How Data Science is Transforming the Meaning of Productivity

Have you ever discussed a topic with a colleague to find you’re all thinking and talking about the same issue in completely different ways?

That’s exactly what happened when some of REPL’s experts got together. The topic was productivity and the discussion revealed that measuring efficiency in retail has moved on. And it’s having a major impact on the way retailers are managing their workforces.

We’ve captured the essence of that discussion in this write-up which features interviews with Rob Bate, Associate Partner for Workforce Transformation and Robin Hayman, REPL’s Solution Director. The article explores:

  • Developments in measuring productivity
  • The relationship between brand and productivity
  • The data science spectrum and what this means for retailers both now and in the future
  • How data science is transforming productivity measurement and shaping retail workforce management solutions

Where Are You on The Productivity Spectrum?

You’d think that productivity is a pretty straightforward concept. According to the dictionary definition, it’s a simple case of effort in, results out, usually tied to human input: “The effectiveness of productive effort, especially in industry, as measured in terms of the rate of output per unit of input: ‘workers have boosted productivity by 30 per cent’”.

But is this really true? Is the definition of productivity changing for retailers?

Robin Hayman certainly thinks so: “10 years ago, we thought of productivity purely in terms of inputs and outputs. But as data storage has become much cheaper, data science is playing a bigger role in helping retailers understand productivity by unlocking the detail in their data.”

This is allowing retailers to move from execution-driven productivity to a new service-focussed model. “The classic tactic is looking at standard metrics, like how many items you can scan through, average price and basket value”, says Rob Bate. “It’s all about finding ways to satisfy and serve customers.”

With the advent of data science, there’s been a move towards service-driven productivity as a differentiator. And this new focus has created a productivity spectrum that’s aligned to brand.

At one end are the no frills retailers where productivity still means pile it high and sell it quick. Then there are the middle to high-end stores who are focussed on service. They’re looking for the sweet spot between moving customers quickly enough through the process to drive profit while maintaining their brand offering.

As with traditional productivity, service-based efficiencies also need to be measured. Which is where data science comes in with the ability to analyse millions of rows of data and generate granular insights.

Puddle or Lake – Which Are You?

Despite the low cost of data storage, only some retailers have captured and stored significant amounts of information. As with productivity, retailers are also spread across a data spectrum.

There are those with data puddles, which can be problematic as Rob explains: “the quality of the output that customers receive really depends on the amount and type of information they can get. For example, if a retailer can only provide data that shows they served 1,000 customers today and they sold 10,000 items, all we can say is that Monday is busier than Tuesday.”

Mobile Device with data coming outIt’s when retailers are sitting on data lakes, which contain huge amounts of information, that data science adds real value. Without it, retailers are unable to pierce the surface and look beneath.

Data science provides the tools and techniques to analyse hundreds of millions of rows, as Robin notes: “a very big retailer shared an eight gigabyte file with us. With this much information, we were able to digest more and see the true trend rather than hoping we were getting an accurate representation from a smaller sample.”

As information becomes even cheaper to store, retailers will inevitably move from lakes to seas to oceans of data. For those retailers who aren’t exploiting the opportunity to record and store their data at the right level of detail, the risk is that the differential will expand between leading and lagging organisations. Leaving those without enough at a serious disadvantage.

The Data Puddle Solution

Whether your business only has a small amount of data or an impenetrable lake, it’s possible to use data science to your advantage.

Depending on where you sit on the data spectrum, REPL works with each retailer differently. By understanding the data each business has, REPL can advise on the level of detailed analysis that can be achieved. However, as Rob says: “if legacy systems simply aren’t capturing data at the right level, unfortunately you can’t recreate history.”

REPL’s data science experts help retailers in this position by revealing how to capture the relevant data at the right level to help them catch up with the competition. Once sufficient data is in place, REPL helps retailers move onto the data analysis stage which we discuss next.  

Measuring Productivity – From Partial to Complete Analysis

Where retailers do have relevant data, the next stumbling block is an inability to leverage the true value of the information. With millions of rows of data available, a single spreadsheet is unable to capture the whole picture. Instead, retailers have two choices:

  1. Model information taken from four or five stores and hope that they’re representative of the entire business
  2. Hold all the data but be unable to analyse it effectively because the information is spread across multiple workbooks

Either way, retailers are unable to interrogate their entire data sets reducing accuracy and insight into the business.

With data science, retailers find technology does what spreadsheets can’t.

Instead of splitting data across multiple workbooks, data science uses powerful computing technology to look for patterns across and throughout the data lake. This means analysis takes place using every piece of data on the complete network of stores in its entirety. “With millions of rows of data in one place, data science unlocks new perspectives into productivity,” says Robin.

Combined with REPL’s pioneering approach, retailers can generate real value from their data. As Rob explains: “we spend a lot of time understanding each retailer’s service vision. Then we look at their current measures and check to see whether they have what they need to tell them whether they’re doing a good job.”

The data is then interrogated – both where gaps are perceived to exist and where the organisation believes it has a good grasp of what’s going on – sometimes with surprising results. As one high-end US retailer found when working with REPL.

How Data Science Supports Retail Productivity

Renowned for providing exceptional customer service, this sports apparel retailer wanted to enhance their service-driven productivity levels. Their existing report used traditional productivity measures and, as a result, only flagged times and days where their stores weren’t hitting the right sales volumes or figures.

When data science was introduced, the retailer experienced a genuine light bulb moment when the analysis revealed that they had a productivity issue on their busiest day of the week.

workforce management solutionsBecause the original reporting focussed on turnover, the issue had never been flagged before. But when customer experience was factored in to the data set, it showed that the retailer had a major problem. More staff were needed to continue to offer the exceptional service levels their brand is known for.

It was only because data science can factor customer experience measures into productivity stats, that the retailer was able to live up to its brand promise.

“Retailers are finding that having all their data stored in one place means they can identify relationships”, says Rob. “For example, we’re looking to see if there’s a link between the queue length, the happiness of the customer, the time that they were being served, the items being scanned and the number of transactions that are being executed over an hour.”

With the move towards clear spaces and a better in-store service experience, retailers are focussed on creating time to train staff and giving employees the opportunity to deliver an outstanding customer experience. The sort of experience that can’t be found online.

How New Productivity Measures Are Transforming Workforce Management Solutions

The insights generated by data science are also underpinning a turning point in retail workforce management solutions. With data analysis revealing powerful insights, for example that a particular mix of skills and experience deliver more and better sales alongside high quality service, WFM solutions can provide the perfect combination of staff on rota at the same time.

Because WFM systems are always learning, they can identify what happens if a manager edits the prescribed rota. And what is the impact on productivity. By adding this intelligence to the existing data set, pivoting all these pieces of information and combining them with artificial intelligence and machine learning, retailers are continuously improving using insights gathered from an ever-expanding set of data.

In Conclusion

WFM solutions and how retailers measure productivity is changing. And it’s all due to the insights revealed by data science.

Those retailers with sufficient data and the expertise to drill down into the detail and interpret the results are able to get an edge on the competition. Those who are lagging behind need to start gathering the right kind of data at scale to be able to leverage the latest technology.

Whether your business needs help filling your data gaps or analysing vast swathes of data, REPL’s team of workforce management and data science experts can help. Get in touch on  +44 (0) 808 200 7375 or at info@replgroup.com.