The American Journal of Social Science and Education Innovations
103
https://www.theamericanjournals.com/index.php/tajssei
TYPE
Original Research
PAGE NO.
103-111
10.37547/tajssei/Volume07Issue05-13
OPEN ACCESS
SUBMITED
19 March 2025
ACCEPTED
28 April 2025
PUBLISHED
23 May 2025
VOLUME
Vol.07 Issue 05 2025
CITATION
Abramau Usevalad. (2025). Implementing Sustainable Digital Design
Principles into Web Product Development. The American Journal of Social
Science and Education Innovations, 7(05), 103
–
111.
https://doi.org/10.37547/tajssei/Volume07Issue05-13,
COPYRIGHT
© 2025 Original content from this work may be used under the terms
of the creative commons attributes 4.0 License.
Implementing Sustainable
Digital Design Principles
into Web Product
Development.
Abramau Usevalad
Senior Art Director, Kargo Jersey City, NJ, USA
Abstract:
This article examines the principles of
sustainable digital design as a holistic approach to web‐
product
development,
wherein
each
interface
optimization considers not only user experience and
business metrics but also the material costs of
computation, data transmission, and rendering. The
relevance of this work is driven by the rapid increase in
energy consumption of data centers and user devices
under the influence of digitalization, and by the necessity
to minimize the carbon footprint of web services amid
constraints of the “green” energy system and social
responsibility for product accessibility. It seeks to
organize and measure important ways of eco-friendly
web design
—
the choices of loading, dark styles, graphic
types, shortening and shaking trees, client or server
rendering options, adaptive delivery of content, and CO
₂
budgets
—
based on their effects on speed, energy used by
devices, and the environment. For this purpose, a careful
look at industry reports, acad
e
mic tests, and real-world
studies was done, plus a metric comparison from the Web
Almanac Core, Web Vitals, and lab tests. The novelty of
this work lies in integrating ecological, economic, and
technical metrics into a unified methodology. For each
technique, comparable numerical estimates of
carbon‐
emission reduction and business‐efficiency gains are
presented, and recommendations are developed for their
combined application within the CO₂ budget of first‐
screen delivery. Key findings indicate that activating lazy
loading with a single HTML attribute can reduce
transferred data volume and improve Time to Interactive
by 20
–
30%; employing a dark theme on OLED displays
under bright lighting can decrease display power
consumption by up to 47%; and replacing PNG images
with SVG sprites can reduce graphic payload by 60
–
80%.
This article will be valuable to web designers, front‐end
developers, product managers, and IT strategists seeking
to marry high interface performance with a minimal
The American Journal of Social Science and Education Innovations
104
https://www.theamericanjournals.com/index.php/tajssei
The American Journal of Social Science and Education Innovations
carbon footprint.
Keywords:
sustainable digital design, web performance,
la
zy loading, dark theme, SVG, tree shaking, server‐side
rendering, adaptive loading, CO₂ budget.
Introduction:
Sustainable digital design conceives every
interface as a chain of material events
—
from CPU cycles
to megawatt‐hours in data centers—
and is thus founded
on the principle that “every byte has mass.” When
designers account for carbon footprint alongside pixels,
and engineers allocate budgets for memory, bandwidth,
and processing power from the outset, the resulting
product is not only “lightweight” but also resilient: it
loads faster, supports a broader range of devices,
requires hardware upgrades less frequently, and better
adapts to technological shifts. In this approach,
aesthetics, code, and economics converge on a single
objective
—
simultaneously reducing costs for users,
organizations, and the planet. The imperative to pursue
sustainability became undeniable once “digital” ceased
to be immaterial. Data production, storage, and
transmission already account for 1.5
–
4% of global
greenhouse‐
gas emissions, and this share continues to
rise [1]. The International Energy Agency forecasts that
data‐center electricity demand will double by 2030,
driven particularly by the surge in generative AI
workloads [2]. It is increasingly clear that, without a
fundamental rethink of digital architecture and design,
digital infrastructure will expand faster than the
decarbonization of the power sector can keep pace,
thereby becoming a significant barrier to global efforts
to reduce emissions.
