The American Journal of Management and Economics Innovations
82
https://www.theamericanjournals.com/index.php/tajmei
TYPE
Original Research
PAGE NO.
82-89
10.37547/tajmei/Volume07Issue06-08
OPEN ACCESS
SUBMITED
22 Arpil 2025
ACCEPTED
24May 2025
PUBLISHED
20 June 2025
VOLUME
Vol.07 Issue 06 2025
CITATION
Dmytro Balan. (2025). A Review of Trends in the Automation of
Advertising Processes on Marketplaces with an Emphasis on ThinkAd’s
AI Solutions. The American Journal of Management and Economics
Innovations,
7(06),
82
–
89.
https://doi.org/10.37547/tajmei/Volume07Issue06-08
COPYRIGHT
© 2025 Original content from this work may be used under the terms
of the creative commons attributes 4.0 License.
A Review of Trends in the
Automation of Advertising
Processes on Marketplaces
with an Emphasis on
ThinkAd’s AI Solutions
Dmytro Balan
CEO of THINKAD INC Boca Raton, Florida, USA
Abstract:
This paper covers the history and current
trends in the automation of advertising workflow
within marketplaces through AI solutions provided by
the ThinkAd platform. The study shall attempt to detail
drivers that have prompted a transition from manual
bidding to intelligent autopilots. It also tries to present
comparative analyses regarding the functionalities of
leading AI services and empirical assessments of their
effectiveness via real-world case studies. Such a work
is meaningful since there has been a strong upward
trend in cost-per-click on Amazon Ads, coupled with
increasing difficulties that come with manual
management regarding hundreds of thousands of
product × keyword × time combinations, which cause
both budget overruns and lost sales. Simultaneously,
first-party data gains value in a cookieless world;
intense competition demands instant strategy changes
via hyper-personalization and goal-based bidding. The
novelty of the study lies in the integration of
descriptive statistics on CPC dynamics and
sellers’ time
savings, content- and case-analysis of AI-platform
technical documentation, and a functional comparison
of semantic-core generation modules, goal-based
bidding, hourly day-parting, and multi-account
mastering. Particular attention is paid to the ThinkAd
platform, which forecasts ACoS 24 hours in advance,
updates bids hourly, and consolidates data across
multiple stores. The main findings indicate that
intelligent automation can reduce ACoS to 22
–
25%
while increasing advertising sales by 82
–
206%, freeing
up to 20 hours of operational time per week, and
The American Journal of Management and Economics Innovations
83
https://www.theamericanjournals.com/index.php/tajmei
ensuring competitiveness for small and medium
enterprises under cookieless conditions and rising click
costs. Integrated via the Amazon Ads API and supporting
multiple regions, ThinkAd sets a new efficiency standard
by combining a semantic module, predictive analytics,
and an autopilot. This article will be helpful to
marketplace
advertising
managers,
e-commerce
analysts, and small and medium business owners when
selecting and implementing AI tools for advertising
campaign automation.
Keywords
: AI automation, marketplaces, ThinkAd, goal-
based bidding, day-parting, first-party data, ACoS, CPC.
INTRODUCTION
Over the past eight years, advertising on marketplaces
has evolved from manual techniques into a
sophisticated system in which each keyword
participates in instantaneous auctions and platform
algorithms apply real-time bid adjustments. A steady
rise in cost-per-click accompanies competition for
customer attention on Amazon. For small and medium
enterprises with limited advertising budgets, the
previous approach of setting a bid and waiting for a
weekly report has ceased to be cost-effective: any
delayed action results in overspending and lost sales.
Simultaneously,
the
informational
burden
has
increased. Today, merchandisers and marketers must
manage bids and maintain synchronization of prices,
inventories, and content across dozens of platforms,
monitor new ad formats, and adjust negative keywords.
The number of variables requiring manual evaluation
has long exceeded human capacity: even a small brand
with a modest portfolio and three to five listing channels
faces over a million daily product × keyword × hour
combinations. Attempting to address this volume
manually is slow and economically inefficient, as
specialist labor costs grow faster than their ability to
reduce ACoS.
