them to tell us what happened, why, and what we should do about it. Almost every analytics tool only does the first part. A data scientist has to help complete the rest of the hierarchy of analytics – descriptive, diagnostic, predictive, prescriptive, and proactive.

Our thanks to our colleagues and friends at the Content Marketing Institute for hosting the chat – and if you’re looking for a more in-depth answer to some of these questions, come see me at the ContentTech Summit in April 2020. I’m doing an entire workshop on the basics of data science for marketers.

Christopher Penn, Chief Data Scientist