Unsurprisingly, this has led to a proliferation of technology providers leveraging broad artificial intelligence (AI) and machine learning (ML) engines to unlock the value that sits within this data.
A scenario not limited to retail.
There is however a common challenge with broad brush strokes: a lack of specificity. And this is exactly what we see a large volume of providers lacking – the appreciation for the art of retail.
As a result, these solutions can be slow to mobilize, validate, and deliver. The opposite of what modern-day retail requires.
We believe that successful solutions are built on a fundamental understanding of the nuances and complexities that come with the retail industry. Without this, solutions are destined for a short shelf-life.
This is why we use cognitive technology.
It is the augmentation of human subject-matter expertise and analytics with advanced mathematical techniques.
It can include the use of AI and ML where appropriate, but it enables a more pragmatic, practical, and impactful approach to decision-making.
This focus on pragmatism is especially important in the retail ecosystem. Technology tools that exist in the store box need to be built from the shelf-edge up.
In doing so, you can provide actionable and transparent insight to associates and head office alike on how to drive performance – for example, improving markdown execution.
Not every solution requires a complex AI model, in fact when you build a ‘black box’ tool that cannot be easily understood, you reduce the integrity and trust in the system from store teams.
With retailers facing labor shortages, supply chain issues, and macro pressure, cognitive technology provides a significant contribution to cost reduction and top-line sales growth within the retail ecosystem.
This blend of human experience and computing power is the approach adopted by Retail Insight, which has repeatedly demonstrated how technology can support and accelerate value for retailers.
To set the scene; retailers are faced with a huge sustainability challenge. They have a need to reduce their food waste bill, while simultaneously managing the impact of waste on their top and bottom line.
WasteInsight perfectly optimizes this balance by providing a dynamic markdown price that is designed to hit the discount ‘sweet spot’.
At the start of a deployment to a new client, we work closely with the retailer’s technology teams to source, extract, and automate daily data feed sources that underpin the solution.
These are foundational data points that all retailers have: sales file, item file, markdown file, and the store file.
Once the initial cognitive model is launched, a daily sell-through and retained revenue feedback loop enable the model to continuously improve the proposed discount levels alongside human hyper-tuning to maximize performance.
This machine learning approach coupled with the analytical input is an important combination to achieve the goal of reducing waste and driving revenue.
In addition, human input allows us to go further than most solution providers with our reporting.
We can provide bespoke weekly reports that offer a consistent view of performance and progress at an estate or product level. An example of this is compliance reporting, highlighting whether stores are correctly executing the markdown process.
We can achieve all of this in just six weeks, and our pedigree speaks for itself. On average, retailers achieve a return on investment within three weeks of launching WasteInsight across their full estate.
We believe such speed to value can only be achieved through the blending of deep retail and process knowledge with a fit-for-purpose mathematical engine, our cognitive technology approach.