5 Challenges of Micro-Fulfillment and How to Rethink Grocery Retail

Posted on January 18, 2021 by John Ripple

Micro-fulfillment is the hot topic for e-commerce. Nearly every major warehouse automation supplier markets a micro-fulfillment solution. VC backed startups are entering the robotics space at an unprecedented pace, investing unfathomable amounts in the hopes of carving out a share of the market. The rapid rise of e-commerce and e-grocery is driving activity and investment. This article will explore the challenges to be solved by grocery micro-fulfillment systems if they are deployed at scale.

When talking about micro-fulfillment solutions, we’re not referring to regional distribution centers that send products to stores, or fulfillment centers that cater exclusively to home delivery with a facility that serves a large metro area. Those are different business models. Micro-fulfillment solutions are deployed at store-level scale servicing the local area.

This is an article about universal challenges that need to be solved and ways the grocery retail environment can be reimagined to take advantage of micro-fulfillment technology. The current micro-fulfillment model fills online orders with backroom solutions (stock room of retail store) and dark store models (location that caters exclusively to online shopping; click-and-collect or delivery) of order fulfillment.

Universal challenges are:

  1. Unpredictable consumer demand & the trust factor.
  2. Impact of stock-outs and consumer choice.
  3. SKU proliferation. Variety drives volumes, not profits.
  4. Replenishment: who, how, where.
  5. Tech-stack reliability. Managing imperfection.

These challenges need to be solved regardless of technology solution. It is wishful thinking to gloss over these challenges with claims of superior performance, algorithms, and equipment. It is easy to think of problems – and harder to provide solutions. In the spirit of contributing solutions, this article will outline a retail approach that improves the viability of robotic micro-fulfillment solutions.

Most systems have been conceived and deployed for dark stores or the backroom. Robotic micro-fulfillment should contribute to the consumer experience to drive revenue and increased profits. They should be inside the store as part of the retail experience. Nobody puts Baby in the corner!

Universal challenges of robotic micro-fulfillment are:

1. Unpredictable consumer demand & the trust factor

Fickle consumers. The bane of a retailer’s existence. That, and black swan events. Did deep-learning AI algorithms predict that COVID-19 and work from home would cause demand for roller skates to increase 12,000%? We all lived through the toilet paper shortage of 2020 and witnessed unpredictable consumer demand, exacerbated by social media and the 24/7 news cycle. It is amusing to look back and chuckle at the toilet paper scare of 2020, but let’s be honest… how many of you stocked up on a few extra rolls “just in case”?

One can dream of (un)imaginable scenarios where supply chain AI captures unpredictable consumer demand as it unfolds. AI algorithms analyzing social media newsfeeds and cable news networks. Frequent mentions of toilet paper linking back to increased pulp and paper processing, toilet paper production and distribution, ending up in your local micro-fulfillment center equally as speedy as the always-on news cycle. The same AI algorithms translate increased online recipe searches into predictions of 800% increased demand for yeast – routing increased supplies to a micro-fulfillment center near you. Chewing gum sales down by 33%? Crap! That came out of left field.

Unpredictable demand is a frequent actor, not a COVID phenomenon. Unpredictable demand is wildfires in the West. Hurricanes in the South. Tornadoes in the Midwest. Flooding and storms in the Northeast. Come to think of it, the only thing unpredictable about consumer demand is that it is predicably unpredictable.

It is human nature to secure scarce resources in the face of potential shortage. In the event of potential scarcity, humans will take action (including the inconvenience of picking their own groceries off a store shelf!) to secure resources necessary for survival. When delivery times for consumer orders are longer than standard – and scarcity is perceived – consumers will bypass ordering through systems that do not offer immediate resource allocation. Delayed resource allocation includes ordering through on-line portals and retrieving orders hours later at the curb. Humans crave control in times of crisis

Dark-stores and backroom micro-fulfillment will no doubt see spikes in volume. But you can be sure consumers will beat a quick path to the nearest grocery store shelf in times of crisis. Transferring goods from dark-stores or backroom micro-fulfillment centers to front-store shelves is impractical. Consumers do business with retailers they trust. Breaching trust by not providing adequate supply in times of crisis is a publicity crisis waiting to happen for retailers. A bad experience, especially in time of emergency is a nightmare for grocery retailers. Consider the following:

  • When customers are unhappy, there is a 91% chance they won't do business that company again (Lee Resources).
  • Dissatisfied customers typically tell nine to 15 other people about their experience; some tell 20 or more (White House Office of Consumer Affairs).
  • A negative customer experience is the reason 86% of consumers quit doing business with a company (Customer Experience Impact Report).

