Techniques for identifying issues related to digital interactions on websites
Techniques are described herein for identifying issues related to digital interactions. For example, a detection system may be provided to analyze suspect sessions to determine if one or more stages associated with the suspect sessions are underperforming compared to corresponding stages associated with similar sessions. The detection system may provide a user interface that allows a user to select one or more attributes that may be associated with one or more sessions. Selection of the one or more attributes may identify multiple sessions (referred to as suspect sessions herein). The one or more suspect sessions may be analyzed to determine whether one or more stages associated with the one or more suspect sessions are underperforming compared to corresponding stages associated with one or more other sessions determined to be similar to the one or more suspect sessions.
1 . A method for assessing one or more webpages of a website, the method comprising:
obtaining, by a detection system, session information for each of a plurality of previous sessions on the website, the session information for a particular previous session including session attributes of the particular previous session, each previous session being associated with one or more stages of the website, the one or more stages having been entered into during the previous session;
identifying, by the detection system, a plurality of suspect sessions from the plurality of previous sessions based on determining that the previous session includes a suspect attribute;
determining, by the detection system, a set of one or more stages associated with at least one of the plurality of suspect sessions;
computing, by the detection system based on the plurality of suspect sessions, a suspect conversion rate for the set of one or more stages;
determining, by the detection system, a correlation between the suspect attribute and suspect sessions being abandoned based on the suspect conversion rate and an expected conversion rate that are computed for the set of one or more stages; and
providing, by the detection system, the suspect attribute and the correlation for determining a likelihood of the suspect attribute causing the suspect sessions to be abandoned.
2 . The method of claim 1 , wherein the suspect attribute indicates a hindrance to users of the website.
3 . The method of claim 1 , further comprising:
automatically inserting a marker into a suspect session based on the suspect attribute and the correlation.
4 . The method of claim 1 , wherein the expected conversion rate is determined using all sessions occurring on the website.
5 . The method of claim 1 , wherein the suspect attribute is selected from a group consisting of:
repetitive clicks from an input device, random key inputs from the input device, use of a coupon, slow response time, a user repeating a stage where a majority of users of the website only interact with the stage once, a specific error message being displayed, a failed login, a JavaScript error, and a connection speed.
6 . The method of claim 1 , wherein the suspect conversion rate for the set of one or more stages is computed by dividing a number of sessions that progress past the set of one or more stages by a total number of sessions including the set of one or more stages.
7 . The method of claim 1 , wherein the expected conversion rate for the set of one or more stages is computed based upon a total number of a plurality of non-suspect sessions and a number of the plurality of non-suspect sessions that included the set of one or more stages, wherein the plurality of non-suspect sessions do not include the suspect attribute.
8 . The method of claim 1 , wherein the expected conversion rate for the set of one or more stages includes a percentage of users that progress past the set of one or more stages.
9 . The method of claim 1 , further comprising:
presenting, by the detection system, a replay of a suspect session with a timeline having indicators of events associated with the suspect attribute that are correlated with the suspect sessions being abandoned.
10 . The method of claim 1 , further comprising:
determining, by the detection system, a drop-off value for the set of one or more stages, wherein the drop-off value is a number of the plurality of suspect sessions that terminated in a stage of the set of one or more stages;
computing, by the detection system, a missed-conversion value for the set of one or more stages, wherein the missed-conversion value is computed based upon an average conversion value, an under-conversion rate, and the drop-off value; and
providing, by the detection system, the missed-conversion value for determining the likelihood of the suspect attribute causing the suspect sessions to be abandoned.
11 . A system comprising:
one or more processors; and
a non-transitory computer-readable medium including instructions that, when executed by the one or more processors, cause the one or more processors to perform operations for assessing one or more webpages of a website, the operations including:
obtaining, by a detection system, session information for each of a plurality of previous sessions on the website, the session information for a particular previous session including session attributes of the particular previous session, each previous session being associated with one or more stages of the website, the one or more stages having been entered into during the previous session;
identifying, by the detection system, a plurality of suspect sessions from the plurality of previous sessions based on determining that the previous session includes a suspect attribute;
determining, by the detection system, a set of one or more stages associated with at least one of the plurality of suspect sessions;
computing, by the detection system based on the plurality of suspect sessions, a suspect conversion rate for the set of one or more stages;
determining, by the detection system, a correlation between the suspect attribute and suspect sessions being abandoned based on the suspect conversion rate and an expected conversion rate that are computed for the set of one or more stages; and
providing, by the detection system, the suspect attribute and the correlation for determining a likelihood of the suspect attribute causing the suspect sessions to be abandoned.
12 . The system of claim 11 , wherein the suspect attribute indicates a hindrance to users of the website.
13 . The system of claim 11 , wherein the operations further include:
automatically inserting a marker into a suspect session based on the suspect attribute and the correlation.
14 . The system of claim 11 , wherein the expected conversion rate is determined using all sessions occurring on the website.
15 . The system of claim 11 , wherein the suspect attribute is selected from a group consisting of:
repetitive clicks from an input device, random key inputs from the input device, use of a coupon, slow response time, a user repeating a stage where a majority of users of the website only interact with the stage once, a specific error message being displayed, a failed login, a JavaScript error, and a connection speed.
16 . The system of claim 11 , wherein the suspect conversion rate for the set of one or more stages is computed by dividing a number of sessions that progress past the set of one or more stages by a total number of sessions including the set of one or more stages.
17 . The system of claim 11 , wherein the expected conversion rate for the set of one or more stages is computed based upon a total number of a plurality of non-suspect sessions and a number of the plurality of non-suspect sessions that included the set of one or more stages, wherein the plurality of non-suspect sessions do not include the suspect attribute.
18 . The system of claim 11 , wherein the expected conversion rate for the set of one or more stages includes a percentage of users that progress past the set of one or more stages.
19 . The system of claim 11 , wherein the operations further include:
presenting, by the detection system, a replay of a suspect session with a timeline having indicators of events associated with the suspect attribute that are correlated with the suspect sessions being abandoned.
20 . The system of claim 11 , wherein the operations further include:
determining, by the detection system, a drop-off value for the set of one or more stages, wherein the drop-off value is a number of the plurality of suspect sessions that terminated in a stage of the set of one or more stages;
computing, by the detection system, a missed-conversion value for the set of one or more stages, wherein the missed-conversion value is computed based upon an average conversion value, an under-conversion rate, and the drop-off value; and
providing, by the detection system, the missed-conversion value for determining the likelihood of the suspect attribute causing the suspect sessions to be abandoned.