How Granular is Too Granular for Assortment Planning?
Creating a retail assortment is like telling a story.
"Every character, location and event in a story needs to be introduced and explained to an audience. Every relevant detail has to be explained, and that explanation is the element of a story which enchants a reader more than any other. It's the essence of storytelling."
The key word in the quote above is “relevant.” But what information is relevant in the assortment planning process?
Great buyers know how to tell a story and express a point of view with only a table of footwear or a single rack of clothes, but how did they make these decisions? How much of it was planned beforehand? What were the relevant details? These are key questions in determining how granular a retailer’s assortment planning processes need to be.
In my experience, when most organizations try to formalize their assortment planning processes they become far too detailed, resulting in excessive time spent planning assortment with only minimal return. And sadly, organizations rarely support those processes with investments in technology. An assortment planning solution can automate the science, providing buyers with a framework for their story, while ensuring a consistent planning process across commodities. However, “consistent” doesn't mean “the same.” For example, a jewelry buyer will have a different assortment plan than a housewares buyer, but they'll use the same processes.
A planning solution should provide visibility to compare — and the ability to reconcile — merchandise and assortment plans. Without the ability to reconcile, assortment planning becomes a lawless, Wild West story, where there are rules but no way to enforce them.
I once consulted for an organization attempting to implement a formal assortment planning process in Microsoft Excel, while maintaining its existing planning processes (financial, merchandise and store plans) in a stand-alone solution. The company introduced new metrics and concepts in an attempt to derive a receipt value at retail based on full priced sell-through (FPST). The intention was rooted in the notion that more full-priced sales units decreases markdowns and increases profitability, while also decreasing the purchases/inventory required. I stressed that gross margin return on investment coupled with gross margin value would be a better measure for the health of the business, but FPST was the route it wanted to take.
FPST is a fine peripheral metric, but to use it to drive receipts at retail requires other metrics. This organization also had to introduce off-price sell-through, average markdown percentage, blend of sales of full-price vs off-price, and average unit retail (AUR) and average unit cost (AUC) to support the conversion between units, retail and cost. These metrics were also introduced to the retailer's merchandise planning process. Finally, merchants had to shift their thought process of receipts driving sales to sales driving receipts, specifically FPST, without regard to store inventory levels, presentation standards or vendor minimums.
The results:
- the introduction of seven new planning metrics;
- the unintended re-engineering of its merchandise planning processes; and
- an assortment plan that provided detailed sales and receipt information by store, style-color, product attributes and month approximately one month before market trips.
As a result, over 90 percent of the time, merchants couldn't follow their detailed assortment plans due to ad-hoc in-market decision making, and planners didn't have the time to manually reconcile merchandise plans and detailed assortment plans post-market as a result of a lack of integration from Excel to their stand-alone solution.
What should have happened?
In speaking with representatives from various teams, the introduction of new metrics and the requirement for assortment plans to be so granular proved the be the downfall of the new, and ultimately scrapped, assortment planning process. The post-deployment recommendations to reduce metrics and make higher-level assortment targets rather than assortment plans was never implemented due to the low adoption rate and high cost of the initial assortment planning process. In the end, planners and merchants went back to their individual processes they had previously been using.
The moral of this story is that assortment planning processes need to be tailored to your current environment and not shoehorned in. Implementation teams need to understand the impact to downstream and upstream activities. Lastly, the need for an integrated planning solution cannot be overstated as the ability to reconcile and monitor plans is the entire point of having a plan!
Jeff Coull is manager of professional services for Retalon, a provider of predictive analytics for the retail industry.
