SHOPIFY-An Interactive Online Shoe Shop System

Abstract

In order to let user experience trustful and efficient online
shoe shopping, SHOPIFY, a system to measure users’ feet
and suggest them shoes that definitely will fit, was
designed. It includes a hardware which is a smart scale
(weight scale), embedded with highly customized mobile or
tablet application solution. The strategy behind this system
is to let users select shoes among a list of shoes that will fit
them. User interviews were conducted, in order to
understand the users’ main issues while shoe online
shopping, their lifestyle and their expectation toward the
system. Based on interviews and analysis, a physical
prototype along with a software prototype were tested, and
evaluated as attractive.

Full text

SHOPIFY- An Interactive Online Shoe Shop System
Elahe Meydani
RWTH Aachen University
elahe.meydani@rwth-aachen.de
elahe.nba2000@gmail.com
Tomris Beril Kurtoglu
RWTH Aachen University
beril.kurtoglu@rwth-aachen.de
t.berilkurt@gmail.com
Lunjie Zhang
RWTH Aachen University
Lunjie.zhang@rwth-aachen.de
lunjiezhang@gmail.com
ABSTRACT
In order to let user experience trustful and efficient online
shoe shopping, SHOPIFY, a system to measure users’ feet
and suggest them shoes that definitely will fit, was
designed. It includes a hardware which is a smart scale
(weight scale), embedded with highly customized mobile or
tablet application solution. The strategy behind this system
is to let users select shoes among a list of shoes that will fit
them. User interviews were conducted, in order to
understand the users’ main issues while shoe online
shopping, their lifestyle and their expectation toward the
system. Based on interviews and analysis, a physical
prototype along with a software prototype were tested, and
evaluated as attractive.
General Terms
Measurement, Documentation, Performance, Design,
Reliability, Experimentation, Human Factors,
Standardization.
Keywords
User Centered Technology Design, HCI, Interactive
Systems, Shopping, Online Shopping, Online Shoe
Shopping, SHOPIFY
1 INTRODUCTION
Online shopping has become one of the most interesting
parts of all internet activities, such that it is now in the third
place after e-mail and web browsing [1]. Shoes are items
that people purchase online these days. The motivations
behind this behavior may be the online sales offers or
variety of items and having a vast variety of choices.
Reaching out to items that are not provided in different
countries is also a good motivation for people in Europe or
America. For many online retailers there are still a lot of
challenges and long ways to reach out to safety zone and
attract vast variety of users [2]. In the online shop system
users’ needs should be paid attention. On the other word
user needs are now the center of study, what they needed
and what we should provide to cover their needs.
Like any other activities, in order to reach out the goal, one
needs to take several steps. The steps of buying shoes via
internet, is slightly different with offline (ordinary)
shopping behavior. There are four general steps. In the first
step customers find out the need or wish to purchase a
product. They refer to internet and start searching for what
they want. Then, in the second step they start to compare
what they’ve seen in different web pages. Third step, they
select one that best fits to their needs, based on size, color,
shipment method, and price. Finally, in the fourth step they
paid with secure transaction methods and posting process
will start. Users may abort their shopping in each of these
steps.
According to our study, there is a high probability that
users cancel their shopping due to reasons like: not finding
the proper size or proper color item is out of stock, item
will not ship to their location or inappropriate payment
method. In order to understand every detail in the process
of online shoe shopping we sat with the users, in two
sessions, while they were trying to buy shoes online. Users
were asked to explain any difficulties and problems that
they face. Also they were asked to explain in each step
what is good and what they prefer to see or to have, that
they were not have it on the time, in the process. Our
attempt was to present what the main concerns and issues
are and to find out the best solution to solve those problems.
At the end of two sessions of observations and recording
the difficulties that users mentioned, we come up with a
result. 16 users out of 20 were complaining in the step that
they have to choose their size. They had problems like:
“they were not certain to choose which size”, “they had
difficulties understanding the size chart” and “they had
difficulties to convert the size unit, for example US
standard to EU standard”. Consequently we were
concentrating on a system providing a method that can
measure user’s feet and compare it to the information that

