Evaluating the Effectiveness of Leading Job Portals: A Cross-Platform

Analysis

Mettu Siddhartha, M Ramya Sree, Poli Vamsi Vardhan Reddy, Balam Ruchith Balaji,

Parlapalli Bhargav Reddy and Radha D

∗

Department of Computer Science and Engineering, Amrita School of Computing, Bengaluru,

Amrita Vishwa Vidyapeetham, India

Keywords:

Job Searching Platforms, SEO Analysis, Website Performance, Usability Best Practices, Core Web Vitals,

PageSpeed Insights.

Abstract:

The modern era has seen widespread adoption of online job platforms, connecting job seekers directly with

potential employers. In the proposed work, various key performance evaluation tools—such as Google’s Page-

Speed Insights, Pingdom, SEMrush, and GTmetrix—have been employed to assess metrics like Performance,

SEO(Search Engine Optimization), Accessibility, Best Practices, Load Time, and Trafﬁc, among others. These

tools provide insights into platform strengths and limitations, independent of developer inﬂuence, thereby

ensuring objectivity and offering substantial recommendations for enhancing user experience and technical

aspects. The analysis concludes that Job Board-1 demonstrates a high level of competence across nearly all

critical factors essential for trafﬁc, audience engagement, and platform availability, and it presents several

technological strengths. Job Board-3, while performing reasonably well in terms of accessibility and SEO,

faces challenges related to high bounce rates and slow loading speeds, suggesting areas for improvement to

increase user engagement.

1 INTRODUCTION

Job seeking in the digital age has transitioned from

ofﬂine to online, with websites serving as a vital con-

duit between job seekers and potential employers.

These platforms, which now include Job Board-1, Job

Board-2, Job Board-3, and Job Board-4, are essential

resources for anyone looking to advance in their ca-

reers. Each platform offers unique features and ser-

vices tailored to meet the needs of its diverse user

base, ranging from job listings and company ratings to

networking opportunities. The effectiveness of these

platforms relies not only on the number of job post-

ings but also on their overall performance, usability,

and accessibility.

Page load speeds and user interface design were

the main emphasis of traditional website performance

monitoring. However, a more thorough study is now

achievable because to the development of increas-

ingly advanced techniques and technology. Conven-

tional approaches frequently involved basic mobile

compatibility tests and broad user satisfaction sur-

veys. Although these approaches offered a wide per-

spective, they were insufﬁciently detailed to compre-

hend the fundamental elements that lead a website to

successfully furnish a smooth job search encounter.

Selected job search platforms will be evaluated

across a range of parameters using industry-standard

tools. Starting the evaluation, the performance, SEO,

accessibility, and other best practices are checked

with Google PageSpeed Insights. This involves con-

sidering factors such as title tags Meta description,

server response time, loading speed of both the mo-

bile and the desk top version of the Website. The

properties like metadata, mobile-responsiveness, and

structured data are analysed in order to understand the

websites’ visibility and their capability of SEO opti-

mization. Moreover, the issues of usability and acces-

sibility are also veriﬁed by such criteria as the exis-

tence of alt text for images, keyboard navigation, and

color perception violations on the platforms. About

performance I use Pingdom by giving tests from a

server in Japan to examine performance in the Region

of Asia. To analyze the trafﬁc of various platforms,

the SEMrush tool is used, including the total visits,

visits by unique visitors, bounce rates, and the average

time spent by a user on a site. This site goes further in

giving different general performance indicators like

616

Siddhartha, M., Sree, M. R., Reddy, P. V. V., Balaji, B. R., Reddy, P. B. and D, R.

Evaluating the Effectiveness of Leading Job Portals: A Cross-Platform Analysis.

DOI: 10.5220/0013598400004664

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 3rd International Conference on Futuristic Technology (INCOFT 2025) - Volume 2, pages 616-623

ISBN: 978-989-758-763-4

LCP, TBT, CLS, total page weight, and the quantity

of requests to give a full report in how well these Plat

forms run. By adopting a multi-dimensional perspec-

tive to analyze each platform, the strategies for how

to optimize them are also shown in detail.

