2. Metrics
across the
customer
journey
2
Amazon advertising drives awareness, consideration and sales
Amazon advertising data reflects
behaviors of billions of customers’
shopping experiences as they
research, consider and purchase
products both on and off Amazon –
all at scale.
3. 3 Source: 1) ComScore, March 2016; 2) ComScore, December 2015
Delivering ads to relevant audiences, empowers marketers to:
Single login
Amazon’s login-based audience segments allow
advertisers to stay engaged with the same shoppers when
they use different screens, thereby supporting holistic,
integrated campaigns
Cross-device attribution
Amazon attributes ads and conversion events―including
when they occur on different devices
Reach and target audiences at scale
181MM unique visitors that shop on Amazon every month1,
that ranks in the top 5 for all publisher sites and has more
visitors than the next 15 retail sites combined2
4. 4
Amazon targeting makes
your message more relevant
Targeting types:
Amazon leverages billions of first party,
real-time shopping behaviors to deliver
relevant ads to your target audience, on
and off Amazon
In MarketLifestyle Lookalike & Remarketing Demo & GeoContextual
5. Pixel-based
Anonymous customer
match
Amazon first-party
data
Lookalike
Types of Targeting at Amazon
Age/ Gender
Location
Income level
Time of day
Browser
Demographic &
Geographic
In-market
Lifestyle
Behavioral
Pixel-based
Anonymous customer
match
Remarketing
5
Contextual
Product category
Segments based on
historical search, browsing
and purchase history
Segments based on the
detail page a consumer
is actively viewing
Segments to extend
campaign reach
Custom segments
created uniquely for
an advertiser
Segments to reach a
specific audience type
6. Lifestyle targeting leverages Amazon first-party data from up to the past year
of habitual search, browsing and purchase behaviors.
Lifestyle Targeting Segments
Health Strivers
Customers who have browsed or purchased health
and/or fitness related products in the past 12
months.
Video Gamers
Customers who have browsed or purchased
products from the Video Game product
category over the past 12 months.
Lifestyle
In-Market
Contextual
Lookalike &
Remarketing
Demographic
7. In-market targeting allows an advertiser to reach customers who have
recently browsed for a product within a specific category
In-Market Targeting Segments
Lifestyle
Contextual
Demographic
In-Market
Lookalike &
Remarketing
8. Contextual product category targeting delivers ad creative based on the
detail page the customer is actively viewing – rather than leveraging a
customers’ historical interactions with Amazon.
Contextual Product Category Targeting
Lifestyle
In-Market
Lookalike &
Remarketing
Demographic
Contextual
9. Remarketing your lost customers
» Customers who have visited
your brand’s website
» Your Lost Customers» Customers who viewed your
products but did not purchase
10. Targeting shoppers of similar products
People who viewed products similar to your products
Shoppers that have
looked at products like
yours
11. Target new people who exhibit similar behaviors to your current customers or site
visitors through audience lookalike targeting
Extend your reach through Lookalike Targeting
Use data collected
from an advertiser’s
website via a
remarketing pixel
Create Advertiser
Audience segments by
matching your CRM
data to Amazon
shopper data
Use Amazon
first-party
shopping data
Lifestyle
In-Market
Contextual
Demographic
Lookalike &
Remarketing
12. 12
How can we use Advertiser Audience segments in campaigns?
Reach NEW customersReach EXISTING customers Reach LOOKALIKE customers
Lookalike &
Remarketing
13. 13
How to create an Advertiser Audiences Customer List segment
1
2
3
Build a list of at least 20,000 email addresses into a .txt file
Your hashed file is matched with a list of
hashed Amazon customer email addresses
Matched records become your Advertiser Audiences segment and
unmatched records are removed. All hashed data is deleted immediately4
Submit list to Advertiser Audiences to hash
Your browser will hash the file before upload
OR
Hash the list yourself using SHA-256
14. 14
Business results of Advertiser Audiences from early case studies
Results over campaign averages in all customer journey stages
Awareness: 2x CTR
Consideration: 9x DPVR
Purchase: 4-8x Purchase Rate
ROAS lift: 100-700% over campaign averages
Est. 5-10% cost savings when focusing on new customers
Lookalike customers have 200% uplift in traffic to detail pages, 4x
more likely to purchase
15. Amazon can reach specific audiences by targeting customers based on:
Demographic Targeting Segments
Gender Age Children
in household
Household
income
Lifestyle
In-Market
Contextual
Lookalike &
Remarketing
Demographic
Amazon can provide visibility into millions of customers’ shopping behaviors as they move through the customer journey. In real-time, we collect browsing and purchase signals from customers across all channels and touchpoints including AAP, Mobile, Kindle, Fire TV, etc. To make this data actionable, Amazon provides targeting capabilities for an advertiser to reach their target audience both on and off Amazon based on customers’ observed shopping behaviors. This means that any time a customer takes an action on Amazon.com, they can be included in targeting segment(s) within a second of that action to enable advertisers to reach the right customers.
Amazon deliver insights throughout the customer decision journey from ad exposure to conversion.