These factors are driving sellers toward intelligent
automation. According to [2], by January 2024, 34% of
Amazon sellers will have already used AI tools for listing
creation and optimization, and another 14% for
marketing and content automation. The appeal lies in
improved targeting accuracy and time savings: research
data [3] shows that sellers adopting AI platforms free up
an average of 15
–
20 hours per week by automating
routine tasks. These hours become a resource for
strategic decisions
—
testing new creatives, expanding
assortments, and negotiating with suppliers
—
instead of
endless manual bid adjustments.
As a result, advertising on marketplaces is no longer
manual labor, not because specialists have become idle,
but because human time has become too expensive for
tasks that machines perform faster, more accurately,
and cheaply. The emergence of specialized AI platforms
such as ThinkAd marks a point of no return: algorithms
assume operational routines, allowing companies of all
sizes to compete with large brands based on reaction
speed and data rather than budget size.
MATERIALS AND METHODOLOGY
The research materials include an extensive review of
publicly available sources and cases demonstrating the
evolution of advertising processes on marketplaces and
the application of AI solutions. Empirical data were
drawn from statistical reports on the dynamics of
average click prices in Amazon Ads [1], figures on the
share of sellers who have already implemented AI tools
for listing and marketing optimization [2, 3], Google Ads
Help materials on the upcoming deprecation of third-
party cookies [4], and studies on the role of first-party
data in a cookieless environment [5]. Additionally,
technical changes in Amazon Dynamic Segments [6],
McKinsey’s findings on the effectiveness of hyper
-
personalization [7], and AWS reviews on the use of
generative AI in retail [8] were analyzed.
To evaluate practical AI solutions, official guides and
case studies were examined: goal-based bidding in
Amazon DSP [9], outcomes of automatic optimization by
Perpetua [10] and Quartile [11], as well as independent
seller surveys on the growth of multi-marketplace trade
[12, 13] and investments in proprietary data [14, 15].
The industrial scale of ThinkAd is documented in the
platform’s public statistics and the Handcraft Blends and
New York Biology case studies [18
–
20]. At the same
time, industry recognition is evidenced by the ECDMA
Global Awards 2025 [21].
Methodologically, a systematic review of secondary
sources was conducted, complemented by content
analysis of technical documentation and case-study
analysis of real AI autopilot implementations.
Descriptive statistics were used to visualize monthly CPC
trends [1] and to assess seller time savings afforded by
AI platforms [3]. The functional comparison of ThinkAd,
The American Journal of Management and Economics Innovations
84
https://www.theamericanjournals.com/index.php/tajmei
Perpetua, and Quartile was based on matching key
features: semantic-core generation, goal-based bidding,
hourly day-parting, and multi-account management [9,
10, 11, 18]. The Handcraft Blends and New York Biology
case studies illustrated the impact of automation on
conversion growth and ACoS reduction [19, 20].
RESULTS AND DISCUSSION
By 2025, competitive pressure on the most significant
marketplaces had reached a level at which even small
bid fluctuations are instantly reflected in profitability.
The average CPC in Amazon Ads, having fallen to $0.89
in March 2024, rose to $1.14 by June and has since
remained around $1.00
—
i.e., 10% higher than the
previous year’s value, as shown in Figure 1 [1]. At such
auction density, any delay in bid adjustment entails a
surge in advertising costs and loss of placement.
Fig. 1. Average CPC on Amazon Advertising by Month [1]
Concurrently, advertisers find themselves in a new
regulatory reality. The phased deprecation of third-party
cookies in Chrome, scheduled to begin in early 2025,
regardless of any subsequent Google adjustments, has
forced platforms and brands to shift their focus to
proprietary customer data [4]. Study [5] records that
fully permitted first-party data becomes a critically
important asset in 2025 due to simultaneous regulatory
pressure and the proliferation of AI personalization
algorithms. For sellers, this means consolidating order,
loyalty, and interaction data and feeding it into
advertising engines; otherwise, precise targeting in a
cookieless environment will be impossible. Thus, rising
bids, a shortage of human resources, and privacy
constraints converge. Without intelligent automation,
campaigns cease to scale, and competitiveness shifts to
those who first learned to convert their data into fuel for
AI systems.
Hyper-personalization has become the logical response
to scarce ad slots and rising click costs: the more
expensive the impression, the more crucial it is that ads
reach users with the highest conversion probability. The
shift from segmentation by gender or age to behavioral
clusters accelerated after the launch of Amazon
Dynamic Segments in November 2024, when the
platform first allowed algorithms to rebuild audiences
on the fly based on the latest search and purchase
signals [6]. The practical effect of this approach is
confirmed by McKinsey [7]: precise personalization
consistently adds 10
–
15% to revenue, and companies
that can apply data swiftly see gains up to 25%, as shown
in Figure 2.