Micro-fulfillment solutions designed for times of normality and convenience mean they will remain novelties in the back rooms of select retail outlets. This model is not favorable to the retailer. Nominal impact to revenue, whilst costs continue to climb. Not an ideal scenario for solution providers hoping to provide their robotic solutions.

Backroom and dark-store micro-fulfillment models do not support unpredictable consumer demand at widespread adoption scale. Retailers cannot risk a supply shortage and breach of consumer trust.

2. Impact of stock-outs and consumer choice

Grocery retailers have a unique challenge compared to non-grocery e-commerce retailers such as Amazon, Alibaba, and others. Groceries must fulfill entire orders from a single location. Orders that include dozens of items. It is a different story for non-grocery e-commerce retailers. According to Statista.com, 70% of ecommerce orders for clothing, shoes, and accessories had 3 items or less per order. There is a big difference between fulfilling large item count orders from single locations vs. fulfilling small item count orders from multiple locations.

Non-grocery retailers can fill orders from multiple distribution locations ending up on the consumer doorstep, grouped or separate. Most consumers don’t care whether their shoes and pants arrive in two different boxes from two different locations, even at two different times of the day. As consumers, whether our 2-day shipment crossed the continent or crossed the city, it tends not to matter and is rarely a factor in the purchase decision.

Grocery is a different business. Orders have dozens of items. The nature of grocery orders is that consumers expect the order to be delivered complete, in a single delivery. If the consumers favorite taco seasoning is out of stock in Detroit, it’s awfully hard to pull from stock in Chicago. Taco Tuesday with the wrong seasoning will not delight consumers. There are some 10 items required to prepare your favorite taco meal. All 10 items need to arrive at once. Is a taco without shredded cheese really a taco?

We’ll continue on the taco theme in terms of the impact of stockouts and consumer choice. A consumer enters their order online and selects Pace Picante Mild Salsa (because that is what their 6-year-old daughter loves and won’t touch anything else and tacos are her favorite!). What if Pace Picante Mild Salsa is out of stock and the online portal indicates as such? Does the customer abort their online order to drive to a (perhaps different) grocery store to get what they need for taco night? Or, what happens if the online ordering system doesn’t indicate the out-of-stock condition and an algorithm selects an “alternative”. Horrors! Never have you heard the wails of a 6- year-old learning that an AI algorithm selected an alternate salsa. Taco night tears. Is that how retailers delight customers?

If the wrong item arrives too many times, customers will be reluctant to use online ordering as standard practice. Out-of-stock at the online portal may result in abandoned virtual shopping carts. Nearly 70% of all online shopping carts are abandoned according to Baymard Institute. How many shopping carts do you see abandoned in your local grocery store? Not many.

3. SKU proliferation. Variety drives volumes, not profits.

A typical grocery supermarket such as Kroger, H-E-B, or Walmart have 20,000 to 60,000 SKUs (stock keeping units). Small format stores such as Aldi, Trader Joes, and Save-A-Lot offer 1,400 to 4,000 SKU’s. A typical supermarket grocery store has average sales of $500 per square foot per year. Aldi’s store moves $650 per square foot per year out the door. Trader Joe’s blows the doors off with $2,000 per square foot per year.

Warehouse style grocery retailers like Costco and Sam’s Clubs maintain 4,000 to 8,000 SKUs and have sales ranging from $700 to $1,300 per square foot per year.

At the low end of quality/price point, and service is German retailer Aldi. They are relentless in driving down costs. A typical store may employ 3-5 people on-staff at any given point with 15 or less on the total payroll. You pay a quarter for your shopping cart and no free bags. With high sales volumes from a limited SKU mix and a no-frills shopping experience, it is likely that retailers such as Aldi will not rapidly adopt micro-fulfillment solutions at scale. Retailers similar to Aldi include Lidll and Save-A-Lot. Check this group off the list for the foreseeable future.