The remaining part of the paper consists of re-

view on the background information and introduction

to the topic is discussed in section 1.and Section 2

presents literature of related studies and prior work

done in similar ﬁeld. Section 3 provides the method

explaining the method used, methods employed and

instruments applied in the study. Section 4 displays

the results in form of ﬁndings, data and any analysis

done on them. Last, Section 5 provides a summary of

the ﬁndings and the implications of the research for

practice.

2 LITERATURE REVIEW

The section outlines prior work done on measuring

the performance of the website, search engine opti-

mization, keyword study and usability studies across

multiple industries. It underlines the role of recog-

nizing the shifts of keywords and users’ behavior and

the inﬂuence made by SEs’ algorithms on the website.

Such as SEMrush, PageSpeed Insight, or GTmetrix

that is used to review the performance and to deter-

mine slips. Largest Contentful Paint and First Input

Delay are among the most important metrics conﬁg-

ured within the Core Web Vitals to address user expe-

rience and site performance needs. Various aspects of

the website should be optimized on a continuous basis

and the results should be compared with industry tar-

gets to ensure that a high efﬁciency with subsequent

improved search engine ranking is attained.

Nanda et al. study is to examine the trend analy-

sis of the keywords, questions, and website domains

most frequently associated with melanoma and skin

cancer search (Nanda, Hay et al. 2021). All the infor-

mation was obtained using different search engines

and ﬁndings indicated higher search terms based on

skin cancer as compared to melanoma. The clini-

copathologic classiﬁcation and diagnosis constituted

the largest group of the ten most frequently used

melanoma keywords. The survival query was the

most popular but general/melanoma or diagnosis type

query type yielded the highest number of searches

per query. To help resolve common issues related

to melanoma and skin cancer, the study suggests ex-

plaining.

Using the SEMrush analytics tool, Afroz et

al. evaluated and analysed the websites of con-

sumer electronics companies (Afroz, Riyazuddin et

al. 2023). It evaluates performance indicators like

visitor attrs, bounce rate, average stay duration, most

popular pages, ctr, backlink, and referring domain.

The AIDA model, developed by St. Elmo Lewis, is

used to evaluate promotive activities of some brands.

The ﬁndings help brands understand consumer behav-

ior trends and their digital footprints, guiding their

marketing approach. Web analytics have developed

rapidly, with SEMrush being a user-friendly tool.

A web-based solution that uses WebpageTest,

PageSpeed Insight, and GTmetrix tools to auto-

matically gather and compare e-commerce site per-

formance metrics was presented by Hossain et

al.(Hossain, Hassan et al. 2021). The application uses

PHP, MYSQL, CSS, and HTML, and allows users to

input the URL of a site. Tests on ten Bangladeshi

e-commerce sites showed site7 had the lowest To-

tal Blocking Time, while site10 had the lowest Load

Time. The application currently supports computer

systems, with future research focusing on mobile ver-

sions and testing limits.

According to Palacios-Zamora et al. assessing

university websites’ effectiveness is crucial to en-

hancing their reputation (Palacios-Zamora, Cordova-

Morana et al. 2023). It discusses various quality as-

sessment models and highlights the impact of parame-

ters like response time, throughput, utilization, exten-

sibility, data transfer rate, concurrency, and reliability

on website performance. Measures like optimizing

images, layouts for desktop, tablet, or mobile devices,

and content caching can enhance site efﬁciency. The

research also suggests improving mobile device per-

formance and using the increasing, usability, and ac-

cessibility brief.

When evaluating user experience, the authors em-

phasized importance of PageSpeed Insights (Web-

PageTest), First Input Delay (FID), and Largest Con-

tentful Paint (LCP) (Dobbala and Lingolu, 2023). It

suggests that web developers should focus on improv-

ing these parameters to enhance website performance,

efﬁciency, and business results. This will ultimately

lead to increased conversion rates, creating a compet-

itive advantage in online business.

