A Systematic Study of the Consistency of Two-Factor Authentication User Journeys on Top-Ranked Websites Extended Version

2025-04-30 0 0 8.17MB 39 页 10玖币
侵权投诉
A Systematic Study of the Consistency of
Two-Factor Authentication User Journeys on
Top-Ranked Websites (Extended Version)
Sanam Ghorbani Lyastani, Michael Backes, Sven Bugiel
CISPA Helmholtz Center for Information Security
Abstract—Heuristics for user experience state that users will
transfer their expectations from one product to another. A lack
of consistency between products can increase users’ cognitive
friction, leading to frustration and rejection. This paper presents
the first systematic study of the external, functional consistency of
two-factor authentication user journeys on top-ranked websites.
We find that these websites implement only a minimal number of
design aspects consistently (e.g., naming and location of settings)
but exhibit mixed design patterns for setup and usage of a second
factor. Moreover, we find that some of the more consistently
realized aspects, such as descriptions of two-factor authentication,
have been described in the literature as problematic and adverse
to user experience. Our results advocate for more general UX
guidelines for 2FA implementers and raise new research questions
about the 2FA user journeys.
I. INTRODUCTION
Would you buy a car where the gas and brake pedals
are interchanged? You would probably be able to learn to
drive this car safely after some acclimatization period. Still,
it would be an experience that is very inconsistent with what
you are used to, and you would most likely not continue
using such an unpleasant car. Like this everyday example,
a consistent user experience is crucial for websites to fit
the mental models that users built and avoid unnecessarily
increasing the users’ cognitive load and friction by forcing
them to learn something new. This important best practice has
been captured in Jakob’s Law of Internet User Experience [41],
[56], [57] as one of several heuristics for user experience [83],
[84] and usability [55] that guide website design. Striving
for consistent user experience has ruled website design for
years, evident in the design of, e.g., online shopping, banking,
forums, blogs, or streaming services. The same best practices
also apply to user authentication as part of the user experience.
When it comes to the incumbent authentication scheme
on the web today, text-based passwords, the user experience
of passwords is highly consistent across different websites,
although recent work [50] discovered inconsistent password
policies for blocklists, strength meters, and composition when
setting passwords on the top websites. Regardless of this incon-
sistency, text-based passwords are notorious for their security
issues. Among the different solutions proposed to strengthen
user authentication on the web, two-factor authentication (2FA)
has been shown to have a very tangible positive effect on
account security [44], [52], [76]. Nowadays, 2FA is frequently
recommended to end users to improve their security hy-
giene [62]. Fortunately, many websites are starting to offer 2FA
options to their users [5], [31]. However, previous work [14],
[65] demonstrated that users struggled with 2FA when their
2FA journey did not match their expectations or previous
experiences and advocated for more standardized procedures.
In a survey with 2FA adopters (see Appendix A), we found
corroborating evidence that inconsistent implementations of
the 2FA user journey caused friction for users that lowered
the usability of 2FA and led users to refuse 2FA or abandon
websites. Unfortunately, up to today, we have only very few
insights about how consistent the user experience of 2FA is
across different websites.
To provide new insights about how websites offer 2FA to
their users and how consistent this user experience is across
websites, we systematically study the 2FA user journeys on
85 popular websites in this paper. More specifically, we want
to determine whether these websites consistently follow the
same design patterns and strategies to offer 2FA to their users.
Or, in other words, we are interested in the external functional
consistency of the 2FA user journeys across popular websites.
To approach our research question systematically, we need
concrete factors based on which we can compare the different
user journeys. Unfortunately, such a list of factors does not
exist for two-factor authentication, and there is no common
guideline or best practice on how to implement the 2FA user
flow on websites. Furthermore, 2FA is a technology that has
only started gaining wider adoption among websites in the
last couple of years and was hence not part of the initial
website design. Additionally, the 2FA ecosystem is fragmented
into various options for 2FA, such as TOTP, WebAuthn, push
notifications, SMS, or custom solutions, each with its own
setup process, dependencies (e.g., hardware token or app),
and benefits/drawbacks in terms of usability and security [10],
[63]. For these reasons, it was not a priori obvious which
exact comparison factors could describe potentially diverse
user journeys on different websites.
To solve this challenge, we devised a methodology to
derive a list of comparison factors from open and axial
coding of existing user journeys on the 85 websites in our
data set. As a result, we created a list of 22 comparison
factors that describe the user journey from discovery of an
offered (promoted) 2FA support during sign-in/registration, to
the education of the user about the available second-factor
options and their setup processes, to usage and deactivation
of the chosen 2FA option(s). Based on those factors, we then
compared the 85 websites to identify common design patterns
and differences and highlight beneficial or detrimental patterns
for user experience.
arXiv:2210.09373v1 [cs.CR] 17 Oct 2022
Our results show that there is no overarching design
pattern for the user journey that most websites follow. Instead,
we found the design space to be clustered into groups of
websites with very similar patterns, some of those favored by
the top websites and others by less popular sites. The only
design aspects that almost all websites agree on about 2FA
are that it is an optional feature, how it should be called
and described, and where it should be found in the account