As the 5th largest publisher Amazon has broad reach and in terms of retail we have a larger audience than the next 15 sites combine
The single Amazon login allow us to gather insights across the customer journey
Which allows for Cross-device attribution from ads and conversion events―including when they occur on different devices (Note we do not have complete multi-touch attribution, we have a last click attribution model and have the “Global Ad Assist Report” to measure assists.
Amazon targeting leverages millions of customers' real-time shopping behaviors across devices, allowing advertisers to reach the right customers with a relevant
and timely message when they're on Amazon, or on hundreds of websites and apps. Amazon algorithms work to instantly define audience segments based on
their browsing and purchase behaviors and optimize campaign targeting towards the most effective audiences.
Lifestyle: reach audiences with specific interests such as favored brands, and products as shown by past year of search, browse and purchase behaviors
In-Market: reach those who show signals that they might be in the market for a product based on recent search and browsing history
Contextual: reach desired audience by targeting with relevant ad content that aligns with the content of the detail page that a customer is viewing
Advertiser Specific: Segments are created using anonymized data sent from an advertiser’s website via a pixel or through a customer list match with Amazon customers.
Demographic: reach desired demographic audience by targeting customers based on their gender, age, children in the household and household income.
Note: Targeting works best when layered together.
Here is a look at our suite of targeting practices, some of which truly differentiate us in the publishing space.
Demographic: reach your desired demographic audience by targeting customers based on their gender, age, children in the household and household income.
To provide this targeting, Amazon uses anonymized, third-party data. We also can use core targeting parameters, such as location, time-of-day, and browser.
Note of caution when using demo targeting: some key audiences can be missed using demo targeting alone.
Contextual: Contextual targeting delivers ad creative to a customer based on the current detail page the customer is viewing. It analyzes the content of the detail page that a customer is viewing to determine the type of ad creative and message to serve while the customer is in a shopping mindset.
Behavioral targeting: Lifestyle and In-Market
Lifestyle:
Reach audiences with specific interests such as favored brands, and products as shown by past year of search, browse and purchase behaviors
Lifestyle targeting leverages up to 365 days of observed, first party search, browsing and purchase behaviors.
In-Market:
In-Market segments include customers who show signals that they are actively shopping for a product based on their recent search and browsing history within in the last 30 days
Unlike other major publishers, these audience segments are based on observed first-party purchase behaviors updated in real-time across devices.
Remarketing and Audience Lookalike
Reach desired audiences by using anonymized advertiser data (with permission of the advertiser) for remarketing, building audience lookalike modules or creating segments of common/overlap customers between the advertiser and Amazon Targeting.
Remarketing allows an advertiser to continue engaging with customers with whom they already have a relationship.
To build remarketing segments, advertisers must agree to place an Audience pixel on their site. Amazon will use information generated by the pixel to create audience segments and target campaigns for the advertiser. Once a pixel is in place, the advertiser can further engage these customers who have already visited their site.
Remarketing segments are only available to the advertisers for which they were constructed.
Remarketing is beneficial for non-endemic advertisers and smaller advertisers whose products are not sold on Amazon or whose category is still growing on Amazon.
Audience lookalike targeting enables advertisers to extend reach by targeting audiences who exhibit behaviors similar to those who have previously purchased products or shown desirable behaviors such as subscribing to an advertiser newsletter or filling out an advertiser’s form. Lookalike targeting allows Amazon to create new unique behavioral segments specific to the advertisers’ success metrics, and these segments are updated in real-time. Lookalike Audiences can be created in a few different ways:
Using information provided via a pixel on the advertiser’s website
From an advertiser’s customer list via Advertiser Audience Segments
Using Amazon first-party data as the basis for the lookalike (i.e., create a lookalike of shoppers who have browsed for socks over the last 90 days)
Lifestyle targeting leverages Amazon first-party data from up to 365 days of habitual search, browsing and purchase behaviors that are used to signal specific interests such as favored brands, products and sub-categories.
In-market targeting allows an advertiser to reach customers that are likely to take action or make a purchase because their behavior shows signals that they are actively shopping for a product based on their recent browsing history within in the last 30 days.
For example, customers who are “In-market for HDTVs” have, in the past 30 days, browsed on detail pages within the HDTV product category. In-Market segments are established at the product category (Electronics), and sub-category (Laptops) levels for many categories and at brand level (Google Chromebooks) for some products.
Contextual product category targeting allows advertisers to deliver relevant ads to users while they are actively browsing on a detail page within the contextually targeted product category or browse node. Unlike Amazon’s other targeting options, Contextual targeting does not leverage a customers’ historical patterns or previous interactions with Amazon; instead, Contextual product category targeting delivers ads based on the detail page a user is viewing in real-time. This ensures that relevant ads are delivered to the right audiences at the right time - while they are actively researching and considering products.
These contextual browse node segments allow advertisers to target customers who are in the middle of their customer journey and are already shopping for specific categories of products, such music software, suspense books, bicycles, strollers etc. Contextual product category targeting is designed to drive consideration and purchase behaviors. For endemic link-in campaigns that have direct response goals, we recommend that campaign managers add contextual targeting segments as a separate line item in the media plan per standard operating procedures.
