Content uploaded by Ramin Abolghasemi Komleh
Author content
All content in this area was uploaded by Ramin Abolghasemi Komleh on Sep 03, 2023
Content may be subject to copyright.
Journal of Information Systems and Informatics
Vol. 5, No. 3, September 2023 e-ISSN: 2656-4882 p-ISSN: 2656-5935
DOI: 10.51519/journalisi.v5i3.518
Published By DRPM-UBD
1002
This work is licensed under a Creative Commons Attribution 4.0 International License.
Investigation of Relationship Between Google Cost-Per-
Click and Search-Volume on Keyword of Chicago Tours
Ramin Abolghasemi Komleh1
Faculty of Business and Management, University of Innsbruck, Austria
Email: 1ramin.abolghasemi-komleh@student.uibk.ac.at
Abstract
Pay-per-click is one of common and important ways in online advertising through Google;
but some keywords are usually expensive, and this issue is visible through Google cost-
per-click in Google keyword planner tool. It has done many research work in Google
advertising formats, but place of studies in field of relationship between Google cost-per-
click and keyword search-volume, especially in tourism area is blank. This paper tries to
answer of this main question that is there a significant relationship between cost-per-click
and keyword search-volume and the number of keyword's word? In other words, does
cost-per-click increase or decrease, based on keyword search-volume and the number of
keyword's word (multi-word keyword or hyper focused keyword phrases)? Chicago
auspicates a mighty increase in tourism 2022 as its announcement an 86 percent growth in
visitors in 2021 compared to 2020 and choosing the right keyword in Google search engine
by travel agencies can lead to more sales. In this research, I sampled 100 search result
through combination of ″chicago+tours″ as a search keyword on google.com (Google
USA) and Google keyword planner daily data (December 26, 2022) based on the last 7-10
days, and via regression analysis with least squares and loess model, I tried to investigate
the relationship between search-volume and cost-per-click and it will help scholars in
future research in this area. This study was carried out within the scope of selected
keywords in tourism area in form of cross-sectional. I find that there is no significant
relationship between short-tail keywords and cost-per-click. The result of this research
shows that some medium or long-tail keywords are more expensive than short-tail
keywords with more search results. Another result that was observed is the lack of
significant relationship between keyword search-volume and its cost-per-click, so that in
some cases, high-search keywords are cheaper than keywords with a low-search.
Keywords: Google Ads, Google AdWords, Pay-Per-Click, Cost-Per-Click
1. INTRODUCTION
Since the advent of the Internet, advertisers found several ways to promote their
productions and services via online advertising [1]. Online advertising has evolved
over the past two decades with flashy changes in use of various online advertising
formats [2]. For advertisers, online campaign decisions are complicated by the
wide variety of advertising formats and advertisers' poor knowledge base about
Journal of Information Systems and Informatics
Vol. 5, No. 3, September 2023
p-ISSN: 2656-5935
http://journal-isi.org/index.php/isi
e-ISSN: 2656-4882
Ramin Abolghasemi Komleh | 1003
their effects [1]. The ever-increasing emergence of highly intrusive ad formats of
advertising has led to the tremendous growth of ad-blocking systems [3]. In
schematization of digital marketing, keyword selection for search engine
deployment is desperate for client attraction and dependent investment is a more
related part of marketing budget [4]. Every second, million searches are performed
on Google search engine and Google Ads can be a very effective way to drive
targeted users to advertiser website, precisely when people are searching for a
variety of products or services [5], [6]. Google AdWords start in year 2000 to
solicitation to light business attendance. Through filling a form and present a bank
account, small business advertises their product or services on Google in rivalry
with big brands. Difficulty with vast cost-per-mile (CPM) or cost-per-thousand
(CPT) rate, led to an auction-based model using cost-per-click (CPC) [7]. Google
transformed online advertising with AdSense contextual aimed ads, based on new
attainment practical semantics technology. Targeted advertising uses text of
articles or stories that a user read and likeness them with relevant service and
product advertising. For example, when someone read an article about system
software might receive ads related to electronics products. Microsoft Bing, MSN
and Yahoo bid contextual advertising, but Google continues to novelize through
Froogle for database of listings, Google maps for geo targeting; Google data base
for lists and YouTube for video. These all-Google additions suggest plenty new
online advertising opportunity [7]. Google present main three kind of goods:
keywords, quantity of keywords and engine search results [8]. Google same as
networking television, swapping content with advertisement, but a main difference
of Google and networking television is that Google perpendicularly integrates
advertising representation and ratings system with search engine [8]. Google
AdWords is layout that advertiser, suggest for keywords on Google search page.
Google employs a major degree of revision of both number of ads and mechanism
following ads appear [8]. Indeed, AdWords is a bid process that Google actuates
to assign paid results to search engine inquest, which perch singly the side or on
top to unpaid result [9], [10].
Google as an advertising representation, sells keywords; as a ratings corporation,
sells census of keywords and as add-up provider, sells search indicator. All these
information have trade value because Google is transforming the information,
which or else does not have value, into valuable market commodities and appends
them to search output and priced them accordance of their popularity [8], [11].
