AWS Prescriptive Guidance - Enabling business reporting and generative AI capabilities for Amazon selling partners

Enabling business reporting and generative AI capabilities for Amazon selling

partners

AWS Prescriptive Guidance

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

AWS Prescriptive Guidance: Enabling business reporting and

generative AI capabilities for Amazon selling partners

Amazon's trademarks and trade dress may not be used in connection with any product or service

that is not Amazon's, in any manner that is likely to cause confusion among customers, or in any

manner that disparages or discredits Amazon. All other trademarks not owned by Amazon are

the property of their respective owners, who may or may not be aﬃliated with, connected to, or

sponsored by Amazon.

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

Table of Contents

Introduction ..................................................................................................................................... 1

Business overview ......................................................................................................................................... 1

Solution overview ......................................................................................................................................... 1

Data available through the SP-API ................................................................................................ 3

Reports API .................................................................................................................................................... 3

Data Kiosk API ............................................................................................................................................... 7

Seller sales and traﬃc data .................................................................................................................. 8

Vendor analytics data ............................................................................................................................ 9

Seller economics data ............................................................................................................................ 9

Integrating the SP-API .................................................................................................................. 11

Register as an SP-API developer ............................................................................................................. 11

Request SP-API roles ................................................................................................................................. 11

Register your application .......................................................................................................................... 11

Select an authorization model ................................................................................................................ 12

Understand application authorization .............................................................................................. 12

Authorize vendor groups for application access ............................................................................ 13

Connect to the SP-API .............................................................................................................................. 13

Building a data ingestion pipeline ............................................................................................... 14

Implementing analytics capabilities ............................................................................................ 16

Implementing generative AI ......................................................................................................... 18

Examples of using Amazon QuickSight Q ............................................................................................ 19

Improving and optimizing ............................................................................................................ 21

Next steps ...................................................................................................................................... 22

Resources ........................................................................................................................................ 23

Amazon Selling Partner API resources .................................................................................................. 23

Data architecture resources ..................................................................................................................... 23

AWS service demos .................................................................................................................................... 23

Document history .......................................................................................................................... 24

iii

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

Enabling business reporting and generative AI

capabilities for Amazon selling partners

Manikanta Gona Grafsgaard (Amazon Web Services) and Hina Vinayak (Amazon)

August 2024 (document history)

Business overview

Amazon is a data-driven company. It provides a wealth of data to both Amazon vendors and sellers

through various oﬀerings, such as Vendor Central, Seller Central, electronic data interchange (EDI)

solutions, and APIs.

In the Amazon Selling Partner API, analytics reports empower sellers and vendors to deeply analyze

their sales performance, inventory health, glance views, and more. Comprehensive reports cover

sales, traﬃc, net pure proﬁt margin (net PPM), forecasting, inventory, and catalog management.

Furthermore, brand analytics reports are a crucial component of this data-driven strategy, oﬀering

invaluable insights to ﬁrst-party and third-party sellers. Third-party sellers also gain powerful

insights through customer loyalty analytics and search analytics.

By providing these robust analytics and reporting capabilities, Amazon helps its selling partners

to make informed, data-driven decisions that can drive their business growth and success on

the Amazon marketplace. However, navigating and analyzing these extensive datasets can be

challenging for some vendors and sellers.

Solution overview

You can use generative artiﬁcial intelligence (generative AI) and analytics services to enhance your

business reporting for the Amazon marketplace. Amazon Q Business and Amazon QuickSight

can help you analyze data from the Selling Partner API and improve your business reporting.

By implementing data analytics and generative AI capabilities, you can unlock deeper insights,

automate repetitive tasks, and enhance your customer experience on Amazon. This ultimately

drives more sales and growth for your business.

The following is an overview of the data analytics, DevOps, and generative AI capabilities that you

can gain by implementing the recommendations in this guide:

Business overview 1

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

• Create custom reports and interactive dashboards that unlock insights from your Selling Partner

API data.

• Develop secure, scalable extract, transform, and load (ETL) pipelines that ingest, transform, and

load the data.

• Combine Amazon Q with other business intelligence (BI) solutions to generate advanced

analytics, forecast, and make data-driven decisions.

