发帖
雷达卡
UG.3
MULTIVARIATEA/B TESTINGOF MOBILE
APPLICATIONS
COPYRIGHT NOTICE
[0001] A portion of the disclosure of this patent document
contains material,which issubject to copyright protection.
The copyright ownerhasno objection to the facsimile repro-
duction by anyone ofthe patent document orthe patent dis-
closure, as itappears in the Patent and Trademark Office
patent files orrecords, but otherwise reserves all copyright
rights whatsoever. The following notice applies to this docu-
ment: Copyright Oc 2014 Thomson Reuters.
CROSS-REFERENCETORELATED
APPLICATION
[0002] This application claims priority to U.S. Provisional
ApplicationNo. 61/874,700filed Sep. 6, 2013, entitled“Mul-
tivariate A/B Testing of Mobile Applications,”the content of
which is incorporated herein in its entirety.
TECHNICAL FIELD
[0003] This disclosure relates generally to conversion test-
ing and conversion rate optimization.More specifically, the
disclosure is directed towards systems and methods for mul-
tivariate testing of mobile applications in order to optimize
outcomes of customer interactions with those applications.
BACKGROUND
[0004] In the United States, the mobile communications
market has become quitecompetitive with the entrance of
new virtual network operators anda diverse array of pricing
options andstrategies (includingprepaid and postpaid
phones,usage-based fees, shared plans, reduced-speed
access, and even handset subsidies, financing and buy-back
programs).Carriers, providers, marketers, developers and the
like arethus constantly seeking new and improved ways to
bring innew users ortohave existing users adoptnew and/or
additional, preferably revenue-generating, services. This
drive will become evenmoreimperative inthe future in view
of the general recognitionthat, at somepoint, the rate of
growth of new users in the mobile space will slow or even
plateau.
[0005]As part of this challenge, carriers, providers, mar-
keters and developers seekto maximizethe impact their ser-
vices and offerings have on current or potential new users. A
common challenge for these carriers, providers and theiruser
interface (UI)designers and developers, is determining which
variation ofadesign or message will be most compelling for
customers. While in-person, qualitative testing might provide
the most accurate metric, adoption of such tests on a large-
scale can be cost and time prohibitive. Alternatively, perfor-
manceof such testing ona smaller scale is typicallyof suchan
insufficient sample size as to be able to drive relevant
enhancements.
[0006]In response to suchchallenges, an increase in web-
basedA/B and multivariatetesting with numerous commer-
cialsolutions for testing direct mail and website configura-
tions has been observed inthe lastseveral years. Today, there
arenumerous commercial products fordirect mail and web
site testing (typically covering test configuration, trial rate,
response measurement and reporting). Although generally
led by marketinggroups asa way of optimizing marketing
spend, multivariate testing isincreasinglybeingused to test
user experience designs.
[0007] As generally known,A/B testing (or split testing) is
a method of website optimizationwhere live traffic is driven
to differing versions ofa web page oremail (version A and
version B) and thenmonitored fortheway the user interacts
with his/her version of the page. The ‘attractiveness’ or con-
version rate (i.e., the proportion ofusers who take an action
that is beyonda casual content view or website visit—e.g.,
click-throughs, registrations, videos watched, successful
completion ofa checkout process) for that componentis then
inferredby response. Analysis of this interaction can help the
web-page owner/developer/designer/marketer determine
which page is more effective. Multivariate testing usesa
similarcore mechanism asA/B testing, butincludesa higher
numberof variables and reveals more information abouthow
eachof these variables interact. Thus, moredetailed informa-
tion is revealed as to the effectiveness of individual compo-
nents or elements of theweb page or emailbeing tested.
[0008] Existing web-based approaches for split and multi-
variate testing, however, often do notprovide support for
mobile web-based applications and, even inthose instances
where they do, significant experimentation and substantial
rework is still neededin order thattheybe made applicable to
a small screen. Further, these web-based solutions are insuf-
ficientlyadaptable for robust testing ofmobile applications
existing inamobile environment.
[0009] Accordingly,a multivariate platform-agnostic test-
ing platform for mobile applications that helps carriers, pro-
viders, marketers, and developers optimize the engagement
of customers, potential customers and other users with these
applications is needed. This, in turn, is expectedto help accel-
erate product development, drive new customer and revenue
growth, increasecustomer retention, and/or furtherthe adop-
tion of new and/or additional revenue-generating products,
services and solutions.
SUMMARY
[0010] The present disclosure is directed towards systems
and methods for multivariate testing. In one aspect, the
method includes, inresponse toa request for one ormore
content items on an access device,identifying one or more
dataitems associated with the access device and identifying
one ofa set of payloads based on theone or more data items
associated with the access device. The identified payload is
thenretrieved and rendered on the access device.
[0011] In one embodiment, the method further comprises
collecting one or more test results in response to aninterac-
tion with the access device. Li another embodiment, the
method furthercomprises determining whether the access
deviceis in an on-line modeoran off-line mode. Wheninan
on-line mode, the set of payloads further comprises one or
more test payloads and the identified payload is retrieved
froma remote server. Bycontrast, whenthe accessdevice is
in an off-line mode, thesetof payloads further comprises one
or more default payloads and the identified default payload is
retrieved from the access device.
[0012] Accordingtoone embodiment, theone or more data
items associated withthe access deviceare collected fromone
ormore sensors of the access device. The one or more data
items associated withtheaccess devicemay comprise at least
one of an access device type, an access device manufacturer,
an access device model,anetwork connection type,a date,a
MULTIVARIATE A B TESTING OF MOBILE APPLICATIONS.pdf
(5.17 MB, 需要: 35 个论坛币)
提升卡
置顶卡
沉默卡
变色卡
抢沙发
千斤顶
显身卡
京公网安备 11010802022788号