settings. In contrast, for the crucial steps of setting up and
using 2FA, we found that websites implement mixed strategies,
such as varying numbers of simultaneously supported 2FA
technologies, inconsistent presentation of device remembrance
options, or varying degrees of feedback to users.
According to UX guidelines, this lack of consistency in-
creases users’ cognitive load and should be avoided. However,
consistency alone does not guarantee a good user experience.
We found that several of the more consistently used design pat-
terns have been described in prior work as problematic for user
experience, including non-encouraging descriptions or missing
possibilities to personalize the 2FA. We also discovered that the
journeys of top websites, like icloud.com, are outliers from
the best practices in the academic literature. Therefore, our
results create a call for action to reinvestigate what constitutes
a good overall 2FA user experience, to study whether there
is a “gold standard” for implementing 2FA user journeys, or
to explore the motivations of website developers to implement
specific design patterns.
II. BACKGROUND
A. Two-Factor Authentication
With two-factor authentication enabled on a website, a user
must successfully provide two authentication factors to verify
their identity. Almost always, the first factor is a traditional
text-based password. For the second factor, there are different
technical realizations of knowledge, possession, and inherence
factors. Most common [5], [12] are one-time codes delivered
via SMS text-message, phone call, or TOTP [53] apps, like
Google Authenticator, Duo, or custom apps that the user
registered with the website; push notifications by sending an
alert message to a dedicated app on the user’s phone that asks
the user to confirm a login attempt; and hardware tokens via
the U2F or FIDO2/WebAuthn [82] standards that rely on public
key cryptography and challenge-response protocols.
Each of these comes with its own set of usability and
security benefits and drawbacks [63]. Important for our work
is that a website with 2FA support can offer one or multiple of
those 2FA options, may even allow users to set one of those
solutions up multiple times, or may enforce a particular order
in which they can be set up or used.
A commonly acknowledged problem with two-factor au-
thentication is account recovery when a user loses access to a
factor (e.g., a mobile device with the TOTP app is unavailable).
Often the strategy to avoid lockout from a 2FA-protected
account is to set up a dedicated recovery option, such as
printed-out one-time passwords that can replace another 2FA
option, or to configure multiple 2FA options, when supported
by the website, e.g., multiple hardware security keys.
B. User Experience
Unfortunately, providing an exact definition of “user ex-
perience” is very difficult, as there is no consensus on the
exact definition [8], [40], [46], [59]. However, a common topic
among the definitions is that UX encompasses the various
aspects of user interaction with a product, such as a website.
Cooper et al. [16] note that there exist three overlapping
concerns for UX: form, content, and behavior. While form
and content (e.g., UI design or phrasing) have an impact on
usability, this work focuses on behavior (i.e., functionality) and
only touches on some aspects of form and content.
To help designers provide the best possible user experi-
ence, various best practices and general guidelines have been
developed (e.g., books [16], [43], [70], [77], [84] or online re-
sources, such as Laws of UX [83], Nielsen Norman Group [3],
or Interaction Design Foundation [2]). Among the earliest are
Shneiderman’s eight "Golden Rules" for interface design [69],
[70] and Nielsen’s "10 Usability Heuristics for User Interface
Design" [55], [58]. Shneiderman’s rules state, for instance,
that one should strive for consistency and provide informative
feedback to users. Of Nielsen’s heuristics, heuristic nr. 4, also
known as Jakob’s law of Internet user experience [57], is the
most important for this work and provides the motivation to
study the consistency of 2FA user journeys across websites.
This heuristic states that “users spend most of their time on
other sites” and that “users prefer a site to work the same way
as all the other sites they already know.” As a consequence, one
should “design for patterns for which users are accustomed.”
Having such conventions and consistency helps users build
upon existing mental models and avoid cognitive friction by
forcing them to learn something new [84]. If an unconventional
website mismatches the user’s mental model, the website will
be difficult to learn, difficult to use, or even rejected [77]. One
way to drive external consistency is to make ample use of
guidelines. For instance, for apps there exist Google’s Mate-
rial Design Guidelines [36] and Apple’s Human Interaction
Guidelines [9]. We are not aware of any general guidelines
for implementers and designers of two-factor authentication
on websites, although there are case-specific guidelines (for
example, FIDO2 [29]) or small collections of best practices
(e.g., [23], [75]).
Although in this work we focus on external, functional
consistency, some of the comparison factors for 2FA user
journeys that we identified (see Section VI) also touch on other
UX guidelines and best practices. Tesler’s law [84] states that
for any system there is a certain amount of complexity that
cannot be reduced, and it is recommended that the product
design ensures that as much as possible of the burden on
the user is lifted. Krug [43] recommends that if a difficulty
for the user cannot be avoided, the design should provide
brief and timely guidance, and Cooper et al. [16] recommend
contextual help and assistive interfaces without the need to
break the user’s flow. If it cannot be avoided that the user has
to learn something new, users learn best from examples (e.g.,
pictures, screenshots, or short tutorial videos) [77]. In addition,
Hick’s law [84] recommends breaking down complex tasks
into smaller steps to decrease the cognitive load. Moreover,
excise tasks, such as navigational excise, should be reduced,
e.g., by reducing the number of places that a user must go
and providing clear overviews [16]. Hereby, it is important
2
to consider that users do not read but scan webpages [43]