• Build custom machine learning (ML) models that analyze your Selling Partner API data.

• Use generative AI to automatically create optimized, high-quality product descriptions for your

Amazon listings.

• Use large language models (LLMs) to generate engaging, persuasive content, such as marketing

copy and customer communications.

• Use machine learning to forecast sales, inventory, and other key business metrics.

To implement these capabilities, you do the following:

1. Integrate the Selling Partner API – Set up secure connections to the Selling Partner APIs to

access your sales, inventory, customer, and other business-critical data.

2. Build data pipelines – Develop robust ETL pipelines to normalize, structure, and prepare your

Amazon data for analysis and modeling.

3. Use Amazon Q and other analytics services – Combine Amazon Q with complementary BI and

data science services to create a comprehensive analytics ecosystem.

4. Explore generative AI services – Evaluate the AWS AI services and integrate them into your

workﬂows to automate content generation, product descriptions, and predictive modeling.

5. Implement AWS best practices – Use AWS services, such as AWS Lake Formation and Amazon

DataZone, to manage and govern your data according to your compliance requirements and

AWS best practices.

Solution overview 2

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

Data available through the Amazon Selling Partner API

The Amazon Selling Partner API (SP-API) is a REST-based API that helps Amazon sellers and

vendors programmatically access their data on orders, shipments, payments, and much more.

The data from this API can increase selling eﬃciency, reduce labor requirements, and improve the

response time to customers. This guide focuses on the data that you can access through Selling

Partner API for Reports (Reports API) and the Selling Partner API for Data Kiosk (Data Kiosk API).

Reports API

The data from the Reports API is designed to help you manage your business on the Amazon

marketplace. The analytics reports provide valuable insights about your products, customers, and

overall performance, such as:

• Sales performance – Analyzing sales data can help you identify trends and customer buying

patterns. You can use this information to optimize your sales strategy.

• Product performance – Analyzing product data can help you identify which products are selling

well and which ones are not. You can use this information to optimize your product oﬀerings.

• Customer behavior – Analyzing customer data can help you understand your customers'

preferences and needs. You can use this information to personalize your marketing and sales

strategies.

The following table explains each brand analytics report. The following reports are available

to Amazon sellers and vendors who have the Brand Analytics Selling Partner API role and are

registered in Amazon Brand Registry.

Report Description Parameters

Market Basket Analysis Report Provides data on the items

that are most commonly

purchased in combinati

on with the items in the

customer's basket (cart) at

checkout

The report can be requested

for diﬀerent reporting

periods (DAY, WEEK, MONTH,

or QUARTER). The data is

provided across those periods.

Reports API 3

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

Report Description Parameters

Amazon Search Terms Report Contains data on the most

frequently selected Amazon

Standard Identiﬁcation

Numbers (ASINs), by search

keyword and department for

a marketplace

The report can be requested

for diﬀerent reporting

periods (DAY, WEEK, MONTH, or

QUARTER). Requests cannot

span multiple reporting

periods.

Repeat Purchase Report Contains data on the quantity

of repeated purchases of the

selling partner's items

The report can be requested

for diﬀerent reporting periods

(WEEK, MONTH, or QUARTER).

The data can span multiple

reporting periods.

The following table explains each vendor retail analytics and seller retail analytics report.

Report Description Availability Parameters

Rapid Retail Analytics

Inventory Report

Contains inventory

data at an ASIN level,

aggregated to an

hourly granularity

Vendors The report can

only be requested

. The data can span

multiple date range

periods, including the

current day.

Rapid Retail Analytics

Traﬃc Report

Contains data on the

customer traﬃc to

the detail pages of

the vendor's items,

with hourly granulari

Vendors The report can

only be requested

. The data can span

multiple date range

periods.

Rapid Retail Analytics

Sales Report

Contains sales data

at an ASIN level,

Vendors The report can

only be requested

. The data can span

Reports API 4

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

Report Description Availability Parameters

aggregated to an

hourly granularity

multiple date range

periods.