The American Journal of Management and Economics Innovations
85
https://www.theamericanjournals.com/index.php/tajmei
Fig. 2. Digitally native companies drive more revenue from personalization than other company archetypes [7]
For the seller, this means not just selecting relevant
keywords,
but
dynamically
restructuring
the
storefront
—
displaying different sets of products, pricing
offers, and images depending on query context and
funnel stage. Generative tools, such as Amazon
Personalize in combination with Amazon Q, further
remove speed barriers: a banner or product card is
regenerated in less than a minute after a user signal
change, turning a seller’s first
-party data into a
continuous stream of hypotheses and A/B tests [8].
The second key evolutionary line is predictive analytics,
which combines algorithmic demand forecasting with
goal-based bidding. By launching goal-based bidding in
DSP in 2024, Amazon enabled advertisers to set a
business metric
—
reach, CPA, or ROAS
—
instead of a
manual bid; the system then manages click price and
budget allocation in real time [9]. The algorithm
calculates a conversion probability for each user ×
creative × time slot combination and bids as high as
needed to achieve the goal. In an environment where
even slight overshooting of target ACoS can erase
margins, this shift from micromanaging bids to
managing outcomes becomes a critical advantage.
ThinkAd employs this logic: the platform forecasts ACoS
for each keyword 24 hours ahead, updates bids hourly,
and enables small businesses to maintain target
profitability without constant human intervention.
The third notable vector of evolution is the emergence
of a complete AI autopilot, in which campaign
generation, keyword selection, bid management, and
budget reallocation occur without direct human
involvement. Technologically, this became possible once
marketplace ad interfaces opened access to streaming
auction events: Amazon DSP introduced goal-based
bidding, allowing the algorithm to pursue a set ROAS or
CPA and adjust click prices in real time based on
conversion-probability models built on billions of user
signals [9]. Independent SaaS platforms rapidly adopted
the same principle. In 2024, Perpetua demonstrated a
15% ACoS reduction through hands-free campaigns,
where the system creates ad groups and updates
negative keywords after each search query report [10].
Quartile moved bid adjustments to an hourly interval:
the neural network analyzes demand peaks and raises
bids only where the sale probability exceeds a historical
threshold, saving up to 18% of budget without losing
impressions [11]. According to ThinkAd, this logic,
reinforced by a 24-hour ACoS forecast, allowed small
sellers to maintain target profitability amid a 200%
increase in sales volume and freed 10
–
20 hours of
operational time per week previously spent on manual
optimization.
The next logical layer atop the autopilot is day-parting,
or micro-regulation of ad serving times. Rising click costs
have made round-the-clock advertising too expensive,
and order statistics show that for most categories,
conversions concentrate in narrow time windows.
Agency case studies confirm practical effectiveness:
AiHello recorded an ACoS drop from 30 % to 22.66 %
over four months after implementing algorithmic day-
parting, which lowers bids on weekends and raises them
during evening purchase peaks. This is especially critical
for small businesses: dynamic hourly budget allocation
The American Journal of Management and Economics Innovations
86
https://www.theamericanjournals.com/index.php/tajmei
turns limited funds into a competitive advantage,
winning auctions precisely when competitors have
exhausted their daily limits. AI autopilot and micro-time
control set a new norm: the seller manages a business
metric rather than bids and schedules, and the system
makes all intermediate decisions based on first-party
data and current demand dynamics.
The more sellers spread across different
platforms, the costlier data fragmentation becomes: in
small businesses where one manager handles multiple
Amazon, Ozon, and Wildberries logins, each price or
inventory change must be manually duplicated. Savvy
players respond by consolidating interfaces: a survey
[12] showed that by 2025, only 11% of small companies
still work in a single channel, whereas 81% already
manage at least two and aim for a unified dashboard.
Marketplaces themselves confirm the effect of multi-
format selling: Mirakl reports a 104% GMV increase for
brands selling on three or more marketplaces, turning
cross-account management from a convenience into a
direct revenue driver [13]. ThinkAd’s modules align with
this trend by aggregating statistics from different
Amazon regions into a single model, enabling algorithms
to perceive inter-product and inter-market relationships
lost in isolated analyses.
