When it comes to workforce management, AI is a tool already routinely deployed across all types of companies, in all sorts of industries.
Be it powerful algorithms that claim to sift easily through CVs, automated scheduling rotas or even forecasting tools that claim to predict staffing levels, AI tools and technologies now boast a dizzying array of capabilities.
For busy organisations looking to improve efficiencies, reduce costs and boost employee engagement it isn’t hard therefore to spot the appeal. These cutting edge innovations promise sweeping solutions too good to overlook after all. One global survey of 34,000 workers found two thirds (64%) reported reduced stress levels and a more manageable workload thanks to the introduction of AI, for instance.
But tempting as it might be, it’s critical that organisations don’t simply leap in blind.
Increasingly, these developments in workforce management tools mean that we’re handing over responsibility for a complex and business-critical system to yet another system. In few sectors is that riskier either than in retail, where workforces combine multiple functions, schedules and demands. But that also makes it more important than ever that in retail companies test the emergent behaviours of software they’re looking to rely upon. Or, in short test these tools to ensure they deliver true business value, rather than simply create disruption.
For many years, of course, the challenge has been reliance on ‘old’ ways of testing. Too often organisations are reliant on manual testers, forcing a compromise between quality, cost and time. Test quickly and cheaply and risk low quality. Invest in high quality at low cost and eat into hours and hours of time. Not to mention that the huge reliance placed on ‘people power.’
In fact, the drawbacks of this way of testing are almost too many to list.
- They include:
- – The creation of regression testing scripts that take significant amounts of time and money
- – The repetitive nature of the work means which means mistakes are often made, and often overlooked too
- – Limited resources that mean fixes can’t be implemented in tandem with testing – adding further time to the process
- – A failure to provide sufficient scale or enable load testing due to the small number of transactions that can be carried out
- – Having such an important piece of work suck up labour, without the guarantee of an accurate end results, has left retailers crying out for a solution for years
Thankfully, as much as AI has created the challenge, it has also presented the solution.
In particular, by using automated software testing instead, it’s possible to drastically reduce the time and effort involved, and bump up quality significantly at the same time. It’s why in many other industries automation testing is already the default option. But in retail the complexity of technology and systems has meant its adoption has been far slower.
That could be about to change though. At REPL, our solution has been to leverage the same new developments that are driving the revolution in workforce management, to create automated testing that reflects the particular demands of retail. That meant using RPA processes that allow us to reduce the time it takes for each test by a factor of 400-500x. Whilst at the same time, massively increasing repeatability.
This approach makes it possible to execute 35 tests of an automated workforce schedule with full end-to-end validation (including shifts, punches, holidays and timetables) in just a minute. With a manual tester, the same process would likely have taken two days. In fact, using this same approach it would be possible to automate a set of scripts involving around 1,000 test scenarios overnight. Manually, it would take six weeks.
This isn’t just about speed though. This approach also delivers unwavering accuracy, deploying underlying technology and able to test across a wide selection of workforce management activities.
For any organisations looking to quickly and accurately validate the effectiveness of configuration changes this shortening of the process is invaluable.
The reality is as workforce management software gets more innovated, automated and data-driven, so too does the testing of these solutions need to step up a gear, delivering equivalent speed, accuracy and rigour. Until now the inherent complexity of a retail workforce has left this hard to achieve. At REPL not only have we shown that it’s now possible, but we’ve also demonstrated that this ‘new’ way of testing is more efficient, more cost-effective and of higher quality. All which leaves organisations free to enjoy the fruits of AI, without coming across any bad apples.