*Contextual targeting is now only available on product detail pages.
Remarketing allows an advertiser to continue engaging with existing customers or prospects. Leveraging a remarketing pixel, advertisers can remarket to people who may have viewed their website or filled out a form.
To build remarketing segments, advertisers must agree to place a remarketing pixel on their site. Amazon will capture data generated by the remarketing pixel to target campaigns for the advertiser. Remarketing segments are only available to the advertisers for which they were constructed. Remarketing is beneficial for non-endemic advertisers and smaller advertisers whose products are not sold on Amazon or whose category is still growing on Amazon.
+ 600 segments
Audience lookalike targeting enables advertisers to extend reach by targeting new people who exhibit behaviors similar to customers who have previously purchased products or showed desirable behaviors such as subscribed to a newsletter or filled out a form. Lookalike targeting allows Amazon to create new unique behavioral segments specific to the advertisers’ success metrics. A shopper is moved in or out of a model-based segment within a second after changing their shopping patterns. Lookalike Audiences can be created in a few different ways:
Using data collected from an advertiser’s website via a pixel
From an advertiser’s CRM database via Advertiser Audience Segments
Using Amazon first-party data about consumers who have previously purchased or interacted with the advertisers’ products and detail pages
Directly reach your existing customers by targeting the matched overlap of people who are both your customers and Amazon’s customers.
Find people who look like your customers by using lookalike targeting.
Only reach new customers on and off Amazon by anti-targeting matched customers.
Put a list of at least 20,000 email addresses into a text file, with each email on a separate line.
2. You have the choice of hashing the list yourself using an SHA-256 utility of your choice OR submitting the raw text file of email addresses for the tool to hash in your browser. If you submit a raw text file of email addresses, the list will be hashed in the browser before being sent to Amazon. The Advertiser Audiences Customer List tool will never submit any unhashed data and Amazon will never be able to view your raw email list.
- If pre-hashing, you must input email addresses in all lower-case with no spaces and only 1 email per line (all emails will be converted to 64 alphanumeric characters)
- If pre-hashing, be sure to select the checkbox specifying “This file contains hashed data”
- Advertiser Audiences will not accept any unhashed data. Even if there is 1 raw email in the pre-hashed list, the tool will not send the data to Amazon
- Hashing is performed using a one-time key that changes each time you use the tool. This is to fully protect all private data
3. Your hashed file is matched with Amazon’s hashed customer email addresses in an encrypted environment
4. The matched emails become your Advertiser Audiences segment, and the unmatched emails are removed. Once your new segment has a status of “Active”, you can use it the segment targeting field when creating line items. There is no audience fee applied to Advertiser Audiences segments. All data is immediately deleted
What happens when I click ‘Create Segment’?
If you have not pre-hashed, the email list is hashed locally on your browser and is transferred to Amazon. Once Amazon receives all of the hashed data, we run a match of it to Amazon’s hashed data. A segment is then created comprised of the overlapping records identified via the matching process.
If your data is pre-hashed and you have selected the pre-hash checkbox, the tool will check to ensure the data has indeed been secured using SHA-256 (if not, it will return an error and will not send any data to Amazon). If successful, the data is transferred to Amazon for the same matching process to find overlapping records.
All results are sourced from 3 case studies available on the advertising.amazon.com website and reflect campaigns in q4 2016 from Nespresso, Burt’s Bees (Clorox), and Ubisoft.
ALL STAGES OF THE CUSTOMER JOURNEY
Relevance applies to any campaign focus, as speaking to the right audience is more likely to achieve the results you want. Including current customer segments into an AMG media plan can help to optimize the use of media for any part of the customer journey.
GREATER RETURN
In early case studies we’ve seen purchase rate increase 4-8 times over. This is because advertisers are able to leverage their existing segments of prospects and loyalists in their CRM database for retargeting and lookalikes. Greater sales for the same advertising spend with no cost to use Advertiser Audiences translate to a lift in ROAS. Segments who are matched using Advertiser Audiences have experienced a ROAS increase of 100-700% over the total campaign average.
NEW CUSTOMERS
In one case, Nespresso focused on new customers by matching current machine owners with Advertiser Audiences and excluded them from the campaign. This created savings since they weren’t serving redundant ads to current machine owners to buy a machine (thus saving money on the CPMs). This meant their ad budget could be better utilized by serving new machine ads to an audience that was not currently in their CRM system.
LOOKALIKES
Creating a lookalike segment from your best customers leverages data science to uncover common traits that are likely to drive more consideration and purchase. In a case study, Burt’s Bees created lookalikes from their most loyal customers, uncovering new prospects with similar attributes and were likely to also be future loyal customers. These lookalikes had a 200% increase in detail page traffic and were 4x more likely to purchase than the campaign average.
Amazon can reach advertisers’ desired demographic audience by targeting customers based on their gender, age, children in the household and household income. To provide this targeting, Amazon matches shopper ids with third-party demographic data sources. We also can use core targeting parameters, such as location, time-of-day, and browser.
Note of caution when using demo targeting: some key audiences can be missed using demo targeting alone because of users who share Amazon login ID’s across a household.