0% found this document useful (0 votes)
9 views13 pages

Amazon Terminology Definitions

Uploaded by

3.dost.001753
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
9 views13 pages

Amazon Terminology Definitions

Uploaded by

3.dost.001753
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Definitions

Table of Contents
1. ASIN
2. Product Type
3. Attribute
4. Amazon Service
5. Conversation
6. Turn
7. Query
8. Conversation Layer
9. Base Layer
10. Context
11. Shopping Goal
12. Query Goal
13. Product Segment
14. Evidence

ASIN

A specific product that has its own Product Detail Page on [Link]. It is a unique product
instance with a barcode, a price, etc., something that a customer can directly click on, add to cart,
or buy. Example: [Link]

(go back to Table of Contents)

Product Type

A category of products, which can be broad (e.g. “furniture”) or specific (e.g. “rechargeable
double a batteries”). A product type is a set of products, it is not a unique and specific product
that a customer can directly click on, add to cart, or buy. Examples: “earrings”, “fishing gear”,
"Catan board game".

(go back to Table of Contents)

Attribute

A feature/characteristic/aspect of a product type or ASIN. Examples: color, brand, weight,


reviews, compatibility, durability, easiness to clean, installation requirements, prime eligibility.

(go back to Table of Contents)

Amazon Service
An Amazon service/program. Examples: Prime, returns, order history, deals, etc.

(go back to Table of Contents)

Conversation

A dialog between the customer and the Amazon assistant. A conversation consists of multiple
turns. Example:

• Turn 1:
o customer: "what do i need to play golf"
o assistant: "To play golf, you will need (...)"
• Turn 2:
o customer: "show me you clubs for beginners"
o assistant: "Here are our best-selling golf clubs for beginners (...)"

(go back to Table of Contents)

Turn

A turn is a query-answer pair in a conversation. Example:

• customer: "what do i need to play golf"


• assistant: "To play golf, you will need (...)"

(go back to Table of Contents)

Query

An action taken by the customer while shopping on Amazon. Exhaustive list: clicking on an
ASIN link, typing a text (command, question, or statement) in the Amazon search/conversation
text field and pressing enter.

(go back to Table of Contents)

Conversation Layer

The text field that contain the customer queries and the assistant answers.

(go back to Table of Contents)

Base Layer

The page behind the conversation layer, which shows search results or product details. It acts in
sync with the conversation layer. When the customer types a query, the assistant can both refresh
the base layer page and write some answer in the conversation layer.
(go back to Table of Contents)

Context

The Amazon page on which the customer is when they make a query. Examples: home page
([Link] search results page
([Link] product details page
([Link]

(go back to Table of Contents)

Shopping Goal

The Shopping goal or mission is the reason the customer starts interacting with Amazon
assistant. We define 6 stages depending on how advanced in the shopping journey the customer
is:

stage description example explanation


The customer doesn't mention any
product. They just mention an
Shopping ideas for indoor
I'm looking for inspiration interest (indoor gardening), and they
- stage 5 gardening.
would like to be shown some
products about that interest.
The customer doesn't know what
products they want. They just have
I have a problem I'd like to
Shopping what do i need to a problem (changing a tire) and
solve, but I don't know what
- stage 4 change a tire? would like to be shown products
the options are
that can help them solve the
problem.
what is important
I know the kind of product I The customer knows what
Shopping to consider when
want, but I don't know product they want. Now they want
- stage 3 buying a washing
much about them information about that product.
machine
The customer knows what
I know the kind of product I do you have 1984
Shopping product they want. Now they try to
want and I have a few in paperback
- stage 2 find the right ASIN by filtering
constraints in mind under $20
based on their criteria.
does it have a The customer knows what
I've narrowed it down to a
Shopping higher refresh rate product they want. Now they try to
few products, I need to
- stage 1 than the previous find the right ASIN by comparing
figure out the right one
one? the candidates.
I know the product I want, energizer double
Shopping The customer knows exactly
now I need to find it and a batteries 32-
- stage 0 which ASIN they want.
purchase/repurchase it pack
(go back to Table of Contents)

Query Goal

The intent or query goal/type is the intent of the customer at the turn level. We define the query
intents as follows:

Example of
Intent Entity Description Examples Comment
expected answer
If a
customer
says "show
me your
- show me christmas
your christmas gifts on
gifts on sale", they
sale<br>- do want to be
you have it in bullet shown
paperback<br points<br><br>(e.g. specific
The customer
>- what's the "Here are our best- products
Recommendat wants
ASIN cheapest selling novels:<br>- that have a
ion recommendati
option<br>- X: blabla<br>- Y: price (a
ons for ASINs
what tennis blabla<br>- Z: discounted
racquet do you blabla) one), so
recommend they want
for a skilled to be
player<br>- shown
winter jackets ASINs, not
product
types like
earrings or
handbags.
If a
customer
asks "what
refresh rate
should i
The customer
get" while
wants - what refresh
Recommendat ASIN on a
recommendati rate should i same as above
ion attribute product
ons for ASIN get
details
attributes
page, they
want to be
shown a list
of refresh
rates with
Example of
Intent Entity Description Examples Comment
expected answer
their pros
an cons.
- what can i
give my
mother-in-law
The customer for her
wants birthday<br>-
Recommendat
product type recommendati things i need same as above
ion
ons for product to play
types golf<br>-
types of
digital
cameras
The customer
wants
Recommendat product type - best coffee
recommendati same as above
ion attribute maker brand
ons for product
type attributes
The customer
wants
- is there a
Amazon recommendati
Recommendat way to
Program/Serv ons for same as above
ion automatically
ice Amazon
re-order stuff
programs/servi
ces
For the
question
"does it
come in
The customer - does it come
red", the
wants an in red<br>- a value<br><br>(e.g.
customer is
Factoid ASIN objective piece how much is "yes", "red", "35
asking for a
of information the refresh grams")
piece of
about an ASIN rate
information
(the color)
for an
ASIN
The customer For the
wants an question "is
- is the weight
ASIN objective piece the weight
Factoid in kilograms same as above
attribute of information in
or pounds
about an ASIN kilograms
attribute or pounds",
Example of
Intent Entity Description Examples Comment
expected answer
the
customer is
asking for a
piece of
information
(the unit of
measureme
nt) about an
ASIN
attribute
(weight).
For the
question
"what's the
most
common
brand of
- does Tylenol washing
require a machines",
The customer
medical the
wants an
prescription<b customer is
objective piece
Factoid product type r>- what's the same as above asking for a
of information
most common piece of
about a
brand of information
product type
washing (most
machines common
brand)
about a
product
type
(washing
machine).
For the
question
"how long
The customer
is the
wants an - how long is
average
objective piece the average
product type wash cycle
Factoid of information wash cycle of same as above
attribute of a
about a a washing
washing
product type machine
machine",
attribute
the
customer is
asking for a
Example of
Intent Entity Description Examples Comment
expected answer
piece of
information
(time)
about a
product
type
attribute
(wash
cycle).
The customer
wants an
objective piece
Amazon
of information - how much is
Factoid Program/Serv same as above
about an prime
ice
Amazon
programs/servi
ces
The customer
wants a - what do a quote<br><br>(e.g.
subjective people think "Most customers find
Opinion ASIN
piece of of this<br>- is these sandals very
information it comfortable comfortable")
about an ASIN
The customer
wants a
subjective - is the graphic
ASIN
Opinion piece of card of this same as above
attribute
information laptop the best
about an ASIN
attribute
- what do
The customer
people think
wants a
of nespresso
subjective
machines<br>
Opinion product type piece of same as above
- how
information
comfortable
about a
are sandals for
product type
hiking
The customer - is a high
wants a refresh rate
product type
Opinion subjective good for the same as above
attribute
piece of eyes<br>-
information what do
Example of
Intent Entity Description Examples Comment
expected answer
about a people think
product type of adidas
attribute
The customer
wants a
subjective
Amazon piece of
- is prime
Opinion Program/Serv information same as above
worth it
ice about an
Amazon
program/servic
e
a descriptive
paragraph<br><br>(e
.g. “This gaming
The customer laptop by ASUS has
wants a a 19" screen, the
definition, latest NVIDIA
- tell me about
Description ASIN description, graphic cards, and a
this product
explanation, or intel i7 processor,
summary of an which makes it
ASIN particularly suited for
high-consuming
games like X and Y
(...).")
The customer
- tell me about
wants a
these shoes'
definition,
ASIN laces<br>-
Description description, same as above
attribute why does it
explanation, or
only have
summary of an
three stars
ASIN attribute
The customer
- what should i
wants a
know about
definition,
gaming
Description product type description, same as above
laptops<br>-
explanation, or
what is a
summary of a
driver
product type
The customer
- what does
product type wants a
Description paperback same as above
attribute definition,
mean<br>-
description,
Example of
Intent Entity Description Examples Comment
expected answer
explanation, or what is a
summary of a refresh rate for
product type
attribute
The customer
wants a
definition,
Amazon description, - what is
Description Program/Serv explanation, or amazon music same as above
ice summary of an unlimited
Amazon
program/servic
e
If a
customer
says "show
me your
christmas
gifts on
sale", they
The customer want to be
wants shown
a recipe/step-by-step
instructions, specific
- tell me how instructions<br><br>
guidelines, or products
it works<br>- (e.g. "First, you
Instruction ASIN procedure to that have a
tutorial for should X, then Y,
achieve price (a
this product (...). Remember to
something discounted
also Z.")
with respect to one), so
an ASIN they want
to be
shown
ASINs, not
product
types like
earrings or
handbags.
The customer - how do i If a
wants repair this customer
instructions, bike's asks "what
ASIN
Instruction guidelines, or wheel<br>- same as above refresh rate
attribute
procedure to how do i should i
achieve install it on get" while
something my computer on a
Example of
Intent Entity Description Examples Comment
expected answer
with respect to product
an ASIN details
attribute page, they
want to be
shown a list
of refresh
rates with
their pros
an cons.
The customer
wants
- recipe for a
instructions,
chocolate
guidelines, or
cake<br>-
Instruction product type procedure to same as above
how does an
achieve
electric bike
something
work
with respect to
a product type
The customer
wants
instructions,
guidelines, or
- i want to
product type procedure to
Instruction change a same as above
attribute achieve
bike's tire
something
with respect to
a product type
attribute
The customer
wants
instructions,
- how do i
guidelines, or
return an
Amazon procedure to
item<br>-
Instruction Program/Serv achieve same as above
instructions to
ice something
sign up for
with respect to
prime
an Amazon
program/servic
e
- what's the a side-by-side
The customer
difference comparison<br><br>
Comparison ASIN wants a
between these (e.g. "X has a lower
comparison of
two dresses A and a thinner B,
Example of
Intent Entity Description Examples Comment
expected answer
two or more whereas Y has a Z,
ASINs which makes it more
suited for W.")
- is it more
The customer
comfortable
wants a
than the
ASIN comparison of
Comparison previous same as above
attribute two or more
one<br>- is
ASIN
the refresh
attributes
rate higher
- compare
1984 and
Fahrenheit
The customer
451<br>-
wants a
what's the
Comparison product type comparison of same as above
difference
two or more
between stick
product types
and upright
vacuum
cleaners
The customer
wants a
product type comparison of - samsung vs
Comparison same as above
attribute two or more LG for TVs
product type
attributes
The customer
wants a
Amazon comparison of
- Prime Video
Comparison Program/Serv two or more same as above
vs Prime
ice Amazon
programs/servi
ces
- Show me
order
history<br>-
coupons <br>-
Any other where can i
Other soft-landing
request see my past
orders<br>-
call amazon
support<br>-
redeem gift
Example of
Intent Entity Description Examples Comment
expected answer
card<br>-
does Target
accept
American
Express<br>-
why are
Amazon
shares so
pricey<br>-
what time is
it<br>- how
old is barack
obama