Vendor Sales Report Contains key retail

sales metrics (such as

ordered and shipped

revenue and units),

reported at the

catalog level and at

the ASIN level

Vendors The report can

be requested for

diﬀerent:

• Reporting periods

(DAY, WEEK, MONTH,

QUARTER, or YEAR)

• Distributor views

(MANUFACTURING

or SOURCING)

• Selling programs

(RETAIL,

BUSINESS, or

FRESH)

Net Pure Product

Margin Report

Contains data on

Amazon's net pure

product margins for

selling a vendor's

items, reported at the

catalog level and at

the ASIN level

Vendors The report can

be requested for

diﬀerent reporting

periods (DAY, WEEK,

MONTH, QUARTER, or

YEAR).

Vendor Traﬃc Report Contains key retail

traﬃc metrics, such

as glance views,

reported at the

catalog level and at

the ASIN level

Vendors The report can

be requested for

diﬀerent reporting

periods (DAY, WEEK,

MONTH, QUARTER, or

YEAR).

Reports API 5

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

Report Description Availability Parameters

Vendor Forecasting

Report

Contains forward-l

ooking mean, prior

70, prior 80, and prior

90 weekly customer

demand forecasts

, reported at the

ASIN level for the

most recent weekly

forecast generation

date

Vendors The report can

be requested for

diﬀerent selling

programs (RETAIL or

FRESH).

Vendor Inventory

Report

Contains inventory

and operational

health metrics (such

as sellable on-hand

units and vendor

conﬁrmation rate),

reported at the

catalog level and at

the ASIN level

Vendors The report can

be requested for

diﬀerent:

• Reporting periods

(DAY, WEEK, MONTH,

QUARTER, or YEAR)

• Distributor views

(MANUFACTURING

or SOURCING)

• Selling programs

(RETAIL or FRESH)

Reports API 6

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

Report Description Availability Parameters

Sales and Traﬃc

Business Report

Contains key sales

performance metrics

(such as ordered

product sales,

revenue, units

ordered, and claim

amount) and page

traﬃc metrics (such

as page views and

buy box percentage)

for the seller's entire

catalog, aggregated

by date and ASIN

Sellers The report can

be requested or

scheduled. The data

can be provided

for diﬀerent date

range aggregation

levels (DAY, WEEK,

or MONTH) and ASIN

aggregation levels

(PARENT, CHILD, or

SKU).

Data Kiosk API

The Data Kiosk API provides a GraphQL-based dynamic reporting suite. It helps you generate

custom GraphQL queries and access bulk data from Amazon's datasets. You can use this API to

access a wide range of data, including:

• Seller sales and traﬃc data

• Vendor analytics data

• Seller economics data

The following are the key capabilities of the Data Kiosk API:

• GraphQL-powered dynamic reporting – The API supports GraphQL query operations, helping

you to create custom queries to retrieve the speciﬁc data that you need.

• Bulk data access – You can access large datasets through the API, facilitating comprehensive

analysis and reporting.

• Flexible data exploration – The GraphQL-based approach allows for dynamic, ad-hoc data

exploration. This ﬂexibility helps you iterate on queries and extract insights.

Data Kiosk API 7

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

• Comprehensive data coverage – The API provides access to a wide range of seller and vendor

data, enabling holistic performance analysis and business intelligence.

The schemas and directives in the Data Kiosk API provide a structured way to access your data

through a GraphQL API, with the ability to aggregate and ﬁlter the data based on your speciﬁc

needs. By using the Data Kiosk API, you can gain deeper, more actionable insights into sales, traﬃc,

and analytics data. This empowers you to make informed, data-driven decisions to optimize your

business strategies.

Note

The retention of a query varies based on the ﬁelds requested. Each ﬁeld within a schema is

annotated with a @resultRetention directive that deﬁnes how long a query containing

that ﬁeld will be retained. When a query contains multiple ﬁelds with diﬀerent retentions,

the shortest (minimum) retention is applied. The retention of a query's resulting documents

always matches the retention of the query.

Seller sales and traﬃc data

In the Data Kiosk API, Analytics_SalesAndTraffic_2024_04_24 is the root type for sales and

traﬃc queries. Two queries are available for this root type:

•

salesAndTrafficByAsin retrieves sales and traﬃc data that is aggregated by ASIN (PARENT,

CHILD, or SKU).