and that this scanning is based on the mental model they
built from past experiences, which creates expectations of
what they want to see and where [77]. Furthermore, part
of Postel’s law [84], similar to Shneiderman’s third golden
rule [69], [70], recommends providing clear feedback to users,
and the Peak-End Rule [84] recommends paying attention to
the final moments of the user journey because people judge an
experience largely based on how they felt at its peak and recall
negative experiences more vividly than positive ones. Lastly,
personalization can enhance the user experience. Although we
did not explicitly investigate websites for their quality of those
additional guidelines, some of our comparison factors indicate
if 2FA settings are found in common places, if additional
information and instructions are provided, if user notifications
are present, or if users can set preferences.
III. RELATED WORK
Several works have studied two-factor authentication prob-
lems and focused on the usability component and user at-
titudes. Bonneau et al. [10] conducted a systematic expert
assessment of various authentication solutions, including many
of the solutions used for 2FA. They concluded that the usability
of these solutions falls very often short compared to text-based
passwords. In contrast to Bonneau et al., most other works
relied on user studies to investigate problems of 2FA.
A focal point of prior user studies was the setup and usage
of different two-factor authentication solutions to understand
users’ attitudes toward 2FA, obstacles for its adoption, and
how to improve the usability and user experience. Previous
works studied two-factor authentication in settings such as
online banking [38], [42], [78], [79] or military [71] services.
Like other studies of 2FA [11], [21], [22], [27] they found
that users consider 2FA often burdensome and slow, that
convenience trumps perceived security, and that users do
not always understand the risks that 2FA tries to remedy.
Several works have studied 2FA problems in organizational
contexts [7], [15], [25], [64], [72], [74] where the use of MFA
can be mandated. While those studies show that many of the
problems overlap with non-organizational settings, they could
also shed new light on the positive influence of features such as
device remembrance [25], [64] or better help and instructions.
Several studies [42], [63], [78], [79] compared different
options for the second factor to identify option-specific differ-
ences in user attitudes and usability, while other works specif-
ically studied security keys [14], [17], [65] or authenticator
apps [20]. An interesting aspect of those works [14], [63], [65]
for our study is that they differentiated between 2FA setup
and login, where users often struggled in the setup due to
unclear instructions/workflows. Strong recommendations from
those works were clearer instructions and guidance for the
setup to avoid user frustration that often leads to non-adoption.
Additionally, improved notification design patterns [34]
have been shown to encourage users to adopt 2FA.
Lastly, recent works [28], [32], [45], [61] studied specifi-
cally FIDO2 single-factor authentication. They found similar
user concerns as for 2FA. However, the weighting of the
concerns shifted (e.g., loss of the authenticator device is ranked
very high) or new concerns were added (e.g., misunderstand-
ing biometric WebAuthn). Relevant to our work, the FIDO
Alliance has recently published UX guidelines for security
keys [30] and implementers of desktop authenticators [29]
that, similar to our methodology, divide the user journey into
different steps and provide recommendations for the design of
each step; however, explicitly tailored to the technical details
of FIDO2/WebAuthn with biometric authenticator devices or
security keys. Nevertheless, those guidelines incorporate many
of the UX guidelines explained in Section II-B.
The key difference of our work is that we do not study
how concrete changes in form, content, or functionality af-
fect the usability and concrete experience of 2FA, but that
we are first to systematically study how consistent the user
experience is across existing popular websites. Our work, in
contrast to previous works, strongly focuses on Jakob’s law of
Internet user experience which states that an inconsistent user
experience across websites increases cognitive friction and can
be detrimental to users’ adoption. Providing first insights into
how well the 2FA user journeys adhere to this law is the core
contribution of this work. Further, we are not aware of prior
studies that measured Jakob’s law across a larger number of
websites but instead, to the best of our knowledge, qualitative
and quantitative testing of websites focuses on single websites
or comparative user studies between a small set of websites
based on general UX best-practices and guidelines. Therefore,
we had to devise a methodology to measure the consistency
of the 2FA user journeys on different websites.
IV. METHODOLOGY
To compare the 2FA user journeys of different websites
and measure their consistency, we require concrete comparison
factors that describe these journeys. Unfortunately, there is
no existing list of such comparison factors, of which we are
aware, or general guidelines for implementing 2FA on websites
from which we could extract such factors. Therefore, a crucial
challenge for our study is to create a list of relevant and
representative factors. We used inductive research methods
(e.g., [48, Chapter 11.4]) to solve this challenge. Figure 1
gives an overview of our methodology, whose data collec-
tion (Section IV-A) and identification of comparison factors
(Section IV-B) we explain in the following. In a nutshell, we
use open and axial coding from grounded theory on the screen-
recorded 2FA user journeys of different websites to identify the
list of comparison factors and to form an agreement about how
each website matches each factor. Using the coding results, we
then compare the different websites and study how consistently
they implement the 2FA user journey and where they differ
(Sections VI and VII).
A. Data Collection
The first part of our methodology is to collect a represen-
tative data set of user journeys recorded on different websites
that we can analyze. Since we are building our knowledge
