Content Marketing World is part of the Informa Connect Division of Informa PLC
This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC’s registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 3099067.
There’s no doubt there is a staggering amount of data out there. When properly analyzed, it can deliver valuable insights to marketers. So, who can help us connect the data to our content? This is where a data scientist comes in.
We hear a lot about the role of a data scientist. It, after all, has been called the “sexiest job of the 21st century.” But what do they do? How can they help us make smarter marketing decisions? We turned to Trust Insights Co-founder Chris Penn to find out. Chris tweeted with us on a recent #CMWorld Twitter chat and gave a data science crash course in one fast-paced hour.
To start the discussion off on the right foot, Chris provided a little background on what data science is (and isn’t).
This discussion reaffirmed what we’ve known all along; our community has the best and brightest marketers around. Don’t believe us? Check out these tweets.
#CMWorld A2: Marketers care about data science for 4 reasons: accelerate decisions, lower costs, beat competitors, and uncover hidden opportunities. pic.twitter.com/jqOtqdDZlV
— Christopher S. Penn (@cspenn) January 28, 2020
#CMWorld A2: Standard analytical practices may do these functions as well, but data science approaches inquiry with these outcomes in mind by design.
— Christopher S. Penn (@cspenn) January 28, 2020
#CMWorld A2: The key is the application of the scientific method – hypothesis testing, reproducible results, statistical validity – to our data and analytics. Not what marketers usually do. pic.twitter.com/yFCHboih8f
— Christopher S. Penn (@cspenn) January 28, 2020
A2: The analytics are able to go a little deeper into the finding to better understand the audience and what they are doing – using math, science and all kinds of approaches that marketers don’t use. #CMWorld
— Bernie Fussenegger #SMOS2020 Speaker (@B2the7) January 28, 2020
#CMWorld A3: Marketers who want to use data science need business skills, scientific skills, statistical skills, and coding skills to approach data science challenges. pic.twitter.com/lifmme6Sev
— Christopher S. Penn (@cspenn) January 28, 2020
#CMWorld A3: In terms of attitude, marketers who want to become data scientists or use data science methods must be curious, persistent, and patient – three things in VERY short supply in business today.
— Christopher S. Penn (@cspenn) January 28, 2020
#CMWorld A3: You’ll specifically need to know a stats language like R, stats and probability, some linear algebra and calculus, and at least one structured data language on top of knowing marketing and business and the scientific method.
— Christopher S. Penn (@cspenn) January 28, 2020
A3: After coming to Agorapulse I realize that an OPEN MIND is super important. It’s amazing what can be done with all the stats. BUT as a marketer you have to be willing to realize their power & be willing to explain the basics to get the whole team on board! #CMWorld https://t.co/8i3n2iXiOd
— Agorapulse (@Agorapulse) January 28, 2020
A3. The best way to leverate data science in #marketing is to develop a
#ContinuousImprovementmindset and not just swing for the fences.
Keep making small adjustments, call them tests, and keep improving. #CMWorld
— Tod Cordill (@todcordill) January 28, 2020
#CMWorld A4: The data science lifecycle can be boiled down to defining a problem and experiment, gathering data, analyzing it, and accepting or rejecting your hypothesis.
— Christopher S. Penn (@cspenn) January 28, 2020
#CMWorld A4: Expanded, it’s probably a 20-step process that can take weeks, months, or years to find the answer – just like every other form of science. pic.twitter.com/Gi4SVLVPne
— Christopher S. Penn (@cspenn) January 28, 2020
#CMWorld A4: The place marketers should start is exploratory data analysis (EDA), which is a subdiscipline of data science in which you explore the data you have without preconceived notions. pic.twitter.com/We8FtloDFl
— Christopher S. Penn (@cspenn) January 28, 2020
#CMWorld A5: Predictive analytics, which is a sub-discipline of analytics, is all about predicting either when something will happen (forecasting) or what makes something likely to happen (regression/classification).
— Christopher S. Penn (@cspenn) January 28, 2020
#CMWorld A5: An example is forecasting out your SEO keyword lists up to a year in advance, then being able to build your content against that plan, rather than scrambling every week to plan content. Here’s an example: pic.twitter.com/0bl3WjKxMH
— Christopher S. Penn (@cspenn) January 28, 2020
#CMWorld A5: The underlying technology to do that came from data science – a hypothesis, tested and proven true, then built into a model that can be use.
— Christopher S. Penn (@cspenn) January 28, 2020
A5: It creates models for future efforts. It happens elsewhere, too. Astrophysicists, for instance, use models to predict future trajectories of bodies and potential interactions. #CMWorld
— Rachel Wendte (@rkwendte) January 28, 2020
A5. Based on past actions that were proven to be successful you can predict your content. However because people are responsible for those actions, it probably isn’t recommended. Look at how Instagram keeps changing things&creators/businesses are being affected. #cmworld pic.twitter.com/IeKJe5kq6k
— SL Thomas (@iamslthomas) January 28, 2020
#CMWorld A6: The most important “tool” in data science is you learning to be comfortable being uncomfortable. The nature of science is that you’ll discover answers you really don’t like, and your company/boss will not like.
— Christopher S. Penn (@cspenn) January 28, 2020
#CMWorld A6: Technology-wise, learn statistics and coding. I recommend the R programming language, and I recommend @IBMWatson Studio as a great learning environment. Free to try out! #IBMChampion #IBMPartner pic.twitter.com/PThqwTea5L
— Christopher S. Penn (@cspenn) January 28, 2020
#CMWorld A6: Third tool is great data governance. My @TrustInsights cofounder and partner @katierobbert talks about this often. Data is no good if you can’t find it or it’s in terrible condition. pic.twitter.com/vsqpFAN4un
— Christopher S. Penn (@cspenn) January 28, 2020
A6: A good friend in the #data world who knows the #dataanalytics tools, will entertain your questions, and is also curious by nature. Give them 🎁☕️ & recognition. #CMworld
— Penny Gralewski (@virtualpenny) January 28, 2020
A6: Tools…we use a bunch. I’ve seen some cool dashboards, but I think TALKING within our team is the greatest tool. It keeps accountable and on track and we can support each other #CMWorld https://t.co/nBELkJH9hY
— Agorapulse (@Agorapulse) January 28, 2020
#CMWorld A7: You need a data scientist if you have data problems you need to solve in a scalable, repeatable way with the scientific method. Data science creates that scalability and reproducibility. pic.twitter.com/fPiOXzPMXw
— Christopher S. Penn (@cspenn) January 28, 2020
#CMWorld A7: Look for people and agencies with all 4 skills: business acumen, scientific skills,coding skills, and math/statistical skills. No one is strong in everything, but a good partner is competent in all, excellent in 1 or 2. pic.twitter.com/rKBfC7jrPN
— Christopher S. Penn (@cspenn) January 28, 2020
#CMWorld A7: With analytics, almost all marketing analytics tools are that – analytics only. Data and maybe an explanation of what happened. Insights you can act on require much more.
— Christopher S. Penn (@cspenn) January 28, 2020
Data can make for better content. How are you thinking like a data scientist? Or are you currently working with one? Let us know about it in the comments below.
Realizing you don’t have the data science expertise you need? Unsure of where to start? Attend Chris’ half-day workshop at ContentTECH Summit. Roll up your sleeves and prepare for a data deep dive.