1) The document discusses lessons learned from DataKind's work using data to solve social problems over the past 3 years.
2) Some key lessons included celebrating all outcomes, whether intended or not; working with social organizations instead of just focusing on data; and ensuring the problem is well-framed and data is relevant.
3) Successful data for good projects require a partnership between social actors and data experts to identify a well-defined problem, relevant data sources, and solutions that are impactful but also understandable.
10. Use anonymized question
data from citizens to
systematically establish new
emerging areas and
categories of societal concern
Set out to We ended up with
Zero hour contracts
What to expect when data for good’ing
11. Use open company ownership
data to understand how
complex company structures
are being used to support
corruption
Set out to We ended up with
What to expect when data for good’ing
12. Use anonymized hospital &
child hospice data with open
demographic data to discover
up to 3,500 children who are
in need of hospice care
Set out to We ended up with
What to expect when data for good’ing
13. How can we use data to have a real impact on
peoples lives?
Units of “good”
16. Seeking out and partnering with social actors
Target real world impact by focusing on joint
interests
Bringing non-data people along with us
Work with not for means