From 2009, Google is superlative visited website and from June 2009 to August
2009, within 32.7 percent of globe internet users visited to Google.com [8]. Google
ads program absorbs small and medium sized businesses that cannot impute TV
or local newspapers expensive advertisements [8]. In Google AdWords’ campaign,
can specify offer entities with set of keywords and related precise, expression or
wide types. These keywords are check versus user inquest, which permits for well
tuning of user intention [12]. The procedure of text-based search ads is various
Journal of Information Systems and Informatics
Vol. 5, No. 3, September 2023
p-ISSN: 2656-5935
http://journal-isi.org/index.php/isi
e-ISSN: 2656-4882
1004 | Investigation of Relationship Between Google Cost-Per-Click and Search-Volume ....
from non-search and non-text-based ads [8]. Google addition to display search
result pages, is showing ads on third party websites and owners of these websites
can link to Google AdSense so Google text ads become visible on their websites
in exchange for money from Google [8]. Indeed, AdSense is Google procedure of
linkage third-party ads to related third-party add-up, same as news sites or blogs,
and display advertisement sidelong selected context in digital billboard [10].
Advertisers, instead of purchase a time gap as in broadcast ads or a space as in
printing ads, suggest the keywords on AdWords program [8]. Advertiser pays
display cost of ad to become visible to user and calculate simultaneous with clicks
on ad and Google allocates a modality score to every keyword for any advertiser
and willingness to pay does not alone specify status of ad [8], [9]. Quality score
and bid rate of keywords assign ad rank, and if user query be more relevant to ad,
it will receive higher state score [8].
Google controls conjunction through inflict a space range on search ad.
Irrespective of budget, any ad has similar longitude limit four script. First line
should be a headline contain 25 characters or less; last line ought to be the
advertiser website URL. Google AdWords homologous to civil media, can
purpose a geographical zone and a population through IP address of internet
connection acquaint Google server recognize the user location [8]. Pay-per-click
(PPC) is based on compete and bid between advertising users [13]. This kind of
strategy in digital marketing is as well as referred to as CPC [13]. PPC is one of
common and important ways in online advertising through Google; but the top
rank keywords are much expensive [14]. When advertiser build its ad, must specify
the amount that is willing to spend each time a user click on advertisement through
setting CPC must set a daily budget, based on a month set by advertiser, divided
to days of month [15].
The results of a study about keywords selection on Google ads, show that keyword
choice problem dissolve through combination of evolutionary computation,
abysmal learning, machine learning, and text processing technique [16]. According
to another research study on keywords click-through-rate (CTR) and average CPC
prediction problems, the random forest to be the best technique for CTR and
average CPC prediction, whenever the gradient boosting gives the most inaccurate
outcome [17]. With considering of all efficacy elements of AdWords campaign,
specific phase of layout should contain keywords choice, design of step-up
explanation with deployment, performance of divested keywords, designation of
a plan and geo location, website design and optimization and campaign parameters
[18]. Suggested model of AdWords campaign depends on two factors. First is
strictly connected with campaign parameters, but second is relevant to aspect of
promoted goods and their firms [18]. Broad match in compare to exact match,
leading to less advertising efficiency. However, negative effect of broad match on
CTR is more outstanding for more particular or less position keywords, and effect
Journal of Information Systems and Informatics
Vol. 5, No. 3, September 2023
p-ISSN: 2656-5935
http://journal-isi.org/index.php/isi
e-ISSN: 2656-4882
Ramin Abolghasemi Komleh | 1005
of broad match on click rate, income and gain are short negative for higher
position keywords, and paid search advertiser campaigns should consider keyword
particularly and location in creating of optimal strategy decision for their keyword
[19]. Keyword longitude in both America and England markets is negative
correlated with CPC, and despite their developed markets, advertisers not prefer
to Keyword search in Google search engine through multiple words as an
advertising trigger [20].
Another finding shows that advertisers spend rather for particular keywords and
less for keywords that contain a brand title, and pay more for keywords that are
searched the highest position than the lower status. This result may be due to this
fact that only top ads seen and even when not being click has advantage [21]. With
changing of matching option from broad to exact, except of cost, metrics of
keyword traffic enhance. lengthy keywords, which are typically related with more
centralize search purpose, produce more clicks and have a more modality scale
and longer keywords are inexpensive for precise match option, but for other
matching options are more expensive [22]. Whenever CTR decline based on
position, conversion rate first rise and next reduce for longer keywords with
position [23]. Google search engine ranking mechanism not take into account the
conversion rate, and for this reason, advertisements that placed in upside position,
have not always maximize income [23], [24]. However, CTR in top of page is much
superior than lower positions, and most users have narrow search [23].
The primary increase based on position shows that users with superior purchase
intention, perhaps evaluate some positions before purchase decisions making.
Users with high purchase intention, pause their searching after few slots and only
information seeking users clicks on advertisements in lower positions [23].
keyword return on investment is much different under various imputation
strategies. For instance, in a focal firm, first click allotment led to less income
return and rather pronounced reduction in rather particular keywords [25].