•

salesAndTrafficByDate retrieves sales and traﬃc data that is aggregated by date (DAY,

WEEK, or MONTH).

Note the following about these queries:

•

Enumerated types (enums), such as AsinGranularity and DateGranularity, specify the

level of aggregation.

• Scalar types, such as Date, are used to represent the data.

Seller sales and traﬃc data 8

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

Vendor analytics data

In the Data Kiosk API, Analytics_vendorAnalytics_2024_09_30 is the root type for vendor

analytics queries. Two views are available for this root type:

•

manufacturingView retrieves metrics about products that you manufacture, aggregated by

the speciﬁed date.

•

sourcingView retrieves metrics about products that you source directly to Amazon, aggregated

by the speciﬁed date.

You access the relevant view and its associated metrics based on the role assigned to you as

a vendor (SOURCING or MANUFACTURING). These queries include the start and end dates, the

marketplace identiﬁer, the aggregated metrics, and the total metrics. Note the following about

these queries:

• The metrics are grouped by various attributes (such as ASIN or brand) by using the

ManufacturingViewMetricsGroupedBy and SourcingViewMetricsGroupedBy types.

•

Metric types, such as costs, customerSatisfaction, orders, and productAvailability,

represent the diﬀerent metric groups.

•

The DateGranularity enum speciﬁes the level of date aggregation.

Seller economics data

The seller economics report provides a comprehensive view of your performance and ﬁnancials on

the Amazon marketplace. It covers sales data, advertising spend, fees, costs, and net proceeds. You

can aggregate the data by date and product identiﬁer.

In the Data Kiosk API, Analytics_Economics_2024_03_15 is the root type for seller economics

queries. The seller economics data for a given product is represented by the Economics type. This

type includes the following key aspects:

•

ads – The advertising spend data for the product within the speciﬁed date range.

•

cost – The oﬀ Amazon cost provided by the seller for this product (per unit).

•

fees – The fees applied to the product within the speciﬁed date range.

•

fnsku – The Fulﬁllment Network Stock Keeping Unit (FNSKU) of the product.

Vendor analytics data 9

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

•

msku – The Merchant Stock Keeping Unit (MSKU) of the product.

•

netProceeds – The net proceeds for this product within the speciﬁed date range.

•

sales – The sales data for the product within the speciﬁed date range.

Seller economics data 10

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

Integrating the Amazon Selling Partner API

In order to access the data through the Amazon Selling Partner API (SP-API), you must complete

the following actions:

1. Register as an SP-API developer

2. Request SP-API roles

3. Register your application

4. Select an authorization model for your application

5. Connect to the SP-API

Before you can register your SP-API application, you must create an Amazon developer account

and register as an SP-API developer. For a comprehensive overview of the developer registration

process, see SP-API Registration Overview in the SP-API documentation.

Request SP-API roles

An SP-API role determines whether a developer or application has access to a speciﬁc operation or

resource. As a developer, you must request and qualify for a particular role, or you will not be able

to access the operations and resources grouped under that role.

Roles protect access to personally identiﬁable information (PII) and other sensitive data. They

limit data access to make sure that developers can access only the data that is required for their

application. This helps protect customer data and preserve customer trust.

Accessing the data available in the brand analytics reports requires that you have the Brand

Analytics role. For more information about requesting access to a role, see How do I request and

qualify for a role in the SP-API documentation.

The registration process varies slightly depending on the application type. For the purposes of

registration, applications are categorized as one of the following types:

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

• Public applications and private seller applications – These are applications that are publicly

available and are authorized by a seller or vendor, or they are seller applications that are

available only to your organization and are self-authorized.

• Private vendor applications – These are vendor applications that are available only to your

organization and are self-authorized.

For more information, see Register your application in the SP-API documentation.

