2. Speakers
#PitneyBowesData / #locationdata
Rob Minaglia
VP, OEM and Platform Sales, Pitney Bowes
rob.minaglia@pb.com
@robminaglia
Michael Villarreal
Senior Data Specialist, Pitney Bowes
michael.villarreal@pb.com
Steve Mateer
Data Channel Executive & Moderator, Pitney Bowes
steve.mateer@pb.com
3. Location-based Marketing Questions from Brands
Can I confirm that
customers visit my
locations during
campaigns?
How can I improve
creative based on
location context?
Where do my customers
live and work? What does
this tell me about them?
How can I understand where
my customers are now and
where they have visited?
How can I deliver the right
messages to the right
customers at the most
relevant times and
locations?
4. About Pitney Bowes
Trusted by these brands
• $3.5B revenues, over 1M clients
• Orchestrate mail, shipping and cross
border e-commerce for 15 billion
items / year
• Customer engagement, retail site
analytics, data quality and single
customer view
• “Address Fabric” – location context
around people, places and things for
marketing, site selection, compliance,
property valuation, risk mgmt.
• Identity resolution – linking physical,
social and digital into one profile
Enabling these platforms
About Pitney Bowes
5. Location Context is Foundational
Insights
&
Demand
analytics
Find
lookalike
customers
Site
selection &
Catchment
analysis
Points of
Interest
Boundaries
&
Areas of
interest
Physical
address
Digital &
Social Identity
Demographics
Geo-location
Geo-targeting
Media
placement
Journey
mapping &
attribution
Enrich
audience
profiles
7. Spatial analytics, data & visualization to enhance insights, activate
media
Optimize media activation based on
catchment areas and competitive locations
Map first party sales data to
brand locations
8. Challenges with audience creation from mobile
34% of advertisers reports challenges
associated with inaccurate location data
• Faulty location signals
• Outdated/incomplete reference data
• “Centroid” data can be deceiving
• Precision vs. accuracy limits (eg, urban density)
27% report obstacles around data collection
and transparency
Source: June 2017 study conducted by Forrester on behalf of Verve
The Mobile Marketing Association (MMA) guidelines for location data quality
(March 2017), highlighted that “measurement vendors must disclose specific
methodologies employed (eg, for geo-fencing or user assignment);” and that it is
“essential that locations are accurate.”
9. For DSP’s…develop audiences more accurately, faster and with greater efficiency…even using drive
times or demographics to enable self serve campaigns
For agencies…deliver richer audience insights to clients
Benefits for agencies…deliver richer audience insights to clients
Lat / longs
Correlate to
brand, category,
what’s near me
Build &
enrich
audiences
Targeting
Measure Match Spatial search
Footfall
Attribution
Insights &
Mapping
Analytics
Location
signals
ActionsActivation
First party
data (eg,
sales)
Media
activation
Match to
identity
graph
Match to
location
10. Match location signals to Points of Interest (POI)
Factors impacting accuracy, ability to provide context and integration
• Approaches to building
• Authoritative sources vs. screen scraping vs. check ins and web
scraping; use of ML
• Validation methods
• Update frequency (monthly)
• Coverage: Number of points, countries covered (166M / 173)
• Validated brand location & parentage (3400+ brands)
• Consumer, leisure, business & geographic places
• Number of categories and attributes per Place (19,000 / 72)
• Consistency of data structure across geos
• Accuracy of addresses and lat/long
• GDPR compliancy
• Integration with polygons; Drive and Walk Times (“isochrones”); and
enrichments such as Financial Vitality scores, demographics
19,000 Categories =
more discrete segmentation
157 Restaurant
types
11. Auto Dealership Boundaries provides content and data
attributes for over 42,000 geofenced auto dealer buildings
and lots for both franchise and independent dealerships.
Auto Dealership Boundaries – confirming on-site visits
Mobile device locations
- GPS lat/long
- device ID
Auto Dealership Boundaries
Confirming on-site visits using
precise polygons against
known mobile device
locations within a given time
frame
12. Geo-targeting using boundaries and spatial search
…DMA’s and zips just won’t do in a mobile world…
Source: IAB Location Terminology Guide
DRIVE / TRANSIT TIME
(embedded within the data)
MICRO-BOUNDARY
Business/building/parcel
• Brand or category
• Store, mall
• Venues
• Auto dealers
• School, campus
• Park, golf courses
• Airports, train stations, etc.
