The document discusses using data to design customer experiences that respond in real-time to different contexts. It emphasizes the importance of collecting data from multiple channels to get a holistic understanding of customers. Effective use of data requires combining user research insights with data-driven insights to fully understand why customers behave in certain ways. Case studies of how Code and Theory has partnered with clients like Foursquare and Essie to build customized data visualizations and strategies that provide actionable insights are also discussed.
2. @hollielubbock
The digital landscape is in flux,
we need to design and manage
customer experiences to
dynamically target audiences
and respond in realtime
The Problem
RESPONDING TO CONTEXT
3. @hollielubbock
Multiple Channels
RESPONDING TO CONTEXT
SOCIAL MEDIA IS ONLY ONE
TOUCH POINT AMONG MANY
CUSTOMERS ARE EXPOSED TO
We are speaking to users across multiple channels we
need to look at these holistically
4. @hollielubbock
Filter Bubbles
RESPONDING TO CONTEXT
USERS VIEWS OF THE INFORMATION ON THE
INTERNET IS EVEN MORE FILTERED
"collectively, these filters will isolate people in information bubbles
only partly of their own choosing, and the inaccurate beliefs they form
as a result may be difficult to correct.”
Paul Resnick and colleagues
University of Michigan's School of Information
5. @hollielubbock
Tailored Content
RESPONDING TO CONTEXT
DURATION PLATFORM
USER TIME
CONTEXT FREQUENCY
LOCATION FORMAT
Our digital experiences have become unique.
Publishers are tailoring the content we see to your
behaviours and touch points
6. @hollielubbock
To effectively target users within their filtered system we need data coming
in from multi channels
RESPONDING TO CONTEXT
BUS. STRATEGY &
FINANCE
MARKETING &
COMMS
CUSTOMER SERVICE
& SALES
OPERATIONS:
SUPPLY-DISTR.
HUMAN
CAPABILITIES
WEBSITE &
SYSTEMS
DATA
WAREHOUSE
SOCIAL
MEDIA
7. @hollielubbock
RESPONDING TO CONTEXT
Creating Impact
DATA = INFORMATION
DATA ≠ KNOWLEDGE
The Purpose of data is to help you make better informed business
decisions, but how do we measure impact within such a large dataset?
8. @hollielubbock
We have so much data
coming in, we don’t have time
to respond to it effectively
So is it actually useful?
Ummm
RESPONDING TO CONTEXT
9. The Risk Of Continuous Optimisation
RESPONDING TO CONTEXT
When looking for insights from data there is a risk we
continually optimise and only reach a certain point.
We need to balance this with looking for innovations
outside of our normal markets and metric areas
10. @hollielubbock
WE NEED TO TAILOR AND DESIGN
OUR DATA USAGE
We have dashboards measuring things we know. Why are we
measuring that things are working? Surely we know it works. If we’ve
planned and built and tested the system with and for users and know
that as a baseline it works well.
Make data work for you
RESPONDING TO CONTEXT
11. @hollielubbock
Peaks and Troughs
RESPONDING TO CONTEXT
0 41 42 43 44 45 46 47 48 49
THIS SHOWS US WHERE THE
SYSTEM HAS GAPS OR BREAKS
AS WELL AS WHEN IT
PERFORMING EXCEPTIONALLY
WELL.
Cut the amount of data – only measure the
exceptions, rather than business as usual.
13. @hollielubbock
DATA TELLS US WHAT, WHEN AND WHO BUT NOT WHY
USERS ARE BEHAVING A CERTAIN WAY
But Why?
RESPONDING TO CONTEXT
14. UX & CX
RESPONDING TO CONTEXT
sharpestpencil.com.au/
wp-content/uploads/
2013/07/talking-
fingers.jpg
In traditional UX we talk to people, this gave us great
feedback, but there was always a lapse in the feedback
loop as we couldn’t find out and action as fast at we’d
have liked.
15. @hollielubbock
User Journey Mapping
RESPONDING TO CONTEXT
Customer or user journey mapping can help you
think through what’s missing from a user perspective
– across the whole lifecycle of engagement with your
product, from discovery to advocacy
16. @hollielubbock
UCD & Data
RESPONDING TO CONTEXT
We need to combine UX and data practices to get the
whole picture. The world is becoming more targeted
and individualised and one can no longer work
without the other
17. @hollielubbock
WE SHOULD ALL BE ABLE TO PLAY WITH
AND ACCESS DATA
Large organisations need to reshape their data offering to have people
with data skills or data analysts embedded into each product team.