about user journeys inductively, the screen recordings must
have as high as possible coverage of all steps and choices
along each journey. To this end, an automated tool, such as
a Web crawler, could be used to explore various websites.
Unfortunately, the need for a priori knowledge about how
websites might implement their user journeys to guide the
3
Literature review
Structure of user journeys:
Discovery\ Education\ Setup\
Usage\ Deactivation
Recording of user journeys
on websites
Coding of comparison
factor and user journeys
Agreement between
researchers
Website comparison and
consistency analysis
code
code
Data Collection Identifying Comparison Factors Analysis
Fig. 1: Overview of our methodology
crawler and the need to use additional authentication devices
(e.g., phone or security key) hamper an automated collection.
Alternatively, we could use a crowd-sourced data collection,
e.g., Amazon Mechanical Turk. Unfortunately, this was not
possible in our setting for ethical reasons. We would need to
ask our participants to use private accounts (or create fake ac-
counts) on different websites and explore security settings for
which they might need to provide a (personal) email address,
phone number, or security key, and risk accidentally locking
themselves out of an account as a result of a misconfiguration.
Instead, two researchers independently explored and
screen-recorded the 2FA user journeys for our study. Their
general instruction was to “thoroughly explore all aspects” of
these journeys. However, this exploration could be informed
beforehand from the literature, which discusses different as-
pects of 2FA user journeys (see also Section III). For example,
recent works (e.g., [19], [34], [45]) and guidelines [29] identify
discovery of 2FA options and user education, different works
studied 2FA setup and login (e.g., [63], [65]) or mandating 2FA
(e.g., [7]), and account recovery is a commonly identified prob-
lem. Based on those insights, we structure the exploration of
user journeys into five steps: The first step is Discovery of 2FA
support on the website. We explore the landing pages, FAQ,
and account registration for information on 2FA and follow
all linked information. To further encourage users, there might
also be nudges and messages about securing the account with
2FA, for which we scan the websites’ interfaces. To use 2FA,
the user must find the corresponding settings in their account
settings, which we explore for the locations and options for
authentication. In the next step, Education, we examine how
a website introduces 2FA and if it gives further explanations,
such as descriptions of how 2FA works and what it offers.
Once the user has decided to use 2FA, they need to Setup
their second factor(s). We explore the workflow of setting
up all supported 2FA options (e.g., TOTP or Security Key).
This exploration includes examining the websites’ instructions,
exploring the different settings choices (e.g., personalization
choices), and feedback from the website on successful setup.
After setting up two-factor authentication, we examine the
Usage of 2FA on the website. We re-login and observe how the
website prompts us to authenticate and whether it provides any
options (e.g., device remembrance), which we explore. Finally,
we explore the 2FA Deactivation procedure in the website
settings and how the website communicates those changes.
For data collection, we maintained identical study con-
ditions. All recordings were made on MacBooks running
macOS 11 in the same network with the latest version of
the Chrome browser when we started our data collection.
Data collection was carried out between 06/2021 and 08/2021.
This fixed setup should minimize the risk [80] of external
factors (e.g., varying geolocation) and possible risk-based
authentication to distort the data.
It is important to note that we focus only on the workflow
for account creation, initial 2FA setup, and 2FA usage. We do
not explore the workflows for account recovery or to change
personal information relevant to 2FA after 2FA setup, such as
a phone number or email address. We consider those follow-
up problems to be studied after we have insights into the
consistency of the fundamental steps that mint the users’ first
impressions about 2FA on a particular website.
B. Identifying Comparison Factors
Since there is no predefined set of factors to compare 2FA
user journeys, we applied emergent coding [48, Chapter 11.4],
in particular open and axial coding from grounded theory, to
identify comparison factors from our recorded user journeys.
These coding techniques are commonly applied in qualitative
data analysis for text content. To still use those established
methods, we treated the screen-recorded journeys like semi-
structured interviews. Semi-structured interviews follow a set
of predetermined questions, but the remaining questions are
made up during the interview based on the interviewee’s
answers. We transferred this idea to our data collection (see
Section IV-A): The exploration of user journeys follows a
set of predetermined questions for discovery over usage to
deactivation but allows the researcher to divert to individually
explore a website in more detail and discover new or unique
aspects of 2FA user journeys. Two researchers separately iter-
ated through the set of recorded user journeys and segmented
the observed journeys into meaningful parts to which they
assigned concepts (i.e., codes). This is followed by axial
coding, where the two researchers combined those concepts
via induction and deduction into categories. For example,
the codes “2FA advertised on the landing page” and “2FA
recommended during account creation” can be combined into
“Promotion of 2FA.” These combined concepts can be used as
comparison factors on all websites. The researchers also noted
whether there exists a functional dependency between factors.
After agreeing on the list of comparison factors, the researchers
discussed how each website matches each comparison factor
(e.g., fully, partially, or not at all). Since the matching of
comparison factors might reveal that the list of factors is
too fine-grained, potentially weighting small differences too
heavily, or too coarse-grained, potentially hiding important
differences, the researchers repeated the axial coding process
until a set of comparison factors and website matching was
