Stop by a cinema to get a ticket for the latest show — what do you see? Is it a row of stands with workers waiving at you, inviting you to get to their booth so you can skip the rest of the line, or is it a row of machines offering you to print a ticket in 2 minutes with the simple swipe of your credit card?
Was it a long time ago since you’ve visited fast food, that now has the row of touchpads, substituting a cashier telling you about the restaurant’s lunch deals? Having an unexplainable outstanding balance on your mobility bill, how many menu buttons did you have to go through before reaching a live operator?
Public transport, hospitality, apparel, agriculture, and a countless number of other industries are shifting towards automatization, leaving the days of personal contact behind. The marketing industry is among the few, pioneering the adoption of automation from the early stages. And it is forecasted to shift towards the use of machine learning and artificial intelligence even more.
Two Sides of a Coin
With each new invention comes a portion of consequences, inevitably tied to its adoption. The Industrial Revolution has changed not only the way we operate, educate and hire, but also became a massive change-point in the entire history. And so will automatization.
Implementation of Artificial Intelligence and machine learning will inescapably change every industry we know today, invent new ones and make us forget about industries that no longer serve. Being the pioneer of machine learning adoption, you’d think Web Analytics would be among industries that will substitute human labour with the machine intelligence completely.
But is it as simple as it seems? Or does this statement include more sides to it than we currently consider?
Automation in Action
Implementing automation in a business has its undeniable perks. Recently, a Chinese factory in Dongguan — a densely populated industrial city in the Pearl River Delta, has decided to take automation to a whole new level. By replacing human hands with robotic ones in making cell phone parts and attributes, the factory owners have discovered a drastic increase in profits and a vast decrease in the occurrence of errors. Robots have increased the amount of produced pieces by three, going from 8 thousand produced parts to 21,000. Which results in a 162.5% productivity increase.
Yet it doesn’t stop there. Not only going from around 650 employees to only 60 brought increased profits, but also helped cut the factory’s losses. It’s hard to imagine the quality of a human hands’ product competing with that of a programmed machine, and winning. And you don’t have to — after switching to mostly robotic staff, the defect rate, previously reaching 25 percent dropped to only 5%.
Substituting machine learning and robotics with human employees has numerous advantages. Increased profitability, lower error rates, and the reduced cost — no need for paying insurance, sick days, human resources, team building, maternity and vacation leaves, and other human-related expenses.
Yet, is it as applicable to the Web Analytics industry? Is it as simple and effective to substitute human employees with an algorithm when it comes to data processing?
People in Web Analytics
The Web Analytics industry is already highly automated. Although, as opposed to any industry with the outcome relying largely on the speed and conformity of the produced results which can be automated, this is not the case when it comes to large data processing and decision-making. The largest aspect of effective Web Analytics execution is not in numbers — it’s in people.
You can use an algorithm to collect all the necessary information, to sort it and organize, yet you can’t force an algorithm to make a decision on the steps that should be carried out after the data has been collected. Human mind might not be as precise and accurate as machine intelligence, yet it is the only resource we can use to conduct a decision.
Making a sound judgement is not the sole reason behind the need in a human vision when it comes to decision-making. With the ever-increasing amount of information, it is necessary to determine which metric is worth being monitored, and which should be discarded to avoid the piling of stats and confusion. Having people participate in the process of determining the validity and reliability of input information is crucial to obtain reliable output results.
The human mind is largely undiscovered and in many ways — intuitive. Putting a picture of a family next to a check-out section of an online store to increase convertibility might not make any sense to a robot. But it is a completely thought-through rational decision to a human analytic, who understands the foundations of human psychology.
A set of ones and zeros also cannot comprehend or measure happiness. Customer Satisfaction rate became one of the most crucial metrics to track & monitor for the past few years. A happy customer is a loyal customer, and the leading industry brands have been largely focusing on keeping their clients happy and customer satisfaction rates high. And you can only do so by having humans involved in the process.
And answering our headline question — Is the future of Web Analytics still people — yes, yes it is. And it will be for an extended period of time, as machines and programmed algorithms can not make decisions, think and feel. Because surprisingly, this is the heart and soul of where the core magic of successful Web Analytics implementation is hidden.
Originally published at insightwhale.com on February 19, 2019.