CPC G06Q 30/0201 (2013.01) [G06F 16/904 (2019.01); G06F 17/18 (2013.01)] | 19 Claims |
1. A computer-implemented method for identifying high value customer journey sequences, the method comprising:
converting, by a processor, interaction data determined from a plurality of users interacting with webpages of a website into a plurality of sequences, each sequence indicating an order of visits in time of one of the users to a subset of the webpages and whether the corresponding sequence resulted in a goal;
determining, by the processor, the visits to each of the webpages that are most frequent in the sequences as length N=1 sub-sequences;
i) determining, by the processor, most frequent length N+1 sub-sequences that start with the length N sub-sequences from among the sequences;
ii) incrementing, by the processor, N and resuming to step i) when at least one length N+1 sub-sequence is determined;
for each sub-sequence determined, determining, by the processor, a first count indicating how often the corresponding sub-sequence appears in the sequences, a second count indicating how often the sub-sequence resulted in the goal, and a probability based on the first and second counts; and
a graphical user interface presenting the sub-sequences with the probability above an upper threshold as customer journey sequences that achieve the goal.
|