Simultaneously, the transition to a cookieless ecosystem
accelerates. After Google finally postponed their
removal to 2025, the market stopped waiting for day X
and switched to proprietary transaction databases: 82%
of marketers already report increased investment in
first-party data, not because of regulators, but because
predictive models lose accuracy without it [14]. In its
report [15], IAB directly links this migration to signal
erosion and identifies data purity and connectivity as the
main barriers to large-scale AI adoption. From two-
thirds to nearly 90% of agencies, brands, and publishers
use accessible AI tools, yet these lack the functionality
required for full-scale implementation, as shown in
Figure 3.
Fig. 3. Types of AI Tools and Platforms Being Used [15]
Finally, the next paradigm shift is already visible on the
horizon: commerce designed for humans is gradually
being complemented by commerce for machines.
Amazon has unveiled Rufus
—
a generative assistant that
responds to customer queries directly in the search bar
and autonomously suggests product assortments [16];
Walmart has followed the same path, testing an AI agent
capable of compiling shopping lists and placing orders
without user involvement. When recommendations and
purchases are executed by algorithms, priority shifts
from the attractive banner to structured, machine-
readable information about price, availability, and
rating, ThinkAd already optimizes feeds for such
scenarios: the system exports product attributes in a
The American Journal of Management and Economics Innovations
87
https://www.theamericanjournals.com/index.php/tajmei
format equally understood by both the marketplace and
third-party shopping agents, thereby preparing sellers
for an era in which their primary client will be not a
human, but an AI intermediary.
Since 2017, the author of this article has worked in e-
commerce and has followed the complete path of an
Amazon seller: from niche selection and supplier
negotiations to elevating brands to top category
positions and maintaining those results amid rising
competition. Concurrently, the author has consulted
dozens of entrepreneurs, helping them launch and scale
their brands by developing placement strategies,
conducting deep competitive analyses, and designing
advertising campaign architectures to reduce ACoS and
increase ROI sustainably. He also serves as a marketing
strategist at Lev Brands
—
a large company with annual
revenues
exceeding
$50
million
—
where
his
methodologies and technologies have become the
foundation for business scaling and enhanced
advertising profitability.
In recent years, the author has focused on creating an AI
platform that automates key Amazon advertising
processes. Thus, ThinkAd was born: a specialized
Amazon advertising automation platform developed
and founded by the author, created by sellers
themselves, and initially oriented toward practical
campaign management tasks. Already at the public
release stage, the service processes 5,597 products,
retains over 31,000 active campaigns in its database,
optimizes more than $3.6 million of historical
advertising budget, and manages 1,162,468 keywords
for 326 brands, demonstrating the industrial scale of the
solution and its readiness to support small and medium
enterprises [18]. The conceptual foundation of ThinkAd
is a platform written by sellers for sellers: its creators
emphasize that the product emerged from their own
seven-figure Amazon sales and therefore addresses real,
rather than hypothetical, PPC specialist pain points [18].
The core functional block is the automatic semantic
module: the system generates a Semantic Core from
niche and competitive data, enabling users to obtain a
professional keyword core in just a few clicks, which
feeds the predictive bidding model. At the optimization
level, ThinkAd combines hourly conversion and price
collection with an Auto Mode: the algorithm pursues a
specified ACoS/ROAS, adjusts bids hourly, and filters out
ineffective keywords, while the built-in Ignore List,
together with the Wasted Spend Management module,
automatically excludes sources of empty spend. The
extended feature set includes Advanced Dayparting
—
scheduling by hour and weekday based on proprietary
statistics
—
Real-Time Keyword Harvesting, which
translates high-conversion search phrases into exact
targets,
and
Multi-Account
Mastery,
allowing
management of multiple stores and regions through a
single dashboard, thus addressing multi-marketplace
expansion without manual duplication of settings.