This load growth carries environmental, social, and
ethical risks. Algorithmic data management raises the
question of “who is accountable for machine errors,”
while information overload in interfaces provokes user
fatigue and erodes trust. Moreover, most users continue
to access the web via older smartphones and slower
networks; when products cater primarily to flagship
devices, they inadvertently exacerbate the digital divide.
In this context, systemic responsibility extends beyond
internal optimization and aligns with the United Nations
Sustainable Development Goals: quality education (Goal
4), reduced inequalities (Goal 10), climate action (Goal
13), and strong institutions (Goal 16) depend directly on
how thoughtfully we design digital environments.
From a business perspective, sustainable design proves
to be not charity but a mechanism for efficiency gains.
Major platforms report double-digit declines in CDN and
request-processing expenditures by reducing image
sizes, implementing lazy loading, and adopting server‐
side rendering. At the same time, interface speed
delivers direct revenue: a study [3] showed that cutting
page‐load time from five to one second increases
purchase conversion by a factor of 2.5. Accessibility
improvements also monetize: Forrester analysts
calculated that every dollar invested in inclusive
enhancements can yield up to one hundred dollars in
cost savings and additional revenue [4]. Beyond direct
returns, there are intangible dividends
—
improved
brand loyalty, reduced legal risk, and enhanced
reputation as a responsible organization.
Finally,
code
minimalism
benefits
developers
themselves. Fewer dependencies and less “fast fashion”
in the interface reduce the need for emergency patches,
lower the barrier to entry for new team members, and
extend the p
roduct’s lifespan without radical redesign.
Ultimately, sustainable digital design emerges not as a
compromise between ecology and usability but as a
strategy in which economic, technical, and social
benefits coalesce into a single, easily quantifiable value.
MATERIALS AND METHODOLOGY
The materials and methodology of this study are based
on a comprehensive analysis of 23 sources, including
international reports, academic publications, industry
benchmarks, technical documentation, and practical
implementation case studies. Principal sources include
the World Bank report on the ICT sector’s share of global
emissions [1] and the IEA forecast on data‐center energy
growth under the influence of AI [2], as well as studies
on the economic effects of web‐performanc
e and
accessibility optimization [3, 4]. To assess the
prevalence and efficacy of specific practices, data from
The Web Almanac (Media and Sustainability sections)
[5, 19] were employed, alongside laboratory
measurements and field‐test results published in
industry blogs and technical articles [6
–
8, 10
–
12, 18].
The
theoretical
foundation
comprises
works
demonstrating
the
linkage
between
interface
optimization and device energy savings: a study of lazy
The American Journal of Social Science and Education Innovations
105
https://www.theamericanjournals.com/index.php/tajssei
The American Journal of Social Science and Education Innovations
loading in WordPress themes [6] and experiments by
Fast Familiar Workroom on energy savings during page
scrolling on budget smartphones [7]; analyses of dark‐
theme effects on OLED screens under high‐ and low‐
ambient‐light conditions [8–
10]. To evaluate vector
graphics and sprite benefits, comparison data from
Vecta.io on PNG versus SVG [11] and Cloudinary
recommendations for optimizing SVG sprites were used
[12].
Correlations
between
JavaScript-bundle
minification and reduced parsing time were examined in
two independent SaaS‐platform case studies by Wagner
et al. [13, 14].
Methodologically, this research combines systematic
review and content analysis of documentation,
comparative technology analysis, and quantitative
efficiency
estimates.
The
systematic
review
encompassed standardized metrics from the Web
Almanac and Google Core Web Vitals [5, 15], while
content analysis covered articles on tree shaking and SSR
versus CSR in Next.js projects [13, 16]. Adaptive loading
implementation details were examined using Client
Hints and the Network Information API in React-Hooks
practices [18], along with Google statistics on the impact
of load speed on user churn [17]. CO₂ budgeting control
relies on Web Almanac page‐weight recommendations
and data from the Website Carbon service [19, 20].
