AN INTRODUCTION — Part 1 of 4
Auto-Replenishment & CPG
The future of purchasing for household essentials.
After a long career as an eCommerce leader in the travel industry, I made the switch to the CPG (Consumer Packaged Goods) industry in 2019. I was fortunate to take on the lead disruptive innovation role within eCommerce at RB (maker of Lysol, Finish, Air Wick and more). This was very different from where I’d been because CPG exists almost entirely in a 3rd party distribution world. Online, they sell through Amazon, Walmart, and Target instead of through Lysol.com. In travel, on the other hand, hotels, airline and car rental companies live for their online DTC (direct to consumer) business and merely tolerate their high commission, online travel agency distributors.
Throughout 2020, auto-replenishment was a major focus of mine. We’ve already seen high profile failures in the market such as the Amazon Dash button and it’s important that any new solution are an improvement that take real life consumer feedback into account.
What is Auto-Replenishment?
Auto-replenishment is the process of automatically re-purchasing products at the moment a consumer needs a refill. An auto-replenishment system forecasts, or detects through real time monitoring, when the inventory of a household product is low and then automatically generates a refill order on behalf of the consumer to ensure that he/she doesn’t run out. The process might be setup to ask for permission to complete the purchase or it may fire automatically. Keep in mind that this was all started pre-COVID which in hindsight seems like a pretty forward thinking strategy...
An example of auto-replenishment would be a product like the Finish (dish detergent) Smart Tub which my team showcased in partnership with Amazon inside their smart home pavilion at CES 2020. The wi-fi connected smart tub uses a time of flight sensor in the lid to measure the depth of empty space within it, thereby calculating how much dish detergent the consumer has remaining. Once the volume gets below a certain level, let’s say 20% remaining, the device fires off an order to replenish the supply and ensure that the customer never runs out and never needs to worry about adding it to their shopping list.
There are three different categories of auto-replenishment, depending on how much of an investment you want to make and how loyal your customers are: hardware driven auto-replenishment, data driven auto-replenishment, and a hybrid model.
Hardware Driven Auto-Replenishment
Hardware driven auto-replenishment is exactly what was described above. Either the product packaging itself needs to have technology integrated into it or some type of aftermarket device needs to be created. Typically, only the most loyal customers (or early tech adopters) will be prospective customers for this type of replenishment. These consumers need to be willing to both spend more to get the device and be willing to have it connected to their home network. Devices can include but are certainly not limited to:
- Connected Containers (for product refills)
- Connected Dispersal Devices (in the case of air care)
- Attachable Sensors (either aftermarket or added to existing packaging on the production line)
- Smart Scales (the base station for a product)
All of these devices are capable of measuring either the vertical product height or the overall product weight to establish the real time volume of inventory remaining. They will trigger the replenishment order when it makes the most sense for the customer to prevent running out and without generating an oversupply.
Auto-replenishment could be considered the consumer implementation of Kanban, the Toyota Product System just in time inventory model. The consumer does not want to dedicate more space than is necessary in the home to storing a large inventory. Instead, store only what is needed to manage the home efficiently in the short term with continual resupplying to meet future needs.
Data Driven Auto-Replenishment
Data driven auto-replenishment could be considered “auto-replenishment lite” but at the same time it may be perfectly good enough for most consumer implementations. In this case, instead of measuring the actual volume of the product with a connected device, a calculation is made based on a consumer’s historical purchase behavior.
Let’s say that a Target customer buys an 8 pack of paper towels once every 5–6 weeks. This same behavior is observed for 3 purchase cycles in a row. 4 Weeks into the 4th purchase cycle, Target could proactively message the customer asking if he/she would like the paper towels to be automatically purchased or at a minimum added to the shopping cart to be included with the next order.
While this model requires far less development up front when compared to hardware, the brand/retailer runs the risk of messaging overload if too many products are pinging the customer on different cycles.
Hybrid Model: Hardware+Data Driven Auto-Replenishment
Many CPG products are used in conjunction with an appliance. The big three are the dishwasher, the washing machine, and the dryer. Some manufacturers have developed auto-replenishment technology directly into their machines. Why is this a hybrid model instead of being included within hardware driven auto-replenishment? For the most part, these smart appliances are estimating the remaining volume of their corresponding household products based on the number of run cycles used. If a smart laundry machine is tied to a 96 count of Tide Pods, it can assume that after about 85 run cycles the customer is down to 11 cleaning pods remaining and can prompt the consumer to reorder.
Similarly, aftermarket products exist that can be attached to the exterior of an appliance to also make the same auto-replenishment inference based on run cycles. Here, the device may use something as simple as detected vibration, noise, heat, etc. to assume that the appliance is going through a run cycle and deduct one unit from the supply, again prompting the consumer to reorder.
The Necessary Changes to Make Auto-Replenishment Mainstream
More than likely, auto-replenishment will evolve away from on-demand, single item orders. The economics simply don’t support that model from a P&L standpoint. I predict that auto-replenishment will evolve into a very intelligent basket builder, calculating multiple product volumes along with shipping considerations so that the consumer receives replenishment bundles instead of each product individually. It’s not going to be a perfect “just in time” system, but there are other CPG eCommerce issues preventing that anyway (again it comes down to the P&L).
Because the individual brands, in most cases, have no impact on the actual sales of their products, these changes are going to need to be introduced by either the retailers themselves or by some innovative 3rd parties. A 3rd party solution makes the most sense because consumers rarely buy all of their household staples from a single store. A new intermediary that can not only interface with the real time inventory from different local grocery retailers but can also calculate the most cost effective way to fulfill your orders will make a huge dent in the universe of online grocery shopping.
Subscription Services are not Really Auto-Replenishment
Subscription services, while definitely a part of the replenishment conversation, are really a completely different category. Subscriptions are controlled by a pre-determined schedule and not by real time demand. Even flexible scheduling is still not automatic in the sense we’re talking about and so subscriptions should really be considered a different element in the relationship commerce arena.
Subscription services have two core problems which will always limit their effectiveness and adoption: undersupply and oversupply. It may seem odd that both of these are problems with subscriptions, but the reality is that human beings are notoriously bad at predicting, especially when it comes to product usage.
Here’s a quick example. You have been using shampoo presumably for your entire life. How often do you run out and need to buy a replacement? I personally have no idea. Now let’s think smaller using the Finish dish detergent example. Let’s say you always buy a pack of 37 dishwasher tablets. What subscription frequency would make the most sense? Would you accidentally run out on a monthly cycle or would you end up with 7 left over when the replenishment arrived? 7 doesn’t seem like much, but after 4 monthly cycles you will have 28 tablets remaining when the next replenishment arrives. After 6 months, you would actually have 3 packages of Finish on hand at the same time. This is a very clear example of why auto-replenishment is a far superior solution to subscription when it comes to raw inventory management.
Moving Forward
If you are looking to explore auto-replenishment for your brand or your product line, there are a number of options to choose. It all depends on whether or not you want to use Hardware, Data, or a Hybrid model. Choosing Hardware or Hybrid will cut down on the number of options because there is a lot of time and resource that goes into developing specialty tech like this. The average up and coming startup just doesn’t have the capital to invest in it, even though there are a few. When it comes to data driven auto-replenishment there are going to be a lot more choices but also a lot more variety in the way it’s done. Some will be retailer specific but others are agnostic as described above. There is definitely a place for both solution styles and this will really depend on the product.
Make sure to test out the model you choose before making a major commitment. The insights you learn will be critical to selecting the best solution for your customers.
If you’re looking for more insights, feel free to reach out to me on LinkedIn.
