The retail industry has always been tough.  Even when retail simply meant a few regional storefronts, proprietors faced hard questions on what to stock and how much, how to price it, and how to staff their stores.  Introducing new technology, like bar-code scanners, presented headaches as well.

In the 21st century retail has grown far more complex.  Many retailers are struggling to catch up with consumer expectations around omnichannel distribution.  On top of that, those same consumers have grown accustomed to enjoying far greater product variety, fast delivery, and always-low prices.  They use the internet to comparison shop while on the store aisle, and they exploit social media to chide poor performers.  Every decision a modern retailer faces must be made in consideration of the many new faces of their business – as well as those same old traditional challenges.

Our team at End-to-End relishes the challenge of helping its retail customers address the following problems:




Supply chain analytics for retailers begins – as in other industries - with creating a demand forecast. This means reconciling historical data with planned demand-shaping actions and using it to project a picture of future demand.

End-to-End Analytics has deep expertise in the particular challenges posed by forecasting in a retail environment, such as how to:

  • Detect and correct for stockouts in the sales history so that past failure to fulfill demand is not construed as the absence of customer demand.

  • Forecast at the SKU-store level, where demand is often sporadic and “lumpy.”

  • Forecast demand for new products, which lack history.

  • Deal with millions of SKU-store combinations.

  • Quantify uncertainty in the forecast.

  • Choose appropriate test stores for new merchandise.


Store Replenishment

Store Replenishment

No task in retail is more central to success than that of deciding how many units of each SKU to send each day to each store.  Yet many retailers still rely on antiquated “min-max” policies to drive store replenishment.  This inevitably leads to excess inventory on some SKUs but stockouts on others.

We implement state-of-the art replenishment logic that allows retailers to:

  • Learn the unique demand profile of each SKU at each store.

  • Find the right target service level, based on costs, for each SKU in each store.

  • Optimize store replenishment frequencies.

  • Adjust replenishment quantities to account for weekends, holidays, and seasonality.

  • Balance distribution center picking and dispatching costs, freight, inventory, and service levels.

  • Minimize stockouts while managing within a given overall inventory budget.




Along with replenishing stores, deciding how many units of each SKU to buy is among the most basic tasks in retailing.  Bad decisions here lead directly to stockouts, or to excess inventory and markdowns.

Yet many retailers still regard purchasing as something of an art, ignoring the sound supply chain management principles that have long been taken for granted in other sectors.

We work with retailers to:

  • Determine the right distribution center service level for each SKU (which may be very different from the desired in-store service level).

  • Calculate the safety stocks and purchase quantities required to achieve those service levels.

  • Adjust purchasing decisions for holidays, promotions, supplier shutdowns, etc.

  • Make smart “strategic buys” – such as buying more to take advantage of a quantity discount, or to hedge a shortage risk.

  • Optimize purchase frequencies by balancing per-order costs against the costs of additional cycle stock.

  • Monitor receiving loads and adjust purchase plans to smooth receipts where required.


Assortment Planning

Assortment Planning

Assortment planning means deciding which SKUs to stock in each store.  This task is common to all retailers.  Unfortunately, many existing approaches to assortment planning are flawed and lead to suboptimal decisions.

For example, some retailers measure category performance using the average return per SKU.  But what matters for assortment planning is the marginal return on an incremental SKU, not the average return on an existing SKU.

Retailers also commonly use GMROI (gross margin return on investment) to drive assortment planning.  But this measures the return on just one constrained resource, cash.  It completely ignores the other key constrained resource: shelf space.

To improve assortment plans we work with retailers to:

  • Estimate sales of categories which have never previously been sold in a given store.

  • Calculate the impact of adding – or deleting – SKUs to the assortment.

  • Find the right number of assortments, balancing complexity against the sales increase stemming from more fine-tuned assortments.

  • Identify which SKUs should be added to each assortment, and which should be deleted.




Retailers commonly have to allocate a scarce product between stores.  The perceived best practice in this area, known as “fair share” allocation, assigns available inventory in proportion to the net requirements of each store.

Unfortunately, this procedure does not always send stock where it is most needed.  Instead, it often under-supplies stores that have stocked out (i.e., experienced strong demand), while still supplying stores that are close to but below their target inventory levels.

We work with retail clients to implement “smart” allocation schemes that:

  • Send each available unit where it has the highest probability of generating a sale.

  • Determine what fraction of available inventory to deploy immediately to stores, and what fraction to retain in the distribution center for resupply purposes.

  • Protect channels such as online sales that are served directly from the DC stock.

  • “Push” just enough inventory of new products to stores to accommodate upside demand while holding back the remainder to be allocated on a “pull” basis.




For retailers with unpredictable demand and long lead times, markdowns on some SKUs at some stores are inevitable.  Unfortunately, current markdown polices are often far from optimal – that is to say, costly.

For example, many retailers wait too long to begin marking items down, leading to larger-than-necessary total discounts to clear stock by the end of the season.  Other retailers have little idea how responsive each product’s demand is to markdowns, resulting in a scattergun approach.

We help our clients improve markdown performance by helping them to:

  • Decide when to start marking down their stock.

  • Determine how much to mark down, both initially and with successive moves, if needed.

  • Incorporate markdown policy into their initial purchase quantity decisions.

  • Evaluate the opportunity cost of the shelf space lost to carrying discounted items.


Price and Promotion

Price & Promotion

Price and promotion serve as two principal levers available to retailers hoping to influence sales.  These, combined with product and placement, serve as the famous “4 P’s” of marketing. 

The overwhelming number of price and promotion options, though, often overwhelm the organization.  Rather than apply data-driven analysis, they rely on time-honored rules of thumb.  And oftentimes as a result they perpetuate poor decisions year after year.

Our clients look to us to help them:

  • Set a profit-maximizing everyday price.

  • Address portfolio pricing decisions to maximize profit without surrendering volume.

  • Account for likely cannibalization of sales when price reductions are anticipated for sister products.

  • Measure the response to different kinds of promotions, highlighting the most cost-effective.