Commentary

How to get a handle on the self checkout data explosion

Self checkout continues to expand, but glitches continue at an alarming pace as data volume rises. Retailers and restaurants need to scrutinize their data monitoring systems at the store level.

Image: Adobe Stock

January 15, 2024 by Ozan Unlu — CEO, Edge Delta

Self-service checkouts and kiosks are rapidly expanding across a variety of industries, especially retail and quick service restaurants. Such rapid growth is a testament to the transformative power of these systems. According to the latest retail statistics from Statista, 47% of Americans regularly use self-service checkouts, citing long lines at traditional checkouts as the most annoying part of shopping — even ahead of high prices!

Within the QSR industry, 65% of consumers prefer self-service and tablet-based kiosks, according to the National Restaurant Association. These systems also enable restaurants to turn tables much faster, with businesses averaging a 35% increase in sales within the first 30 days after implementing self-serve ordering, according to Square data.

No doubt, customers love the ease and speed of self-service checkouts and kiosks. Some innovators are even going a step further with mobile kiosks, helping loyal customers avoid irritating lines at peak hours by calling in their orders in advance.

But such high expectations for convenience can spell major problems when self-service checkouts and kiosks experience a performance glitch — and this unfortunately happens quite often.

Self-checkout mishaps continue

In fact, 67% of consumers that regularly use self-checkout have had at least one fail while using it, according to research from Raydiant. If a retailer cannot ensure near-perfect performance for self checkout, slow and unreliable systems will inevitably alienate customers, endangering ROI from these investments. The same applies to QSRs, as routinely malfunctioning self-service kiosks run the risk of driving customers away.

Why is it so hard to solve performance issues with self-service checkouts and kiosks? The answer lies in the volume of data they generate.

It's not that these systems cannot hold up under heavy traffic loads. Thankfully, advances in modern computing have taken care of that. Rather it's that as the volume of transactions increases, so too does the volume of observability data, or event data that helps IT teams gauge the internal health of applications and infrastructure by detecting anomalies.

Observability data has traditionally provided a treasure trove of information, enabling faster identification and resolution of hotspots, ideally before performance is impacted in the first place.

The challenge is that self-ordering checkout and kiosk systems are generating massive volumes of event data per day. In the QSR realm, it's not uncommon for a restaurant chain's self-ordering kiosks to generate terabytes worth of data per store, per day.

There is so much data being generated that it becomes nearly impossible to harness and leverage it to its fullest potential, to understand what is broken, why it broke and how to fix it.

The performance issue practice

Let's take a closer look at the QSR example. Today, when a self-service kiosk performance issue occurs, such as unplanned downtime, a customer may report the malfunctioning kiosk and someone working in the individual store either calls the chain's main IT department or submits a ticket.

From there, the observability team within the IT department accesses the particular store's storage target and manually searches the data to determine the root cause of the issue. As you would expect, all of this can result in significant delays between the time an issue occurs to the time it is identified by the store to the time the IT team is notified and able to fix it. Meanwhile, lines are growing longer, customers are getting aggravated, employees become harried and sales are dropping. This is clearly not a good approach.

Given that data growth is showing no signs of slowing down, there needs to be a more efficient approach to observability. The answer lies in processing and analyzing data as it's being created, at each individual store.

By implementing data analytics and machine learning at the edge, teams are able to actually detect production issues as they occur without building granular monitors in advance.

Establish baselines

For example, you are able to create baselines of what is "normal" and automatically alert on anomalies, such as an abnormal spike in error or fail logs. In this way, individual stores can detect issues as they occur without relying on humans (including customers) to identify that there's an issue, or submitting tickets and making calls.

A chain's main IT department can be automatically alerted, and they can then proactively access an individual store's local storage on their own for deeper investigation.

In addition, the data can be dramatically streamlined (i.e., only the data tied to anomalous events and production issues kept) in order to both reduce the cost associated with trying to store everything as well as the unnecessary "noise" that IT teams must sift through in order to identify the root cause.

The end result is greater automation and agility, reductions in cost and tedious manual labor and ultimately and reduced downtime which helps the restaurant meet its revenue targets.

Self-service checkouts and kiosks are clearly on the way to widespread acceptance. This is a very positive development from a business perspective, but it also comes with the potential downside of conditioning customers to expect the utmost in convenience, all the time. That's why performance problems must be avoided at all costs, and ultimately this comes down to changing the way we handle our data.

About Ozan Unlu

Ozan Unlu is the CEO and Founder of Edge Delta, an edge observability platform. Previously he served as a Senior Solutions Architect at Sumo Logic; a Software Development Lead and Program Manager at Microsoft; and a Data Engineer at Boeing. Ozan holds a BS in nanotechnology from the University of Washington.

Connect with Ozan: