top of page

Cracking the code of retail work with AI Solutions

by Anh H. Nguyen

There is a certain tendency among society to denigrate retail work. People often consider it unskilled and non-essential. With that come lower wage, lower status, and limited career prospects. Retail workers are often boxed in a vulnerable situation, too: when the economy is shaky, they are often the first to be let go. In the US, for example, about 1.1 million store-based retail workers have been laid off since the beginning of this year due to the pandemic. Even within the retail business, the workers' contributions are often judged unfairly and assessed incorrectly. Yet retail workers are an indispensable part of the retail industry, the work force, and our lives in general, making up about 10% of total employment in developed countries. In this article, we seek to address three points:

1. What exactly does "customer service" in retail entail?

2. How should retailers leverage the variety of workers' skills to boost their sales?

3. Why is AI the perfect solution to that end goal?

A brief summary of retail work

What retailers look for in customer-facing candidates differ greatly based on the sub-sectors they belong to. In general, it could be roughly divided into two groups: hard skills such as technical expertise and relevant product knowledge, and soft skills such as charisma and interpersonal aptitude. Depending on the field, the range of skill requirements could heavily lean towards one group or be a mixture of both. For example, electrical retail workers should have sufficient pre-existing grasp of electronics and computers. A retail pharmacist who gives general healthcare advice and supplies medication to the public needs even more training and qualifications. Fast food workers, on the other hand, could start their jobs almost right away. The higher end the retail business is, however, the more demanding both types of skills become. A retailer that focuses exclusively on ultra-luxury watches, for instance, would require its sales staff to meet rigorous standards. They ought to be able to speak expertly on the provenance and mechanics of the watches they sell, be charming and obliging, yet still exude authority and confidence.

Customers regard retail work in another manner, though. While they might be aware, to some extent, of the impacts of retail workers' soft and hard skills, the differentiation may not matter much to them. From the shoppers' point of view, retail work rarely goes beyond the boundary of direct interaction with customers. That is the primary lens through which they regard the effectiveness of customer services. Within this area, there are three forms: routine service, personal service, and self service, according to On the materiality of service work (2006) by Pettinger. Routine service is simple and self-explanatory: a good example is a cashier processing transactions at the till. Self service is when the staff largely leave customers alone to browse their products, essentially serving themselves. Personal service is the most evolved form of customer service. Sales staff may assist customers by consulting, giving recommendations, and generally taking care of them in any way possible. The more upmarket the retailer is, the more engaging and personalized it tends to be. In many cases, the line between serving and selling becomes blurry.

There is an elusive "golden ratio" that retailers need to master

Retail, in its broadest sense, is multidimensional to the max, since literally anything could be offered as a goods to be sold. Not only that, retailing could occur either in retail stores or via vending machines, door-to-door sales, online and radio and TV channels. In this article, we only discuss the sales staff in traditional brick-and-mortar stores. In previous articles, we have touched upon the importance of allocating an ideal number of personnel based on real-time data. However, we have not talked about two things:

1. The skills and attributes of retail workers


2. The balance of self service and personal service.

Every retailer is unique. Even retailers of the same vertical or brand will differ widely based on size, location, target demographics, and their relative position on the food chain. Before trying to raise the conversion rate of sales staff, retailers need to make sure their workers possess the right mix of soft skills and hard skills. Do they have the caliber to educate and guide customers? Do they possess mastery in the art of selling? Do they even have the correct look, or "brand fit"? These delicate nuances may be lost at a mass-market F&B establishment, but will absolutely make a difference at a deluxe jewelry retailer.

Next, retailers should test to see if their workers are following an ideal pattern of accommodating customers and letting them be. There are several rules of thumb, like high-end shoppers usually love to be waited on hand and foot, but they are not set in stone. Not deducing this ideal pattern results in unsuitable training and procedure, and of course loss of sales. Too attentive staff may suffocate and scare off customers, while a hand-off method may leave customers feeling neglected. In short, retailers need to find (or curate) the model workers, then coach them on the model system of customer services. Since every retailer is different, there is no one-size-fits-all solution to this quandary.

For retailers seeking to optimize sales staff's performance, AI may prove to be a godsend

In order to test any hypothesis, there is a critical need to quantify all applicable metrics. A particular retailer may be pleased with the current level of sales, but in case that is not true, it may choose to swap out and test one element of its sales staff (A/B testing). For example, the workers may be instructed to increase the customer interaction rate by 25%. The resulting sales growths (or lack thereof) from all experiments will then be recorded and compared. The retailer may keep tweaking until it derives at a satisfactory decision.

The only reliable way to measure all necessary metrics is to employ a AI analytical software that draws upon factual data. The much touted digital transformation, in a nutshell. And video footage from security cameras installed in stores is about the most genuine and concrete data there is. In the above example, the only way the retailer could determine the actual interaction rate would be to use the camera footage. It could also pinpoint the most receptive customer demographics, the best timing and practice to approach customers, etc.

Quantifying customer interaction rates

In Vietnam, Palexy is a pioneer in the field of utilizing the existing surveillance systems in retail stores to improve all manners of the retail operations. Particularly, Vietnam's business owners could access fact-based insights that help raise the productivity of their staff without any additional hardware cost. In uncertain times like right now, the need for data-based optimization is even more urgent. For a more comprehensive checklist of solutions, contact us at



bottom of page