Select an authorization model for your application

The authorization model for the Selling Partner API is based on Login with Amazon, an

Amazon implementation of OAuth 2.0. Your application is authorized through interactions with

pages displayed by Amazon and your website. The web browser is the user-agent that passes

parameters between your website and Amazon at each selling partner action. To implement

OAuth authorization, you must conﬁgure your website to accept and process the parameters that

Amazon passes to it. You must also conﬁgure your website to redirect the web browser and pass

parameters to Amazon. For more information about authorization, see Authorizing Selling Partner

API applications in the SP-API documentation.

Understand application authorization

For the purposes of authorization, there are three types of applications:

• Public applications for sellers – These applications are publicly available and are authorized by

sellers. You can choose one of the following authorization workﬂows:

• Selling Partner Appstore authorization workﬂow – An OAuth authorization workﬂow that is

initiated from the Selling Partner Appstore detail page.

• Website authorization workﬂow – An OAuth authorization workﬂow that is initiated from your

own website.

• Public applications for vendors – These applications are publicly available and are authorized

by vendors. You can use the Website authorization workﬂow. This is an OAuth authorization

workﬂow that is initiated from your own website.

• Private applications for sellers or vendors – These applications are available only to your

organization. These can be seller or vendor applications. You can use the Self authorization

approach. When you create a private application for your own organization you can self-

authorize it to access your account information. You can self-authorize your application in draft

Select an authorization model 12

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

status; there is no reason to publish a private application. For information about revoking self-

authorization from seller and vendor applications, see Revoke self-authorizations in the SP-API

documentation.

Authorize vendor groups for application access

When you authorize your Selling Partner API application to access your data, you are granting

access to the vendor group that is associated with the sign-in credentials for your Vendor Central

account. By extension, you are granting access to all vendor codes that are present in the vendor

group. Therefore, it's important to use the right Vendor Central credentials and vendor group for

your Selling Partner API integration.

The vendor group is the account you log in to. Depending on your business agreements, operation

models, and other factors, your vendor group can include one or more vendor codes. Each vendor

code allows you to list products in a speciﬁc category, or it includes the necessary business

agreements, such as one vendor code for a particular brand.

You can have multiple authorizations for each vendor group, or you can create a single vendor

group that contains all of your vendor codes. The option to use multiple vendor groups that are

associated with your proﬁle gives you the ability to use an application with the same vendor code

in diﬀerent vendor groups. With this option, you don't have to submit multiple vendor developer

applications for each vendor group.

For more information, see Authorize vendor groups for application access in the SP-API

documentation.

Connect to the SP-API

After you have registered and authorized your application, you can start making requests. For more

information, see Connecting to the Selling Partner API in the SP-API documentation.

Authorize vendor groups for application access 13

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

Building the data ingestion pipeline for your Amazon

selling partner data

This section provides a strategy to ingests Amazon vendor and seller data from the Amazon Selling

Partner API (SP-API) to a data lake in your AWS account. This data pipeline architecture is designed

for agility. After the data is available in your account, you can implement analytics and generative

AI capabilities to obtain advanced business insights from this data. This data helps you understand

your business, inventory details, and analytics at scale across all marketplaces.

The following architecture diagram shows how you use AWS Lambda functions in an AWS Step

Functions workﬂow in order to ingest data from the SP-API into a data lake in your AWS account.

The data is stored in Amazon Simple Storage Service (Amazon S3) and in Parameter Store, which is

a capability of AWS Systems Manager.

The architecture diagram includes the following components:

1. Step Functions is used as a serverless orchestration service to centrally manage the workﬂow for

integrating with the SP-API.

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

2. The Selling Partner API for Reports (Reports API) supports notiﬁcations to automate the

report workﬂows. For this, you use an SP-API notiﬁcation Lambda function to subscribe the

application to the REPORT_PROCESSING_FINISHED notiﬁcation type.

3. In order to make calls to the SP-API, you use an Authentication Lambda function to obtain a

4. The LWA access token from the authentication function is passed to a Report creator Lambda

function. This function makes a createReport call to the SP-API by using the LWA access token

and the regional endpoints, marketplace IDs, and report conﬁgurations data that is stored in

Parameter Store.