RESULTS AND DISCUSSION
Lazy loading is a simple technique, but its effects extend
beyond local speed optimization. When the browser
only fetches those images, video frames, or iframes that
fall within the user’s viewport, it reduces the initial
volume of data that must be transmitted and decoded.
According to the Web Almanac 2024, the native
loading="lazy" attribute is already used on one-third of
all pages, making it the fastest-growing media-handling
practice [5]. Laboratory measurements on popular
WordPress themes show that after enabling lazy loading
of visual resources, Time to First Interaction is reduced
by 20
–
30%, and the total weight of initial HTML plus
critical styles and scripts decreases by approximately
one third [6]. The savings are particularly pronounced on
mobile networks: a study [7] on several budget
smartphones recorded lower battery discharge during
scrolling a typical landing page when images were
loaded lazily.
The technical implementation requires no heavy
libraries: for most browsers, it suffices to add the
loading="lazy" attribute to the <img> or <iframe> tag,
and for fine-tuning the visibility thresholds, one can use
the
IntersectionObserver
API.
An
example
implementation is shown in Fig. 1.
Fig. 1. Example of implementation of lazy loading (compiled by author)
The American Journal of Social Science and Education Innovations
106
https://www.theamericanjournals.com/index.php/tajssei
The American Journal of Social Science and Education Innovations
Thus, a single line of HTML becomes a tool that can
significantly reduce page weight
—
by 30
–
60% when
working with heavy media files
—
while simultaneously
accelerating Time to Interactive by 20
–
40%, decreasing
device energy consumption (especially on mobile
devices), and reducing the carbon footprint by
approximately 10
–20 g CO₂ per thousand loads. Along
with other sustainable design principles, lazy loading is
the first barrier against interface bloat. It allows
resources to remain where they are genuinely needed
and confirms that user and planetary savings in web
development coincide.
After optimizing traffic and computation, the next level
concerns care for pixels: the dark theme turns
background “black” into the literal deactivation of
subpixels on OLED screens, so each unlit diode saves
fractions of a watt and slows battery discharge. The
principle is the same as for lazy loading: a minor code
change shifts resource allocation across the entire chain,
and again, visual aesthetics can simultaneously reduce
carbon footprint and enhance usability.
The physics of the process is simple: in an organic light-
emitting diode, only those subpixels that must be
brighter than absolute black emit light; therefore, an
interface dominated by #000000 consumes less energy
than a traditional light background. Laboratory
measurements of six popular Android applications on
four generations of OLED smartphones showed that, at
brightness automatically increased to 100% under
sunlight, switching to dark mode reduces display power
consumption by 39
–
47%, and in maximally contrasted
scenarios, savings reach 63% of total screen power [8].
However, the same experiment recorded only a 3
–
9%
benefit indoors at average brightness, and a recent
study [9] found that 80% of users compensate for the
dark palette by increasing brightness, thus nullifying or
even reversing the effect [9]. The energy balance,
therefore, depends not only on color but also on
context: dark mode is useful under high-brightness
conditions, whereas in evening lighting, the gain is
determined by the user’s discipline.
Beyond electricity, dark mode affects comfort: reduced
overall screen luminance decreases glare and retinal
irritation, which is especially noticeable at night. It is no
coincidence that Flurry analytics showed that after
22:00, 82.7% of smartphone users switch to a dark
interface [10]. Increased reading comfort prolongs
session length; consequently, news feeds and
messaging applications record higher usage times,
achieving direct conversion returns without additional
ad units.
Implementation in the browser requires one line: the
media query @media (prefers-color-scheme: dark)
allows palette overriding without duplicating styles.