Bernine et al. provided an analytical model

grounded on Petri nets in assessing the effectiveness

of a Web services structure if requests and services

adhere to an exponential server (Bernine, Nacer et al.

2020). This mainly applies since the arrival of user as

well as the web service requests follows the Poisson

distribution only. The model is solved analytically,

and the response time and mean number of clients

for the system are determined. The limit number of

clients in the system is derived from which the sys-

tem becomes congested.

Evaluating the Effectiveness of Leading Job Portals: A Cross-Platform Analysis

617

According to Kumar et al. website perfor-

mance has a signiﬁcant impact on user experience,

search engine optimisation, and overall business suc-

cess(Kumar, Kumar et al. 2021). Key automated

tools such as Google PageSpeed Insights, GTmetrix,

Pingdom, WebPageTest, and Apache JMeter are re-

viewed by the authors, who highlight their useful-

ness in locating performance bottlenecks and stream-

lining web applications. The authors support a me-

thodical approach to performance evaluation that en-

hances site efﬁcacy and efﬁciency by using a variety

of instruments, ongoing monitoring, and benchmark-

ing against industry norms.

Akg

ul et al. highlighted the signiﬁcance of public

value, usability, and readability in improving user ex-

perience and public service delivery of Turkish gov-

ernment website performance (Akg

ul, 2024). The

authors evaluate how well these websites meet the

needs of citizens, encourage participation, and guar-

antee that all users can access the material. Using

a mixed-methods approach and making use of auto-

mated technologies such as Google PageSpeed In-

sights, GTmetrix, and WebPageTest, the study ad-

vocates for a user-centred strategy in e-government

projects and offers useful suggestions for enhancing

government websites.

The idea of SEO and the methods by which stan-

dard tools can be used to evaluate a website’s search

engine optimisation were described by Simec et al.

(Simec and Kri

zani

c, 2023). As an example, it pro-

poses to apply three tools to evaluate the same site

and compare their functions and options. In the same

token, SEO is a very ﬂuid ﬁeld where one is always

in assimilating new information in the ﬁeld. Semrush

is an online tool that helps analyze SEO, content, a

market, web advertising, and social networking. It

gives a percentage bound to each item where the im-

portant mistakes ﬁxed and the warnings given are said

in terms of a percentage.

Simunic et al. conducted a study on SEO factors

to improve internet visibility (

Simuni

c, Stifanich et

al. 2023). The research involved literature analysis

and an analysis model. The study found no signiﬁ-

cant correlation between website quality and search

query points. However, there were qualitative deﬁcits

in variable optimization. The study suggests that de-

tecting and optimizing important Google factors for

ranking can lead to higher online direct sales, enrich-

ing new scientiﬁc data knowledge and creating poten-

tial for further research.

The authors Considers the rise in trafﬁc from ser-

vices like HTTP, FTP, and SMTP, network trafﬁc clas-

siﬁcation is crucial for monitoring and managing data

ﬂow in networks (Archanaa, Athulya et al. 2017). In

this work, the performance of using different sorts of

supervised learning algorithms: Ensemble learning,

Decision tree and Bayesian classiﬁers for detecting

network trafﬁc. When applying the wrapper method

in feature selection, then the research highlights that

the Decorate Algorithm an ensemble classiﬁer is efﬁ-

cient and reliable than other algorithms.

Sujee et al. explained Modern educational

databases have expanded, and it is possible to ﬁnd

many hidden resources that can contribute to improv-

ing the results achieved by students (Sujee, Padma-

vathi et al. 2021). This work explains how predictive

modeling, grouping, and association rule mining can

be employed in order to discover information beneﬁ-

cial to students and tutors. RBF model enables one to

predict which students are performing well and which

group could beneﬁt from a boost in instructions. It

helps the instructors in the settings to present teaching

methodologies that ﬁts every student’s needs success-

fully.

