The Little Black

Book of
Audience Data

jouncemedia.com
Jounce Media’s Little Black Book series is your
quick reference for all things ad tech. Each book
contains fundamental concepts that underpin the
advertising technology ecosystem. Master the
content in these books, and you’ll be armed to
tackle any digital marketing problem.

For more information about Jounce Media’s ad

tech consulting services, visit jouncemedia.com

© 2020 Jounce Media, LLC All Rights Reserved

 Must-Know Concepts Must-Know Terminology
This book aims to be your audience data reference guide. It Throughout the book, we use the following industry-specific
contains descriptions of the fundamental concepts that underpin terms to describe the programmatic buying and selling of
the audience data collection ecosystem and provides an overview advertising inventory:
of the landscape of data management vendors.

Ad Tech Foundation Publisher A seller of ad inventory – either a website or

an app. Also referred to as the sell side or the
q Identity Management supply side.
q Cookie IDs and Ad IDs
Advertiser A buyer of ad inventory. Also referred to as
q Mobile Contexts the buy side or the demand side.

Data Collection
Ad Exchange A piece of technology that conducts real-time
q Client-Side Pixels auctions for ad inventory. Supply-side
platforms (SSPs) all operate ad exchanges,
q ID Syncing
but also have additional pieces of publisher
q Data Syndication technology.

q App SDKs
Bidder A piece of technology that buys inventory on
q Server-Side Data Transfer ad exchanges. Demand-side platforms (DSPs)
all operate bidders in addition to other pieces
of advertiser technology.
Data Augmentation
q CRM Onboarding

q Transaction Syncing Client Side Everything that happens on a consumer’s

device.
q Deterministic Cross-Device Matching
Server Side Everything that happens remote from a
q Probabilistic Cross-Device Matching
consumer, either in the network or a data
q Segment Design center. Server side activities happen in the
cloud.
q Vendor Landscape
Identity Management
Depending on the advertising context, a single consumer might Cookie IDs are used for browser-based advertising (both
be identified by a variety of labels. Personally identifiable desktop and mobile). Ad IDs are used for in-app advertising
information (PII) is used for email, direct mail, and phone. within mobile devices and connected TVs.

PII Cookie IDs Ad IDs

name@email.com Connected TV
Desktop
Browser

Tablet
1 Main Street
Town, State Zip
Mobile
Browser

+1-555-555-5555 Smartphone
Cookie IDs and Ad IDs
Cookie IDs and ad IDs are the connective tissue between the Each time an ad tech system has access to a client, it
client (the user’s device) and the server (everything happening reads the client’s ID and then finds the matching server-
in the cloud). Ad tech systems assign an ID to each client and side user profile, which contains information about who
then create matching server-side user profiles. the user is and what advertising he/she has seen.

Client Side Server Side Client Side Server Side

User Profile User Profile

Cookie ID: 123 User 123 Ad ID: 456 User 456

Male Female
Age 40 Age 32
Minivan Intender New Parent

Cookie IDs Ad IDs

• Used for advertising within a browser (desktop & mobile) • Used for advertising within native apps (mobile & TV)
• Each ad tech system sets a different cookie ID • All ad tech systems use the same ad ID
• Cookie data is typically accessed via a tracking pixel • Device data is typically accessed via an app SDK
Mobile Contexts
When using mobile devices, user behavior is tracked using a Within native apps, most activity is recorded using the
combination of cookie IDs and ad IDs. All browser activity is device’s ad ID. But when a user loads a webpage that is
recorded using a cookie that is set on the user’s browser, similar hosted within an app (called a “webview”), the user’s
to desktop advertising. behavior is tracked using a new cookie ID that is different
from the cookie set on the user’s primary browser.

Browser Native App Webview

website.com App Newsfeed website.com

Click

Tracking based on a cookie ID Tracking based on an ad ID Tracking based on a cookie ID, though
different from primary browser cookie
Client-Side Pixels
Ad tech companies track web-based activity through client-side The ad tech company then finds the matching server-side
pixels that fire when websites load. Each time a pixel fires, the user profile and records a new event to that profile. Over
ad tech company reads the ID of the cookie it has previously time, pixel fires produce a user profile that contains the
set on the user’s browser. user’s web browsing history.

Client Side Server Side

Website.com
User Profile
User 123

Male
Age 40
Minivan Intender

Website.com visit
Container
Tag

Tracking Pixel
Cookie 123

If no cookie is present on the user’s browser, the

One or more additional ad tech company sets a new cookie, creates a
tracking pixels matching user profile, and then records the visit
in the new user profile.
ID Syncing
Because each ad tech system operates on a different web By establishing that company A’s cookie 123 and company
domain, they have different cookie IDs for the same web B’s cookie 789 are set on the same browser, the two
browser. ID syncing allows ad tech systems to share browser- companies can merge together the matching user profiles
based audience data with each other. to form a more complete understanding of the consumer.

Client Side ID Sync Server-To-Server Data Transfer

Company A Company B

User Profile User Profile

User 123 User 789
Company A Cookie ID: 123

Male Male
Age 40 Age 40
Company B Cookie ID: 789

Minivan Minivan
Intender Intender

The process of ID syncing is necessary only for

browser-based advertising. In-app inventory does
not require ID syncing because each ad tech
company uses the same ad ID.
Data Syndication
Establishing ID syncs between each pair of ad tech companies is By forming ID syncs with each partner in its data network,
an unmanageably large task. Instead, data management a data management platform can facilitate data transfer
platforms (DMPs) act as ID syncing hubs. between any pair of network participants.