Common search acting has positive affect on brand searching activity through
increment of information, but brand search has no effect on generic searching
[26]. Google average CTR in one top listing search result was 19.3 percent;
however, CTR of listed in second place was half that [27]. In another study, six-
month data collection of more hundred keywords from a big retailer show that a
click monetary value is not same among every position, because conversion rate is
highest on top and reduce rank with down search engine results page; however,
google search engine to consider the current bid and previous click-through rates,
before calculating of an advertisement final rank in current cycle; but the running
bid has more effect than former click-through-rates. However, the score rises of
landing page quality depend with increase in conversion rates and reduction in
advertiser CPC [24].
Journal of Information Systems and Informatics
Vol. 5, No. 3, September 2023
p-ISSN: 2656-5935
http://journal-isi.org/index.php/isi
e-ISSN: 2656-4882
1006 | Investigation of Relationship Between Google Cost-Per-Click and Search-Volume ....
Today there are a major number of studies on online marketing area [28]; but many
theoretical literatures only focus on bid strategies and optimal mechanism layout
of search engines [29] and it seems that there are few articles on tourism topic.
Studies on keyword ads in tourism appeared from 2010, and few have intended to
data analyzing tourism keyword ads [28] and not exist multitude studies on how
tourism enterprises apply Internet in marketing [30]. Marketing strategies
formulization become a crucial issue for traveling agencies that wishing to bond
their superiority in the market and up to now frequent using of humorous and
attractiveness subject was the main orientation in this area advertising [31].
However, SEO invest and paid search, is as the most effective strategy for the
digital marketing of as a tourist destination [32].
In many research that has done so far on Google advertising formats, the place of
studies in field of relationship between Google CPC and Keyword search-volume,
especially in tourism area is blank. This paper is mostly related to two broad
streams of research. First, the role and effect of suitable keywords in Google
search engine. Second, relationship between created keywords and CPC through
Google keyword planner tool and tries to answer of this main question that is there
a significant relationship between CPC and keyword search-volume and the
number of keyword's word? In other words, does CPC increase or decrease based
on keyword search-volume and the number of keyword's word (multi-word or
hyper focused keyword phrases)?
2. METHODS
This quantitative research, is a cross-sectional study and because of estimation and
focusing on the relationship between dependent (keyword and search-volume) and
independent (CPC) variables, regression analysis done, and for curve fitting and
robust estimation, used the least squares and loess model, respectively. The
imported data have been interpreted as cross-sectional. From 24 most visited
places in the United State in 2022 [33], city of Chicago due to a mighty increase in
tourism 2022 as its announcement an 86% growth in visitors in 2021 compared to
2020 [34] it has selected as a main search keyword of this research. The statistical
population of this research is 100 search keywords that received from Google
keyword planner tool. Keyword tool is one of AdWords features, and introduces
the possible keywords for ads campaign and reporting the Google statistics for
selection keywords, containing searching performance and seasons trends, and
produces and propose keywords via keyword variation function or site related
keyword function [15]. These selected keywords took through combination of
main keyword of Chicago, as a city name with word of tours in form of ″chicago
tours″, on google.com (Google USA) from keyword planner daily data (December
26, 2022), based on the last 7-10 days, received from https://www.thehoth.com.
Journal of Information Systems and Informatics
Vol. 5, No. 3, September 2023
p-ISSN: 2656-5935
http://journal-isi.org/index.php/isi
e-ISSN: 2656-4882
Ramin Abolghasemi Komleh | 1007
Figures 1 & 2 show the Google search box with Google autocomplete predictions
and first result page of the keyword.
Figure 1. Google search box and autocomplete predictions of main keyword
Figure 2. Google first result page related to main keyword
Google keyword planner tool according to input the main keyword, introduces
search-volume, cost-per-click in USD, Competition and keyword difficulty. Table
1 shows the rank of the selection keyword in 100 offering keywords by keyword
planner that be comparable with other offering keywords. These keywords contain
minimum two words, as a short-tail keywords, up to several words as a medium
and long-tail keywords. These variables have been given numeric codes as follows
and is sorted in search-volume order through gretl software. These keywords
contain from two word up to four with or without conjunction or preposition.
Journal of Information Systems and Informatics
Vol. 5, No. 3, September 2023
p-ISSN: 2656-5935
http://journal-isi.org/index.php/isi
e-ISSN: 2656-4882
1008 | Investigation of Relationship Between Google Cost-Per-Click and Search-Volume ....
Table 1. Search keywords offering by Google keyword planner
String code table for variable 1 (Keyword)
It is visible that the keywords ranking of Google autocomplete predictions in
search box (Figure 1) is different from keywords ranking, based on search-volume
in Google keyword planner tool. It may seem simple to choose a suitable tool to
check the rank of words, but users to choose a tool need to consider the support
criteria of the search location. Search results vary based on location. For this
reason, one of the important features that any word rank control tool should have
been, the ability to ranks checking from a particular country or city. This feature
is especially required for websites where local SEO is important to them and in
this situation, the ranking of keywords in other regions should be checked. So, the
result of Google USA (this study) may be different with another countries.