2
shoe retailers provide for each item. This idea serves the
purpose of providing proper size of shoes for the customers.
To achieve a user oriented design, two important methods,
DIA Cycle and Usability Engineering Process were used
[26]. We combined our ideas with users’ feedback to refine
the system. This was done repeatedly with the cycle
provided by the methods mentioned above. In this paper we
tried to present all the works related to this project. We will
first go through the analysis of design and technology, user
groups and technology requirements to cover the user needs.
All of these will be showed by detailed overview of various
prototypes and user interviews in structured and semi
structured way. Finally we will demonstrate the final
version of our system.
Preliminary Study and Research
In order to understand all the terms and conditions,
standards, measurements and orthopedic issues, we went
through lot of scientific papers. We gained related
knowledge:
Understanding user behaviors with focus on online
shopping and key factors of successful online shopping
[9].
Discovering deficiency on footwear online stores [8].
Standards and guidelines for designing, designing
effort and opportunities [4]
Attributes that contribute to consumers having a satisfying,
high quality online shopping experience [10, 11].
From all gained information we understand that there are
still challenges in order to improve the online shoe
shopping process. Therefore with the focus on solving the
“size” issue that we mentioned in the previous section and
know the design principles and standard term, we started to
prototype and interview.
In the next section, we will describe different systems that
exist in order to measure human feet.
2 RELATED WORK
2.1 Existing Systems for Feet Measurement and
related topics
Here we will describe 5 systems that measure feet with the
use of different methods and technologies:
2.1.1 Yeti
Yeti [12] 3D foot scanner is a precision optical device,
employing eight cameras and four lasers. It is a high speed
and high resolution device that can scan the shape of feet
with very detail parts and with the help of Vorum’s
FootWare [12] system, people can use all data for
designing shoes. YETI Shape Builder (figure 1, left image)
is scanning software that is bundled with the YETI scanner
(figure 1, right image), it rapidly can process all raw data
that came from scanning and create images and
measurement information in multiple views. This system
has advantages and disadvantages. The advantage is that it
is fast and has high accuracy. However it is a designed
system for medical uses and it is not suitable for every day
using.
Figure 1: YETI scanner (right image) can scan foot with laser
and Vorum’s FootWare system (left image) can compute
information and show a 3D result
2.1.2 Podia
Podia [13] laser foot scanner scans the feet with using
multiple points of view. Functions of this system like the
previous one YETI [12]. It allows generating a full 3D
reconstruction of foot and also performing measurement on
it.
Figure 2: Podia laser foot scanner system scans the shape of feet.
2.1.3 MetascanTM
MetascanTM [14] pressure mat system can gather
information by force and pressure from bottom of feet
while walking or just standing on the mat. Each part of
bottom of the foot is contacting the ground, so pressure
analysis will help to get useful information. This method
also is good in measurement of foot’s arch. With the help
of visual analysis and standards in foot calculation, the
system can calculate the size. This system provides the

3
inspiration for us to show foot arch information for the
users in the system that we
implemented.
Figure 3: Left image shows Metascan mat that extracts foot shape
depending on pressure. Right image shows an example of foot
arch.
2.1.4 FootFairy
Footfairy [16] is an iPad application and its method for
measurement is different from YETI [12], Podia [13] and
Metascan[14]. These focus on measuring feet in order to
create orthopedic shoes with having accurate measurements.
FootFairy is a system which only works for children. It
includes an iPad application that can measure foot size by
having only a picture of bottom of the feet. In this system
iPad will act as a scanner, so that it is enough to stick iPad
to the bottom of the foot and wait until the scanning
process is done by showing the “done” message. After
scanning the foot, the algorithm will calculate the foot size.
This is fast and accurate method, but disadvantage is that it
only works for children.
Figure 4:left image shows scanning a child’s foot with the help
of an iPad which works as a scanner. Right image shows the
result after scanning.
2.1.5 RightShoes
RightShoes [17] is a new application running on
smartphones and turn the phone in to a 3d scanner. With
the help of camera on the phone, users take 23 pictures in
circular way in order to compute the feet measurements.
For this purpose user should stand still on a specific sheet
and let someone take those pictures. Then the application
will compute the size. This application connected to a
special shoe shopping stores. However as a major
disadvantage we can say that taking 23 pictures with
correct height and angles is a difficult job and system
guideline for this process is not strong enough and users
need somebody helping for taking the photos and system
does not have ability to provide size information for user
even if the user makes all instruction steps correctly.
Figure 5: RightShoes is a mobile application that use mobile
camera to measure the foot. It only provides detail information
about foot, could not provide the accurate size calculation.
3 DESIGN WITH USER CENTERED STUDY
After studying about potential problems and interviewing
users, we understood that selecting the proper size is the
main issue for users. We started brainstorming with
possible solutions that we could have created. We created
initial scenarios with a common goal on them- to help user
to find the best fit shoes during online shopping-.
According to our brainstorming we created primary
storyboards, which were the mixture of our all ideas related
to the online shoe shopping assistant.
After observing all options, conducting three interviews
(two semi structured, one structured) and two online
surveys with 462 interviewers, three detailed shadowing in
offline shoe shopping and online shoe shopping
environment, talking to shoe assistant and Orthopedic
doctor; in general we came up with five different ideas for
designing a system, which will be discussed in sections 5.1,
5.2, 5.3, 5.4 and 5.5.
For every idea, we went to the users and asked their
opinion with the system and prototype; the main purpose
was to put the users in the center of study with help of
interviewing and shadowing for related idea. Based on
eight interviews and five different prototyping and 57
participants result, we came up with the finalized idea of
“magical scale” to work on further. Complete description
will be on section 4.5.