4
found to which all involved researchers agreed. The focus of
coding was on the functional aspects of the websites, and less
on the elements of the content or user interface since this study
focuses on the consistency between websites and not rating the
quality of each website’s user journey.
V. DATA SET
To gather a set of websites for our study, we rely on the
open source project 2fa.directory [5], [6] that maintains a list
of websites with 2FA support, which almost 1,000 contribu-
tors currently curate. The websites are assigned to different
categories, such as social, communication, or retail. Since
2fa.directory distinguishes websites at the level of subdomains,
we merged subdomains into their domain when we were aware
that they use the same account for authentication. For example,
drive.google.com, cloud.google.com, and mail.google.com are
in different categories but rely on the same Google account,
while amazon.com and aws.amazon.com have separate ac-
counts. For merged entries, we chose the category we thought
end users most likely knew the domain for (e.g., mail for
google.com). Since we rely on a manual investigation of the
user journeys of each website, we needed to reduce the set
of all websites listed on 2fa.directory to a feasible number.
First, we excluded categories for which we cannot create
an account, for example, almost all websites in the banking
and government categories. Second, we used the Tranco [49],
[73] data set to rank websites according to their popularity.
We selected the top websites from each category, where we
selected the number of websites from each category based
on the category’s weight in the 2fa.directory data set. For
example, there were only four VPN provider websites in
the 2fa.directory set but 45 Gaming websites. This initially
resulted in 120 websites. Unfortunately, we had to exclude 35
websites that we could not study for different reasons, such
as language barriers, geo-restrictions, or the need for financial
expenditures. In the end, we recorded the 2FA user journey on
85 websites with 2FA support from 26 categories.
VI. COMPARISON FACTORS
In this section, we explain the comparison factors that we
identified in our analysis of 85 popular websites following
the methodology of Section IV and describe informally how
we categorize websites according to these factors. We apply
the methodology of Bonneau et al. [10] by categorizing every
website if it matches ( ), partially matches ( , ), or not
matches ( ) a factor. However, in our categorization, some
factors are dependent on other factors, and we denote it ex-
plicitly when a conditional factor’s prerequisite is not fulfilled
( ) and this factor does not apply to a website. Further, in
contrast to Bonneau et al., we do not use the categorization as a
ranking to determine if a website is better than another website,
but we use the categorization to identify patterns in how
websites realize their 2FA user journey and to study whether
websites realize this journey in a consistent way. Although,
for some of the factors described below, this categorization
overlaps with a scale from known best practices to known
poor practices from the literature. We found 22 comparison
factors; 8 are conditional and depend on other factors to
be applicable. Appendix H provides various examples of the
different comparison factors.
A. Factors for Discovery
Promotion: The website promotes its 2FA support in a
clear and obvious way during account creation or immediately
after login (e.g., through a banner, pop-up, or highlighted
message) ( ). If the website does not clearly promote but only
mentions the 2FA support in a way that could be easily missed
by the user (for example, only a quick link in the footer of the
landing page), we categorize this as quasi-promotion ( ). If
the service does not promote its 2FA support and the user has
to discover it themselves (e.g., browsing the settings pages),
we categorize this as not matching ( ).
Non-Optional: The website mandates setting up 2FA for
user accounts ( ). For instance, without setting a 2FA option
up, the account registration cannot be completed; or after
account registration, core functionality and features of the
website are not available to the user until the user sets up
2FA for their account. Otherwise, using 2FA is optional and
not mandatory for the website ( ).
Common-Naming-and-Location: The website denotes its
2FA settings with a commonly used name, and the 2FA settings
are in a commonly used location in the account settings
( ). We identify commonly used names and locations in our
analysis of our selected websites and summarize the results in
Section VII-A.
If either the name ( ) or the location ( ) is uncommon,
we categorize this as quasi-common-naming-and-location.
If the naming and location are uncommon, we categorize
the website as not matching this factor ( ).
B. Factors for Education
Descriptive-Notification: The website briefly describes
what 2FA is in general or why it is important to users. The
description is provided to the user before the user clicks to
enable 2FA ( ), e.g., located together with a notification about
2FA availability or within the settings page; or the description
is only provided after the user starts the 2FA setup process
( ) at which point the user can still abort the setup. If the
website does not present a description of 2FA, we categorize
this website as not matching ( ).
Additional-Information: The website provides more de-
tailed information through a link (e.g. “learn more”) to help
users understand 2FA ( ). If no such information is provided
or the link is broken, the factor does not match ( ).
C. Factors for Setup
Option-Specific-Information: The website provides spe-
cific information about all 2FA options it supports ( ). For
instance, it informs the user that TOTP or Push-notifications
require the installation of an app or that WebAuthn requires
a hardware authenticator. If the website does not provide this
information but directly starts the setup process (e.g., asking
users to scan a QR code or to use a security key without further
explanation), this factor does not match ( ).
Step-Wise-Instructions: The website gives an overview
of the steps involved in setting up a specific 2FA option (e.g.,
linking a device or app, verifying the link, setting a recovery
5
摘要：