According to SEO keyword match types, the selected keyword in this research is
″Modified broad match″, contain ″chicago″ as a city name, ″+″ sign and the word
String code table for variable 1 (Keyword):
1 = 'chicago architecture boat tour'
2 = 'chicago architecture tour'
3 = 'chicago boat tours'
4 = 'architecture tour chicago'
5 = 'architecture boat tour chicago'
6 = 'chicago architectural boat tour'
7 = 'chicago architecture tours'
8 = 'chicago boat tour'
9 = 'chicago tour'
10 = 'chicago tours'
11 = 'boat tour chicago'
12 = 'boat tours chicago'
13 = 'chicago architectural boat tours'
14 = 'chicago river architecture tour'
15 = 'chicago river tour'
16 = 'chicago river tours'
17 = 'architectural boat tour chicago'
18 = 'chicago tour 2022'
19 = 'chicago band tour'
20 = 'chicago bus tours'
21 = 'chicago river boat tours'
22 = 'architectural tour chicago'
23 = 'chicago helicopter tour'
24 = 'chicago food tours'
25 = 'chicago ghost tours'
26 = 'chicago river boat architecture tour'
27 = 'chicago river boat architecture tours'
28 = 'chicago tour bus'
29 = 'wendella boat tour chicago'
30 = 'will wood tour chicago'
31 = 'architectural tour chicago river'
32 = 'big bus tours chicago'
33 = 'chicago architectural tour'
34 = 'chicago gangster tour'
35 = 'chicago the band tour'
36 = 'helicopter tour chicago'
37 = 'river tour chicago'
38 = 'tours chicago'
39 = 'tours in chicago'
40 = 'best chicago architecture boat tour'
41 = 'chicago architecture boat tours'
42 = 'chicago architecture tour chicago'
43 = 'chicago band tour 2022'
44 = 'chicago brewery tours'
45 = 'chicago bus tour'
46 = 'chicago crime tours'
47 = 'chicago food tour'
48 = 'chicago on tour'
49 = 'chicago river boat tour'
50 = 'chicago segway tours'
51 = 'river boat tour chicago'
52 = 'architectural river tour chicago'
53 = 'chicago architectural river tour'
54 = 'chicago boat architecture tour'
55 = 'chicago ghost tour'
56 = 'chicago mob tour'
57 = 'chicago pizza tour'
58 = 'chicago river architectural tour'
59 = 'gangster tour chicago'
60 = 'walk chicago tours'
61 = 'architectural boat tours of chicago'
62 = 'best architecture boat tour chicago'
63 = 'bus tour chicago'
64 = 'chicago walking tours'
65 = 'river architecture tour chicago'
66 = 'river boat chicago tours'
67 = 'tour of chicago'
68 = 'architecture river tour chicago'
69 = 'architecture tour chicago river'
70 = 'band chicago tour'
71 = 'boat tour chicago river'
72 = 'bus tours chicago'
73 = 'chicago boat tours chicago'
74 = 'chicago fireboat tours'
75 = 'chicago gangster and ghost tour'
76 = 'chicago gangsters and ghosts tours'
77 = 'chicago haunted tours'
78 = 'free walking tour chicago'
79 = 'gangsters and ghosts tour in chicago'
80 = 'ghost tour chicago'
81 = 'ghost tours chicago'
82 = 'no cap comedy tour chicago'
83 = 'segway tours chicago'
84 = 'smartless tour chicago'
85 = 'tour bus chicago'
86 = 'tours and boats chicago'
87 = 'best chicago boat tour'
88 = 'big bus chicago - panoramic night tour'
89 = 'big bus tour chicago'
90 = 'boat tours in chicago'
91 = 'boat tours of chicago river'
92 = 'bus tours in chicago'
93 = 'chicago architecture river tour'
94 = 'chicago architecture tour boat'
95 = 'chicago city tour'
96 = 'chicago donut tour'
97 = 'chicago pizza tours'
98 = 'chicago tour dates'
99 = 'chicago tours 2022'
100 = 'chicago tours and boats'
Journal of Information Systems and Informatics
Vol. 5, No. 3, September 2023
p-ISSN: 2656-5935
http://journal-isi.org/index.php/isi
e-ISSN: 2656-4882
Ramin Abolghasemi Komleh | 1009
of ″tours″. Table 2 shows the range-mean statistics for search-volume, using 10
sub-samples of size 10.
Table 2. Search-volume range-mean
Range
Mean
01 - 10
10400.0
7900.00
11 - 20
1200.00
3220.00
21 - 30
800.000
1740.00
31 - 40
300.000
1270.00
41 - 50
51 - 60
61 - 70
71 - 80
81 - 90
91 - 100
0.00000
120.000
130.000
0.00000
110.000
0.00000
1000.00
892.000
681.000
590.000
546.000
480.000
Figure 3 shows the range-mean plot for search-volume with least squares fit, the
slope of range against mean is equal 1.36453.
Figure 3. Range-mean plot for search-volume
Table 3 shows the range-mean statistics for CPC, using 10 sub-samples of size 10.
Table 3. Cost-per-click range-mean
Range
Mean
01 - 10
1.60000
2.39600
11 - 20
2.61000
2.19600
21 - 30
3.34000
1.87400
31 - 40
2.61000
1.80600
41 - 50
51 - 60
61 - 70
2.47000
2.16000
2.61000
1.51200
2.07100
2.13100
Journal of Information Systems and Informatics
Vol. 5, No. 3, September 2023
p-ISSN: 2656-5935
http://journal-isi.org/index.php/isi
e-ISSN: 2656-4882
1010 | Investigation of Relationship Between Google Cost-Per-Click and Search-Volume ....