The domain, domain age, web impact variables,

and Alexa trafﬁc rank of Indian universities were

compared with the NIRF ranking 2021 by Meghwal

et al. (Meghwal, Joshi et al. 2022). It discovered that

all the websites in the study employed SEO tools. Out

of the universities, Amrita Vishwa Vidyapeetham uni-

versity had the oldest domain registration in Decem-

ber 1988 and the Indian Institute of Science university

had the highest Domain Authority score of 62. A Na-

tional level survey found it that three Universities of

Karnataka were in the top ten Universities in India.

Subbulakshmi et al. explained how a framework

for extending the assessment of Web portal pertinence

can be used to measure the reliability of Web portals

and to calculate their credibility score (Subbulakshmi,

Gopika et al. 2019). Main parameters taken into ac-

count include the content, links, spam information,

the frequency of updating the web portals under con-

cern. The credibility score is determined using met-

rics such as page rank, blacklist status, average page

hits, and two key quality factors: , namely credibility

and relevance. A web crawler collects page links from

different websites in response to speciﬁc queries and

in the end, presents the Web sites in order of reliabil-

ity so that users can easily determine which sources

contain reliable information.

The Deep Cyber Threat Situational Awareness

Framework (DCTSAF), developed by Soman et al.

uses deep learning to detect malicious domains

and URLs . Traditional methods like blacklisting

and signature-based strategies don’t work, especially

when faced with fresh or more sophisticated threats.

Deep learning is the foundation of the framework,

which uses character-level embeddings to work with

INCOFT 2025 - International Conference on Futuristic Technology

618

the raw data and automatically determine the most rel-

evant features. The use of both hierarchical features

and long-distance linkages in domain names/URLs

makes CNN-LSTM networks better than other mod-

els. The system is extremely powerful and can handle

two million events per second, which leads to early

threat detection and an alert.

Malathi et al. pointed out In the current genera-

tion to form social network they must family social

network site such as Linked In, Google Plus and face

book (Malathi and Radha, 2016). Social networks

can be represented by graphs and their most signiﬁ-

cant ﬁgures, connections, and interaction. Other uses

of graphs include the representation of relations and

processes of different systems, chemical and physi-

cal, biological, information as well. This work ex-

amines a database of US politics books as a network

with such elements as betweenness, eigen vector, de-

gree and closeness.

3 EXPERIMENT

In conducting the analysis of website performance,

several industry-standard tools were used to gather in-

sights across key metrics such as Performance, SEO,

Accessibility, Best Practices, Load Time, and Trafﬁc

To assess performance, SEO, accessibility, and

best practices, Google PageSpeed Insights was used.

This tool provides detailed insights into the speed

characteristics of a website, scoring and offering sug-

gestions for improvements on both desktop and mo-

bile versions.

3.1 Performance

This metric complies with how quickly and active the

website is once opened. The tool breaks down fea-

tures such as response time of the server, resource

loading and rendering speed.

3.2 SEO

SEO score is deﬁned as the capability of the website

to appear in search engines results and evaluate meta-

data, mobile-friendliness, and structured data.

3.3 Accessibility

This metric deals with the accessibility of users with

complications for example; this measures the number

of images that have the alt text; users who can only

use the keyboard to get round the website; users who

have problems with color perception.

3.4 Best Practices

Google PageSpeed Insights assess the current state of

website according to modern web technologies and

standards like security, response, and coding.

3.5 Load Time

The loading time for this website was determined by

Pingdom Tool. Namely, the tests were performed

from an Asia-Japan-Tokyo server to consider the load

effect in this region. when using Pingdom, which de-

livers speciﬁc results based on several aspects such as

server response time, resource loading time, and total

page loading time. This assists in ﬁnding out some of

the constraints in performance that might hamper on

user experience within certain espoused zones.

3.6 Trafﬁc Analysis

For the assessment purpose of monitoring website

trafﬁc and its behavior index, the SEMrush Tool was

used. It also offers additional trafﬁc details such as the

visits, the visitors, the bounce rates and the average

session time. The following metrics were analyzed:

3.6.1 Total Visits

The over all the number of time the site was visited in

the given time of the study.

