The assortment solution field is expansive and offers a wide range of possibilities; however, a true assortment optimization solution is beyond any one company’s ability to solve.
The influx of technologies and data, coupled with challenges from omnichannel, has rendered many of the available assortment solutions in the market obsolete or diminished their effectiveness. These factors are exacerbating the increasingly complex challenge of managing assortments.
Imagine having a smartphone with all the bells and whistles, but you need a highly skilled assistant following you around to operate it for you. That’s simply nuts. A new way of thinking is needed.
The assortment process was defined as a key tactical component of category management by Dr. Brian Harris more than 30 years ago. It’s still relevant today and followed by many as the standard way to operate and manage categories for fast-moving consumer goods (FMCG) products in retail.
The technology to deliver on the process has seen considerable innovation. Many of these tools continue to present yet another promise to deliver a turnkey solution of tailored, localized assortments that meets the needs of the shopper on and off the shelf. The focus of this article is on solution technologies relating to assortment.
As Harris reaffirms: “Of all tactical decisions that retailers and manufacturers make, the assortment decisions are the most critical. If assortment in a category is not correct, all other tactical decisions – shelf management, promotion, pricing, purchasing, buying – will be suboptimal.”
From a one-size-fits-all assortment for a chain, to cluster-level assortments, list/delist, new product introductions, the use of historical point-of-sale/market data, and product attributes, complexity and data size kept increasing, and so did the time and effort to manage and maintain assortments.
Various assortment management solutions in the industry are built on the strength and core capabilities of the companies building them. They achieved varying levels of success – it was evolutionary. They were highly fragmented in their approach because nothing existed prior; it was a learning process. Integration was primitive. Those solutions claiming to have an end-to-end solution lacked depth.
As a result, more savvy users opted for one-off solutions from various providers, and others decided to embark on a journey to build internal solutions. Many considered it to be intellectual property that needed to be kept within the firewall of the organization.
The Current State
Out of what can be considered chaos, we can hope for order. With the latest trend of artificial intelligence (AI) and ChatGPT news, expectations are running rampant. Legacy solutions continue to partly deliver more of the same, but in a slightly different way.
Over the past decade, we have grown accustomed to an “overpromise and underdeliver” mindset, and this became the standard operating model. Higher management demanded more for less, and leaders were heavily penalized when their calculated risks didn’t pan out. All of this resulted in timid behavior. You see this mainly from mature companies with considerable market share.
This is one of the main reasons we see the number of startup companies increasing rapidly to try to fill the gap. The startup challenge is that more established solutions have cut off the air to any newcomer, coupled with a risk-averse mentality busy with the crisis of the day, resulting in a shortsighted workforce opting for the safe solution. This has caused stagnating innovation from established leaders in the category management solution space.
The assortment solution field is expansive and offers a wide range of possibilities too numerous to mention. My focus here is mainly on solutions that are closely adjacent to planograms or space-aware, from established leaders like Blue Yonder to promising newcomers like Hivery.
With data being key to the process, some provide flexible data integration capabilities and powerful formula-based rules to ingest and optimize. Some are process-driven to provide organization-wide discipline that can be sustained. Others – those new to the game – have built their solutions from the ground up to accommodate highly granular data and introduce AI into the process. What are the correct answers? All of the above.
Outside-the-box thinking is needed, beginning with the end in mind. Best would be a moonshot approach that covers the end-to-end process and builds a solution for the future, not the past.
A true assortment optimization solution is beyond any one company’s ability to solve. Assortment needs access beyond traditional data sources from the likes of Nielsen and IRI, to include alternative data in the mix such as online purchase behavior and social media trends. Therefore, this is driving the need for a real data platform that’s accessible in a marketplace with multiple data sources, like what Snowflake and DataBricks are building.
Only when you have the right foundation, with the raw ingredients in place, can solution companies – and there are many – start to design and build capabilities that deliver actionable, measurable results while they differentiate on AI and user experience, specializing for various verticals.
Until then, we’re running in circles. This is what led us to a skeptical customer base.
For example, Amazon, with its immediate access to data, can provide a targeted assortment to any shopper/device approach with considerable success. Amazon has significantly raised the bar in the area of managing assortment, being able to see shopper behavior, and responding in real time.
An “assortment as a service” model needs serious consideration. This would require a selfless collaborative effort to realize, bringing traditional and nontraditional, structured and unstructured data to an easily accessible platform. It would facilitate the building of new solutions with scale to leverage AI and machine learning to tackle all shelf-related problems, including assortment and more.
This is achievable. We see all of the needed components from current established solutions and new ones that are becoming mainstream, as well as future-oriented startups that are exploring possibilities. Data companies are slowly expanding coverage to nontraditional data sources, solution companies looking to leverage new technologies to compete and stay viable, and new data platforms capable of ingesting large amounts of data. These three pillars need to work together for a viable solution.
We hear of new solutions daily in our newsfeed. The ingredients and talent are available. Lacking is the collaborative leadership to sponsor and shepherd it to fruition. This is an area of high interest and a key part of innovating category management. It will be interesting to see which solution providers rise to the challenge.
About the Author
Georges Mirza has been ahead of trends developing retail/CPG market-leading industry-changing solutions. He led the charge and established the roadmap for robotic indoor data collection, image recognition and analytics for retail to address out-of-stocks, inventory levels and compliance. Having previously managed portfolios of space and category management solutions at Nielsen, Blue Yonder and SymphonyAI Retail CPG, Mirza currently advises companies on how to strategize and prioritize their roadmaps for growth. Follow him on LinkedIn or Twitter.