4
4 DESCRIPTION OF CONTEXT OF USE
Context of use [27] consists of description for user group,
user needs and system requirements.
4.1 User Study and User Groups
To reach out to the two main goals in user centered design:
Ensure that all the factors that related to use of
system are identified before design work starts, in
order to design what user want, not what we think
user wants.
Provide a basis for designing later usability tests.
We need to analyze the context of use. It means that we
should collect and analyze information about our users
precisely, what they want to do, and their technical and
environmental situations and constraints. We gathered our
data with help of interview, workshop, survey, site visit,
observational study and focus group. In next section we
present our user groups and process of finding out the main
user groups and their characteristics.
4.2 Identifying User Groups
In order to find out best characteristics of system, user
groups need to be described. According to [21] [22], we
tried to find out our potential users and tried to focus on
ones who are our main users. In this way, we first
brainstormed a list of possible users. Then we tried to
describe the main characteristics of each. And according to
these characteristics we grouped them.
4.2.1 Users and Their Main Characteristics
After interviewing and studying the market, we decided to
identify the main characteristics of users. With knowing
these characteristics and attributes, separating user groups
became much helpful.
Our user groups are categorized in Table 1:
Label
Task
Knowledge
Physical
Environment
Preferences,
Disliking, Desires
Advance
online
shoppers
-Have a lot of
experience in
online shopping.
-Know a lot of
online retailers.
-Familiar with
all kind of
payment method
and return
policy.
-Know about
material (based
on experience).
-Home (on the
sofa,chair,
ground, in the
living room)
-Office
-Prefer fast and free
deliveries.
-Prefer pictures with
more detail.
-Prefer more
description about
item
Ordinary
Online
Shoppers
-Sometimes shop
online based on
need or curiosity
-Return to
specific retailers
each time.
-Less knowledge
about material
and size
selection
Home ( in the
living room,
behind the
desk, in the
private room)
-Prefer free returns.
-Prefer pictures
with more details
-Prefer size
measurement help
or feedbacks.
-Prefer common
payment options
Offline
Shoppers
Do not shop
online .
-No trust to
online retailers.
- Less dare to
try.
Home
-Prefer to shop
outside.
- Like walking and
touching stuff
Table 1- describes user groups and their characteristics
4.2.2 Main User Groups
According to our interviews and result analysis, we
prioritized our user groups as showed in table 2:
User Group
Percentage based on
interviews(out of 45 users)
priority
Advanced online
shopper
30%
1
Ordinary Online
Shopper
47%
1
Off line shopper
23%
-
Table 2- describes the prioritized user groups.
The user groups Offline shoppers who do not prefer to shop
online are not in our interest. We were going to focus
mainly on advanced online shoppers and ordinary online
shoppers which present users that either do online shopping
regularly or do it randomly. These two would be our focus
group.