Mobile frameworks
—
from Swift UI to Android Jetpack
—
inherit the system theme, and
persisting the user’s
choice via localStorage or SharedPreferences consumes
no more than a dozen bytes. Like lazy loading, the effect
happens without outside libraries. It does not make the
setup more complex: the lighter the static assets, the
better they work on slow devices.
Sustainable design gives context to dark mode, verifying
the central thesis of this article: a design decision is
“green” not because it looks minimalist, but statistically
it reduces energy consumption, prolongs battery life,
and lowers cumulative emissions throughout the
device’s lifecycle. Proper brightness configuration and
consideration of user scenarios are other examples of
how caring for the planet aligns concurrently with
business and user interests.
Once lazy loading and dark mode have reduced code
and rendering costs, the next logical step is to lighten
the graphics. Replacing raster PNG and JPEG with vector
SVG shifts the interface from pixel-based descriptions to
geometry, so that the same icon weighs tens of kilobytes
less and renders without loss of clarity even on Retina
displays. In a laboratory test by Vecta.io, the difference
reached 60
–
80% in favor of SVG
—
8 KB versus 82 KB for
the PNG counterpart
—
directly saving bandwidth and
accelerating first content rendering [11]. Additional
gains come from the sprite approach: when dozens of
icons are combined into a single file, the browser
requires only one HTTP request. The study recommends
SVG sprites as a request-optimization practice [12].
Ultimately, vector graphics reduce the data transmission
carbon footprint and spare the product from storing
multiple raster copies for different DPIs. An example of
usage is shown in Fig. 2.
The American Journal of Social Science and Education Innovations
107
https://www.theamericanjournals.com/index.php/tajssei
The American Journal of Social Science and Education Innovations
Fig. 2. Example of implementation of SVG images (compiled by author)
However, even a “bare” SVG can be wrapped in
superfluous JavaScript code. Minification and tree
shaking, built into production modes of Webpack, Vite,
and other bundlers, eliminate this excess by removing
whitespace, unused variables, and entire modules. A
demonstration SPA on web.dev, after enabling tree
shaking, “slimmed down” from 20.8 KB to 8.5 KB—
the
main bundle’s reduction was almost 60%—
with the
savings showing not only in traffic but also in script
parsing and compilation times [13]. Larger projects
confirm the scalability of this result: in a 2025 case, a
SaaS platform team reduced its production bundle by
70%
—
from 4.8 MB to 1.6 MB
—
through import
refactoring, thereby restoring acceptable speed on 3G
networks [14]. Reduced computation directly lowers
energy consumption on user devices and cuts CDN-
traffic costs.
The following optimization layer concerns where the
browser executes the remaining code. Client-side
rendering is convenient for single-page applications but
requires full JavaScript loading before the first pixel
appears, thus shifting computational costs to the user’s
device. Server-side rendering, by contrast, delivers
ready-to-render HTML and allows the browser to begin
rendering immediately upon response receipt. This is
why Core Web Vitals guidelines recommend SSR to
improve LCP: having markup in the initial HTML reduces
main-content load latency and diminishes main-thread
blocking [15]. Practical measurements in Next.js projects
show that server-side or hybrid rendering accelerates
Largest Contentful Paint by 30
–
50% compared to classic
CSR and simultaneously increases store conversion rates
by tens of percent due to the shorter user-exit window
[16]. The weaker the device and the less stable the
network, the more pronounced the difference: on
budget smartphones under 4G conditions, SSR offloads
CPU work, reducing Total Blocking Time and,
consequently, energy consumption, which aligns with
sustainable-design goals.
Thus, transitioning to SVG, aggressive bundle cleansing,
and considering rendering-mode choice from another
savings profile in which every kilobyte and millisecond
converts into tangible monetary and ecological
dividends. Combined with lazy loading and dark mode,
this creates a sustainable architecture in which visual
quality, accessibility, and carbon-footprint reduction
reinforce one another, and the product remains “light”
for both the user and the planet.
After code and graphics have been lightened via SVG
and minification, the next logical level of sustainability is
to adjust what and to whom the server delivers flexibly.