At the high end of quality/price point and low-end of service are retailers like Costco. A walk through your local warehouse retailer will tell you micro-fulfillment may not be their best play. Products arrive in bulk and leave in bulk. They move a HUGE volume of goods and have a small number of SKU’s. Many products are sold in quantities that don’t physically fit within the handling constraints of most micro-fulfillment solutions on the market. Retailers such as BJ’s and Sam’s Club are similar. Check this group off the list for mainstream adoption of micro-fulfillment technology in the near future.

At the high end of quality/price and high-end of service is Trader Joe’s. They offer 4,000+ SKU’s and move an astounding $2,000 per square foot out the door each year. Trader Joe’s offers a premium experience at competitive prices. Retailers like Trader Joe’s could use micro-fulfillment to expand their SKU offerings while maintaining their current retail model. However, convincing a retailer a change their retail model that has proven widely successful may prove to be a challenge. And… Trader Joe’s is owned by Aldi, a company known to be reluctant to embrace the latest automation fads. If that isn’t enough, Trader Joe’s recently stated that, “they would rather invest in staff than in digital infrastructure”. Check Trader Joe’s off the list…

Stuck in the mainstream category offering mass market quality/price and ostensibly higher service levels are grocery supermarkets across North America. These chains offer a broad assortment of merchandize but struggle with the lowest sales per square foot. They have large numbers of employees to stock shelves, provide checkout, and corral shopping carts. This category includes Walmart, H-E-B, Kroger, Publix, Albertson’s, Meijer, Safeway, Hannaford, and many others.

Many mainstream grocery retailers probably wish they could be more like Trader Joe’s or Aldi’s (lean, mean, profit generating machines) whilst maintaining their gorilla-sized market share. Micro-fulfillment in the backroom or building dark stores won’t turn these megastore mainstream grocery retailers into profit busting juggernauts. The mainstream grocery retailers need to rethink the retail approach, providing a differentiated consumer experience, leveraging the power of technology to simultaneously wow customers, improve service, and reduce costs.

The solutions section of this article will describe how micro-fulfillment solutions can be reimagined for the masses. Keep reading about the other challenges or skip ahead to the solutions.

4. Replenishment: who, how, where.

Replenishment of micro-fulfillment systems is a challenge. The non-grocery e-commerce world has hours to move products from reserve storage in regional distribution centers to forward-pick locations inside the same building. Orders need to make it out the door by the end of day, not in the next hour.

The more volume that is handled by grocery micro-fulfillment systems, the more replenishment is a challenge. Micro-fulfillment systems add “touches” – with every touch costing money (whether automated or people, “touches” aren’t free).

Conventional Grocery: goods to replenish stores are picked at the case-level at regional distribution centers for delivery to the store. Much effort has been expended to ensure picked products are store-ready (products grouped by aisle location) and to ensure density of product during shipment. Shipping density from DC to store is optimized and an important part of the efficiency equation.

Suppliers such as WITRON provide sophisticated warehouse automation systems that perform this function. They have provided such systems to grocery retailers for the last 20 years, including: Walmart, Kroger, Supervalu, Target, Sobey’s, Meijer, and others. Once the products arrive at the store, it is 1-touch to get the products on the shelf ready for the consumer. Products are shipped to stores on densely packed pallets or carts, with product density maximizing the “cube” of the trailer delivering products to the store.

Micro-fulfillment Grocery: goods to replenish micro-fulfillment centers are picked at the case-level at regional distribution centers for delivery to the store (as-if picking for the store). However, once at the store, an associate has 1-touch to put the product into the micro-fulfillment center and 1-touch at a later-time to pick the product for a consumer. Every product touched twice inside the 4 walls of the store. Yes, the retailer may have by-passed the “touch” at checkout – but they just invested millions in automation and didn’t move the needle on costs, revenues, or provide a service that delights the customer. And… they still need to get the goods in the customers hands to complete the transaction.

An alternate replenishment model would have retailers pick products at the DC directly into totes that would be handled by micro-fulfillment systems, but the increased costs of tote handling at the DC and increased costs of shipping (due to reduced density) makes this model impractical at anything other than experimental scale.