Product Teams
RESPONDING TO CONTEXT
Jer Thorp, “Data (v.),” OCR Journal #001, 2015
via https://medium.com/@blprnt/data-v-da0e0d24777c
Illustrations by Sukjong Hong - http://sukjonghong.com/
UX / UI
DEVELOPMENTDATA
18. @hollielubbock
Intersection
RESPONDING TO CONTEXT
SOCIAL MEDIA
Confirms it is happening
USER CENTRED THINKING
Shows us why its is
happening
DATA DRIVEN
Shows us something is
happening
It’s the intersection which allows to react effectively
19. How do we use data at
Code & Theory?
So then…
RESPONDING TO CONTEXT
20. @hollielubbock
I WILL PRACTICE MY PROFESSION WITH
CONSCIENCE AND DIGNITY;
I WILL USE DATA TO FIND AND TELL THE TRUTH;
I WILL USE WORDS THAT MEAN SOMETHING;
I WILL STRIVE TO MAKE MY METHODS
TRANSPARENT AND REPRODUCIBLE;
I WILL USE DATA TO CLARIFY, NOT OBFUSCATE;
I WILL PRESERVE THE INTEGRITY AND HYGIENE
OF MY DATA;
I WILL GET TO A GOOD QUESTION;
10 Commandments of Data at C&T
RESPONDING TO CONTEXT
I WILL ANSWER QUESTIONS USING THE BEST FORM
POSSIBLE, EVEN IF IT REQUIRES EDUCATING MY
AUDIENCE;
I WILL NOT HIDE UNWANTED RESULTS;
I WILL USE ROBOTS TO DO ROBOT WORK, NOT HUMANS;
I WILL MAINTAIN THE UTMOST RESPECT FOR PRIVACY;
I WILL NOT USE MY KNOWLEDGE TO VIOLATE HUMAN
RIGHTS AND CIVIL LIBERTIES;
I MAKE THESE PROMISES SOLEMNLY, FREELY AND UPON
MY HONOR.
https://medium.com/@vndrewlee/the-analyst-s-oath-9ea7a56eb266#.xei2i08f1
21. @hollielubbock
REGULAR FLOW OF
~5.4 GBOF SOCIAL DATA
PER MONTH
CODE AND THEORY DATA WAREHOUSE
- REGULAR REPORTING
- INSIGHT DISCOVERY
- AD HOC ANALYSIS
CLIENTS
CREATIVES
STRATEGISTS
ACCOUNT TEAMS
ANALYSTS
REPORTING DASHBOARDS
CUSTOM WEB APPS
EMAIL ALERTS
DATA DOWNLOADS
>500 BRANDSMONITORED CONTINUOUSLY
DOZENSOF AD-HOC SELECTORS
SOCIAL APIS AND 3RD PARTY DATA PROVIDERS
How We Use Social Data At Code And Theory
RESPONDING TO CONTEXT
22. LIVES IN BROOKLYN
SPEAKS ENGLISH
DOESN’T TWEET
SO OFTEN
WORKS AT BUZZFEED
ON TWEET DECK
ON HIS IPHONE
LATE FOR WORK
FROM NEW JERSEY
DESCRIBES HIMSELF
JOINED TWITTER IN 2008
LIKES THE
COLOR BLUE
TWEET AUTHOR
RETWEETER
DISPLAYED TWEET
50 CHARACTERS
RAW TWEET
>3,000CHARACTERS
Why do we like raw data?
The full tweet is a huge, rich
packet of data just waiting to
be analysed… Way more than
140 characters
23. @hollielubbock
OFF-THE-SHELF TOOLS ARE BUILT TO ANSWER
SPECIFIC QUESTIONS.
In an agency context, data needs to be
flexible and adaptive in order to address
unexpected questions and insights that arise in
the creative process
https://medium.com/@vndrewlee
Why We Built Our Own Tools
RESPONDING TO CONTEXT
MORE EYEBALLS + MORE INFORMED
JUDGEMENTS = BETTER DECISIONS
Locked in a database, a powerpoint, or a
proprietary tool your data is not doing anything
for you.
24. @hollielubbock
DATA STRATEGY = MEASUREMENT STRATEGY
Data is the measurable digital footprint of every part of your business.
