BE DATA INFormeD 
without being a data scientist
Pamela 
Pavliscak 
(not meant to be spoken) 
@paminthelab 
Change Sciences
I <3 Data 
but I’m still 
learning
Today’s JOURNEY 
data overload 
the analyst 
the threshold 
poor signals 
the abyss 
return with elixir 
data thinking 
supertechnological aid 
trophy metrics
3 million emails 
2 million Google searches 
684,000 Facebook shares 
$272,000 spent online shopping 
100,000 tweets 
3,600 photos to Instagram 
2 million blog posts
1 baby born 
5 mobile phones activated 
@paminthelab
Each person carries an average of 
3 devices
375 megabytes consumed 
Per household each day
For now, data is focused on 
Selling and surveillance
DATA 
Being 
Nothingness
BIG DATA 
is a little scary
Big data is like teenage sex - everybody 
talks about it, nobody really knows how 
to do it, everyone thinks everyone else 
is doing it, so everyone claims they are 
doing it. 
-Dan Ariely 
“ 
@paminthelab
INSIGHT VACUUM
Think of data science like archeology 
traces left behind
They came 
<something, something> 
They left
It’s the every person story
There are really three ways to use DATA
PROVE 
something you know
Proving is about the bottom line
Numbers 
KPIs & scores 
@paminthelab
Strategic rather than tactical
TOO HIGH LEVEL for experience design
IMPROVE 
something you can change
A lot of consistency 
and a little experimentation
DISCOVER 
something you don’t know
Look for patterns
Discovering is about relationships
How to Think 
like a data scientist
curiosity 
THINK WITH DATA 
1
CONSIDER 
measurable moments
Expectations (before) 
Moments (during) 
Memories (after)
The Experience Continuum 
the long 
before 
just a 
memory 
weeks 
days 
hours 
minutes 
now 
now 
now 
minutes 
hours 
days 
almost 
now 
now 
the 
here after
remember 
it’s people first
Behaviors 
Interactions 
Feelings 
Attitudes
Start with buckets, 
and count the things
hamburger icon test case
Our gut says no
PROBABILiTy 
BE A BIT BAYESIAN 
2
P (B | A) P (A) 
P (A | B) = 
P (B) 
TRANSLATION 
Belief + new evidence 
= updated belief
Evaluate 
the evidence
Analytics 
Piwik, Clicky 
Google 
Chartbeat, Crazy Egg, 
Mixpanel, KISS metrics
Social Media Data 
Shoutlet 
Reachli 
Trackur 
Topsy 
Social Mention 
Open Status 
Google Trends
Other data sources 
1. A/B tests 
2. Surveys 
3. Heatmaps 
4. Quantified qualitative studies 
5. Biometrics & eyetracking 
6. Sales data 
7. CRM data 
8. Call center logs 
9. Billing systems 
10. Business intelligence systems
It’s really all behaviors & words
USE 
degrees over absolutes
Pay attention to variation 
@paminthelab
Correlation ≠ causation
It’s never 100% CERTAIN
Source: Exis 
Subtle changes 
make a difference
TRansparency 
DESCRIBE YOUR BIAS 
3
There is always a filter 
HEISENBERG UNCERTAINTY 
PRINCIPLE (small data) 
SIGNAL BIAS 
(big data)
Even a big data source is not everyone
By studying something, we change it
Nice people skew studies 
HUMAN CONTACT 
NICE
There is a nice factor 
TV Network Site ONLINE Study 
0 20 40 60 80 100 
TV Network Site LAB Study 
0 20 40 60 80 100 
Source: Change Sciences
gut feel 
<bias> 
one site
CoNTEXT 
EMBRACE COMPLEXITY 
4
COMbine forces 
Quantitative & Qualitative
Balance subjective & objective
Not Curious 
74% tapped right past 
help of any kind
meaning 
TELL A STORY 
5
SKIP TO MEANING 
5 
4 
3 
2 
1 
0
Select 
actionable detail
OVerview 
zoom 
filter
Audience 
put a human face 
on it
C-Suite 
Team
CREATE 
emotional impact
MENU IS 
CLEARER 
20% more people 
clicked Menu than the 
the icon alone 
Source: Exis
I’ve been seeing that little doodad on 
the site, but I just thought it was part of 
the design. 
-M, Millennial 
“ 
Source: Change Sciences
Hamburger awareness 
Q1 2013 
Q2 2013 
Q3 2013 
Q4 2013 
Q1 2014 
Q2 2014 
Source: Change Sciences 
is on the rise
Data 
thinking 
1. Get curious 
2. Think Bayesian 
3. Explain bias 
4. Embrace context 
5. Show & tell
UX + DATA SCIENCE
BE DATA INFormeD 
without being a data scientist 
Pamela Pavliscak 
@paminthelab 
ChangeSciences.com

Be Data Informed Without Being a Data Scientist