Real-world users access content from dozens of devices
and network types, and Google statistics show that 53%
of mobile sessions drop off if a page loads for more than
three
seconds
[17].
Therefore,
rather
than
indiscriminately sending the same “heavy” package to
all users, adaptive loading assesses current conditions
and proportionally reduces data volume: a low-end
smartphone on 3G receives a streamlined layout
without video backgrounds, while a flagship on Wi-Fi
receives the full media version.
Technically, the solution is built on Client Hints and the
Network Information API. The browser reports effective
connection type, the save-data flag, number of cores,
and amount of memory, and the script or server decides
which images, fonts, and scripts are essential. Google's
method is detailed using React Adaptive Loading hooks,
where “slow
-
2g” connections receive thumbnails
instead of posters, and budget “quad
-
core” devices skip
heavy animations [18]. Such load distribution is
especially beneficial in regions with unstable internet:
even without changes to business logic, the interface
becomes noticeably faster, conserves data and battery,
The American Journal of Social Science and Education Innovations
108
https://www.theamericanjournals.com/index.php/tajssei
The American Journal of Social Science and Education Innovations
and thus reduces emissions associated with modem and
processor operation.
To prevent this additional logic itself from becoming
ballast, it is simultaneously necessary to reconsider what
qualifies as “essential” visual content. According to the
Web Almanac 2024 Sustainability chapter, more than
half of a page’s total weight is attri
butable to images
[19]. By replacing decorative PNG icons with text or
system emoji, and raster logos with a typographic
brandmark in the default font, one can instantly
eliminate dozens of requests and tens of kilobytes
without loss of meaning. Text requires no decoding,
adapts to dark mode, and is immediately available to
screen readers, enhancing accessibility. Practice shows
minimal typography and color accents are perceived as
expressively as static illustrations, especially on small
screens.
However, even after all local optimizations, a project
risks reverting to bloat without clear quantitative
benchmarks. That is why more companies are
introducing a CO₂ budget: in addition to traditional time
and pixel metrics, they set an upper limit on the mass of
the first screen. The Web Almanac recommends staying
below 1 MB, with an ideal threshold around 500 KB [19].
Page-weight statistics are presented in Fig. 3. Such a
limit helps eliminate superfluous elements at the
design-mockup stage: every new library or background
video must pass the test of “is it worth its carbon cost.”
External services like Website Carbon show that today’s
average page emits about 0.8 g CO₂ per view, providing
a clear savings target [20].
Fig. 3. Page weight by percentile [19]
Adaptive loading, eliminating unnecessary images, and
proactive weight control complement the previously
discussed techniques of lazy loading, dark mode, and
bundle optimization. Together, they form a framework
where every additional byte must justify itself by
benefiting the user. Thus, design, performance, and
climate responsibility inevitably converge into a single
engineering discipline, making web products faster,
more accessible, and more durable.
A striking example is Google Chrome: browser-level
optimization, implemented under the names Data Saver
and Lite Mode, takes adaptive-loading principles to their
logical extreme
—
compressing and simplifying the entire
page before it reaches the device. When a connection is
classified as “2G” or “slow
-
2G,” Chrome p
roxies the
request through Flywheel, replaces heavy images with
WebP thumbnails, removes unnecessary JavaScript, and
delivers a “Lite page.” Field measurements showed that
The American Journal of Social Science and Education Innovations
109
https://www.theamericanjournals.com/index.php/tajssei
The American Journal of Social Science and Education Innovations
this scheme reduced data transfer volume by up to 90%
and halved load time; most importantly, on slow
networks, the share of pages that fully render before
timeout increased significantly [21]. This result
demonstrates that traffic optimization is not merely a
technical enhancement but a direct factor in audience
retention.