Data is the substance that allows us to effectively measure and
optimise. To be effective, your data strategy must align to the
measurement strategy and in turn be clearly mapped to the business
strategy
Strategy
RESPONDING TO CONTEXT
26. The Background
FOURSQUARE
In early 2015, Foursquare approached Code
and Theory about a data visualization project
for the release of their Pinpoint ad product.
WHERE YOU GO IS WHO YOU
ARE
Utilizing Foursquare’s data (user check-ins and wireless signals), they
were able to determine (or “pinpoint”) real people’s movement noting
“the places you go are the best indicator of who you are.” Improving upon
technology that filtered out inaccurate location data, Pinpoint accurately
allowed advertisers to know their audience–and their tastes–much more
precisely. With this information, they could deliver very specific
marketing and advertising campaigns.
27. REPRESENTING THE DATA
With mounds of information at their disposal and a clear business case
for turning the data into a stand alone business, there was one question:
How could Foursquare clearly present this complicated information to
mid-level marketers in an elegant and intuitive way?
Code and Theory worked with Foursquare to
create a product that would present this
information through a suite of data
visualizations.
The Business Problem
FOURSQUARE
28. PINPOINT THE STORY
Working with our data analysts, visual designers and developers, we
designed and built a system to demonstrate Pinpoint’s accuracy by
projecting visual layers onto an interactive map.
Utilizing WebGL and the Mapbox API, we created a 3D world where you
can see check-in data, GPS data, and Foursquare's own venue shape data
for a given location. Through a set of controls these different layers are
turned on and off at different moments, camera sequences are created to
zoom and rotate around venues, and custom transitions can be applied.
Individual venue sequences are easily placed into a queue to show an
animation moving through the world from location to location, telling
compelling stories.
From the High Line and Empire State Building
to the Mall of America, the visualizations were
sophisticated ways to hone in on geographic
areas and display Pinpoint's accuracy.
The Solution
FOURSQUARE
29. DATA
WARE HOUS E
DATA
SERVER
DATA
SERVER
DATA
SERVER
APACHE
HADOOP
CLUSTER
APACHE HIVE
QUERY FOR
SPECIFIC
DATASETS
BASED ON
PARAMETERS
X,Y,Z
DATASET
FOR V IZ A
DATASET
FOR V IZ B
DATASET
FOR V IZ C
FORMATTED
JSO N FILES
VISUALIZATION
CREATIO N/
MANIPULATION
TO OL
WE B APP
VIZ
A-1
VIZ
A-2
VIZ A-
*
VIZ
C-1
VIZ
C-2
VIZ C-
*
VIZ
B-1
VIZ
B-2
VIZ B-
*
VIDEOS
PRESENTATION
PRESENTATION
PRESENTATION
PRESENTATION
PRESENTATION
PRESENTATION
PRESENTATION
PRESENTATION
PRESENTATION
KE YNOTE
PRESEN TATIONS
33. We look at how our brands’ posts spread across the
network and identify the communities that are engaging
34. WE LEARNED THAT A SMALL NUMBER OF KEY BLOGS
SET OFF EACH CASCADE
The amount of social activity that a user can generate and how central
they are within an active content-sharing network is more important
than their reach
35. @hollielubbock
COMMUNITIES RE-BLOGGING ESSIE’S TUMBLR POSTS (Q4 2013)
64.69% nail polish blogs
25.07% “preppy” style blogs
6.47% “girly” style blogs
1.26% inspirational quotes
TOP ENGAGING ESSIE FACEBOOK POST (Q3 2014)
We use the knowledge of how brands posts spread and use
what we learn to inform creative executions that bring the
underlying consumer truths to life on the brand’s social
channels (in this case lifestyle messaging)
Shifting The Message
RESPONDING TO CONTEXT
36. Key Takeaways
FOCUS YOUR ANALYSIS – MEASURE THE EXCEPTIONS
DATA STRATEGY = MEASUREMENT STRATEGY = BUSINESS STRATEGY
CREATE CROSS FUNCTIONAL PRODUCT TEAMS WHO USE AND REACT TO
CUSTOMER NEEDS AS QUICKLY AS POSSIBLE
WATCH OUT FOR CONTINUOUS OPTIMISATION - FOCUS ON THE FUTURE
RESPONDING TO CONTEXT