Ad
Data Broker Exchange

Data Publisher
Management Ad Server
Platform

Attribution
Bidder Provider
App SDKs
Ad tech companies track app usage through app SDKs. As the Select data events are then syndicated server-side by the
user engages with the app, the SDK keeps a client-side record of app SDK provider to the advertiser’s media buying and
all activity. On a periodic basis (typically when the app is attribution partners.
backgrounded), the cache of recorded events is pushed to the
matching server side user profile via an HTTP connection.

Client Side Server Side

App Analytics Media Buying
Provider Company
Native App

User Profile User Profile

User 456 User 456

[Prior events]

App Open App Open

App Login
App Close

Advertising SDK
IDFA 456
Because all app activity is recorded using a commonly
App Open understood ad ID (either an IDFA or Android ID), ID
App Login syncing is not necessary. App data can be easily
App Close syndicated across ad tech companies.
Server-Side Data Transfer
When it is not possible for an ad tech company to directly track Server-to-server data transfer can be executed in batches
user behavior through client-side data collection, data can be via log files or in real-time user server-side pixels.
transmitted server-side from one ad tech company to another.

Batch Files Server-Side Pixels

A log file containing user data is generated by the data owner Each time the data owner tracks an event, it sends a
and transmitted to the receiving ad tech company. Each notification to the receiving ad tech company. This
record in the log file contains a user identifier, and the notification specifies a user identifier, and the receiving
receiving company updates the corresponding user profiles. company updates the corresponding user profile.

Batch files are most commonly used when transmitting Server-side pixels are most commonly used when
very large data sets that do not need to be ingested by transmitting smaller data sets that must be immediately
the receiving party in real time. Demographic data is ingested by the receiving party. Retargeting data is
typically transmitted via batch files. typically transmitted via server-side pixels.
CRM Onboarding
CRM onboarding enables offline-to-online data transfer. Data With the offline-to-online match established, advertisers
onboarding networks establish matches between offline can make their offline customer data actionable online.
identifiers like an email address and online identifiers like a Everything that advertisers know about the customer
cookie ID. joe@email.com can inform the ads that are delivered to
cookie ID 123.

Offline CRM File Match Network Online DMP

User Profile User Profile User Profile

PII joe@email.com joe@email.com

(personally identifiable
information)

Anonymous Cookie ID 123 Cookie ID 123

Online ID

Customer Male Male

Age 40 Age 40
Information Minivan Owner Minivan Owner

PII + customer information PII + anonymous ID Anonymous ID + customer info

Transaction Syncing
Advertisers can also activate offline customer data by syncing The ad tech company looks up the matching user profile
transaction IDs with ad tech companies. When a sale or other and stores this transaction ID. The advertiser can then
transaction event happens, the advertiser passes a unique supply additional transaction information through an
transaction ID to its ad tech partner. offline batch file, and the ad tech company can update
the appropriate online user profile.

Client Side Server Side

Ad Tech Vendor Advertiser CRM File

Website.com

User Profile User Profile

User 123 Customer 456

[Prior events]

Transaction ABC Transaction ABC

• New Balance
Runner • Nike Dri-FIT
• RaceReady

Tracking Pixel
Cookie 123
Transaction ABC Runner
Deterministic Matching
Because consumers use multiple browsers and multiple devices, Deterministic matching systems use a verified cross-device
ad targeting and attribution systems need a unified way to identifier like a user login to link multiple online IDs. This
identify each consumer across screens. “Device graphs” match approach offers highly precise matches, but limited scale.
multiple online user profiles to each other, providing a more
complete view of the consumer’s digital footprint.

Browser 1
User Profile
Cookie ID 123

jane@email.com jane@email.com

Browser 2 Deterministic
Match
User Profile
Cookie ID 456

jane@email.com jane@email.com
Probabilistic Matching
Probabilistic matching systems link multiple IDs using inferred Each matched user ID is assigned a confidence level, and
data points like a shared location or a shared IP address. advertisers can increase or decrease the desired
Probabilistic matching offers high scale, but is less precise than confidence threshold in order to control match accuracy
deterministic matching. vs. scale tradeoffs.

Most Common IP Address 166.171.250.93 166.171.250.93 166.171.250.93 166.171.250.93

Most Unique Site adexchanger.com jouncemedia.com jouncemedia.com adexchanger.com

Second Common IP Address N/A 65.206.95.146 65.206.95.146 65.206.95.146

Segment Design
Data management platforms (DMPs) and demand side platforms By using Boolean logic (AND / OR / NOT), advertisers can
(DPSs) give advertisers tools to combine more than one audience specify custom definitions of the consumers who should be
into a targeting segment. targeted with advertising.

Audience A
Audience B

OR
Audience A
Audience B

AND NOT

Audience C Audience C
Audience Data Ecosystem
1st Party Data 3rd Party Data
Content Consumption
Data Syndicators

Location History

Browser Tag Managers

Offline Behavior

App SDKs

Identity Resolution
© 2018 Jounce Media, LLC All Rights Reserved