The effect is especially pronounced in contexts with
limited data plans and unstable connectivity. When
Google launched Lite Mode for news in India, the project
targeted 200 million mobile users. It reduced data
consumption to less than one-third of the standard
mode while preserving key information
—
headlines and
text [22]. The practice proved that “sustainability
through inclusivity” works in two ways: on one hand,
saving kilobytes and grams of CO₂; on the other, making
the product accessible to a broader audience, thereby
boosting loyalty and expanding the market in countries
with emerging digital infrastructure.
It is important to note that since 2022, Lite Mode in
Chrome has been disabled: mobile networks became
cheaper, and the browser’s architecture learned to save
data by default [23]. Nevertheless, the case remains
illustrative for sustainable design: it shows how focusing
on real usage contexts spurs creative rethinking of the
interface and the entire content-delivery chain. When
teams are given room for such solutions, ecological and
economic benefits are complemented by intangible
value
—
their product becomes harder to replicate
because it is built upon deep user understanding and
responsible choice of every byte.
CONCLUSION
It has been demonstrated in the above analysis that
sustainable digital design principles in web-product
development can yield sustainability for the user,
business, and environment at different levels of
participation. Lazy loading reduces the volume of
transmitted data and TTFI (Time to First Interaction) with
the user, thereby reducing loads on both data centers
and devices, subsequently reducing the carbon footprint
by tens of grams of CO₂ per thousand loads. A dark
interface theme on OLED screens shows optimized
display energy consumption. It enhances user comfort,
particularly under high-brightness conditions, reflecting
the direct interrelation between aesthetics and
efficiency.
Further graphics optimization through vector SVG
images and sprite assembly reduces visual-resource
weight by 60
–
80%, eliminating the need to store
multiple raster copies and speeding up rendering.
Minification and tree shaking in modern bundlers allow
JavaScript bundles to “slim down” by nearly half,
reducing CPU load on devices and CDN traffic. The
choice between client-side and server-side rendering
adapts to device and network capabilities, ensuring
maximal speed of the first painted content.
Adaptive loading based on Client Hints and the Network
Information API generates a personalized data package
according to connection quality and device capabilities,
which is especially relevant for users of budget
smartphones and regions with unstable internet.
However, the effectiveness of all these techniques
depends on precise page weight control and introducing
a CO₂ budget: maintaining the first
-screen mass within
500
–
1000 KB becomes a criterion for including new
libraries and visual elements.
Thus, sustainable digital design comes not as a part of
disparate techniques but as an integrated part of an
engineering discipline where every decision made, from
a single line of HTML to architectural choices,
measurably influences performance, accessibility, and
climate responsibility. Lazy loading, dark theme, vector
graphics, aggressive bundle cleansing, adaptive content
delivery, and strict weight control constitute the durable
and inclusive digital environment that simultaneously
achieves economic, technical, and social objectives. This
comprehensive
approach
confers
competitive
advantages to companies and advances global efforts
toward decarbonization and reducing digital inequality.
REFERENCES
“Measuring the Emissions & Energy Footprint of the ICT
Sector,” World Bank, 2024. Accessed: Apr. 11, 2025.
[Online].
Available:
“AI is set to drive surging electricity demand from data
The American Journal of Social Science and Education Innovations
110
https://www.theamericanjournals.com/index.php/tajssei
The American Journal of Social Science and Education Innovations
centres while offering the potential to transform how
the energy sector works,”
IEA
, Apr. 10, 2025
—
M. Wiegand, “Site Speed is (Still) Impacting Your
Conversion
Rate,”
Portent
,
Apr.
20,
2022.
https://portent.com/blog/analytics/research-site-
speed-hurting-everyones-revenue.htm
13, 2025).
M. Bervell, “The Business Case for Digital Accessibility: A
Revenue-
Generating Investment,”
Test Party
, Mar. 05,
2025.
https://testparty.ai/blog/the-business-case-for-
S. Judis and E. Portis, “Media,”
The Web Almanac by
HTTP Archive
, Dec. 2024, Accessed: Apr. 15, 2025.
[Online].