Summary statistics, using the observations 1 - 100
for the variable 'CPC' for the variable 'SearchVolume'
(100 valid observations) (100 valid observations)
Mean
Median
Minimum
Maximum
Standard deviation
C.V.
Skewness
Ex. kurtosis
5% percentile
95% percentile
Interquartile range
Missing obs.
1.8785
1.6700
0.0000
3.3400
0.90606
0.48233
-0.11371
-1.3445
0.35000
2.9600
1.5800
0
Mean
Median
Minimum
Maximum
Standard deviation
C.V.
Skewness
Ex. kurtosis
5% percentile
95% percentile
Interquartile range
Missing obs.
1831.9
1000.0
480.00
14800.
2577.1
1.4068
3.5818
13.764
480.00
5350.0
1010.0
0
71 - 80
81 - 90
91 - 100
2.51000
2.33000
2.70000
1.59600
1.47000
1.73300
Figure 4 shows the range-mean plot for CPC, the slope of range against mean is
equal -0.492363.
Figure 4. Range-mean plot for cost-per-click
As visible in Table 4, sum of the values of CPC variable data set is 1.8785 USD in
1831.9 search-volume from 100 keywords. The median from numerical value
located in the middle of 100 keywords for CPC is 1.6700 and for search-volume
is 1000. The minimum CPC for a keyword is zero and minimum search-volume is
480. The maximum CPC in the population of these 100 keywords is 3.34 USD
and top search-volume is 14800. Standard deviation shows that how far is the data
from the mean value. If the standard deviation of a set of data is close to zero, it
is a sign that the data are close to the mean and have little dispersion (0.90606 cost-
per-click); while the large standard deviation indicates the significant dispersion of
the data (2577.1 search-volume).
Table 4. Summary statistics, using the observation 1 – 100
Cost-per-click
Search-volume
In Figure 5, CPC variable (X-axis) and left axis shows keyword amount and right-
side axis is search-volume quantity and keyword variable axis shows the dispersion
Journal of Information Systems and Informatics
Vol. 5, No. 3, September 2023
p-ISSN: 2656-5935
http://journal-isi.org/index.php/isi
e-ISSN: 2656-4882
Ramin Abolghasemi Komleh | 1011
ratio of three variables relative to each other and in range of 2.5-3 USD, the
keywords of row number 50 onwards also have density, up to value of 20 keyword
in 2.5-3 USD and almost 3000 search-volume, but the amount of outlier dispersion
is clearly visible.
Figure 5. X-axis (CPC), Y-axis (keyword, search-volume)
3. RESULTS AND DISCUSSION
It is quite clear that the fitting of the curves for a particular data set is not always
unique. Therefore, it is necessary to find the curve with minimum deviation from
all measured data points. This curve is known as the best fitting curve and is found
using the least squares method. In Figure 6, CPC as an independent variable (X-
axis) and Keyword as a dependent variable (Y-axis) with least squares shows the
dispersion ratio of two variables relative to each other and in range of 2.5-3 USD,
the keywords of row number 50 onwards also have density.
Figure 6. Keyword versus cost-per-click with least squares fit
Journal of Information Systems and Informatics
Vol. 5, No. 3, September 2023
p-ISSN: 2656-5935
http://journal-isi.org/index.php/isi
e-ISSN: 2656-4882
1012 | Investigation of Relationship Between Google Cost-Per-Click and Search-Volume ....
In Figure 7, CPC independent variable (X-axis) and search-volume as another
dependent variable (Y-axis) shows the dispersion ratio of two variables relative to
each other and the amount of search-volume has sparse density with CPC. The
least squares regression line shows the variables relationship in below scatterplot.
Here can draw a line that minimizes the total distance between the line and the
points while ensuring that there are approximately the same number of points
above and below the line.
Figure 7. Search-volume versus cost-per-click with least squares fit
In Figure 8, keyword variable (X-axis) and search-volume variable (Y-axis) show
the dispersion ratio of two variables relative to each other and it has a regular
interval and density; because (X-axis) has been sorted in search-volume order.
Figure 8. Search-volume versus keyword with least squares fit
Least square fitting and Loess in general, are non-parametric strategies for fitting
a smooth curve to data points. Parametric means that the data in advance conform
to some kind of normal distribution. Since some distribution is assumed
Journal of Information Systems and Informatics
Vol. 5, No. 3, September 2023
p-ISSN: 2656-5935
http://journal-isi.org/index.php/isi
e-ISSN: 2656-4882
Ramin Abolghasemi Komleh | 1013
beforehand, parametric fitting can result in fitting a smooth curve that
misrepresents the data [35]. In these cases, nonparametric smoothers can be a
better option, without assuming the data adaption with normal distribution. Here
for robust estimation, loess attempts to find a curve of best fit. In Figure 9, CPC
variable axis, relative to search-volume axis outlier dispersion is clearly observable
and this line purpose to showing the non-random part of the association between
CPC and search-volume in this two-dimensional scatter plot.
Figure 9. Search-volume versus cost-per-click with loess fit
In Figure 10, CPC axis, relative to keyword axis, outlier dispersion is obviously
visible and through loess fit line, demonstrates of non-random component of the
association between CPC and Keyword in scatter diagram.