The SP-API generates the report. Upon completion, a REPORT_PROCESSING_FINISHED

notiﬁcation event is sent to an Amazon Simple Queue Service (Amazon SQS) queue, which

provides information when report processing is CANCELLED, DONE, or FATAL. This triggers a

Notiﬁcation processing Lambda function to process the event. If the notiﬁcation event has a

status of DONE, a reportDocumentId is included.

6. The notiﬁcation event is passed to a Data processing Lambda function in the Step Functions

workﬂow. This function uses the reportDocumentId to make a getReportDocument call to

the SP-API. The SP-API returns a pre-signed URL for the location of the report document and

the compression algorithm used, if the report document contents have been compressed.

7. This response is passed to a Storage Lambda function, which downloads the report document,

decompresses it (if applicable), and stores the report document in Amazon S3.

8. AWS Key Management Service (AWS KMS) is used to centrally manage encryption keys, which

can be used to encrypt the secrets in AWS Secrets Manager. Data is stored in Amazon S3 and

Parameter Store.

9. SP-API requests are limited by using the token bucket algorithm. Therefore, an API client is

recommended for rate limiting.

10.AWS CloudTrail and Amazon CloudWatch are used for monitoring and logging across the AWS

services. These logs provide traceability.

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

Implementing an analytics strategy for your Amazon

selling partner data

This section provides a detailed strategy for how Amazon vendors and sellers can perform

advanced analytics on the data ingested from the Amazon Selling Partner API (SP-API). These

analytics capabilities can provide:

• Insights into sales performance, inventory management, brand analytics, and other key metrics.

• The ability to create custom calculations, ﬁlters, and visualizations to address your speciﬁc needs.

The following architecture diagram shows how you use AWS Glue to discover, prepare, move, and

integrate the data in the data lake so that you can use it for analytics and insights.

The architecture diagram includes the following components:

1. AWS Lake Formation is used to build the scalable data lake and to centrally manage the security,

access control, and audit trails.

2. Amazon Simple Storage Service (Amazon S3) is used as the data lake storage.

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

3. AWS Glue is used to catalog, transform, enrich, move, and replicate data across multiple data

stores and the data lake. AWS Glue simpliﬁes complex, manual, and expensive traditional data

integration processes, and it supports increased data volumes and data diversity.

4. Amazon DataZone helps you catalog, discover, share, and govern data across the organization.

5. Amazon Athena provides interactive querying, analyzing, and processing capabilities.

6. Amazon Redshift is used as a cloud data warehouse. With zero-ETL integration, you can perform

near real-time analytics on petabytes of transactional data, or you can use Amazon Redshift ML

capabilities to derive real-time insights.

7. Amazon QuickSight provides ML-powered business intelligence. Amazon QuickSight Q, powered

by machine learning, uses natural language processing to answer your business questions

quickly.

8. Amazon EMR is a managed cluster platform that simpliﬁes running big data frameworks to

process and analyze vast amounts of data on AWS. Using these frameworks and related open-

source projects, you can process data for analytics purposes and business intelligence workloads.

9. Amazon OpenSearch Service can be used for operational analytics. It also provides vector

database search capabilities.

10.Amazon SageMaker can be used to build, train, and deploy ML models, and to add artiﬁcial

intelligence to your applications.

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

Implementing a generative AI strategy for your Amazon

selling partner data

Amazon vendors and sellers can use generative artiﬁcial intelligence (generative AI) services and

features on AWS to gain deeper insights, automate repetitive tasks, and enhance your customer

experience on Amazon. For example, you can use generative AI to automatically create optimized,

high-quality product descriptions for your Amazon listings, generate marketing copy or customer

communications, or forecast sales, inventory, and other key business metrics. Generative AI can

provide insights into the data that you've ingested from the Amazon Selling Partner API.

The following architecture diagram shows how you can use AWS services, such as Amazon Bedrock

or Amazon Q Business, to kick-start your generative AI journey on AWS. Using this architecture, you

build a generative AI pipeline that uses a modern data analytics approach to derive insights from

the data.

The architecture diagram includes the following components:

1. AWS Lake Formation is used to build the scalable data lake and to centrally manage the security,

access control, and audit trails.