Please note: Because of the context (the customer is on an Amazon shopping app), we should
always bias for a shopping meaning. For example, “I want to go fishing” should be interpreted as
“I want to go fishing, what products do you have that are relevant for that” (Recommendation -
product type).

(go back to Table of Contents)

Product Segment

One of the product types from this list Inp_file_05082023_MT.xlsx. Examples:


CELLULAR_PHONE_CASE, BACKPACK, PET_FOOD.

(go back to Table of Contents)

Evidence

A URL where we can find information used to build the Amazon assistant answer. Here is a
prioritized list of internal and external sources that can be used:

• Internal:
o product details pages
o customer reviews
o mentions
o customer Q&As
o seller information
o Amazon help forums
• External:
o Wirecutter
o CNET
o Trustpilot
o Wikipedia
o Online forums like Quora, Reddit, or Twitter
o Tech blogs
o REI

Common questions

Powered by AI

A 'Product Type' is a category encompassing a broad or specific range of products, such as 'furniture' or 'rechargeable double A batteries', and represents a set of items rather than a unique product that a customer can directly purchase. Conversely, an 'ASIN' is associated with a specific product with its own Product Detail Page, indicating a unique product that can be purchased. The distinction affects categorization by allowing customers to search either broadly by type or specifically by ASIN, facilitating different stages and methods of shopping .

