We all know about the rapid growth and popularity that e-commerce websites have gained in recent times. One of the primary reasons behind this popularity and surge of the e-commerce websites is the continuous digitization of the retail industry and constant change in consumer behavior.  However, surprisingly, as per the industry statistics released by Forbes, only 10 percent of sales happen online, and the rest happens through stores. When it comes to grocery sales, just 5 percent of sales happen online.

As per a recent American Customer Satisfaction Index (ACSI) report, satisfaction levels of the customers visiting retail stores have also declined. Customer satisfaction is of utmost importance for every retailer, and the parameters mentioned below help to measure customer satisfaction from a retailer’s point of view:

  • Quality of service
  • Speed of transaction

Test automation is one of the solutions that may facilitate the overall improvement of client satisfaction by up the coverage and quality of applications.

Gartner has also predicted that by 2023, 75% of IoT system integrators will have added comprehensive end-to-end IoT solution testing services to their portfolios. The advancements in technology and test automation tools are superfast to cope with the constant challenges and client requirements. Most of the customers spend a massive amount of time and effort in building solutions that use Artificial Intelligence (AI) and Machine Learning (ML). Nowadays, many organizations adopt Agile and DevOps but face many challenges in automation. The entire benefits of Agile and DevOps can be obtained only if the organization has enough test automation coverage. In the case of the retail domain, it is a bit worse due to the challenges in hardware. “Automation of testing method will facilitate to save almost 65% of cycle time throughout regression testing”.

Multiple Devices To Handle Automation For POS System

Today’s retail stores have different types of devices like mobile self-checkout solutions, which encourages shoppers to purchase quickly through mobiles. To automate testing of Point Of Sales (POS), one has to deal with different types of devices like display screen, debit card readers, and receipt printers, etc. that are present in a retail store.

In spite of plenty of automation tools like Selenium, Appium, TestComplete, etc. available in the market, they still do not directly support POS automation for sales. This can be one among the explanations behind many purchasers not having the ability to attain success in automating of POS systems.

POS software system runs on differing types of operation systems. For instance, NCR runs on Windows, Toshiba runs on its custom-made software package. Devices like positive identification readers, receipt printers, scanner scales, barcode scanners, monitor, and hand-held scanners conjointly interface with the POS software system. So, retailers must make sure the businesses choose automation tools that may run on these operation systems.

Testing Ought To Be Aligned With Real-World Situations

The best method is to try and follow testing be it machine-driven or manual is by aligning with real situations, similar to what a client may knock off a store. For example, simulators that replicate real-world situations of employing magnetic strip/ chip-based cards scan a barcode or weigh an item on the machine can be used.

In the case of APTOS, the user uses varied interfaces with in-house and third-party systems like memberships, promotions, pricing, reporting, inventory, and payments. For example, the rating engine may be a mainframe-based system, and the payment system may be a mobile/tablet-based application. A tool stack that would work on both platforms is vital to have a seamless integration.

There are different layers in POS systems which consist of buying/procuring and items, payments, promotions, etc. Each system needs to be in sync with the other to make even single order work seamlessly. This has brought complexity in testing the entire application layer bringing in more constraints to automating the entire suite. A test methodology that conceives engineering with technology by bringing in innovations like OCR based tools, robots and leveraging microcontrollers is essential.

Shift Left Testing Approach

The shift left testing approach starts to focus on testing right from the requirements phase. This testing approach introduces testers at the initial stage of development to help testers understand the requirements.  This helps the testing teams in identifying high-level test scenarios and start working on the detailed test cases in the development phase. In addition, it helps them in detecting the defects earlier in the software development lifecycle. This approach allows resources to invest less time in finding defects, thereby increasing quality and improving efficiency.  Precisely, implementing the shift left approach saves both time and resources and streamlines the development process and ensuring faster delivery. Hence, the shift left testing approach is an easy and economical way of reducing the cost, efforts, and time in improving the software in a complex state.

Digital Commerce quality automation framework (DCqaf)

DCqaf is a secure and flexible automation-testing framework built to empower retailers’ e-commerce and POS systems. From collecting test data to generating custom test reports, this framework ensures that the entire testing process is accurate and careful to suit every operation in retail. DCqaf helps in integration of the third-party tools that allows to customize even the last piece of code for the customer.

The robotic arm performs payment terminal screen actions, inserts/removes chip and PIN card, etc. Receipt validation is essential, and once printed on the POS machine, it must be validated for the accuracy of details. A combination of a robotic arm and a camera is required to accomplish this. The robotic arm will hold the receipt in such a manner, so an image of the receipt can be captured for the validation purpose. Then the image is saved in JPEG format and the OCR comparison is made using computer vision APIs.

Along with all these solutions, there is a need to develop a mechanism to report the overall status. There are different types of inputs received, like incoming data from the Automation Framework, test cases from the Test Management Tool, and defects from the Defect Management tool.

Automation Helps To Improve Customer Experience And Time To Market

First of all, by implementing automation for customers, a lot of time can be saved during regression testing which is approximately 80%. Secondly, the development and testing teams can contain the defects from going into the production phase. This reduces the regression testing cost and improves test coverage. Manual testers can execute about 40% regression test cases in a day, whereas a single robotic arm can execute 100 test cases during the same time. Hence, one cycle of regression testing can be completed in one day.

Although test automation plays a key role in building a robust system, manual testing cannot be ignored. As a tester, one needs to know what needs to be tested. Certain sections of the code cannot be tested just with an AI-driven framework, and interpretation of a few data points still needs manual intervention. Even in the case of regression, around 90% of the process can be only be automated, but for new enhancements and business level functionalities, you still need the involvement of manual testers. Test scenario written manually defines a test case which in turn defines a test script.

Ravikumar Venkates

He has more than 13 years of experience in the software industry which includes Web-Based and Client-Server Applications. He has extensively worked in e-commerce and payroll domains.

Latest posts by Ravikumar Venkates (see all)