ASystematicStudyoftheConsistencyofTwo-FactorAuthenticationUserJourneysonTop-RankedWebsites(ExtendedVersion)SanamGhorbaniLyastani,MichaelBackes,SvenBugielCISPAHelmholtzCenterforInformationSecurityAbstract—Heuristicsforuserexperiencestatethatuserswilltransfertheirexpectationsfromoneproducttoanother.Al...

展开>> 收起<<
A Systematic Study of the Consistency of Two-Factor Authentication User Journeys on Top-Ranked Websites Extended Version.pdf

共39页,预览5页

还剩页未读， 继续阅读

声明：本站为文档C2C交易模式，即用户上传的文档直接被用户下载，本站只是中间服务平台，本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私，请立即通知玖贝云文库，我们立即给予删除！

相关推荐

更多
立即下载
分类：图书资源 价格：10玖币 属性：39 页 大小：8.17MB 格式：PDF 时间：2025-04-30

开通VIP享超值会员特权

  • 多端同步记录
  • 高速下载文档
  • 免费文档工具
  • 分享文档赚钱
  • 每日登录抽奖
  • 优质衍生服务

作者详情

相关内容

更多

热门标签

人际关系 配电装置 动力学连接体 力的合成 高考理综 全宋诗作者索引 公务员考试
/ 39
评分 收藏
分享
立即下载
客服
关注