3.6.2 Unique Visitors

The number of people who visit this website without

repetition of any user.

3.6.3 Purchase Conversion Rate

This was measured wherever possible showing the

rate which visitors made a purchase or achieved a de-

sired end result.

3.6.4 Pages per Visit

The quantity of unique web page visits by each user

on average per session.

3.6.5 Average Visit Duration

The speciﬁc time that the users stay connected to the

site/ portal.

Evaluating the Effectiveness of Leading Job Portals: A Cross-Platform Analysis

619

3.6.6 Bounce Rate

The rate at which visitors leave the site after visiting

only one of its pages.

GTmetrix [] was also used, offering additional

performance insights. The tool includes metrics such

as GTmetrix Grade, Performance Score, Structure

Score, Largest Contentful Paint (LCP), Total Block-

ing Time (TBT), Cumulative Layout Shift (CLS), To-

tal Page Size, and Total Number of Requests.

3.7.1 GTmetrix Grade

This is a composite score that reﬂects the overall per-

formance and structure of your website. It’s broken

into two components:

Performance Score (70% of the grade) Reﬂects

how well the website performs based on loading

speed and interactivity, derived from Core Web Vitals

metrics.

Structure Score (30% of the grade) Measures

how well your site follows best coding and optimiza-

tion practices to ensure faster load times.

3.7.2 Performance Score

This is the segment of the GTmetrix Grade that looks

at how effectively users can access and interact with

what your website has to offer. While it also incorpo-

rates data from other performance indicators such as

Core Web Vitals to provide a transparent indication of

user experience scores.

3.7.3 Structure Score

This metric emphasizes various aspects of a website’s

architecture or design. Often it shows how well your

site can actually perform and details like too much

JavaScript, ineffective CSS, and uncompressed im-

ages discover. The greater the structure score, the

more your site prepares appropriately regarding load-

ing speed and maintainability.

3.7.4 Largest Contentful Paint (LCP)

LCP determines the page load time on the screen tak-

ing the biggest factor that is the size of the largest vis-

ible content element (image, video or large text block)

into consideration. This is an essential UX value, as

increased LCP can give a user a sensation that a site

is slow. Google advises LCP should happen within a

time frame of 2.5 seconds from the page time.

3.7.5 Total Blocking Time (TBT)

TBT quantiﬁes the time the browser is occupied by

tasks that hinder or completely halt user interaction

(such as script loading). It measures the time a web-

site’s JavaScript or any other resource takes to hamper

interactivity and quantify the amount of interactivity

lost due to a speciﬁc resource. The site’s TBT values

must be lower since this represents a more responsive

website.

3.7.6 Cumulative Layout Shift (CLS)

CLS gives the total scrambled layout/ﬂush that hap-

pens inadvertently when you load a site. For instance,

if images or ads take time to load that they across text

and make it shift about it will contribute to the CLS

score. A CLS score below 0.1 guarantees that users

have reliable and consistent client-side experience.

3.7.7 Total Page Size

This metric demonstrates the sum of the sizes of all

ﬁles required for rendering the webpage – images,

scripts, stylesheets and others. Factoring for larger

page sizes is that greater page size decreases the load-

ing time, particularly for internet users with low band-

width. Therefore the best page size is derived from

minimizing its resources since a decrease in page size

has a positive effect on page performance.

3.7.8 Total Number of Requests

The number of HTTP requests or browser clicks an

HTML page makes to load all required elements, in-

cluding images, java, css, and fonts, is shown by this

measure. The longer it takes for a website to load, the

more requests the page receives. The website loading

time can be extended if script requests can be made

more frequently or at a later time.

Several objective tools were chosen because they

are universal and can give more or less exhaustive in-

formation about various aspects of a website taking

into account technical parameters and users’ experi-

ences. These metrics provide a birds-eye view of the

website and this not only considers such factors as

speed of loading of the website, ease of use of the

website, availability of Search Engine Optimization

and identiﬁed best practices but also factors such as

structure or organization of the website, responsive-

ness of the website, and optimum use of resources

amongst others. In any case, it integrates supports the

identiﬁcation of speciﬁc areas of improvement, pro-

viding a fair assessment based on both, technology

and users.