'Product Segments' help organize Amazon's offerings by grouping items into specific categories like 'CELLULAR_PHONE_CASE' or 'PET_FOOD'. This categorization aids in streamlining inventory management, allows tailored marketing strategies, and facilitates more efficient search algorithms. For customers, it means a more structured navigation experience where they can quickly identify and access products from a given segment, thus enhancing the overall shopping experience by aligning with consumer needs and preferences .

Integrating 'Amazon Services' like Prime and returns into customer conversations offers benefits such as streamlined service access, improved customer experience, and increased user engagement by providing contextual assistance specific to service-related queries. However, challenges include ensuring seamless integration across different service types, maintaining up-to-date information, managing service-specific queries, and ensuring that the conversation remains personalized and relevant. Success in this integration demands robust backend systems that support real-time updates and context-aware interaction management, all of which promote effective service delivery within customer conversations .

The customer's shopping journey progresses from a vague need for inspiration in 'Shopping - stage 5', where a customer is simply interested in ideas (e.g., indoor gardening), through various stages that increase in specificity and informed decision-making. At 'Shopping - stage 4', the customer has a problem but is unaware of specific solutions. By 'Shopping - stage 3', they know the product type but require information on it. In 'Shopping - stage 2', they have product familiarity along with specific constraints. 'Shopping - stage 1' involves comparing a few options to select the best one, while 'Shopping - stage 0' represents the most decisive phase where the customer knows the exact ASIN to purchase. This progression illustrates increasing precision in intent and progressively narrowed information needs .

The ASIN, or Amazon Standard Identification Number, uniquely identifies a product by corresponding to a specific Product Detail Page on Amazon.com. It includes a barcode and a price, allowing a customer to directly click, add to cart, or purchase the product. This specificity simplifies the shopping experience by enabling precise targeting of the product a customer wants. Consequently, it streamlines product searches and reduces the chance of confusion between similar items .

'Query Goals' can be categorized into types such as Recommendation, Factoid, Opinion, Description, Instruction, and Comparison. Each category reflects distinct customer intentions; for instance, 'Recommendation' indicates a need for suggestions on products or attributes, 'Factoid' seeks objective information, 'Opinion' requests subjective insights, 'Description' demands elaborate overviews, 'Instruction' needs procedural guidance, and 'Comparison' desires evaluative contrasts between items. This categorization facilitates the Amazon assistant in accurately interpreting and responding to diverse customer needs, thereby supporting targeted assistance and enhancing user satisfaction .

Attributes, such as color, brand, weight, and compatibility, provide detailed information that helps customers make more informed decisions when shopping. By highlighting specific characteristics of products, attributes enable customers to filter options and focus on what matches their preferences or requirements. This richer contextual understanding aids in narrowing down choices effectively and ensures that products meet specific criteria, ultimately facilitating a more personalized and efficient shopping experience on Amazon .

In Amazon's framework, 'Evidence' refers to sources used to substantiate the information provided by the assistant, such as product detail pages, customer reviews, and external sources like Wirecutter and CNET. By relying on prioritized and diverse evidence sources, Amazon ensures that the information shared by their assistant is not only accurate but also credible and up-to-date. This approach enhances trust and reliability for customers, as they can be confident about the validity of the information they receive, thereby fostering a positive customer experience .

'Context' refers to the specific Amazon page where the customer is located when making a query, such as the home page, search results page, or product details page. This context shapes interactions by allowing situational specificity; for example, the types of results returned or details provided by the assistant can be customized depending on whether the customer is browsing on a search page or looking at a product detail. Within the assistant's framework, context informs the relevance of displayed information, thus enhancing the efficiency of user interactions by ensuring responses are pertinent to the customer's immediate environment .

The Base Layer and the Conversation Layer interact to improve user experience by synchronizing search results with customer queries, thus dynamically updating the interface based on user input. When a customer types a query, the assistant can refresh the Base Layer page while providing answers in the Conversation Layer. This layered interaction ensures real-time feedback and relevant information, enhancing the efficiency and appeal of the shopping experience by keeping both product details and conversational assistance immediately accessible .

You might also like