2. About Near
2
Enabling you to understand consumers better with real-world data, and engage with them.
SaaS products for
real-world data enrichment
& data-driven marketing
World's largest source
of intelligence on
People and Places
Bring massive data into an
AI-powered unified platform to
understand consumer behavior
1.6 Billion
Users
44
Countries
GDPR compliant | No PII data |
Consent driven incoming data
Scale/Data Privacy-led design
70 Million+
Places
Processing petabyte of data on monthly basis
3. Corporate Overview
3
Marquee Investors
San Francisco
New York
Bangalore
Tokyo
Singapore
Sydney
Office HQ
2012
Established Year
USD $134mn to date
Capital
Trusted by
London
Product / Tech / Core Biz
France
4. 4
Agenda
Near - Improving Return on Investment
Probabilistic Mapping of People and Places
Reaching Out to Consumers Based on Real-world Signals
Classifying Staypoints and Waypoints
Curating and Analyzing Audience Profiles
Store Footfall Attribution
5. 5
Near - Improving Return on Investment
Staypoints /
Waypoints
Identify if the consumer is at a brick & Mortar
store or walking/commuting/flying while
emitting location pings from the device
Audience profile
Create bespoke audiences using ML/AI
algorithms, which learn from both online
and offline signals
Ping-to-POI
assignment
Accurate assignment of people
visits to brick & mortar stores to
understand offline behavior
Store footfalls
attribution
Measure campaign efficacy based on
store footfalls using large scale
geofenced polygon database
Improved
ROI
6. 6
Near - Improving Return on Investment
Staypoints Waypoints
● Footfalls
● Offline user behavior
● User profiles
● Insights
○ Dwell time
○ Distance travelled
○ Brand affinity
○ Brand propensity
● Historical visit behavior
● Home location
● Work location
● OOH advertising
● Commute patterns
● Commute time
● Navigation search
7. 7
How Near Extracts Staypoints?
Location Pings from a consumer Broken Sessions Staypoints
Staypoints
Place Matrix
Retrieve endpoint of a session
(staypoints) by overlaying places
Time interval
between
pings
Distance
between
pings
Speed
Dwell time
at a place
Post applying below conditions, algorithm extracts multiple sessions for the user
8. 8
How do we probabilistically map people to places?
Place Matrix
Place Metadata
Historical Visit
App History
Transaction
Wifi Signal
Bayesian Inference
Model Framework
Staypoints
Ping-to-POI
Mapping
Example of Distance-based Mapping
Probabilistic Mapping of people to places
using online and offline consumer behavior
9. 9
How Near Segments Audience Profiles?
We have successfully surpassed the Gender/Age Group on-targeting
benchmarks in USA and extending to other markets such as
ANZ, SEA, MEA and APAC
Near ML / AI Platform
Ping-to-POI
Mapping
Place DB
Historical Visits
Device Apps /
Usage
Online Behavior
Device / Signature
Offline / Online
Data store
Entity Resolution
Supervised Learning
Ground Truth
Labeling
Semi-supervised
Learning
Natural Language
Understanding
Reinforcement
Learning
Transactions
People Properties Place Properties
Cloud Engineering MLOps Feature Engineering
Jobs Orchestration
Container
Orchestration
SQL/NoSQL
Databases
Gender
Age Group
Profiles
Interests
Brand Affinity
Affluence
Ethnicity
Home Location
Work Location
Household
Place Boundary
Brand
Category
Store Hours
Gender Distribution
Age Group Dist.
Profile Dist.
Footfalls
Dwell Time
Distance Travelled
Frequent Visitors
Attribute
Store
Consumer attribute
Store
Places attribute
Store
Nielsen Digital Ad Ratings
10. 10
How Near does Store Footfall Attribution?
Digital Campaign
Footfalls
Exposed Group
(from Digital
Campaign)
Control Group
(ActAlike)
Footfalls are measured for pre,
during and post-campaigns
1. Footfalls are measured with
the most accurate
information from location
pings and places DB
2. Near Places DB is one of
the largest source of
Building Place Boundaries
and we are adding more
every day
3. Location Pings are curated
for staypoints and removing
any anomalies in the data
4. We use distributed
technologies to couple the
two sources over large scale
(several TBs every day)
5. Time series model
estimates / offsets any
missing data on a given day
Attribution
Footfalls during
COVID-19
Digital Campaign and Offline Attribution
11. 11
Use Case: Marketers / Advertisers can now target super-refined segments
With Allspark, you can analyze and reach out
to consumers based on
Gender &
Age Group
Profiles
Interests
Events Weather
Income /
Affluence
Home
Location
Brands
Affinity
Stores
Proximity
Curate Activate Measure
12. Allspark is SaaS Platform for
curating and activating
audiences through your choice
of a DSP and measure
campaign effectiveness
Simplified marketing
powered by the world’s
first AI audience assistant
Allspark
Carbon is an data enrichment
platform where we augment the
customer 360 view by using ML/AI
approaches for matching user
attributes both in offline and
online world
The best software
platform for data
enrichment
CARBONTM
Curate, Activate, MeasureData Enrichment
13. 13
Get Campaign Insights
Daily Offline
Attribution Report
Delivering Performance
Deliver Ads
Programmatically
Activate Campaigns using
Custom Creatives
Reaching Audience
Estimate
Campaign Reach
Curate Bespoke
Audiences
Media Planning
Pre-Sales Market
Research Reports
Export Audiences
To DSP
Advanced
End-of-campaign Report
Activation MeasurementAudience Curation
Allspark Supports the entire Marketing Campaign Lifecycle based on Real-world Behavior
16. THANK YOU
Connect with us at near.co
From inception, the Near Platform has followed a privacy-led design. The Platform never stores or deals with PII (Personally Identifiable Information)
and all incoming data streams are consensual. We are GDPR compliant, and the platform has built-in processes to forget and purge user data on
requests. Read the complete privacy policy at near.co/privacy