5
4.3 User needs
In the following the user needs are presented which are based on
the user interviews:
1. As an experienced online shopper, the user needs
to know other people’s experience or comments in
order to make up her minds.
2. As an experienced online shopper, the user needs a
suggestion system depending on his/her previous
choices in order to tell the user what s/he could
chose.
3. As an experience online shopper, the user needs a
fast and easy shopping process in order to prevent
confusion.
4. As an ordinary online shopper, the user needs a
3D view of item in order to observe item in detail
as user does in reality.
5. As an ordinary online shopper, the user needs a
high quality zoom in feature in order to observe
items quality.
6. As an ordinary online shopper, the user needs to
have a measurement tool, in order to buy the best
fit shoes.
4.4 System Requirements
4.4.1 Functional Requirements
Here are some of the functional requirements:
1. The system should have a list of shoe types.
2. The system should supply the user with
measurement of her/his feet.
3. The system should supply the user with feedback
of her/his feet size.
4. The system should supply the user with different
unit to show the size.
5. The system should supply a standard comparison
between user’s measurement and shoe
measurement.
6. The system should supply user with list of shoes
which fit better.
7. For each shoe type there should be a picture
describes it.
8. The system should supply the user with a list of
shops.
9. The system should supply the user with a ranking
by average price in an increasing way.
10. The system should supply the user with a ranking
by average price in a decreasing way.
11. The list should show the average price of each shop.
12. The system should supply the user with a high
resolution model for each shoe.
13. The system should supply the user with possibility
to rotate the shoe model.
14. The system should supply users with possibility to
read others comment about each item.
15. System will show a demo after first time
installation.
16. System should supply user to write comment for
each item.
4.4.2 Non Functional Requirements
Here are some of the non-functional requirements:
17. System should have quick calculation (less than 45
seconds) for feet measurement
18. The system should supply the user to write their
comments <500 words
19. The system should supply the user readable font
and font size for comment.
4.4.3 Description of some requirements
1. Based on our online survey 80% users (out of 44
users) would like to see other people’s opinion
about an item, in order to see the popularity of item
or feedbacks of using the product. Requirements
14 and 19 are result of this observation.
On the other hands, 65% of users would like to
have this chance to share their experience with
others as well. The result is in requirements 16 and
18.
2. Requirements 2, 5 and 6 are core points of our job
here. Actually a certain size of a specific model
(e.g. 42) has no geometrical relationship with the
number 42 of another model of different
manufacturer. That is the reason why we select the
strategy that contains two main factors: first know
the internal volume and format of each footwear
and sizes established by manufacturer; and second
know exact volume of the right feet and left feet of

6
user, which has been derived by calculations on
measurements.
Aim here is not to represent list of any shoes, but
to represent list of shoes that will fit best,
according to comparison between computed feet
measurements and shoes volume information, as
described in requirements 5 and 6.
The complete list of user needs and system requirements is
available in BSCW with Volere format [28].
4.5 Technology determination
To decide whether to go around technological solutions, is
one issue and to decide what technology should be used is
another. We tried to search for characteristics that make
input methods simpler.
Simple interactions between user and system
Precision on result
Availability
According to our understandings, these factors are most
important. However the fact that precision is highlighted
among others is undeniable, since our system is a
measurement and comparison base system. Based on what
we got through our research on potential option on
technology, we would like to divide this in two main parts.
Hardware part: the part that normally is the
interaction between user and system.
Software part: the part that all calculations and
comparisons take place like algorithmic or
programming parts.
4.5.1 Technology in Use
Based on section 2.1 and more other examples, there are
exists different technology items and gadgets that can help
to gain the main goal. There are advantages and
disadvantages for each of them. 3D scanning apparently is
most used in medical field, but since it is much expensive,
we do not see it around a lot. This technology is used for
people who have disease like Diabetes. On the other hand
the fact that this kind of technology will have more precise
result is obvious.
According to our survey, 79% of internet users (out of 120
interviewers) either agree (49%) or strongly agree (23%)
with the fact that online shopping is convenient for them.
That is the reason that market owners (classic shop owners)
accept the fact that they should have something new to
attract more customers in market. Since users are eager to
use technology in their everyday life, market owners show
more enthusiasm to choose a solution in technology
innovation field and add it to their own business. However,
according to our observation in the shoe market, there are
still empty spaces to solving the size issue. On the other
hand, great manufacturers in shoe field tried to attract
user’s attention on the product itself [4] and tried to enchant
them to buy the product by conducting technology
solutions like converting shoe picture in 3D view with help
of tools like www.web3d.co.uk or recording a video for
each item.
On the other hands, other famous shoe shops like
Deichmann [29] has a visually well designed web site and
high resolution pictures for every item. Deichmann tried to
conduct mobile phone technology and its camera as tools to
measure the users feet size, near to solution mentioned in
section 2.1.5. According to our study and interviews; 85%
(out of 120 users) of users use or like to use technology in
their everyday life, but they think that using a camera on
phone cannot give them a precise result. To wrap up the
hardware part, 3D scanning, lasers scanning, cameras,
pressure mat, heat mat are tools that can help us to reach
out the goal.
The core functionality for all of those systems is to provide
an algorithm that can compute wanted information and
transfer them to the mobile or tablet and show it to the user
in a fast and accurate way. Using professional
programming skills with well-known programming
languages based on a low time computed algorithm, using
network knowledge for communication part and ability to
access to databases and extract useful information out of
them and the ability to compare feet information to the
shoes information according to ISO 9407:1991 standard are
the main software parts.
5 PROTOTYPING & IMPLEMENTATION
After identifying who our users are, what they need and
what they want to do with system and what system should
provide for them, we started the creativity path. During the
course of our project we came up with different ideas and

