Tractor Supply just spent more than $200 million to own an 865,000-square-foot automated warehouse in Idaho. Sixteen days later, a Brooklyn startup raised $62 million to keep renting out packing robots, so its customers never buy one [1][2][3].

The big picture:

Both moves answer the same question: how do you push more orders through the same four walls? They just put the cost on opposite sides of the ledger.

Tractor Supply bought the whole stack. The 116-acre site, the building, the 80 truck docks and the 125-shuttle storage grid all sit on its own books [1][2]. It even pays a local team to put artificial intelligence to work on daily tasks [1].

To earn that fixed cost back, the retailer must hit a demand forecast it wrote itself. The site opens serving 123 stores across nine states. It has room to reach more than 200 [1].

About 81% of what its stores receive already moves through this network [8].

Walmart made the same heavy bet. It just opened its fifth next-generation center, in Stockton, California. The site covers more than 900,000 square feet, and Walmart says it cuts a 12-step order flow down to five [5].

Ultra’s buyers take the opposite path. They pay a setup fee, then a monthly charge for hardware and software [3]. The cost lands on the operating budget instead of the long-term capital budget.

What strikes me here is that renting moves the machine, not the building. Ultra’s packing robots sit inside sites the renter already leases [3]. The floor space stays tied up.

The part I keep circling is the price risk. Ultra’s CEO, Jon Miller Schwartz, told Fortune that strong demand already let the company raise its prices [3]. And the software comes from Physical Intelligence.

That firm is worth $5.6 billion, and the renter never signs with it [3][4].

Harris Williams says operators want more output from the buildings they already have. The bank expects the market to double, from $23.5 billion in 2025 to $47 billion by 2030 [6].

A packaging study reported the day before says the money is flowing to the last manual steps of the line [7]. Collaborative robots are winning those spots because they cost less up front [7].

By the numbers

  • $200 million-plus — The Nampa build: The retailer’s cost for the 865,000-square-foot Idaho site, which opens serving 123 stores and has room for more than 200 [1].
  • 125 — Owned shuttle units: The battery-powered robots inside Tractor Supply’s first storage grid, one of eleven warehouses and the only one with it [1][2].
  • $62 million — Ultra’s funding: A $50 million Series A and an earlier $12 million seed round backing its monthly robot rental model [3][4].
  • 12 to 5 — Walmart’s flow squeeze: The step reduction Walmart reports at its 900,000-square-foot Stockton center [5].

What I’d watch:

I’ve been watching the contract details that decide who holds the risk once the newness wears off. None of these hinge on the headline price.

  • The renewal clause: A monthly term with a price-change right is not the same as a fixed rate. Ultra already raised prices, so the renter’s real risk sits in the fine print [3].
  • The support boundary: When a robot drops an order, operators need to know who to call. Ultra builds the machines. Physical Intelligence supplies the brains. That leaves two vendors sharing one promise of uptime [3][4].
  • The site commitment: The robot still eats up the renter’s floor space. Renting does not move the real estate cost. That is why fitting robots into old buildings matters most here [3][6].
  • The second site: Nampa is the only Tractor Supply warehouse with this specific grid [2]. I’d want to see the retailer install the same kit at a second site before treating the payback model as proven [2].

The catch

No clear payback figure exists for either choice. I could not find one hiding in the filings.

Tractor Supply’s $200 million-plus price tag lumps the land, building, docks and system together [1]. The company has not split out the automation spend, so any per-shuttle cost would be a guess.

Ultra has not published a rate at all. Its CEO said the business has booked real revenue without giving an amount, so the monthly bill stays a secret [3].

My read is that this business has no clear answer yet. It only has a clear question: who can change the price after the ink dries?

At a glance

  • The Big Shift: Tractor Supply just poured more than $200 million into its first storage grid, across eleven warehouses, while Ultra raised $62 million to rent packing robots by the month instead of selling them [1][3].
  • Why It Matters: Buying fixes your costs, but you still hold a demand forecast you wrote yourself. Renting moves the big expense off the books, while your floor space, monthly rate and software plan sit in someone else’s hands [1][3][4].
  • What I’d Watch: The contract terms that carry the real risk once the system is live.
  • The renewal clause: Whether the monthly rental rate is fixed or open to sudden hikes [3].
  • The support boundary: Which vendor actually owns the uptime promise when the hardware and software come from two different companies [3][4].
  • The second site: Whether Tractor Supply expands the grid to another warehouse, proving the model works [2].
  • The Catch: Neither side has published a clear payback figure, and Tractor Supply has not separated its automation spend from the total building cost [1][3].

Related reading

Sources

[1] Tractor Supply Company, “Tractor Supply Celebrates Grand Opening of 11th Distribution Center in Nampa, Idaho” (23 September 2026) — https://corporate.tractorsupply.com/newsroom/news-releases/news-releases-details/2026/Tractor-Supply-Celebrates-Grand-Opening-of-11th-Distribution-Center-in-Nampa-Idaho/default.aspx [2] FreightWaves, “Tractor Supply inaugurates $200M distribution facility in Idaho” (24 September 2026) — https://www.freightwaves.com/news/tractor-supply-inaugurates-200m-distribution-facility-in-idaho [3] Fortune / Term Sheet, Jeff John Roberts, “Ultra raises $62 million for fast-growing ‘robots as a service’ business, announces tie-up with AI research firm Physical Intelligence” (9 October 2026) — https://fortune.com/2026/10/09/ultra-raises-62-million-fast-growing-robots-service-tie-up-ai-research-firm-physical-intelligence [4] Dealroom, “Ultra raises $50M Series A to scale its ‘robots as a service’ warehouse business” (9 October 2026) — https://dealroom.co/news/161151-ultra-raises-50m-series-a-to-scale-its-robots-as-a-service-warehouse-bus [5] Walmart, “Walmart Opens Next-Generation Fulfillment Center in Stockton, Expanding Faster Delivery Across the West Coast” (8 October 2026) — https://corporate.walmart.com/news/2026/10/stockton-next-generation-fulfillment-center [6] Harris Williams, “Warehouse Automation Report: AI, Robotics, and the Future of Fulfillment” (9 October 2026) — https://www.harriswilliams.com/our-insights/ind-warehouse-automation-report-m&a-trends-ai-robotics-fulfillment [7] Modern Materials Handling / PMMI, “PMMI report: Warehouse automation continues to drive innovation” (8 October 2026) — https://www.mmh.com/article/pmmi_report_warehouse_automation_continues_to_drive_innovation [8] Supply Chain Dive, “Tractor Supply opens Idaho distribution center with automation, dedicated AI team” (7 October 2026) — https://www.supplychaindive.com/news/tractor-supply-opens-idaho-distribution-center-with-automation-dedicated-a/831883