GEO-FENCING
(Point & radius)
Simplest but can be
inaccurate:
• “cross river”
• “false alarm”
• “overlap” across brand
locations in urban areas
POLYGON BOUNDARY
Admin & trade areas
• Postal code, DMA, census
block
• Neighborhood, school zones
• Trade areas
13. Community & Venue Boundaries
For Area of Interest (AOI) Geofences
• World Postcode & Admin
Boundaries
• Community Boundaries
• Neighborhoods (220,000)
• Schools & Colleges
• Cities
• Metros
• Property (Parcel) Boundaries
(141M)
• Building Footprints
• Buildings (61M)
• Points of Interest (24M in US )
• Auto Dealers (42K) – (branded)
• Venue Footprints
• Golf Courses
• Train Stations
• Stadiums
• Ski Areas, Casinos
• Commercial neighborhoods
14. Matching Location Signals for Enriched Audiences
Match first and third party data for enriched context
Establish dwell time and density within a place, and travel
patterns (home/work/shop)
Establish location type
– Commercial: Points of Interest (POI)
– Communities: Neighborhoods, Schools, Cities
– Social Venues: Shopping Malls, Stadiums, Golf Courses,
transit centers, etc …
Establish location position
– Business and Building (US only)
– Parcel (US/UK only)
– Post code
– Same side of road
– Within distance and travel time to target
Establish who lives in the location
– demographics, socio-economic and segmentation data
15. • Data science
• Machine
Learning
• AI / Insights
• Modelling
• Workflows
• Visualization
Consumer
spending
Demo-
graphics
Property
Attributes
pbKey
First party data
(eg, sales)
Other third-
party data
Location Data + Spatial Processing = Geo-enrichment for
16. Differences between Demographics and Geodemographics
Demographics
Raw census data e.g. what is the count of 18 to 25
year olds in an area
Used primarily for retail analysis
Geodemographics
Consumer segmentation based on the types of
people who live in a neighborhood. e.g. Category
1A high society families
Used primarily in marketing, market research
and advertising
18. 19
PB Key 200 Million NA addresses
Geocoded overlays for any
enterprise
Utmost accuracy and complete
reach for offline persona
House holding data
Demographics
Behavioral Data
Transaction Data
1.3 Billion Digital Personas
10s of Billions of potential match
keys to enter the graph
Utmost accuracy and complete
reach for online personas
Psychographics
Professional & Educational history
Location Data Social Data
Joining Location with Digital & Social Identity
19. Benefits for DSP’s…develop audiences more accurately, faster and with greater
efficiency…even using drive times or demographics to enable self serve campaigns
Benefits for agencies…deliver richer audience insights to clients
Step 1
Obtain high
quality location
signals /
addresses and
reference data
Step 2
Match to in / near
POI and/or
boundary (AOI)
for location
confirmation &
context
Step 3
Enrich audiences
to improve
context
Step 4
Model campaigns
leveraging
different attributes
Visualize results
Enable footfall
attribution
20. Recommendations
1. Stay current on best practices and innovation
IAB Mobile Location Handbook: https://www.iab.com/wp-
content/uploads/2017/04/Mobile_Location_Data_Handbook_April.2017.pdf
Mobile Marketing Association: http://www.mmaglobal.com/programs/location
2. Due diligence
• Inspect methods of producing data and analytics
• Update cycles, benchmarks, consistency
3. Choose suppliers with whom you can grow
Data enrichment roadmap
Ease of integration/data layering and documentation
Software for data quality, spatial analytics and visualization
Register on Software and Data Marketplace for sample data & visualization
https://signup.pitneybowes.com/signup/sdm?purl=rminaglia
Trial API’s at https://www.pitneybowes.com/us/developer/geocoding-apis.html
Contact us for benchmark data and custom data sets
21. THANK YOU FOR ATTENDING
Insights
&
Demand
analytics
Find lookalike
customers
Site selection
& Catchment
analysis
Points of
Interest
Boundaries
&
Areas of
interest
Physical
address
Digital & Social
Identity
Demographics
Geo-location
Geo-targeting
Media
placement
Journey
mapping &
attribution
Enrich
audience
profiles