7
each time we prototype them with help of storyboard, paper
prototype, video prototype and tried to distinguish the
weakness and strength of each methods in order to produce
one system which has the lowest weakness and highest
strengths.
We came up with 5 different ideas. We prototype them, and
tested those with our focus user groups. Then we gathered
information by user study like:
what was good about the system.
what was not good about the system.
why users want or don’t want to have such system.
Here is the list of our different ideas.
•3D Scan Booth
•Paper, Ruler, Pen (PRP)
•Shopify mobile application
•Klug Strümpfe (Smart Socks)
•Magical Scale
After naming these ideas, we tend to explain them, and
show the analysis of users’ feedback, and the reason why
we discard it or keep it.
5.1 3D Scan Booth
3D scan booth was our preliminary idea inspired by
FotoBooth that are well-known in malls or in railway
stations. The core idea was to have a 3D scan booth in
every big and crowded malls or shopping centers in order
to provide foot scanning feature for users who tend to shop
online. So the procedure was like every user takes off their
shoes and socks and puts her/his right foot and left foot in
the scanner and waits till the device scanned her/his feet.
Then the users could use a USB device or by using the
cloud technology, transfers feet information and has it
during online shopping.
We interviewed users using storyboard to see what they
think about this system. After we interviewed 7 users, we
discarded this idea according to following reasons:
•During offline shopping, people usually will not
think about online shopping.
•It is inconvenient to take off the shoes and socks
in a public place.
•Malls and shopping centers are places which
normally people do off line shopping
However all of seven interviewers were unanimous on high
precision on this method, five out of seven confess that if
they see such a device, they won’t see the need to try it.
After the analysis, we came along with two results: first we
need to solve the problem during the process of online
shopping not before it and second think about tools that are
much simpler and much more available for everyone.
5.2 PRP
After being technological about finding a solution for
sizing problem, we decided to go back to classical method.
Then we decided to observe our user groups to work with
Pen Ruler Paper (PRP). The instruction was easy, to put the
feet on the blank white paper and round the feet on paper
and with help of ruler they should measure the length and
the width of their foot and enter the data to the computer,
for computing the size.
Figure 6: PRP method measure the user feet.
We interviewed the users using prototype to see what they
think about this method. After we interviewed 10 users, we
discarded this idea according to following reasons:
• Using old fashion methods are a little complex
nowadays.
• Since the main purpose is online shopping, it does
not satisfying to switch to mechanical methods
and switch back to compute again.
• Mistakes could be happened when users tend to
enter data to the computer, like entering width
information in length field instead of the width
field.
5.3 Shopify Mobile Application
Then we came up with a plan to create an application
which employs mobile or tablet camera to measure feet
information. Instruction was easy, with wearing black

8
socks, users should put their foot on a blank A4 size paper
and took pictures. For convenient in taking the pictures, we
came up with a grid that could turn to green if user took the
camera in correct way and turn to red otherwise. We
interviewed six users with paper prototype. Surprisingly
users like the system because they did not need to do
anything extra. Being simple in use was the comment that
all the six users mentioned. On the other hand two main
important issues came along, first the precision could have
been different according to precision on camera in each
different phone and the risk of getting inaccurate
measurement results appeared. Also, the time of
computations in devices like mobile phones which
obviously do not have high speed processors like the ones
PCs have, could be long.
So we put this thought on hold, and started to think in
creative way.
5.4 Klug Strümpfe (Smart Socks)
After deciding to go beyond and searching more, we find
out the sensory textile. Textiles that have sensors on them
can capture pressure. So we invented the idea of socks
which used these sensory textiles, in order to give the shape
of the foot of the person who wears them. Users just put the
socks on and wait until the sensors measure their feet and
transfer the data to their adapter or connector and then users
could connect the adapter to the computer.
We created a simple paper prototype and using a pair of
socks. In the test of the system, we used the wizard of Oz
method, for showing that these socks are made out of
sensory textile. We observed three users working with this
system. Users were happy about the precision of this
system alongside with its simplicity and ability to share
between other people who have nearly same foot size and
shape, (ex: brothers with shoe size 42 and 44). On the other
hand users asked the importance of socks size which are
used in measurement.
So we had two solutions that users liked: Shopify mobile
application and smart socks. In order to be reasonable to
select one over the other, we went to the users again to ask
which one they prefer. With the help of five Whys method
[23] [24], we dug in to the root. We interviewed 39 users in
a semi structured method, to see which method we can
discard. 20 users like smart sock methods based on root
reasons like being more precise and usable, 19 users like
mobile phone based on the root reasons like mobile is an
everyday life thing and pretty much everyone has it, and
people know how to worked with it or reasons like no need
having extra device. Since we were not satisfied, we
decided to talk with people who are more professional in
this area. Four shoe shop assistants and one orthopedic
doctor, they were all unanimous on precision is most
important thing here, since everybody knows foot are
important part of human body, if you can not provide
something that will not hurt the health, sooner or later, your
system will be discarded automatically. Orthopedics doctor,
Nuri Erel, proved that using smart socks are more accurate.
He measures his patience’s feet by using the pattern which
made of pulp pad and this pattern is used especially for the
elderly and very elderly amputees. The Pedilan-Axis Foot
1G9 [25] is particularly suitable for wearers who want a
damped motion of the foot at the ankle. These feet are 23 to
27 inches in sizes and can be used up to a weight of 75
kilograms. The reason of his like for the idea of smart socks
is getting the exact results at the end of the measuring
process. The functionality of the socks and pattern work
same, he thinks.
Since the interview results (19 to 20) are not statistically
meaningful, we thought over the reason of the users’ choice.
Something was repeating in the back of our minds, and that
was the main reason why users preferred mobile
application on socks. The main reason was that, mobile is
an everyday life thing, nearly everyone have it and there is
no need to have an extra tool. On the other hand users
preferred socks on mobile phone, because taking pictures
gives sometimes inaccurate results. So a question came to
our mind which led us to our final idea. The question was:
“Why not design a measurement tool for an everyday thing
with accurate results”.
5.5 Magical Scale
When we decided to find an everyday thing which has
contact with foot, a scale came into our sight. If there are
pressure sensors on the surface of a scale, when the user
stands on it, not only size and shape of feet can be
measured, but also foot pressure distribution can be showed.
To confirm a scale is an everyday thing and to understand
the user behaviors of using a scale we made a questionnaire
with 370 users. According to our survey, 314 out of 370
users have a scale at home and for those 56 people who do