Figure 10. Keyword versus cost-per-click with loess fit
Journal of Information Systems and Informatics
Vol. 5, No. 3, September 2023
p-ISSN: 2656-5935
http://journal-isi.org/index.php/isi
e-ISSN: 2656-4882
1014 | Investigation of Relationship Between Google Cost-Per-Click and Search-Volume ....
In Figure 11, the degree of dispersion of data outliers can be seen on search-
volume variable axis, relative to keyword axis and loess fit line provides the
possibility to predict the value based on the response of the explanatory variable.
Figure 11. Search-volume versus keyword with loess fit
A scatter plot can provide different types of correlation between variables with a
certain confidence interval. The pattern of dots is from the bottom left to the top
right, it indicates a positive correlation between the studied variables and the
pattern of points slopes from the top left to the bottom right, it indicates a negative
correlation. So, a line of best-fit (trend line) drawn to examine the relationship
between variables via robust estimation loess fit which shows the correlation of
both positive and negative variables within the selected cross-sectional statistical
population of this research, which can be seen in the loess fitted line of graphs and
indicates that it was not observed significant relationship between search-volume
and CPC in this statistical population.
First, according to Table 5, in investigation of relationship between search-volume
and CPC, main selected keyword of this research ″chicago tours″ in ranking 10
(based on search-volume) with 4400 searches, is 1.02 USD. Other keyword
″chicago tour 2022″ in ranking 18 with 2900 search-volume is 0.40 Cent and
another same keyword, ″chicago tours 2022″ only contain of plural ″S″, in ranking
99 with 480 search-volume, is 1.70 USD; but keyword of ″chicago tour dates″ in
the same search-volume 480 and ranking 98, is only 0.26 Cent and keyword of
″chicago tours and boats″ with same search-volume 480 in ranking 100 is 1.96
USD. In these comparisons, it was not observed any significant relationship
between search-volume and CPC in range of selected samples. The amount of
outlier dispersion is clearly visible in statistical graphs and summary statistics of
two variables of search-volume and CPC.
Journal of Information Systems and Informatics
Vol. 5, No. 3, September 2023
p-ISSN: 2656-5935
http://journal-isi.org/index.php/isi
e-ISSN: 2656-4882
Ramin Abolghasemi Komleh | 1015
Table 5. Relationship between search-volume and CPC
Ranking based on
search-volume*
Keyword
Search-
volume
CPC
(USD)
10
chicago tours
4400
1.02
18
chicago tour 2022
2900
0.40
98
chicago tour dates
480
0.26
99
chicago tours 2022
480
1.70
100
chicago tours and boats
480
1.96
Second, as stated by Table 6 in examining of relationship between the number of
keyword's word and CPC, keyword of ″chicago boat tours″ with three words in
rankings 3 and 12100 search-volume, is 2.80 USD and basic keyword of ″chicago
tours″ with two words in rankings 10 and 4400 search-volume, is 1.02 USD.
Another keyword ″chicago boat tours chicago″ with one more word in rankings
73 and 590 search-volume with price of 3.29 USD, is 85% more expensive than
same previous keyword and another keyword, ″chicago city tour″ with three words
in rankings 95 and 480 search-volume, is 1.42 USD. It is visible that also here too
is not significant relationship between search-volume and CPC.
Table 6. Relationship between the number of keyword's word and CPC
Ranking based on
search-volume*
Keyword
Search-
volume
CPC
(USD)
3
chicago boat tours
12100
2.80
10
chicago tours
4400
1.02
73
chicago boat tours chicago
590
3.29
95
chicago city tour
480
1.42
* This ranking is according to Table 1, based on search-volume of 100 keywords
4. CONCLUSION
The criterion for dividing SEO keywords based on search-volume is the number
of times searching by google search engine and not the users’ intention. According
to the analysis, it was observed that there is no significant relationship between
short-tail keywords and cost-per-click. Considering that users are looking for quick
results in google search engine and prefer to type short words, so it can be assumed
that shorter keywords should cost more; but the results are contrary to this idea,
and in range of keywords used in this research, it was observed that some medium
and long-tail keywords are more expensive than short-tail keywords with more
search results. Another result that was observed is the lack of significant
relationship between keyword search-volume and its cost-per-click, so that in
some cases, high-search keywords are cheaper than keywords with a low-search.
Journal of Information Systems and Informatics
Vol. 5, No. 3, September 2023
p-ISSN: 2656-5935
http://journal-isi.org/index.php/isi
e-ISSN: 2656-4882
1016 | Investigation of Relationship Between Google Cost-Per-Click and Search-Volume ....
Short-tail keywords or head term, as a basic and short phrase that are widely
searched on the internet and usually indicate a specific category or general concept
and since these types of keywords are highly searched, it is difficult to rank them
on the first search results. Also, the problem with this type of keyword is that the
intent of the users is not clear and when someone searches these words in Google,
it is not obvious whether he or she is looking for them. Medium-tail keywords
display more specific results than short-tail keywords, however, these words are
still general and a little unclear. In this search method, users have more specific
goals, but they are looking for the best offer from search engine and do not have
any specific ideas. Long-tail keywords can be a good way to the success of high-
traffic websites and can cover more the traffic web. Meanwhile, the field of
competition in long-tail keywords is far less than other types of keywords. Long-
tail keywords are usually made up of three or four words, and users who use them
are looking for a specific product or topic with clear specifications. This research
was carried out within the scope of selected keyword in tourism area in form of
cross-sectional and its results and observations are directed to its statistical
population; so, research in other keywords in other subjects and different time
frames for the future research is recommended.