INCOFT 2025 - International Conference on Futuristic Technology

620

4 EXPERIMENTAL RESULTS

Analyzing the performance of four job board web-

sites, their strengths and weaknesses differ signiﬁ-

cantly. However, Job Board-1 stands out by excelling

in multiple categories, performing superbly with high

scores in trafﬁc, audience engagement, and easy ac-

cess. It claims the highest visit count, unique visi-

tors, and pages per visit, which are strong indicators

of higher user engagement and a lower bounce rate.

Job Board-2 receives an average number of visitors

but experiences slow loading times, which may lead

to potential user loss. Job Board-4 shows good per-

formance in terms of accessibility and compliance,

but overall visits have decreased, which could be im-

proved by enhancing SEO and reducing load times.

Job Board-3, the smallest platform by trafﬁc, boasts

the highest accessibility and best practice rates. How-

ever, it has a relatively high bounce rate and lower

interaction rates, indicating room for improvement in

user experience (UX). The ﬁndings from this analy-

sis highlight the importance of technical performance,

search engine optimization, and user-centric metrics

for driving effective trafﬁc ﬂow.

According to the data presented in Table 1, the

following important aspects of website performance

can be evaluated. Regarding performance, both Job

Board-1 and Job Board-3 posted faster loading times,

allowing customers to engage in quicker interactions,

unlike Job Board-2, which took 3.30 seconds to load.

The overall SEO report is almost perfect for most

sites, with Job Board-1 and Job Board-3 scoring 100

points, while Job Board-4 and Job Board-2 scored rel-

atively lower. In terms of accessibility, Job Board-

1 achieved a perfect score of 100%, indicating full

compliance with accessibility standards for disabled

users, while Job Board-4 received the lowest score.

Job Board-3 performed moderately well. For best

practices, Job Board-1 had the lowest score of 74,

whereas other platforms, such as Job Board-4, Job

Board-3, and Job Board-2, scored above 90. When

considering visits, Job Board-1 led with 1.8 billion

visits, followed by Job Board-2 with 43.9 million, Job

Board-4 with 29.5 million, and Job Board-3 with 7.1

million. Engagement rates show that Job Board-1 had

the highest page views per visit at 8.8, with an average

visit duration of 12 minutes and 1 second. However,

Job Board-3 had the highest bounce rate at 58.41%,

indicating that users leave the site quickly.

Fig 1 represents the trafﬁc analysis of Job Board-

1, Job Board-2, Job Board-3, and Job Board-4, reveal-

ing distinct performance metrics. Job Board-1 leads

signiﬁcantly with 1.88 billion visits (up 3.62%) and

423.6 million unique visitors (up 6.44%). Job Board-

2 follows with 43.9 million visits (up 9.14%) and 24.7

million unique visitors (down 4.49%).

Table 2 provides a comparative analysis of four

job boards (Job Board-1 to Job Board-4) based on

performance and structure metrics from GTmetrix,

covering six key aspects: GTmetrix Grade, Perfor-

mance Score, Structure Score, Largest Contentful

Paint (LCP), Total Blocking Time (TBT), Cumulative

Layout Shift (CLS), and Total Page Size. Job Board-1

achieved the highest overall performance with an ’A’

GTmetrix Grade (91%), the best Performance Score

(89%), and a strong Structure Score (93%). It also

had the fastest LCP (770ms), minimal blocking time

(260ms), and the smallest page size (585KB), indi-

cating quick load times and efﬁcient resource usage.

In contrast, Job Board-3 performed the worst, with

the lowest GTmetrix Grade (’D’, 61%), the slowest

LCP (4.5s), the highest TBT (1.2s), and relatively

large page size (2.43MB). Job Board-4 also lagged

in some areas, but performed well in blocking time

(172ms). Overall, Job Board-1 stands out as the best-

performing website in terms of both speed and user

experience.