9
not have a scale at home, 39 of them want a scale at home.
Over 95% users think that there should be a scale at home
which shows that a scale is an everyday thing. 286 out of
353 who have or want a scale at home (81%) use a scale
with barefoot. For the measurement of foot size it is
necessary with barefoot. Since most of the participants use
a scale without shoes or socks, this result shows that a scale
can measure foot size without changing users’ behavior.
315 out of 353 (89%) users like a digital scale more than an
analog scale. It is more natural to design more function for
a digital thing than an analogical thing and most of the
users accept a digital scale. This fact provides possibility of
the implementation of foot size measurement with a scale.
258 out of 353 (73%) users like a scale with additional
functions, which means that adding foot size measurement
functionality will not make the users lose their willingness
to have a scale.
To perfect this idea, we made several iterations of
prototypes:
0.0.1. Prototyping cycle 1
We created a scenario and made a video as a low-fidelity
prototype which was evaluated by eight users. The main
problems we got from them were:
How could the scale communicate with smart
devices (smart phone/tablet)?
It is better to measure the height of feet as well.
How long will it take to show the result?
According to the feedback, we decided to add a 3D scanner
in the scale. Since the information we need to get from the
3D scanner is only height of instep. Also it is similar 3d
scanner method that we explain in section 2.1.1 but the user
could not feel it because the everyday scale hides the idea
of scanning and it does not disturb our users
0.0.2. Prototyping cycle 2
We created paper prototype [31] and interviewed, semi-structured
way, with six users (three for each user groups.) by using the
thinking aloud method in order to get feedback from our user
groups. We documented all feedback [32].
0.0.3. Prototyping cycle 3
We created the software prototype version 1.0 [33] and
interviewed, semi-structured way, with five users (four of
them are advance online shoppers and one of them is
ordinary online shopper.) by using the thinking aloud
method. We gave series of tasks [34] to the users that help
users to interact the system and make comment on that. We
documented all the feedback.[35]
0.0.4. Prototyping cycle 4
Depending on our user groups’ and our assistant Svetlana
Matiouk’s feedback we created the software prototype
version 1.1 [36], we eliminated all critical incidents that
violate Usability Requirement ISO 9241-110 and
documented all [37].
0.0.5. Prototyping cycle 5
Finally we created software prototype version 1.2 [38]. In
this iteration we interviewed 10 users (5 from each user
group) with thinking aloud method
As a consequence Magical Scale is the every home tool
(different from 5.1), easy to work with (different from 5.3)
and can be used from every family member (different from
5.4).
6 STUDY OF SYSTEM USABILITY
6.1 Attrakdiff Result
Figure 7: Average values of pragmatic quality and hedonic
quality, as well as the confidence rectangle for the system
'SHOPIFY'.
20 of our users, 10 in “Advanced online shopper group”
and 10 in “Ordinary online shopper group” were asked to
fill out an Attrakdiff questionnaire [30]. The whole result
document can be found in [20]. Considering the portfolio