The main challenges for micro-fulfillment replenishment: speed of replenishing the systems and the tradeoff of additional in-store labor vs. shipping costs. The greater the volumes, the greater the replenishment challenge. The focus should be on providing micro-fulfillment solutions that handle 80% of the store variety (SKUs), but 20% of the total store volume. That is not the current model but should be! The solutions section of this article will describe this further.

5. Tech stack reliability (and service)

Reliability of electro-mechanical systems is a tough challenge. There are a range of suppliers on the market with solutions ranging from simple to complex. Proven tech to vaporware. Which solution is better is not so relevant as the fact that none of the solutions are 100% reliable. Electro-mechanical systems conducting tens of thousands of cycles per day may have high availability ratings, but still suffer from an unpredictable 15-minute service outage every day.

One of the leading suppliers of micro-fulfillment systems, AutoStore, claims the global average availability of AutoStore systems is 99.6%. AutoStore has hundreds of systems in operation and thousands of robots. The best in the business has 42 minutes of downtime every week!

*** Note: there is a big difference between reliability and availability. Be sure to get terminology and expectations sorted out prior to purchase. If you don’t know the difference, consult an expert.

Availability ratings vary by supplier, but they all stop short of perfection. In distribution centers with large fleets of equipment and various processes, at any given time a piece of equipment could be in a fault-mode, but that doesn’t always stop orders from going out the door to the store. Consumers aren’t waiting at the dock for their order. With micro-fulfillment systems, it is frequently a single “technology stack” that stores and retrieves products for orders. When that single tech-stack has an issue, order fulfillment stops. And consumers are (impatiently) waiting.

Imagine, heaven forbid, a 45 minute service interruption in the lead up to a wildfire, hurricane, winter storm, or other unpredictable event. Breaching trust by not providing adequate supply in times of crisis is a publicity crisis waiting to happen for retailers.

Some of the technologies offered have yet to be proven in production environments. By production environment, we mean systems installed at full-scale, fulfilling orders night and day, year in, year out for multiple years. Only after extensive usage can we conclude durability and viability of solutions. Test and demonstration system scale (both physical size and volumes) do not provide realistic results.

There are 3 ways to mitigate the reliability challenge: 1) Provide consumer visibility. Visibility of the problem will increase empathy and patience to a degree. 2) Don’t allow service interruptions to be the reason consumers can’t fill their shopping cart, especially in times of unpredictable demand. 3) Select suppliers with proven solutions. Don’t put unproven solutions in the line of fire.

Re-Thinking Grocery Retail:

Before moving on to rethinking the retail experience and how to integrate micro-fulfillment in the mainstream grocery experience, we should address a major barrier of automated dark store retailers that are focused exclusively home-delivery market. Price.

Automated Dark Store
Local Grocery Store
1 Gallon 2% Milk
1 Box 18 oz Cheerios
1 Lbs 85% Lean Ground Beef

*** Prices from major grocery retailer websites as of January 13, 2021.

The cost of home delivery from an online-only dark store retailer is 60% more than purchasing from your local grocery store. While there are certainly populations in large urban markets where convenience outweighs price, widespread adoption of home delivery will be constrained by the price factor. A family of 4 has a grocery budget ranging from $800 to $1000 per month. Increasing that number by 60% is out of consideration for most consumers.

Services such as Shipt or Instacart offer shopping services with monthly or annual subscription fees. In addition to subscription fees, the cost of groceries ordered through those services is about 10% more than the in-store price. Subscription costs and grocery premiums don’t count the costs of additional service fees (peak times, alcohol, etc.) and tipping.

Much like online dark store home delivery; there is a market for such services from high-end consumers. However, many consumers are not willing to spend an additional 20% or more on their grocery bill for the convenience of home delivery (in the absence of a global pandemic). Micro-fulfillment systems deployed at scale need solutions that cater to the middle 80% of the market.

This is a good opportunity for a moment of self-reflection. You, the reader of this article may find that spending 20% more on groceries is worth the convenience. That is not a privilege of many.