REFERENCES
[1] K. Tutaj and E. A. van Reijmersdal, “Effects of online advertising format
and persuasion knowledge on audience reactions,” Journal of Marketing
Communications, vol. 18, no. 1, pp. 5–18, Feb. 2012, doi:
10.1080/13527266.2011.620765.
[2] H. Li and J. D. Leckenby, “Examining the effectiveness of internet
advertising formats,” Internet advertising: theory and research, pp. 203–224,
2007.
[3] D. Belanche, “Ethical limits to the intrusiveness of online advertising
formats: A critical review of Better Ads Standards,” Journal of Marketing
Communications, vol. 25, no. 7, pp. 685–701, Oct. 2019, doi:
10.1080/13527266.2018.1562485.
[4] A. Erdmann, R. Arilla, and J. M. Ponzoa, “Search engine optimization: The
long-term strategy of keyword choice,” J Bus Res, vol. 144, pp. 650–662,
May 2022, doi: 10.1016/j.jbusres.2022.01.065.
[5] S. Brin and L. Page, “Reprint of: The anatomy of a large-scale hypertextual
web search engine,” Computer Networks, vol. 56, no. 18, pp. 3825–3833,
Dec. 2012, doi: 10.1016/j.comnet.2012.10.007.
[6] S. Z. Za and I. Tricahyadinata, “An Analysis on the Use of Google
AdWords to Increase E-Commerce Sales,” International Journal of Social
Journal of Information Systems and Informatics
Vol. 5, No. 3, September 2023
p-ISSN: 2656-5935
http://journal-isi.org/index.php/isi
e-ISSN: 2656-4882
Ramin Abolghasemi Komleh | 1017
Sciences and Management, vol. 4, no. 1, pp. 60–67, Jan. 2017, doi:
10.3126/ijssm.v4i1.16433.
[7] G. Boone, J. Secci, and L. Gallant, “Emerging trends in online advertising,”
Doxa communicacion, vol. 5, no. 2, pp. 241–253, 2010.
[8] M. Lee, “Google ads and the blindspot debate,” Media Cult Soc, vol. 33, no.
3, pp. 433–447, Apr. 2011, doi: 10.1177/0163443710394902.
[9] O. J. Rutz and R. E. Bucklin, “A Model of Individual Keyword
Performance in Paid Search Advertising,” SSRN Electronic Journal, 2007,
doi: 10.2139/ssrn.1024765.
[10] R. Graham, “Google and advertising: digital capitalism in the context of
post-Fordism, the reification of language, and the rise of fake news,”
Palgrave Commun, vol. 3, no. 1, p. 45, Dec. 2017, doi: 10.1057/s41599-017-
0021-4.
[11] H. Kang and M. P. McAllister, “Selling You and Your Clicks: Examining
the Audience Commodification of Google,” tripleC: Communication,
Capitalism & Critique. Open Access Journal for a Global Sustainable Information
Society, vol. 9, no. 2, pp. 141–153, Jun. 2011, doi: 10.31269/triplec.v9i2.255.
[12] M. Raffinot and R. Rivière, “Optimizing Google Shopping Campaigns
Structures With Query-Level Matching,” Aug. 2017.
[13] K. K. Kapoor, Y. K. Dwivedi, and N. C. Piercy, “Pay-per-click advertising:
A literature review,” The Marketing Review, vol. 16, no. 2, pp. 183–202, Sep.
2016, doi: 10.1362/146934716X14636478977557.
[14] E. D’Avanzo, T. Kuflik, and A. Elia, “Online Advertising Using Linguistic
Knowledge,” in Information Technology and Innovation Trends in Organizations,
Heidelberg: Physica-Verlag HD, 2011, pp. 143–150. doi: 10.1007/978-3-
7908-2632-6_17.
[15] S. Y. Lim, “Can Google be Liable for Trademark Infringement? A Look
at the ‘Trademark Use’ Requirement as Applied to Google AdWords,”
UCLA Entertainment Law Review, vol. 14, no. 2, 2007, doi:
10.5070/LR8142027100.
[16] J. Liang, H. Yang, J. Gao, C. Yue, S. Ge, and B. Qu, “MOPSO-Based CNN
for Keyword Selection on Google Ads,” IEEE Access, vol. 7, pp. 125387–
125400, 2019, doi: 10.1109/ACCESS.2019.2937339.
[17] L. Shi and B. Li, “Predict the click-through rate and average cost per click
for keywords using machine learning methodologies,” In Proceedings of the
International Conference on Industrial Engineering and Operations
ManagementDetroit, Michigan, USA, pp. 694–700, 2016.
Journal of Information Systems and Informatics
Vol. 5, No. 3, September 2023
p-ISSN: 2656-5935
http://journal-isi.org/index.php/isi
e-ISSN: 2656-4882
1018 | Investigation of Relationship Between Google Cost-Per-Click and Search-Volume ....