The workflow follows the principle connect
—
set a
goal
—
enable autopilot: the seller authorizes Seller
Central, initiates Semantic Core generation, specifies
target ACoS and bid limits, after which the AI fully
manages campaigns, including negative keywords and
budget reallocation; in parallel, a Keyword Tracker is
available that displays real-time organic and paid
rankings for selected queries. The service is built on full
integration with the latest Amazon Ads API, so format or
attribute updates appear without delay, and the
platform scales to high traffic volumes and budgets for
multiple brands simultaneously. ThinkAd supports all
American and major European Amazon marketplaces
(USA, Canada, Mexico, UK, Germany, France, Italy,
Spain) as well as Japan and Australia; at early access, it
supports Sponsored Products, with Sponsored
Brands/Video and Sponsored Display modules
announced as coming soon. The commercial model is
transparent: a fixed fee of $49.99 and 1% of advertising
spend for the first three hundred Early Access
participants, after which the price remains unchanged
forever [22].
The effectiveness of this approach is illustrated by
published case studies: the cosmetics brand Handcraft
Blends increased advertising sales by 82%, raised clicks
by 125%, and reduced ACoS to 22% after implementing
ThinkAd’s AI bidder [19]; another client
, New York
Biology, achieved a 206% increase in ad-sales within just
a few months of using the platform [20]. Thus, ThinkAd
positions itself as an end-to-end tool combining
semantic generation, hourly bid optimization, intelligent
scheduling, and a multi-account console in a single
interface, enabling small and medium sellers to compete
for traffic with larger players without expanding
headcount or maintaining their own BI systems.
In 2025, ThinkAd was awarded Gold in the Best SaaS E-
Commerce Platform category at the international
The American Journal of Management and Economics Innovations
88
https://www.theamericanjournals.com/index.php/tajmei
ECDMA Global Awards, where the jury noted its
contribution to reducing advertising costs and increasing
ROAS for small sellers, thereby confirming the industry
significance of the solution [21]. Additional recognition
comes from ThinkAd’s inclusion in the
Amazon Ads
Partner Network, the official registry of accredited tools
and agencies, demonstrating the platform’s compliance
with Amazon’s stringent requirements for optimization
quality, reporting transparency, and managed budget
volume.
It combines an autopilot based on first-party data,
precise hourly dayparting, consolidated multi-account
management, and feed preparation for machine agents.
ThinkAd sets a new benchmark for efficiency in
automating advertising processes on marketplaces. The
platform not only relieves sellers of the routine of bid
and schedule adjustments but also elevates campaign
management to the level of business metrics, from
which the system autonomously builds an optimal real-
time strategy. Thanks to deep integration with the
Amazon Ads API and simultaneous operation across
multiple marketplaces, ThinkAd converts fragmented
data into a cohesive model capable of predicting
demand peaks and adapting to the cookieless era. The
combination of intelligent semantic-core algorithms,
advanced day-parting, and multi-account consolidation
renders the platform indispensable for large brands
seeking scale and small businesses pursuing competitive
advantage on limited budgets.
CONCLUSION
Based on the foregoing review, the evolution of
advertising tools on marketplaces inevitably leads to the
complete replacement of manual management by
intelligent AI systems capable of operating under high
competition, rising click costs, and cookieless-
environment constraints. The increase in variables
—
from bid and inventory dynamics to user behavior
—
has
exceeded the capacity for human oversight, rendering
routine tasks inefficient in terms of time and resource
expenditure. Under these conditions, ThinkAd
demonstrates a practical solution: it integrates the
collection and consolidation of sellers’ first
-party data,
predictive analytics, and an algorithmic autopilot,
enabling minimization of ACoS alongside increases in
sales volume and liberation of up to 20 hours of
operational time per week for strategic tasks.
ThinkAd’s
functional
architecture—
which
includes an automatic semantic module for Semantic
Core generation, goal-based bidding with a 24-hour
ACoS forecast, advanced algorithmic dayparting, and
multi-account mastering
—
fully meets the needs of small
and medium businesses for a scalable tool. Practical case
studies confirm that after deploying ThinkAd’s AI bidder,
clients’ ad sales increased by 82–
206% and ACoS fell to
22
–
25% without additional staffing. Integration with the
Amazon Ads API and support for key regions ensure
timely updates. At the same time, ThinkAd’s
participation in the Amazon Ads Partner Network and its
international awards underscore the solution's high
quality and industry relevance.