Increasing the viability of robotic micro-fulfillment:

Plugging a small automated system into the backroom doesn’t move the needle. Solutions that cater exclusively to home delivery are niche market. We need solutions that work inside existing grocery retail stores that improve the shopping experience, lower costs, and increase profits.

Rethinking the retail model is a hybrid solution: take the best of Aldi, Trader Joe’s, Costco and apply micro-fulfillment technology to provide the broad assortment of mainstream supermarkets. Specialty items picked by consumers directly from in-store micro-fulfillment systems. Home delivery remains an option for the fiscally well-endowed.

Dairy, meat, and fresh product remain as they are. The grocery aisles are reformatted such that they provide a full range of product types, but not the full range of SKU’s. For example, the ketchup section may have a store brand and two branded selections. The shelf space can now hold more product and you can bet there will be stiff competition to be in the branded position. When a shopper is in the “ketchup section” an interface (screens) offer a wider range of related specialty items available in the in-store micro-fulfillment center. The screens could be in-store, mobile app, or cart mounted.

Perhaps 80% of consumers will select a store brand or curated branded selection. The consumers that really want chili-flavored ketchup in a glass bottle… they can pick that up over at the micro-fulfillment pick station. The grocery aisles have a narrower selection of the most popular items – available on the shelf in greater quantities. The grocery aisles may have some 2,000 to 4,000 SKU’s. This setup makes the store simpler to navigate. Critically, the shelves hold sufficient quantity of product to accommodate unpredictable demand events. Non-technology inclined can still complete their shopping technology-free, albeit with a narrower range of products.

A typical flow of shoppers would start in the grocery aisles (limited assortment on shelf, specialty items available to select). After completing grocery selection, if no “specialty items” from micro-fulfillment, proceed directly to dairy, meat, and fresh produce. If there are specialty items, the shopper wheels up to an open pick station at the micro-fulfillment system. There they “activate” their order for specialty products. The micro-fulfillment system begins delivering totes that contain products requested and the shopper selects their own products. When complete, the shopper continues to dairy, meat, and produce as usual.

There are several advantages to a hybrid model:

  1. Brands will compete to be on prime shelf space.
  2. Full range SKU assortment provided through micro-fulfillment.
  3. Shelves hold sufficient product for unpredictable demand.
  4. Consumer is the labor for picking; reduces costs (that touch is “free”).
  5. Replenishment of store shelves is simpler with less labor.

In terms of replenishing micro-fulfillment systems, this is best done off-site. The variety is greater, but the volumes are lower. This is a good application for robotics technology to empty and fill totes with randomized product assortment for the specialty totes. Robotic systems in distribution systems will place products in totes, ideally with mixed SKU’s so a single tote doesn’t carry all the inventory of a low-volume item and so that the same item can be available at multiple pick station simultaneously. Yes, transport costs to the stores is higher in totes (reduced density), but the density penalty only applies to a small subset of the total volume. Existing robotics systems at distribution centers remain utilized.

In store vision systems scan totes for anomaly conditions after picking by consumers. On a regular basis, partially filled totes will be removed from the system and returned to centralized distribution centers. Robots will remove products from totes, vision systems will scan for anomalies, and robots will place merchandize in newly cleaned totes, top up with addition products, and send back to the stores. Companies such as companies like Fizyr provide the intelligent software for robotic picking functions.

This model could be implemented based on technology available today and retail store assets already existing. Rethinking the retail model addresses the universal challenges: Unpredictable consumer demand mitigated with full shelves of product. The store carries the full range of SKU’s, but only stocking shelves with products in high demand. Replenishment takes advantage of existing investments by retailers. Tech-stack reliability only applies to fraction of total store volume.

We need solutions that are practical in terms of deployment and affordable for retailers and consumers. Fulfillment models focused on dark stores and back rooms are niche solutions. Utilizing existing assets and simplifying in-store selection while incorporating in-store robotics is a viable path forward.

About Esoteric

Esoteric was founded to solve the challenge of matching robotics, engineering, and supply chain professionals with the companies that need their services. Using industry knowledge and technology, Esoteric provides a better way for specialized experts to market their talent – and for companies to find that talent. Be an Expert. Find an Expert.

#FindanExpert #robotics #microfullfillment #egrocery #groceryfulfillment #darkstore