2. Amazon Simple Storage Service (Amazon S3) is used as the data lake storage.

3. Amazon Bedrock is a fully managed service that oﬀers a choice of industry-leading foundation

models (FMs). It also provides a broad set of capabilities that you need to build generative AI

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

applications, simplifying development with security, privacy, and responsible AI. You can store

the model output data in an Amazon S3 bucket and can integrate it to Amazon QuickSight.

4. Amazon QuickSight provides ML-powered business intelligence. Amazon QuickSight Q enhances

business productivity with generative BI capabilities to accelerate decision making. With new

dashboard-authoring capabilities, you can use natural language prompts to quickly build,

discover, and share meaningful insights.

5. Amazon Q Business is a generative AI–powered assistant that can answer questions, provide

summaries, generate content, and securely complete tasks based on data and information in

your enterprise systems. It helps you be more creative, data-driven, eﬃcient, prepared, and

productive.

6. You can create a custom web application that consumes the data from Amazon Q Business.

Examples of using Amazon QuickSight Q

The following examples show how Amazon QuickSight Q helps you understand data with executive

summaries, a context-aware data Q&A experience, and customizable, interactive data stories that

help drive decisions from insights.

Example 1

You can ask Amazon QuickSight Q "What are the top sold items in 2023?” Amazon QuickSight Q

gathers and analyzes the data in order to provide an executive summary.

Example 2

You can ask Amazon QuickSight Q what are the top products that an enterprise sells on Amazon.

Examples of using Amazon QuickSight Q 19

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

Example 3

You can ask Amazon QuickSight Q to provide sales metrics for a product, based on month or year-

to-date.

Examples of using Amazon QuickSight Q 20

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

Continuously improving and optimizing your analytics

and generative AI strategies

To continuously improve and optimize your analytics and generative AI strategies for your Amazon

seller and vendor data, we recommend that you do the following:

• Continuously monitor performance – Regularly review the usage and performance metrics

of the analytics and generative AI solutions to identify areas for improvement. Monitor key

indicators, such as user engagement, data processing throughput, API response times, and error

rates. Use this data to optimize system conﬁgurations, identify bottlenecks, and make informed

decisions about enhancements.

• Use advanced ML techniques – Explore the application of more sophisticated ML models and

techniques to further enhance the predictive and analytical capabilities of your application. This

might include experimenting with neural networks, time series forecasting, anomaly detection,

and other advanced algorithms.

• Prioritize user feedback – Actively gather feedback from the Amazon seller, vendor, and brand

community to understand evolving needs and pain points. Incorporate this user input into

your application's development roadmap to make sure that the solution remains relevant and

continues to deliver maximum value.

By continuously monitoring performance, using advanced analytics and ML techniques, and

prioritizing user feedback, you can make sure that the application remains a robust, adaptable,

and indispensable tool. This commitment to ongoing optimization and evolution helps you make

increasingly data-driven decisions, drive measurable business impact, and stay ahead of the

competition.

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

Next steps

This guide is intended for Amazon sellers and vendors who want to use analytics and generative

AI capabilities to drive growth for their business. Using the recommendations and strategies in this

guide, you can:

• Build the cloud application in house

• Engage with AWS Professional Services

• Select an AWS Partner from the AWS Partner Network

• Contact your AWS account team

For more information and additional resources that can help you learn more about AWS analytics

and generative AI capabilities, see the Resources chapter of this guide.

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

Resources

Amazon Selling Partner API resources

• Selling Partner API documentation

• Selling Partner API data models

• Data Kiosk Schema Explorer

Data architecture resources

• Ingest Amazon Retail Data into a Serverless Modern Data Architecture (AWS blog post)

• Amazon Seller and Vendor Central Data Producer (AWS Solutions Library)

AWS service demos

• Amazon QuickSight Q demo (Demo Central)

Amazon Selling Partner API resources 23

AWS Prescriptive Guidance Enabling business reporting and generative AI capabilities for

Amazon selling partners

Document history

The following table describes signiﬁcant changes to this guide. If you want to be notiﬁed about

future updates, you can subscribe to an RSS feed.

Change Description Date

Initial publication — August 9, 2024