Three Must-Read Data and Analytics Books with Tim Harford, Zhamak Dehghani, and Brent Dykes

Key Moments:

• Making sense of the world through a data lens with Tim Harford (05:18)
• Tim’s favorite of the ten data commandments (07:11)
• Guard against using data as a control mechanism (15:14)
• How does curiosity create a healthier relationship with data? (19:57)
• Data Mesh with Zhamak Dehghani (32:31)
• Blending centralized and decentralized schools of thought (36:39)
• Data Mesh isn’t for everyone (48:11)
• How do you begin your data mesh transformation? (53:28)
• Brent Dykes and the importance of data storytelling (1:02:17)
• Why do data scientists need to be better communicators? (1:11:13)
• The Venn diagram of data storytelling (1:15:14)
• What is an insight and why should it cause change? (1:17:56)

Key Takeaways:

Tim Harford

Why is curiosity such an important part of interpreting data? (21:08)

Curiosity is an especially important trait to develop as a data scientist. As Tim explains, true, investigative curiosity can help you find important insights you may have otherwise overlooked.

“When people are curious, they are processing data in a different way. So they're treating new information as, kind of satisfying a hunger. It's filling a hole in the model of the world in a very pleasing way. And if the new information is sort of surprising, if it doesn't quite fit well, that's even better. That's really interesting. So while most of us, when we get data that contradicts what we believe most of us would say, ‘Well, that can't be right. That's probably fake news.’ The curious person is like, ‘Oh wow, interesting. There are whole new worlds to conquer, whole new worlds to explore.’ So that's a much healthier, more interesting way of engaging with data.”

Zhamak Dehghani

How do you create personalized experiences without compromising user privacy? (17:18)

Sharing data can unlock powerful insights across organizations. A decentralized approach speaks to the modern business landscape and is a model that companies can use to scale.

“Data mesh is a decentralized sociotechnical approach for managing, accessing, and sharing data for analytical use cases. Those analytical use cases can range from training machine learning models in a decentralized way, or writing reports and generating reports and insights in a decentralized way. It's designed for complex and large environments, whether we are sharing and accessing data across an organization between the business units or sharing and accessing data across organizations across trust boundaries. The objective of the data mesh is creating a model of responsible data sharing that can scale out in step with organizational growth and complexity as organizations become larger as the touch points with customers and the systems that generates data diversifies as a number of use cases, we want to use that data in machine learning in analytics increases. It's a model that is designed to really scale and meet this expansion of complexity and uncertainty need.”

Brent Dykes

Are valuable insights being lost due to poor communication? (12:54)

The way you present data is everything. Sometimes the opportunity to make a meaningful business change comes down to convincing a group of people in a boardroom of certain insights. With a little coaxing and some careful packaging, you can help your stakeholders understand how to take action.

“People just aren't packaging their insights in a good way; either there's too much noise or the managers who are looking at it don't understand the significance of it. It's not fully flushed out. It's not presented in a way that's going to really connect with people. And so we're just basically wasting these nuggets of valuable information because we simply don't know how to communicate it correctly.”

Key Quotes:

Tim Harford

Well, I've mentioned already the importance of curiosity; when people are curious, they are processing data in a different way. So they're treating new information as, satisfying a hunger. It's filling a hole in the model of the world, in a very pleasing way. And if the new information is sort of surprising, if it doesn't quite fit well, that's even better. That's really interesting. So while most of us, when we get data that contradicts what we believe most of us would say, ‘Well, that can't be right. That's, that's probably fake news.’ The curious person is like, ‘Oh wow, interesting. There's whole new worlds to conquer whole new worlds to explore.’ So that's a much healthier, more interesting way of engaging with data.

Zhamak Dehghani

Data mesh is a decentralized sociotechnical approach for managing, accessing and sharing data for analytical use cases. Analytical use cases can range from training machine learning models in a decentralized way, or writing reports and generating reports and incentive insights in a decentralized way. It's designed for complex and large environments, whether we are sharing and accessing data across an organization between the business units or sharing and accessing data across organizations across trust boundaries. The objective of the data mission is creating a model of responsible data sharing that can scale out in step with organizational growth and complexity as organizations become larger as the touch points with customers and the systems that generates data diversifies as a number of use cases, we want to use that data in machine learning in analytics increases. It's a model that is designed to really scale and meet this expansion of complexity and uncertainty need.

Brent Dykes

People just aren't packaging their insights in a good way; either there's too much noise or the managers who are looking at it don't understand the significance of it. It's not fully flushed out. It's not presented in a way that's going to really connect with people. And so we're just basically wasting these nuggets of valuable information because we simply don't know how to communicate it correctly.

Mentions:

• The Data Detective: Ten Easy Rules to Make Sense of Statistics by Tim Harford

• Data Mesh: Delivering Data-Driven Value at Scale by Zhamak Dehghani

• Effective Data Storytelling: How to Drive Change with Data, Narrative, and Visuals by Brent Dykes

Bio:

About Zhamak

Zhamak Dehghani is a technologist, focusing on distributed systems and data architecture in large and complex environments. She’s a member of multiple technology advisory boards. Zhamak is an advocate for the decentralization of all things, including architecture, data, and ultimately power. She is the founder of data mesh.

About Brent

Tim is an economist, journalist, and broadcaster. He is the author of How To Make the World Add Up / The Data Detective, Messy, and the million-selling The Undercover Economist. Tim is a senior columnist at the Financial Times, and the presenter of BBC Radio’s “More or Less,” “How To Vaccinate The World,” and “Fifty Things That Made the Modern Economy,” as well as the podcast “Cautionary Tales.” Tim has spoken at TED, PopTech and the Sydney Opera House. He is an associate member of Nuffield College, Oxford and an honorary fellow of the Royal Statistical Society. Tim was made an OBE for services to improving economic understanding in the New Year honors of 2019.

About Tim

Brent Dykes has spent more than 15 years in the analytics industry, consulting with some of the world’s most recognized brands, such as Microsoft, Sony, Nike, Amazon, and Comcast. As a former analyst, manager, and technology evangelist at companies such as Omniture, Adobe, and Domo, Brent witnessed first-hand the challenges of communicating data effectively. With an educational background in marketing (SFU BBA, BYU MBA), he brings a unique perspective to data storytelling that blends both qualitative and quantitative skills. Brent is a regular Forbes contributor with more than 30 published articles on various data topics. After publishing two books on digital analytics, he is now excited to share his passion for numbers and narrative in his new book, Effective Data Storytelling: How to Drive Change with Data, Narrative, and Visuals.