By Will Hunter, Founder of Rent with Thred • Updated August 2026 • Thred proprietary rental analysis
A conventional apparel sale records a checkout. A rental business can record a longer sequence: selection, repeated use, wishlist interest, inventory exposure and return to the same product. That creates a different way to study demand.
Rent with Thred’s inventory-adjusted ranking matched 14,692 eligible rental selections to reconstructed product inventory between October 2024 and July 2026. Here are the clearest lessons from that analysis.
1. Inventory Exposure Changes the Meaning of “Popular”
A product with more units and more months on the platform has more chances to be selected. Ranking only by raw rental count can therefore reward inventory abundance rather than true demand. Unit-months let us compare demand with the amount of inventory required to produce it.
2. The Highest Raw Rental Volume and the Highest Efficiency Were Different
The highest raw-selection item in the 150-product sample was the REIS CORD SHIRT/JACKET IN CREAMWOOD from Ottway The Label with 77 selections. The strongest stable inventory-efficiency result, using a 12-unit-month minimum, was the Alex Crane Canvas Cham Pants in Chai at 2.85 customer turns per unit-month.
3. Repeat Selection Identifies Products That Survive the First Wear
The Sunnei Bomber Jacket in Denim led the sample with 8 repeat selections. Several pants, blazers and shirts also generated five or six repeats, showing that repeat use was not confined to one category.
4. Wishlist Interest and Rental Behavior Are Not the Same
The Marine Layer Wyatt Sweater Button-Down led wishlist adds with 12, while the top raw-selection product was different. That gap is useful: intent tells us what catches attention, while rental tells us what wins a limited wardrobe slot.
5. Brand Scale and Brand Efficiency Can Point in Different Directions
Faherty led the major-brand sample on rental selections at 338, while Todd Snyder produced higher selections per unit-month at 1.28 versus 1.10. Both signals matter when deciding how much inventory to buy.
6. A Multi-Brand Wardrobe Makes Substitution Visible
When multiple premium labels sit in the same rental wardrobe, every selection is relative. A Todd Snyder shirt is not only being evaluated against other Todd Snyder shirts; it is competing with Corridor, Faherty, Buck Mason, Alex Crane and everything else available in that customer’s size.
7. Rental Is Useful for Separating Trial From Adoption
A first rental can mean curiosity. A repeat rental is closer to adoption. A buyout can mean the customer decided the garment was worth owning. Those are distinct stages that ordinary one-time sales data often collapses into a single transaction.
8. The Best Operational Metric Depends on the Question
- Selections answer what generated the most rental activity.
- Unique customers answer how broadly a product reached.
- Repeat selections answer what customers came back to.
- Wishlist adds answer what created saved intent.
- Customer turns per unit-month answer how efficiently inventory reached customers.
9. Data Quality Matters More Than a Pretty Ranking
The source analysis explicitly excluded products with negative reconstructed inventory months and documented mapping coverage. Only 14,692 of 17,625 eligible selections could be matched to reconstructed inventory. That is strong enough for useful operational analysis, but not a reason to pretend the dataset is complete.
10. The Next Step Is Repeating the Analysis by Season and Category
Demand changes as the assortment and calendar change. The useful version of this work is not a one-time ranking. It is a repeated seasonal process that asks where inventory dollars are working hardest now.
Methodology and What This Data Can—and Cannot—Tell Us
This analysis uses Rent with Thred’s inventory-adjusted product ranking covering October 2024 through July 2026. The source model matched 14,692 of 17,625 eligible rental selections to reconstructed inventory exposure, or 83.4% of eligible selections. Known subscription products and non-garment items were excluded.
The ranked dataset contains 150 eligible products with reliable non-negative reconstructed inventory history. Inventory exposure is measured in unit-months, meaning one unit held in inventory for one month. Because this is a ranked operational sample rather than 100% of every garment ever carried, totals in these articles should be read as evidence from the matched sample, not as full-catalog market share.
Frequently Asked Questions
How many selections were matched to inventory?
14,692 of 17,625 eligible selections, or 83.4%, were matched to reconstructed inventory in the source analysis.
How many products are in the ranked dataset?
150 eligible products appear in the inventory-adjusted ranking used for these articles.
Why not publish customer-level data?
The useful insights can be produced in aggregate. These articles analyze product and brand behavior without exposing individual customer identities.
The Bottom Line
The biggest value of rental data is not that it creates more numbers. It is that it makes the sequence between interest, trial, repeated wear and ownership more visible.