Available:
https://almanac.httparchive.org/en/2024/media
Vasile Crudu, “Why You Should Use Lazy Loading
Alongside Minification in WordPress for Optimal
Performance,”
MoldStud
,
Apr.
16,
2025.
J. McAlister, “How we made our website use 50% less
energy,”
Fast Familiar Workroom
, Mar. 23, 2021.
https://workroom.fastfamiliar.com/how-we-made-our-
website-use-50-less-energy/
Y. Charlie Hu, “Shedding light on dark mode to save
energy,”
Purdue
University
,
2022.
https://engineering.purdue.edu/ECE/News/2022/shed
ding-light-on-dark-mode-to-save-energy
15, 2025).
M. Turner, “Popular ‘power saving’ mobile feature is
sucking your battery,”
The Sun
, Feb. 20, 2025.
https://www.thesun.co.uk/tech/33472529/power-
saving-mobile-dark-mode-sucking-battery/
(accessed
Apr. 16, 2025).
Princewillton,
“Dark
Mode
Usage
Statistics,”
Helpfultech
, Apr. 28, 2024.
mode-usage-statistics-2024.html
2025).
N. Palombi, “SVG files: what are they and how to use
them?,”
Webflow
,
2025.
https://webflow.com/blog/svg-file
2025).
“A Developer’s Guide to SVG Optimization,”
Cloudinary
,
Jan. 15, 2025.
https://cloudinary.com/guides/image-
formats/a-developers-guide-to-svg-optimization
(accessed Apr. 17, 2025).
J. Wagner, “Reduce JavaScript payloads w
ith tree
shaking,”
web.dev
,
2018.
https://web.dev/articles/reduce-javascript-payloads-
with-tree-shaking
“How We Reduced Our JavaScript Bundle Size by 70%
While Adding Features,”
Medium
, Apr. 21, 2025.
“The most effective ways to improve Core Web Vitals,”
web.dev
,
2024.
https://web.dev/articles/top-cwv
(accessed Apr. 20, 2025).
Maharaf Hossen, “Next.js vs React: Why Next.js is the
Future of Web Development,”
Medium
, Feb. 24, 2025.
Google, “Mobile Site Abandonment after Delayed Load
Time,”
Think
with
https://www.thinkwithgoogle.com/consumer-
insights/consumer-trends/mobile-site-load-time-
statistics/
M. Mihajlija, “Adaptive loading: improving web
performance
on
slow
devices,”
web.dev
https://web.dev/articles/adaptive-loading-cds-2019
(accessed Apr. 22, 2025).
Laurent Devernay Satyagraha, Burak Güneli, I. Akrap, A.
Dawson, M. Gifford, and T. Frick, “Sustainability,”
The
Web Almanac by HTTP Archive
, Nov. 2024, Accessed:
Apr.
24,
2025.
[Online].
Available:
The American Journal of Social Science and Education Innovations
111
https://www.theamericanjournals.com/index.php/tajssei
The American Journal of Social Science and Education Innovations
https://almanac.httparchive.org/en/2024/sustainability
“How is your website impacting the planet?”
Wholegrain Digital
https://www.websitecarbon.com/
(accessed Apr. 24, 2025).
“Chrome Lite Pages
- For a faster, leaner loading
experience,”
Chromium Blog
, Mar. 12, 2019.
https://blog.chromium.org/2019/03/chrome-lite-
pages-for-faster-leaner.html
J. Morehead, “Introducing Google News Lite mode —
faster news for slower networks,”
, Sep. 27, 2016.
https://blog.google/outreach-initiatives/google-news-
initiative/introducing-google-news-lite-mode/
(accessed May 03, 2025).
C. Welch, “Google is ditching Chrome’s data saver mode
on
Android,”
The
Verge
,
Feb.
23,
2022.
https://www.theverge.com/2022/2/23/22947441/goo
gle-lite-mode-data-saver-chrome-android-discontinued
(accessed May 06, 2025)