10
result which shows in the figure 7, we can see that the
medium value of the dimensions hedonic and pragmatic
quality is located between the character-regions 'self-
oriented' and 'task-oriented', and between 'desired' and
'neutral'. The rectangle of confidence around it has a
relatively small size which indicates that the users' answers
were quite similar. As the confidence interval is situated in
two character zones in both the hedonic as well as in the
pragmatic quality dimensions, there is room for
improvement concerning usability and quality.
In figure 8, a detailed diagram of results can be seen, in
which the word pairs are the ones that had been asked in to
evaluate the Attrakdiff questionnaire. All average values
range between 0 and 2, except pairs ‘cheap- premium’
which is between -1 and 0. However, the attractiveness
word pairs have values between 1 and 2. The best averages
were achieved for the word pair ‘bad – goo’ in the
attractiveness category (ATT), with the results clearly
leaning toward the second word of the pairs which is
‘good’. The worst result was obtained for the word pair
‘cheap – premium’ from the hedonic quality - identity
category (HQ-I), with the tendency closer to -1.
In Figure 9, the average values for all product dimensions
can be seen. Values range between just under 1 and less
than 2, with the value for hedonic quality concerning
identity being the smallest around 0.7 and the value for
attractiveness with a value of around 1.5 being the highest.
Hedonic quality concerning stimulations (HQ-S) is around
0.9 and Pragmatic quality (PQ) has an average value near
to 1.1. According to the result of HQ-I, there still room for
improvement to help the user work better with the system.
The result of HQ-S indicates that improvement can be done
toward motivating and captivating the users. However, the
attractiveness of the SHOPIFY is above average.
7 CONCLUSION AND FUTURE WORK
In conclusion, the SHOPIFY system was introduced. It
links a scale that can measure feet information based on the
pressure to an application that can gather information and
compare it to information that retailers provided. Before
designing the system, some field research was introduced
in the section 2. The context of use was analyzed using
ethnographic methods. After understanding lifestyle of
users, we divided them in two main user groups as
‘advanced online shopper’ and ‘ordinary online shopper’.
Interviews were conducted with focus group to understand
user need on which the requirements are based. With
several DIA cycles, the final prototype was designed. And
again with using ‘thinking aloud’ method we improve our
final prototype after five iterations.
Further user observation in the end of the project lead us to
a new idea of further developments. Since Attrakdiff results
showed that the idea behind the system was liked and that
the system would therefore be attractive to users, we can
implement this in the future for the famous shoe companies
and therefore we encourage the ordinary online shoppers to
shop more frequently shoe in online.
Figure 8: Average values of all word pairs from the
Attrakdiff questionnaire for the system ‘SHOPIFY’
Figure 9:Average values of all Attrakdiff dimensions for
the system ‘SHOPIFY’

