Data to Intelligence to Engagement

Words by: David Lancefield, Director of Software Solutions, EMEA, Zebra Technologies

Most people don’t realise that the humble barcode came about thanks to the earnest requests of a front-line worker (a supermarket manager), who wanted a way of getting shoppers through his store more quickly. The delays and the regular stocktaking were costing him his profits.

Though the barcode has been around for decades, its role has shifted from useful to business critical since e-commerce platforms started taking off in the mid-90s. The ease of click-to-buy models have complicated inventory management and fulfilment workflows, as supply chains are no longer linear. The business case for barcode-based track and trace solutions has grown quickly.

For example, in 2020, total European e-commerce grew to €757 billion euros, up 10% over the previous year, and the pandemic has driven even more people to shop online. In Zebra’s 14^th Annual Global Shopper Study, over three-quarters (77%) of consumers said they had placed a mobile order in the previous three months, with over half of Boomers (those in their late 50s to 70s) confirming they’re now taking advantage of mobile commerce (m-commerce) shopping options.

This profound shift in commerce is putting pressure on retailers to deliver more goods in more ways – and to look beyond the barcode to be sure they can deliver what customers want. They don’t want to forfeit their competitive advantage or financial health because they can’t keep up with demands.

Building on the barcode-based retail system

As consumers, we’ve become familiar with QR codes in the last few years as companies started offering digital loyalty cards and payment options. But did you know it has actually been around for over 25 years?  The next innovation came about thanks to Hara Masahiro’s love of the strategy game Go and the challenge retailers and others in the supply chain had scanning multiple barcodes.

Once the QR code was invented, workers could instantly report the status of every item they handled, as well as inventory sitting on the shelf or stockpiled at the receiving dock, with a single scan. The labour and operational costs of ‘inventory management’ dropped, even as technology spend climbed. Of course, a few years later, barcode technology caught up and multiple barcodes could be scanned with a single trigger pull. But technology companies that develop solutions for supply chain operations never stopped looking for a better way to track, trace, count or process goods.

Over time, radio frequency identification (RFID) applications started emerging and the technology matured, quickly, proving that data capture – and track and trace – could be further automated. Thousands of tags could be read each second by fixed readers strategically placed throughout facilities or handheld readers operated by workers, and data could be fed in bulk into inventory management systems with increased accuracy.

As European Commission Thierry Brenton noted in a 2020 article, by 2025, 80% of data will be generated and processed at the edge of enterprise operations, and only 20% in the cloud. Now, even as more of the data processing is occurring in the cloud, data capture remains predominantly at the edge – as does data utilization.

As industry expert Suresh Menon, Senior Vice President and General Manager for Software Solutions at Zebra Technologies has noted, this influx of data originally created a need for software solutions that could enable data scientists to efficiently sift through the data and make it useful. But as the demand and hiring of skilled data scientists increased, so did the realization that such analytics processes needed to be automated – especially if we want to be able to sense, analyse and act on both supply and demand trends in real time.

Today, even as more and more data and workflows go digital, complete confidence in the numbers reflected in retail inventory management systems is lacking. Most organizations are not yet reporting 100% inventory accuracy or the ability to sense demand. As such, inventory planning has not yet been perfected. Why?

Data fragmentation remains an issue for most organizations. Information systems are siloed both within and outside the four walls even though operational functions and supply chain organizations have become more codependent than ever. Issues around data integrity (incompleteness, duplication, and inaccuracy to name a few) remain.

In other words: whether retailers choose to capture edge data via barcodes, QR codes and/or RFID tags doesn’t matter if they can’t find a way to efficiently analyse that data relative to real-time operations. And even the best analytics platforms will lose value unless they can also prompt real-time action by the workers best positioned or capable of maximizing opportunities and mitigating issues.

That said, I think the democratization brought about by low code/no code software platforms, and the greater availability, affordability and adaptability of software-as-a-service (SaaS) platforms are changing things for the better. It is finally becoming possible for retailers to sense, analyse and act fast.

Assigning Value – and Actions – to Inventory Data

Real-time inventory status is key to making the right labour, procurement, merchandising, pricing, and promotion decisions, which barcode, QR code and RFID systems technically provide. However, hardware components don’t analyse or action captured data. That’s where software comes in.

Ever since SaaS platforms became available at scale, we’ve seen a quantum leap in inventory management capabilities. Structured and unstructured data generated by Internet of Things (IoT) components can now flow freely through a data pipeline or directly to a data lake. As a result, application programming interfaces (API) and machine learning algorithms can be leveraged more extensively to access and mine data in the context of a specific operation or function in a low-cost manner.

Everyone – merchandising managers, operations leaders, procurement planners, loss prevention experts – can plug into the same information systems via APIs and extract actionable insights most relevant to their roles. Workflow apps can also be built quite easily to help drive best next actions for operations managers, associates, and delivery drivers. A prescriptive analytics platform, for example, can be taught to detect certain patterns in the data and ‘prescribe’ tasks to employees when inventory-related issues or opportunities arise.

Similarly, an intelligent demand sensing platform can aggregate inventory data from multiple business systems and analyse it alongside contextual alternative data – weather, traffic, holidays, and other demand-influencing events. It can then prescribe specific procurement, merchandising, pricing or promotion actions that can right-size supply against demand.

The Take Home Message: Creating an Ecosystem Leads to New Solutions

In short, this software-led shift from ‘systems of record’ to ‘systems of intelligence’ and, ultimately, ‘systems of engagement’ has been key to progressively improving inventory availability and performance in the last decade. SaaS solutions have even automated decision-making to a certain extent, taking the last bit of manual work – and risk – out of the inventory planning and management equation.

However, we must do more to ensure all stakeholders have full transparency into inventory status from the first mile through the last – or from the stockroom to the store floor. We must tear down solution development silos. Technology vendors and software developers must commit to building and utilizing open platforms when designing inventory-related solutions. And we must ensure these solutions openly share, actively analyse and intelligently action data so all supply chain entities can effectively predict, sense, and shape inventory demand.