Fig 2 presents a comparison of key performance

metrics across ﬁve different platforms, focusing on

areas such as Performance, SEO, Accessibility, and

Best Practices. Two platforms show the highest per-

formance scores, reﬂecting quicker load times, while

one lags behind with the lowest performance score

and a longer load time of 3.30 seconds. In the SEO

category, two platforms achieve a perfect score, in-

dicating strong optimization, while others have room

for improvement. Accessibility is another area where

one platform excels with a perfect score, suggesting

full accessibility for disabled users, while another has

the lowest score. Regarding best practices, three plat-

forms consistently score high, but one unexpectedly

has the lowest score in this category. The chart of-

fers a visual comparison that highlights strengths and

areas for improvement, particularly in SEO and best

practices for some platforms.

Job Board-4 sees a decline in visits to 29.5 mil-

lion (down 13.36%) and 13.1 million unique visitors

(down 5.61%). Job Board-3 shows modest growth

with 7.1 million visits (up 2.38%) and 3.5 million

unique visitors (down 3.63%). In terms of engage-

ment, Job Board-1 has the highest pages per visit at

8.8 and an average visit duration of 12:01 minutes.

Job Board-4 and Job Board-2 show increases in pages

per visit, while Job Board-3’s metrics decline. The

bounce rate is highest for Job Board-3 at 58.41%,

while Job Board-1 has the lowest at 41.11%. Over-

all, Job Board-1 outperforms its competitors across

all key metrics, highlighting its effectiveness and user

Evaluating the Effectiveness of Leading Job Portals: A Cross-Platform Analysis

621

Table 1: Performance analysis of various webpages

Job Board-1 Job Board-2 Job Board-3 Job Board-4

Performance 100 64 60 93

SEO 100 85 100 77

Accessibility 100 90 90 75

Best practices 74 96 100 93

Load Time 767ms 3.30 s 853 ms 112 ms

Table 2: Performance and Structure Metrics of Various Job Boards as Analyzed by GTmetrix

Job Board-1 Job Board-2 Job Board-3 Job Board-4

GTmetrix Grade A (91%) C (76%) D (61%) C (78%)

Performance Score 89% 71% 39% 76%

Structure Score 93% 83% 93% 82%

Largest Contentful Paint 770ms 686ms -83ms 4.5s +3.7s 3.0s +2.2s

Total Blocking Time 260ms 1.0s +775ms 1.2s +919ms 172ms -88ms

Cumulative Layout Shift 0 0.01 +0.01 0 0.01 +0.01

Total Page Size 585KB 3.98MB +3.41MB 2.43MB +1.86MB 2.98MB +2.41MB

engagement.

5 CONCLUSION

In conclusion, analysis indicates that Job Board-1 sur-

passes other job platforms in performance, user en-

gagement, SEO, and accessibility, positioning it as

a benchmark job platform. Another popular plat-

form, Job Board-2, exhibits slower loading times,

suggesting potential improvements in user experi-

ence. Job Board-4, recognized for strong accessibil-

ity and adherence to best practices, may enhance traf-

ﬁc through targeted SEO optimizations. Job Board-

3 ranks highly in accessibility and best practices,

though its bounce rate remains elevated, with moder-

ate user engagement relative to larger platforms. Con-

sistent with prior ﬁndings, this study emphasizes that

a balance of efﬁciency, user satisfaction, and ﬁndabil-

ity is essential for the success of online platforms.

REFERENCES

Nanda, J.K., Hay, J.L. and Marchetti, M.A., 2021. Analysis

of keywords used in Internet searches for melanoma

information: observational study. JMIR dermatology,

4(1), p.e25720.

Afroz, S., Riyazuddin, Y.M. and Jadda, V., 2023, December.

Website Trafﬁc Trends and Performance Evaluation of

Selected Consumer Electronic Company Websites. In

2023 Global Conference on Information Technologies

and Communications (GCITC) (pp. 1-4). IEEE.