[18] G. Szymanski and P. Lininski, “Model of the Effectiveness of Google
Adwords Advertising Activities,” in 2018 IEEE 13th International Scientific
and Technical Conference on Computer Sciences and Information Technologies (CSIT),
IEEE, Sep. 2018, pp. 98–101. doi: 10.1109/STC-CSIT.2018.8526633.
[19] S. Yang, J. Pancras, and Y. (Amy) Song, “Broad or exact? Search Ad
matching decisions with keyword specificity and position,” Decis Support
Syst, vol. 143, p. 113491, Apr. 2021, doi: 10.1016/j.dss.2021.113491.
[20] A. V Plotnikov, P. A. Kuznetsov, A. A. Urasova, and E. M. Akhmetshin,
“Correlation analysis of the data on the UK and US market for contextual
advertising,” International Journal of Civil Engineering and Technology, vol. 9, no.
11, pp. 1630–1639, 2018.
[21] G. Rich, “An Empirical Analysis of the Auction Bids for Keyword Types
in Sponsored Search,” Northwestern University, 2008.
[22] K. K. K. M. Ramaboa and P. Fish, “Keyword length and matching options
as indicators of search intent in sponsored search,” Inf Process Manag, vol.
54, no. 2, pp. 175–183, Mar. 2018, doi: 10.1016/j.ipm.2017.11.003.
[23] A. Agarwal, K. Hosanagar, and M. D. Smith, “Location, Location,
Location: An Analysis of Profitability of Position in Online Advertising
Markets,” Journal of Marketing Research, vol. 48, no. 6, pp. 1057–1073, Dec.
2011, doi: 10.1509/jmr.08.0468.
[24] A. Ghose and S. Yang, “An Empirical Analysis of Search Engine
Advertising: Sponsored Search in Electronic Markets,” Manage Sci, vol. 55,
no. 10, pp. 1605–1622, Oct. 2009, doi: 10.1287/mnsc.1090.1054.
[25] H. (Alice) Li, P. K. Kannan, S. Viswanathan, and A. Pani, “Attribution
Strategies and Return on Keyword Investment in Paid Search
Advertising,” Marketing Science, vol. 35, no. 6, pp. 831–848, Nov. 2016, doi:
10.1287/mksc.2016.0987.
[26] O. J. Rutz and R. E. Bucklin, “From Generic to Branded: A Model of
Spillover in Paid Search Advertising,” Journal of Marketing Research, vol. 48,
no. 1, pp. 87–102, Feb. 2011, doi: 10.1509/jmkr.48.1.87.
[27] S. Wenning, “Use and E ffect of Google Ads Effect of Google Ads for M
arketing S uccess,” International Journal of Management and Humanities, vol. 6,
no. 3, pp. 23–26, Nov. 2021, doi: 10.35940/ijmh.B1402.116321.
[28] M. Yamamoto, “How can we attract tourists more effectively?,” Journal of
Global Tourism Research, vol. 1, no. 1, pp. 9–12, 2016, doi:
10.37020/jgtr.1.1_9.
Journal of Information Systems and Informatics
Vol. 5, No. 3, September 2023
p-ISSN: 2656-5935
http://journal-isi.org/index.php/isi
e-ISSN: 2656-4882
Ramin Abolghasemi Komleh | 1019
[29] T. Y. Chan, C. Wu, and Y. Xie, “Measuring the Lifetime Value of
Customers Acquired from Google Search Advertising,” Marketing Science,
vol. 30, no. 5, pp. 837–850, Sep. 2011, doi: 10.1287/mksc.1110.0658.
[30] J. Pesonen, “Tourism Marketing in Facebook: Comparing Rural Tourism
SME’s and Larger Tourism Companies in Finland,” in Information and
Communication Technologies in Tourism 2011, Vienna: Springer Vienna, 2011,
pp. 537–546. doi: 10.1007/978-3-7091-0503-0_43.
[31] W. Chang and I. Chang, “The Influences of Humorous Advertising on
Brand Popularity and Advertising Effects in the Tourism Industry,”
Sustainability, vol. 6, no. 12, pp. 9205–9217, Dec. 2014, doi:
10.3390/su6129205.
[32] S. R. Castro, S. C. Silva, and P. Duarte, “Does Digital Marketing really
boost city tourism? Evidences from Porto’s Experience,” European Journal
of Applied Business & Management, vol. 3, no. 3, pp. 84–100, 2017.
[33] L. Law, “24 Best Places to Visit in the United States.”
https://www.planetware.com/usa/best-places-to-visit-in-the-united-
states-us-ny-21.htm (accessed Mar. 30, 2022).
[34] J. Klein, “Chicago predicts strong tourism growth in 2022.”
https://travelweekly.co.uk/news/tour-operators/chicago-predicts-
strong-tourism-growth-in-2022 (accessed Jun. 20, 2022).
[35] W. G. Jacoby, “Loess: a nonparametric, graphical tool for depicting
relationships between variables,” Elect Stud, vol. 19, no. 4, pp. 577–613,
Dec. 2000, doi: 10.1016/S0261-3794(99)00028-1.
