Events are a fundamental building block in journeys, so it is not surprising that organizations feel a sense of accomplishment when they arrange their data capture to turn most everything into an event. There is reasonable analysis that can be done on events. If there is a limited number of choices, you can take unordered events—peanut butter in a shopping cart—to make a prediction—the customer will probably want jelly. Companies like events because they are narrow and feel solvable, like improving a particular customer touchpoint. Individuals whose bonus compensation is tied to specific, narrow results also like events for that reason: it looks like you changed something important.
But business problems are more complex, and the value in journeys is to shed light on the nuance and complexity that events alone miss. Consider this tale of two sequences:
- A customer opens an account, an agent coaches them on use of the mobile platform, the customer tries what they learned, and a positive CSAT is returned
- The customer opens an account, struggles to get the mobile platform to perform according to expectations, calls an agent to accomplish their goal, then gives a negative CSAT.
From the aggregated viewpoint, we have identical sets of events leading to the satisfaction survey, but dramatically different CSAT results. Without considering the order of the events, the two scenarios are completely indistinguishable and just add to the noise of the analysis. How can businesses correct a problem they cannot detect? In the absence of a sequenced journey view, the problem is invisible and results in the customer (or customers) taking a higher cost, less enjoyable journey through a live agent that could well be avoidable. Events are just time slices of reality without context.
Don’t get me wrong; events are important and helpful. Solving problems, based on unconnected data or data that is not in event-based format, is very difficult. But events are only a beginning, not an end. The old IT reporting way of looking at single events or SQL-type point-to-point mappings is not as sophisticated as examining a time-ordered progression of events. When you look at events, you’re looking at fixed points in time that are isolated and typically locked within the borders of a single channel.
However, understanding sequences does not have to be a more challenging research-grade problem that requires elaborate internal expertise. The change ClickFox tries to encourage companies to make is to move from the comfort of an event-based view to a rich, journey view that puts events in their proper context. Without sequences, you can only ask time-independent questions that are aggregates of specific event occurrences. Having tools to visualize data and explore what it has to say puts you on solid ground for exploring increasingly complex, sequence-based challenges.
The journey view, including both the events and the time in between them, reveals the patterns between channels, has the potential to be predictive, and creates the opportunity for more impactful changes. Instead of a snapshot view of an event, you get a cinema experience that shows what the customer experiences from the beginning to a given point in time. It makes it possible to review broad range variables, independent of channels, to build a fuller picture of what is happening with your products, offers, customers, and the reality of customer satisfaction metrics. Taking a journey view changes the types of questions that can be asked and the problems that can be solved.
Prediction is very high on the list of capabilities made possible. A single event is a poor predictor of future outcomes. Data sets of events are noisy, and it is not readily apparent which events in the set matter. Some events may be suggestive of what comes next, but as a group they are weak. Sequences, and preferably long sequences, make it easier to surface predictions of subsequent events.
Sequences also allow for better segmentation. Using traditional events, organizations might pinpoint a group of users by a key event attribute such as “these users all clicked this offer at midnight.” Such a grouping is rough at best. By stringing even a few events together, we can segment that group to understand their behavior better and target them in more appropriate and specific ways. This analysis does not require complex computational models. It’s merely taking tried and true analytics and enhancing them with the sequence to provide more value.
We invite you to check out the newly published white paper. In it, we dig more deeply into the benefits of combining events in sequence by looking at the types of problems that can and cannot be solved with standard events when compared with sequenced journeys.