2. 1: Understanding the Datavores
1. Rise of the
Datavores
2. Inside the
Datavores
3. Skills of the
Datavores
…
• A three-year programme of research
• Aim: to generate robust, independent evidence to inform
policy and practice enabling UK businesses to create value
from their data
• Research examines business data practices, effect on
performance, and skills implications
2
3. Rise of the Datavores
Published November 2012: Survey of 500
UK companies commercially active online
Data
Insight
Action
Impact
Collection?
Analysis?
Use?
1. Rise of the
Datavores
2. Inside the
Datavores
3. Skills of the
Datavores
…
3
4. Datavores in the minority; organised differently
0%
10%
20%
30%
40%
50%
Datavores Dataphobes
Decisions based
on experience +
intuition
Decisions based on data
and analysis
4
5. Inside the Datavores
1. Rise of the
Datavores
2. Inside the
Datavores
3. Skills of the
Datavores
…
Looking at the
link between
data activity and
productivity and
profitability
16% more data-active = 8% more
productive
Analysis has the highest impact on
productivity (+11%) and EBITDA
(+3,180 per employee)
Positive synergy between
employee empowerment and data
activity (4x boost)
5
6. 2. Skills of the Datavores
The US will have a shortfall of
‘deep data talent’ of up to
190,000 by 2018.
McKinsey, 2011
The sexy job in the next
ten years will be
statisticians.
Hal Varian
Going from technology and data requires the right
skills… but what are those skills?
Data scientists: a new occupation? a new
capability? A rebranding?
What does this mean for educators, policymakers
and managers?
6
7. Model Workers
Audience Questions
Everyone What are the skills of productive data
analysts?
Educators Is the education system producing enough
of them?
Managers How can managers organise their data
talent to create value?
We interviewed managers of data
analysis teams, HR managers, data
scientists and CTOs. We targeted
companies where data plays an
important role in production and/or
operation.
7
8. Data landscape: Four Data modes
Variety
Volume
Only 1 in 4 of the
companies in our
sample in this data
modeBusiness
Intelligence
(Analytics)
Data intensive science
(Com bio, epidemiology)
Web Analytics
(digital marketing)
Big data (data
scientists)
8
9. One mode to rule them all?
Supply (better tech and
more data) & demand
(competition) driving
firms into the ‘big data
corner’
Variety
Volume
Big data (data
scientists)
9
Business
Intelligence
(Analytics)
Web Analytics
(digital marketing)
Data intensive science
(Com bio, epidemiology)
10. The perfect analyst
Analysis +
computing
Domain
knowledge +
Business savvy
Storytelling +
team-working
Creativity +
curiosity
Theprofilemostofourrespondentslookfor
4 in 5
firms
report
difficulties
recruiting
Talent lacks skills
+ experience
Not enough talent
Talent without the
right mix of skills
Internal capacity
issues
10
11. Future trends…
L
w
Supply
Demand
Better tools Education adapts
More sectors
become data-driven
Better tools lower
barriers to entry for
SMEs
Education
adapts too
slowly…
? In the short-term, data
talent crunch + some
instances of offshoring
11
12. Policy implications: skills
1. Develop workforce skills
• Upskill existing professions
• Make this part of cluster development
programmes?
1. Build up the data analyst
profession
• Develop training and certification
standards?
• Raise awareness and share good
practice
1. Ensure access to overseas talent
• Including students & entrepreneurs
12
13. Policy implications: education
1. Better university-industry
communication
• Sector skills councils
communicate, universities
innovate, NCUB broker links?
• CDEC, Imperial College data
institute
2. Promote inter-disciplinarity
3. Improve teaching of math +
stats in schools…and after
schools
13
15. Implications for managers
Data talent is often innovative and creative. This is a source
of opportunities (innovation) and management challenges
(motivation, organisation, predictability).
15
16. The companies we interviewed are… going out
to where the talent is
16
20. 3. Conclusions
1. Big data companies are in minority, but everyone looking
for talent with data scientist profile
2. Data analysis is creative work -> good for innovation,
but management (and education) challenges
3. Blockages in data talent pipeline echo situation with
coding. What can we learn from Next Gen campaign?
4. Autumn 2014: Next report based on new skills survey +
HESA data.
20