Founder of datalab.cc, author for Lynda.com/LinkedIn Learning, associate professor of psychology at Utah Valley University. Specializes in demystifying data science and helping people do good with data.
Length: 53:08 Minutes
Overview: In this video you will see the distinction between big companies and the 99% of the rest of the companies that counts for 48% of the jobs in the US. It is described the way small and medium companies work internally, what are they needs and how you can connect with them to apply data science in these businesses.
Learn how most of the companies are organized and work in the US.
Learn the basic analysis you need to cover most of the information needs of Small and Medium companies.
List some of the alternatives to work in data science for nonprofit or charities organizations.
0:00 - 1:23 Introduction and Overview
1:23 - 2:49 The field of Data Science
- Most innovative companies in Data Science
- Data Science Jobs that pay 100K or more
- The map of the data science
2:49 - 7:09 The existence of the 99%
- Number of jobs in the US, like Walmart, IBM, Amazon, Apple
- 99.7% of the companies are small or medium, less than 500 employees
- 79% of the companies in the country are really small, less than 10 employees.
- Small and Medium Enterprises (SME) account for the 48% of the job market.
- Data Science is not all the time like in big companies.
- SME exists, they have data, and they need data professionals, also there is a lot of nonprofits.
7:09 - 14:10 Learn who are the SME
- What are their goal, their methods and their limits
- Not designed for rapid growth, they provide a service for a particular group, it changes the kind of analysis you can do for them
- They work a lot with Excel.
- The problem with the Drew Conway Data Science Venn diagram. Too much focus on coding and machine learning, forgetting the domain expertise, which is a key element for data science.
- Small business work in different businesses models
- What is the data of SME. Spreadsheets, CRM, CMS, social media, emails, web analytics & SEO.
14:10 - 41:59 Limits to be aware of
- Severely limited in money, time, personnel.
- As a consultant adapt to SME to work with them. Simple quick & cheap. Answering to Yes or No questions.
- What you need to do. Use common tools, be fluid in Excel. Use minimum sufficient analysis. Clarity & applicability, focus in what the organization can control.
- What analysis you should know for them. Sums, counts, & percentages. Tables and pivot tables. Subgroup and filters in Excel.
- Black boxes are not allowed. Human-in-the-loop analysis and decision making.
- Data visualization needs to be really really simple. Use bar charts, maybe scatterplots, and line charts. Keep it as simple as possible. Users don’t particularly think like you do. Don’t include distractions.
- Example of modern dance. Only a poem can translate a poem. Nothing for annotating and visualizing dance. Its an open problem work now. You can solve that problem.
- Once you know how they work then you can connect with them, some examples of nonprofits like DataKind.org, Data Science for Social Good, Data world tour, or DIY.
- Few suggestions to organize your own event: Ask for clear questions to answer. Prepare analysis templates. If possible, prepare data. Follow up. Adopt an organization, they all need the benefits of data science. What you get is: good practice, good networking, possible charitable donation.
41:59 - 45:42 Enable SME
- Teach them, handing off, document thoroughly, explain decisions; plan for common tools, and for tools that doesn’t cost money; save in generic formats; save in accesible locations. Support self sufficiency, stay in contact, create materials.
45:42 - 53:08 Questions and Answers