Hossain, M.T., Hassan, R., Amjad, M. and Rahman, M.A.,

2021. Web performance analysis: an empirical analy-

sis of e-commerce sites in Bangladesh. International

Journal of Information Engineering and Electronic

Business, 11(4), p.47.

Palacios-Zamora, K., Cordova-Morana, J., Mendoza-

Cabrera, D. and Pacheco-Mendoza, S., 2023. Mea-

surement on University Websites: A Perspective of

Effectiveness. JOIV: International Journal on Infor-

matics Visualization, 7(3-2), pp.1995-2006.

Dobbala, M.K. and Lingolu, M.S.S., 2022. Web Perfor-

mance Tooling and the Importance of Web Vitals.

Journal of Technological Innovations, 3(3).

Bernine, N., Nacer, H., Aissani, D. and Alla, H., 2020. To-

wards a performance analysis of composite web ser-

vices using Petri nets. International Journal of Mathe-

matics in Operational Research, 17(4), pp.467-491.

Kumar, N., Kumar, S. and Rajak, R., 2021, December. Web-

site Performance Analysis and Evaluation using Au-

tomated Tools. In 2021 5th International Conference

on Electrical, Electronics, Communication, Computer

Technologies and Optimization Techniques (ICEEC-

COT) (pp. 210-214). IEEE.

Akg

ul, Y., 2024. Evaluating the performance of websites

from a public value, usability, and readability per-

spectives: a review of Turkish national government

websites. Universal Access in the Information Soci-

ety, 23(2), pp.975-990.

Simec, A. and Kri

zani

c, M., 2022. Comparison of tools for

analyzing the degree of optimization of website search

engines.

Simuni

c, M., Stifanich, L.P. and Car, T., 2023. Hotel

web site SEO analysis: Segmentation and valoriza-

tion as a precondition for discovering and understand-

ing insights to improve online visibility. Ekonomski

vjesnik/Econviews-Review of Contemporary Busi-

ness, Entrepreneurship and Economic Issues, 36(2),

pp.299-311.

Archanaa, R., Athulya, V., Rajasundari, T. and Kiran,

M.V.K., 2017, January. A comparative performance

analysis on network trafﬁc classiﬁcation using super-

vised learning algorithms. In 2017 4th International

INCOFT 2025 - International Conference on Futuristic Technology

622

Figure 1: Trafﬁc analysis of various platforms

Figure 2: Visualization plot for website analysis

Conference on Advanced Computing and Communi-

cation Systems (ICACCS) (pp. 1-5). IEEE.

Sujee, R. and Padmavathi, S., 2021, October. Performance

Analysis and Prediction of Students Results Using

RBF Networks. In 2021 2nd International Conference

on Smart Electronics and Communication (ICOSEC)

(pp. 333-338). IEEE.

Meghwal, J., Joshi, K., Chaparwal, N. and Rajput, P.S.,

2022. NIRF Ranking 2021: A Webometric Analysis

of Top 10 University Websites of India. International

Journal of Research in Library Science (IJRLS), 8(2),

pp.191-205.

Subbulakshmi, S., Gopika, A. and Thomson, L., 2019,

July. Enhanced Ranking of Websites Based on Cred-

ibility Evaluation. In 2019 International Conference

on Communication and Electronics Systems (ICCES)

(pp. 2035-2040). IEEE.

Soman, K.P., Poornachandran, P. and Menon, V.K.,

Vinayakumar R Center for Computational Engineer-

ing and Networking (CEN), Amrita School of Engi-

neering, Coimbatore Amrita Vishwa Vidyapeetham,

India vinayakumarr77@ gmail. com.

Malathi, A. and Radha, D., 2016, October. Analysis and

visualization of social media networks. In 2016 In-

ternational Conference on Computation System and

Information Technology for Sustainable Solutions

(CSITSS) (pp. 58-63). IEEE.

Google PageSpeed Insights

SEMrush

Pingdom Tool

GTMatrix

Evaluating the Effectiveness of Leading Job Portals: A Cross-Platform Analysis

623