21 Feb The Digital Customer Experience

My previous post lamented the loss of the information that flows between an individual customer who enters a shop, and the shopkeeper who is able to see the immediate experience of that customer and use it to maximize his business, both to increase the probability of immediate sales as well as customer satisfaction and future customer loyalty. In a digital transaction on a website, that shopkeeper is no longer present, but the business enterprise running the website but still needs that customer experience information to make the business better. How does the business get this?

Using customer surveys, and asking for reviews are two widely used methods, but these are after the fact, intrusive, and will generate some form of subjective bias (knowing that you are answering a survey will affect the answers that you give). Poorly constructed surveys will collect poor data, which will lead to poor decisions, and looking at publicly posted reviews it is clear that people have trouble separating their experience on a company website from their experience with product that was purchased. For example, the company website might be perfect, but if there was a shipping problem, or the purchased item was broken when it arrived, or doesn’t work, this gets reported as a website issue when it is not, and the information from that review has no value.

If we could measure and quantify the digital equivalent of that interaction, i.e. an individual customer’s experience on a website, and do it conveniently in real time and in a standardized manner, then the value of that customer experience could be recovered.

And we can. Think about what happens when you go to a website and buy something on-line. You spend time on the home page, move between webpages at varying rates and in varying patterns, move the cursor, click here and there, again with patterns of timing, scroll, swipe, and type. Think of swiping as a signature; your finger moves slightly differently than anyone else’s when you flick it across the screen. And typing is probably the most mundane but revealing of all; the length of time you take to press a key, how long you hold it down, how much time it takes you move to the next key, and what words you type when you have a choice, i.e. the vocabulary do you use.

With these simple actions, you have provided a wealth of uniquely personal information – just like you did when you walked into the real store in the earlier scenario. It is your personal data, acquired in real time directly from the digital interaction itself. These data are nothing less than your personal mark and signature in the digital world. By collecting this individually, rather than as a large aggregation (for example, the summary of a survey), the patterns of individual behavior can be elucidated and merchandizing resources better targeted.

Note that in all of this no demographic information has been collected. No questions about your race, gender or age, where you live, your family size or income or anything else that others already collect for merchandizing. Only the information that was specifically part of the digital transaction that you performed, and which you directly provided yourself, is collected. Only the behavior you have exhibited from when you entered the website until you made a transaction (or left the website without making a transaction) is of interest. The details of your credit card, shipping address and other data that are already collected by others reveal nothing about your experience that led up to your digital transaction, or caused you to abandon it.

So in the digital consumer world, how can we find that information that the savvy shopkeeper used to get, and would use to improve his business? What is the digital equivalent of all that data which formed the human customer experience, and how might it be collected? And with that data collected, how could we understand that experience, and express it in a useful manner? This is the problem that led us to start SriyaDXI, and develop our Digital eXperience index – DXi.

We started with the simplest thing we could imagine; after arriving at a website, what is the actual communication that takes place, and in which the customer engages? There are actually not that many different things that are communicated and these were mentioned earlier – the clicks, cursor movements, scrolls, keys used and swipes, and the time between all the elements that define these actions. These actions control the time spent on a webpage, the movement between webpages and how that movement was caused, e.g. by clicking on a specific image or link on the webpage. Other communication events that occur, such as latency or dropped communications are available, and while the customer does not control these items, they will certainly have and effect on the customers experience. And again, notice that in all of this, there is no “intrusive enquiry” made of the customer; we are simply watching the small, individual acts of communication. Since the data is simple, it can be easily examined using statistical methods.

Further, this data is all from a single individual customer, and will reveal something about that individual customer’s experience. Of course, it is the combined experience of all customers that will determine the success of a business, but it is essential to keep the individual experiences separated. So we calculate a DXi score for every individual customer, rather than a total DXi value for all of the data collected. While the data collected can easily be aggregated, e.g. total page visits per hour on the entire website, this is already accomplished by other analytical services and it looses the patterns of behavior that the individual customer exhibited.

Our DXi solves part of the problem and returns the shopkeeper/customer interaction to the digital world. But there is more that can be done to enhance the customer digital experience; both improving business for the shopkeeper and making the customer feel valued and satisfied.

Analyzing data that is given off by individuals in the course of normal, public actions is already being performed. Two areas of highly personal and unique data from individual people that we know are being explored and for which a number of algorithms exist are facial recognition and voice recognition. So why not apply these algorithms to the highly personal data we just collected during that digital transaction? This is the core our more advanced treatment of the data collected in which machine learning (ML) algorithms are used; this is the “Machined Learned eXperience index”, or MLXi. Although neither facial nor voice recognition form part of the current online customer experience, they could in the future and our system of using individual data can accommodate this directly.

But not everyone is that highly experienced, savvy shopkeeper of my original post. That shopkeeper possessed a wealth of knowledge that allowed him to measure and use the immediate customer experience. The simplest analogy to the digital environment would be a massive “expert system” containing all the information gained from shop-keeping experience, combined with a very rapid search algorithm for making decisions. What could we bring to bear on this issue, and to improve the digital experience of a customer and enhance the digital shopkeeper’s business?

This data, and the opportunity it presents to e-commerce, is being overlooked in the world’s current efforts at merchandizing to the digital customer; quite literally, data (and thus money) is being left on the table. Attention to the details of “small data” in addition to our rush to understand big data is required if we are to move beyond the standard approach to digital merchandizing. This is the purpose of our efforts at Sriya DXI.


The author Dr.David Dodds is the co-founder and CSO of SriyaDXI LLC.

No Comments

Post A Comment