‘Time’ Dimension of Data
Quite a few start-ups in recent years have been valued significantly higher than what their value might have been if arrived at purely looking at their revenue and profits. This is particularly true for new age internet start-ups that claim to derive value from the data they capture as a reflection of the consumer behaviour which can be leveraged by them and other third-parties using data analytics techniques. While some part of the data is useful for predicting consumer behaviour and associated personalization, an important aspect is the validity of the captured data over time.
Behavioural economics has already started exploring some of the aspects in its researches on temporal discounting. Including the Time dimension, with suitable weights assigned to the different attributes of Time such as recency, time between cause and effect, subjective nature of time, etc., can significantly improve the targeting and personalization by providing a ‘nudge’ at crucial moments during decision-making process thereby improving customer experience.
Time is a critical variable that needs to be kept in mind when analysing the data. Recency of data is one factor because the older the data gets it is less likely to be a predictor of consumer behaviour as most consumer decision-making has a strong temporal dependence with a variety of interrelated influencing factors. These factors and their correlation at the time of decision-making set the context in which the ‘Go No-go’ call is taken.
Quantum of data versus its relevancy
For the last few years since storage has become cheaper and ubiquitous, most companies have been collecting humongous amounts of data. What often gets lost in the process is the time relevancy of the captured and stored data. Data management and data analytics professionals need to regularly revisit the reasons for collecting and storing the data and question the relevancy of using the stored data in the organization’s strategic decision making. Stored data can be good for historic analysis on a post-hoc basis but may lead to faulty conjectures and decision making in the present and for the future. For example, a single Black Swan event such as the current COVID pandemic can render the entire historical consumer behaviour data irrelevant and obsolete.
How modern E-commerce and tech-enabled start-ups capture and use Time related information to analyse consumer behaviour?
Most of the e-commerce players capture massive amounts of consumer behaviour data which allows them to predict purchase patterns, give personalised suggestions, and make it easier for the consumers to make a purchase decision. All databases tend to record the timestamp but it is usually one of the least used fields during the data analysis process; particularly when analysing consumer behaviour.
On the other side of this spectrum lie the important celebratory occasions like birthdays, anniversaries, festivals, etc. where time is of essence and can be leveraged effectively to reap returns. While a cake/flower delivery site such as ‘Ferns N Petals’ does a great job of capturing the occasion details in your profile and prompting you annually to make a repeat purchase, the larger e-commerce players such as Amazon and Flipkart fail to capture such basic information. Depending on the nature of business, personalization of this nature becomes a great opportunity for engaging with your customers reminding them of key occasions and prompting purchase for the same thereby driving repeat sales.
Uber’s algorithm captures the temporal dimension to deliver customer experience in a great way. If you have been a regular user of Uber like me, you might have noticed that based on your location and time of the day, Uber tends to show the possible destination choices based on your previous trips. This helps to reduce additional typing and searching of location as you can simply tap on the destination and be on the move. What I have found particularly interesting is that Uber doesn’t just do this for your regular commutes to office, school, home, etc. but it extrapolates it to other locations and cities too. For example, I travel to a few cities on business trip regularly and my destinations in those cities and place of stay are also somewhat constant. So when I land in one of these cities and open the Uber app, based on the time of day Uber’s automatic suggestions allow me to start out in just a couple of clicks. In a few of these cities, the mobile network may also not be strong sometimes but since Uber prompts me for possible locations I don’t need to spend precious time waiting for location search to get completed for starting my ride.
How COVID has made the existing consumer behaviour hypothesis irrelevant?
The world is currently grappling with a pandemic of epic proportion with practically every business and individual having been impacted in some way or the other. While on one side, the data from various regions that were affected significantly till a few weeks back is being used as a yardstick and predictor for the impact of the virus in other regions, some important variables that are difficult to measure and incorporate in the models are the behaviour patterns of the people in the different geographies, the socio-economic setup, the geo-political situation, the weather conditions (some factors are being considered but weather is a complex phenomena comprising a multitude of factors which have strong temporal dimension), and so on. Also, each of these factors is itself changing continuously and therefore extremely difficult to measure or predict. Today’s data in such a context becomes of limited relevance tomorrow and generalizations derived from such datasets are short-lived.
Another aspect related to COVID is that, with the drastic impact of the pandemic on everyone’s lives, the entire consumption patterns and consumer behaviours have changed. Therefore, the pre-COVID era data related to consumers and their behaviours is not going to be relevant in a post-COVID world. Changes in consumer behaviour are a given phenomenon in such major global events and it is well-documented that consumers don’t revert to their previous consumption patterns as the stressors and memories associated with such events never really get completely removed from the minds and environment of the affected people.
“I wish it need not have happened in my time,” said Frodo.
“So do I,” said Gandalf, “and so do all who live to see such times. But that is not for them to decide. All we have to decide is what to do with the time that is given us.”
- JRR Tolkein in ‘Fellowship of the Ring’
