What Factors Make or Break Your Listings? B-Stock Uses Data & Predictive Modeling to Find Out

B-Stock regularly provides its sellers with insights into the purchasing behavior of the secondary Market, and this helps them understand current trends and the implications for their business.

But what is it about a given listing that makes it valuable to buyers? What exactly is the difference between a lot that sparks a drawn out bidding war and one that simply flops? If you’re a seasoned inventory management pro, you might have guessed that there are many factors that play into a listing’s value, some obvious and some incredibly subtle.

The countless variables that comprise these three main families make it quite difficult to determine what really leads to a high final sale price, but with over 10 years of network-wide sales data and some sophisticated modeling techniques, B-Stock has been able to identify the factors that matter most and reliably predict most auctions’ recovery rates.

Measuring Liquidation Auction Success

While our solution can be evaluated on many criteria—ease of use, features, cycle time, and so on—this discussion will focus on money. Specifically, the metric we’ll use here is recovery rate—the percentage of a given lot’s original MSRP that the client makes back through the secondary market.

Here, we’ll discuss a handful of ideas that sellers might want to experiment with while fine tuning their reverse logistics operations and inventory reduction efforts. 

Do Recovery Prices Really Matter?

In a word, yes. Our sellers want high prices for the surplus items in their auction lots, in order to recover as much of their value as possible. B-Stock’s earnings are a percentage of this recovery, so we’re equally incentivized to max out our clients’ return on investment. Depending on the size of the seller, even a small variance small variance in recovery rate equate to millions of dollars captured or lost.

That’s why B-Stock set out over a year ago to develop a new analytical model that could accurately predict the recovery rate of any given lot based on known variables, like example category, condition, brand, and many others. After thorough study of variables and outcomes, we were able to determine which affect recovery the most.

The model itself was built on data from over 100,000 online liquidation auctions, more than 100 unique variables, and two different machine learning models.

The Top 5 Factors That Determine Recovery Rate

1. Category

The first and perhaps most obvious variable to cover is  product category. B-Stock assigns a high-level category every auction, and it’s the easiest way for a buyer to find relevant auctions.

Some examples of these categories include:

• Apparel, footwear & Accessories
• Furniture
• Groceries
• Office Supplies & Equipment
• Sports & Outdoors

If we look at the recovery rate across all of B-Stock’s default categories, you’d find that it varies widely, with the top two categories recovering three to four times as much as mixed lots. Therefor, the first takeaway, of course, is that sorting lots by category will increase recovery. This is a recommendation that we have made to many sellers, and we commonly observe that this practice increases from 30-100% and, on some occasions, even more.

Even within our default high-level categories, there can be large variance in recovery rate. For example, we have found that handbags outperform both clothing and shoes within the Apparel, Shoes & Accessories category.

Incorporating this nuance into our model has further increased the accuracy.

2. Brand

Going beyond just categories, we’ve also observed that certain brands consistently have a higher recovery rate than others. While this should come as no surprise, our broad body of data allows us to see specific top performers and where they lie in relation to others. Even on the secondary market, we commonly see clothing from well-known luxury brands outperforming smaller names or store-brands of the same category by ten or twenty times.

This phenomenon is certainly not limited to items in the apparel space. Take for example, Computers, Equipment & Software. Here too, well-known brands handily outperform house brands. There’s even a clear hierarchy within the upmarket brands, with certain upmarket brands dominating even other well-respected brands. 

Just as sorting by category benefits overall recovery, grouping these high-value products together results in meaningful overall recovery gains for sellers.

3. Condition

Similar to its approach to item categories, B-Stock provides sellers with a range of product conditions that they must choose from when listing their merchandise. The condition of the listed goods is another main factor to consider when predicting a listing’s recovery rate—and while this is unsurprising in itself, a closer analysis of sales patterns in a certain category reveals an unexpected phenomenon.

Returned and overstock products can go by a variety of names depending on the seller, and come in a variety of states and conditions—with tags,  without tags, open box, non-original packaging, and more. To analyze the data, B-Stock maps these different characteristics to our overall grading system, generally classifying them as Brand New, Like New, Used-Good, Used-Fair, Mixed Condition, or Salvage.

In almost all categories, “Brand New” and “Like New” items command the highest prices. But interestingly, we sometimes see Used-Good and even Used-Fair items outperform the “better” conditions. It turns out that in some limited contexts, an item might have achieved its “used” grade by being appealing enough for someone to purchase in the first place. On the other hand, an undesirable, item might not have been appealing enough to catch any shopper’s eye to begin with.

The benefit of a comprehensive, network-wide predictive model is that B-Stock can take anomalies like this into account when forecasting sale prices and tweak this model following new discoveries.