Thus, ThinkAd emerges not merely as an automation
technology but as a system that transforms marketplace
advertising into a manageable business process, where
the priority shifts from manual adjustments to real-time
attainment of business metrics. Through deep platform
integration, intelligent algorithms, and a focus on first-
party data, it lays the foundation for further
development of commerce for machines, providing a
competitive advantage to large brands and companies
with limited budgets. In the future, such comprehensive
AI solutions will define new standards of efficiency and
responsiveness to continuously changing digital-market
conditions.
REFERENCES
Amazon Advertising Stats,
Ad Badger
, Mar. 11, 2025.
https://www.adbadger.com/blog/amazon-advertising-
stats/
AI In ECommerce Statistics,
Sellers Commerce
, Sep. 04,
2024.
https://www.sellerscommerce.com/blog/ai-in-
T. Buckland, Best AI Tools for Amazon Sellers,
Amazon
SEO
Consultant
,
Nov.
14,
2024.
https://amazonseoconsultant.com/best-ai-tools-for-
amazon-sellers/
Frequently asked questions related to third-party cookie
deprecation in Chrome - Google Ads Help,
https://support.google.com/google-
ads/answer/14762010?hl=en
S. Millman, Artificial Intelligence, Signal Loss, and
Consumer Personalization: The Rising Importance of
First-Party
Data
in
2025,
Research
World
The American Journal of Management and Economics Innovations
89
https://www.theamericanjournals.com/index.php/tajmei
(accessed Apr. 23, 2025).
N. Uthaiah, Key Amazon Advertising Updates for
November 2024,
SellerApp Blog
, Dec. 05, 2024.
https://www.sellerapp.com/blog/amazon-advertising-
updates-november-2024/
The Value of Getting Personalization,
McKinsey &
Company
,
Nov.
12,
2021.
Apr. 25, 2025).
A. K. Nagaraja, How generative AI and data are
redefining retail experiences,
Amazon Web Services
,
Oct.
23,
2024.
Maximize your brand goals with goal-based bidding in
Amazon DSP,
Amazon Ads
, Sep. 19, 2024.
https://advertising.amazon.com/resources/whats-
new/amazon-dsp-goal-based-bidding
2025).
Perpetua uses AI to help reduce ACOS by 15%,
Amazon
Ads
,
Aug.
06,
2024.
https://advertising.amazon.com/library/case-
studies/perpetua-tsuiteru-acchan-honmono-honpo
(accessed Apr. 26, 2025).
Quartile
for
Amazon
PPC,
Quartile
,
2025.
https://www.quartile.com/amazon-ppc
27, 2025).
R. Bragg, Just In! Survey Shows Rise of Multichannel
Marketing in 2025,
Taradel
, Jan. 17, 2025.
https://www.taradel.com/blog/just-in-survey-shows-
rise-of-multichannel-marketing-in-2025
27, 2025).
E. Berne, 2025 trends in multichannel selling: what
marketplace sellers need to know,
Mirakl
, Feb. 19, 2025.
https://www.mirakl.com/blog/ecommerce-seller-
trends-2025
D. Andersen, Data-Driven Marketing Trends for 2025:
Why You Need to Update Your Strategy,
Invoca
, Aug. 05,
2024.
https://www.invoca.com/blog/state-of-data-
driven-marketing-update-your-strategy
29, 2025).
State of Data 2025, IAB, 2025. Accessed: Apr. 30, 2025.
[Online].
Available:
content/uploads/2025/03/IAB_State_of_Data_2025_M
arch__V2.pdf
R. Mehta, Amazon’s Rufus AI assistant now available to
all US customers,
Amazon
, Sep. 18, 2024.
https://www.aboutamazon.com/news/retail/how-to-
use-amazon-rufus
I. Bousquette, Walmart Is Preparing to Welcome Its Next
Customer: The AI Shopping Agent,
The Wall Street
Journal
,
May
15,
2025.
ThinkAd - AI Amazon Optimization Platform,
ThinkAd
Handcraft Blends Boosts Conversions with ThinkAd AI-
Powered
Bidding,
ThinkAd
,
Aug.
20,
2024.
New York Biology Achieves +206% Increase in
Advertising Sales,
Thinkad.ai
ECDMA Global Awards 2025 Winners,
ECDMA
, 2025.
https://awards.ecdma.org/winners-2025/
(accessed
May 16, 2025).
“Pricing,” ThinkAd, 2025. https://thinkad.ai/pricing
(accessed May 18, 2025)