11
8 ACKNOWLEDGMENTS
We would like to thank the User-Centered Technology De-
sign team at Fraunhofer Institute for Applied Information
Technology for giving us the opportunity to work on such
an interesting project topic.
And our special thanks to all the users who participated in
all the testing sessions and gave us invaluable feedback that
helped us to improve our system.
9 REFERENCES
[1] Na Li and Ping Zhang. Consumer online shopping
attitudes and behavior: an assessment of research.
University of Syracuse, 2002
[2] Arvind Rangaswamy and Gerrit Bruggen. Opportunities
and challenges in multichannel marketing: an introduction
to the special issue. March 29,2005
DOI: 10.1002/dir.20037
[4] The augmented reality shoe store (21 July 2012).
Retrieved Jan 6, 2015 from
http://www.digitalbuzzblog.com/goertz-augmented-reality-virtual-
shoe-fitting-store-installation/
[6] Georgios Kotsolios. Feb 12 2014. Back to the future of
marketing.(2nd ed.) Partridge publishing
[8] Michael J. Cassidy. Footwear Identification, Canadian
Government Publishing Centre, Ottawa 1980
[9] Moez Limayem, Mohamed Khalfia and Anissa Frini.
What makes consumers buy from internet? A longitudinal
study of online shopping, July 2000
[10] Sonja Grabner. The role of consumer’s trust in online
shopping. Journal of business ethics, 2002
[11] Mary Gilly and Mary Wolfinbarger. Shopping online
for freedom, control and fun. University of California, June
2000.
[12] Retrieved Jan 4, 2015 from www.vorum.com
[13] Retrieved Jan 4, 2015 from www.stt-systems.com/foot
[14] Retrieved Jan 4, 2015 from www.painfreefeet.ca
[16] Retrieved Jan 4, 2015 from www.footfairy.com
[17] Retrieved Jan 4, 2015 from http://rightshoes.ch/
[20] T.Beril Kurtoglu, Elahe Meydani ,and Lunjie Zhang,
(Jan 29, 2015). Shopify-ATTRAKDIF-results. Retrieved
from
http://fit-bscw.fit.fraunhofer.de/bscw/bscw.cgi/43746088
[21] Clayton Lewis and John Rieman.1994. Task-Centered
User Interface Design: A Practical Introduction. Retrieved
Jan 5, 2015 from
http://hcibib.org/tcuid/tcuid.pdf
[22] Amanda and Bill foundation. May 2007.Tips for
conducting focus groups. Retrieved Jan 7, 2015 from
http://www.insites.org/CLIP_v1_site/downloads/PDFs/Tips
FocusGrps.4D.8-07.pdf
[23] Mike Sondalini.2014. Understanding how to use the 5-
whys for root cause analysis. Retrieved Jan 7, 2015 from
http://qualitycommunityschools.weebly.com/uploads/4/1/6/
1/41611/how_to_use_the_5-
whys_for_root_cause_analysis.pdf
[24] Retrieved Jan 4, 2015 from
http://en.wikipedia.org/wiki/5_Whys
[25] Retrieved Dec 27, 2014 from http://www.ottobock.de
[26]Erion Elmasllari,Svetlana Matiouk (Oct 13, 2014).
Lecture 1- User Centered Process. [PowerPoint slides].
Retrieved from
http://fit-bscw.fit.fraunhofer.de/bscw/bscw.cgi/43193847
[27]Erion Elmasllari, Svetlana Matiouk (Dec 4, 2014).
Lecture 6 -Participatory design. [PowerPoint slides].
Retrieved from
http://fit-bscw.fit.fraunhofer.de/bscw/bscw.cgi/43193847
[28] T.Beril Kurtoglu, Elahe Meydani, and Lunjie Zhang,
(Feb 4, 2015). Requirement Volere v1.32374. Retrieved
from
http://fit-bscw.fit.fraunhofer.de/bscw/bscw.cgi/43757630
[29] Retrieved Jan 4, 2015from
http://www.deichmann.com/DE/de/shop/welcome.html
[30] Marc Hassenzahl, Dr. Micheal Burmester, and Franz
Koller. Mensch&Computer 2003: AttrakDiff:
EinFragebogen zur Messung wahrgenommener
hedonischer und pragmatischer Qualität, 2003, pages187-
196.
DOI: 10.1007/978-3-322-80058-9
[31] T.Beril Kurtoglu, Elahe Meydani, and Lunjie Zhang,
(Feb 5, 2015). Prototype1. Retrieved from
http://fitbscw.fit.fraunhofer.de/bscw/bscw.cgi/43761048
[32] T.Beril Kurtoglu, Elahe Meydani, and Lunjie Zhang,
(Feb 5, 2015). Interview-result. Retrieved from
http://fit-bscw.fit.fraunhofer.de/bscw/bscw.cgi/43761067
[33] T.Beril Kurtoglu, Elahe Meydani, and Lunjie Zhang,
(Feb 5, 2015). Generate. Retrieved from
http://fit-bscw.fit.fraunhofer.de/bscw/bscw.cgi/43761082
[34] T.Beril Kurtoglu, Elahe Meydani, and Lunjie Zhang,
(Feb 5, 2015). Experimentprotocol-3. Retrieved from

12
http://fit-bscw.fit.fraunhofer.de/bscw/bscw.cgi/43761082
[35] T.Beril Kurtoglu, Elahe Meydani, and Lunjie Zhang,
(Jan 29, 2015). Week13 (pages 9-17). Retrieved from
http://fit-bscw.fit.fraunhofer.de/bscw/bscw.cgi/43746099
[36] T.Beril Kurtoglu, Elahe Meydani ,and Lunjie Zhang,
(Feb 5, 2015). Generate. Retrieved from
http://fit-bscw.fit.fraunhofer.de/bscw/bscw.cgi/43761123
[37] T.Beril Kurtoglu, Elahe Meydani, and Lunjie Zhang,
(Feb 5, 2015). Feedback P3. Retrieved from
http://fit-bscw.fit.fraunhofer.de/bscw/bscw.cgi/43761123
[38] T.Beril Kurtoglu, Elahe Meydani, and Lunjie Zhang,
(Feb 5, 2015). Generate. Retrieved from
http://fit-bscw.fit.fraunhofer.de/bscw/bscw.cgi/437615