4. Manifest Design

The first of the major value factors that sellers can control, smart manifest design is a critical step in maximizing sale price and recovery.

First off, what’s a manifest? Before listing, sellers need to submit a manifest—a list of what they plan to sell—to help potential buyers make informed bids.

That said, what does it mean to have a well-designed manifest? Consider these three listings from a hypothetical furniture liquidation auction:

1. An expensive couch and 30 cheap dining room chairs
2. An expensive couch and three medium-priced file cabinets
3. An expensive couch and two medium-priced armchairs

Which of these do buyers value most? While these lots are all believable combinations of merchandise with similar total retail values, B-Stock has observed that Auction 3 will consistently outperform the others. But why? Essentially, it’s a matter of grouping similar things together.

Auction 1 offers a big-ticket item—a couch—alongside many inexpensive chairs. They are simply too dissimilar in price and purpose. A small business might not want many low-margin products that can be purchased by consumers as one-offs or in odd numbers, or perhaps they simply don’t have space to store 20 chairs.

Auction 2 pairs the same couch, with three file cabinets—technically furniture, but a completely different flavor of it. These would be much harder for a home furnishing store to sell to its existing customer base. Such items might pair better with a conference table or desks

Auction 3—the couch and two semi-expensive armchairs—represents the best balance between item similarity and price parity. While certain buyers may be open to lots with items of disparate values and uses, they clearly feel that it should be their prerogative to decide that, not the seller’s.

This example illustrates two takeaways, both backed by B-Stock’s proprietary data: First, that auctions with items that are similar in retail prices tend to recover more. Second,  lots that contain items from fewer subcategories—ideally just one—recover more value than highly varied lots.

Essentially, the more subcategories that exist per lot, the higher the chance that a buyer won’t need every item, and will either bid modestly or simply skip that auction.

5. SKU Depth

Another way to evaluate listings is by their SKU depth. Items of a given SKU will always be of the same category and subcategory. But for many types of products, there are multiple characteristics beyond subcategory that differentiate one item from the next.

Apparel provides a good example of this. A “shallow” SKU lot can contain many different products, including shirts, pants, dresses, jackets, men’s and women’s clothes, and so on, all of various colors and sizes. The items in a “deep” SKU” lot will be far more uniform. In an extreme example, such listings can be as specific as “200 mens polyester blend dress shirts, blue, size large.”

But is any particular zone of the SKU depth continuum better for recovery than the other? Consider these three listings:

1. A lot containing pants, shirts, dresses, and socks of all different colors and sizes
2. A lot containing only shirts of all different colors and sizes
3. A lot containing only shirts all of the same color and size

Which will recover most? Buyers generally seem to like specificity, right?  Actually, Apparel buyers faced with this choice  will regularly choose Auctions 1 and 2 over the highly specific Auction 3. 

Often, we see that the Deeper the SKU, the lower the recovery rate. Here’s the tricky part, though. This relationship is not true in all categories however. Take Luggage and Sporting goods, for example. These two types of merchandise benefit from high SKU Depth.

The bottoms line? It’s not safe to make sweeping generalizations about the direction of this relationship without examining plenty of other data. And B-Stock is the ideal partner to help.

Takeaways & Implications

If you’re wondering what exactly you and your business should leave with, we can lay it out.

First of all, you can have confidence that B-Stock knows the result of most auctions before they even close. By plugging these major variables—and many smaller ones—into our machine learning model, we’re able to accurately guess the recovery rate of nearly any sale.

Next, keep in mind that these insights can help you no matter where you are in your recommerce journey. In the early stages, you can watch the three foundational factors—category, brand, and condition—to better forecast your recovery and plan your business accordingly. Once you’re up and running, we encourage you to take control of flexible factors like manifest design and SKU depth to extract even more value from your surplus goods.

As far as actions to take once you begin selling in the B-Stock marketplace, there are a few important things for both new and longtime sellers to keep in mind.

• Be particular and precise. Always provide images, listing titles, and manifest data that are as descriptive and detailed as practically possible.
• Trust in our expert guidance. The best practices and recommendations of our dedicated account managers are always-numbers backed.
• Always challenge the status quo and stay on the lookout for ways to refine your sales strategies—B-Stock will be happy to help you tweak and experiment.

One final note: even the world’s top manufacturers and retailers can struggle to manage mountains of returned and unsold goods that clog warehouses and tie up cash value. Thanks to B-Stock’s fully managed recommerce solution, businesses like yours no longer need to write off that locked-away value as the cost of doing business. 

Reach out to us today to learn how B-Stock can help your retail or manufacturing business reduce its inventory and reclaim value.