What Warehouse Automation Vendors Need to Know About How Their Buyers Actually Decide

Primary Frameworks: Robotics ROI Model · Vendor Evaluation Framework · Pilot-to-Scale Failure Framework · Vendor Economics Framework Hub: Insights Decision Question: Why are our warehouse automation proposals getting rejected at financial review, and how do we build the business case the CFO actually runs? Evidence Window: 2020-2026 Author: Deepak Gupta, Founder & Principal Analyst, Autonomy Bridge

Core Question

The intralogistics sector is the most commercially mature robotics market. It is also producing the most documented commercial failures.

Autonomy Bridge’s primary research across 69 intralogistics companies identifies the dominant failure pattern: 21 companies are stalling at sales cycle length. [C1] (Sourced fact , primary research) The proposals are reaching buyers. They are not closing.

The sector research documents the specific mechanism: “Perpetual 10-robot pilots. Ops VP approves pilot, CFO never sees the ROI deck. Fleet decision requires a different buyer than the pilot.” [C2] (Sourced fact , primary research)

The problem is not the product. The problem is not technology skepticism at the buying organization. The problem is that the vendor’s ROI model and the CFO’s ROI model are different documents , and the vendor only produces one of them.

This article draws on Autonomy Bridge’s complete Phase 1 buyer-side research , the Robotics ROI Model, the Vendor Evaluation Framework, the Pilot-to-Scale Failure Framework, and the operator evaluation insights , and translates them to the vendor perspective. The question it answers is: what is the CFO actually calculating when the proposal arrives at financial review?

Why the Question Matters Now

Commercial model failure in this sector does not happen only to early-stage companies. Four named events between 2024 and 2026 demonstrate the scale at which it occurs.

Locus Robotics. Locus reached $230 million in annual revenue and 6 billion total picks processed , the largest RaaS fleet in warehouse picking. The company experienced significant layoffs in 2024. Path to profitability and IPO remains unclear despite those operational metrics. [C12] (Sourced fact) Revenue at this scale confirms product-market fit. The RaaS model requires continuous fleet financing that grows ahead of the revenue base. Commercial scale and economic model viability are separate questions.

Dexory. Dexory raised $165 million in a Series C in October 2025. Its customers include GXO, Maersk, DHL, Stellantis, and GE Appliances. The research entry identifies the commercial constraint precisely: “converting pilots at major 3PLs into wall-to-wall deployments across their entire warehouse networks. Each enterprise customer has hundreds of facilities , the sales cycle to expand from 1-2 pilot sites to 50+ is long.” [C13] (Sourced fact) Enterprise customer logos and Series C capital do not automatically convert pilots to fleet contracts.

Exotec. Exotec has 10,000-plus robots deployed across 200-plus customer sites, with $1 billion-plus in cumulative sales. The company is valued at $2 billion-plus. Its research entry states: “each deal is a multi-million dollar capital project requiring 12-18 month sales cycles. At $2B+ valuation, they need to accelerate deal velocity to justify growth expectations.” [C14] (Sourced fact) Market leadership in ASRS does not eliminate the cycle length problem.

Attabotics. Attabotics raised approximately $200 million, filed Chapter 11 in July 2025, and was acquired by LaFayette Systems in September 2025. The research entry identifies the cause: “hardware costs exceeded what customers would pay , classic pricing/unit economics mismatch.” [C17] (Sourced fact) The new ownership launched an integrator partnership program in April 2026 with SAVOYE as the first partner , structurally addressing the distribution problem that contributed to the bankruptcy.

These cases are not outliers. They are the sector pattern at scale. The commercial failures happen because vendors are solving the wrong problem. They are optimizing for pilot approval when the blocking decision is at fleet approval, and they are building ROI models the operations team finds compelling when the blocking reviewer is in finance.

What the Evidence Shows

Autonomy Bridge’s primary research covers 69 intralogistics companies with identified commercial failure patterns. The distribution is: [C1] (Sourced fact , primary research)

(Autonomy Bridge proprietary analysis, 2024-2026) [C1]

SCS is the largest named-company category. The high-confidence SCS-coded companies include Dexory, Exotec, GreyOrange, Locus Robotics, Mujin, Ocado, Path Robotics, Seegrid, Vecna Robotics, and Witron. [C1] These are not companies with weak products or insufficient market presence. They are commercially established companies whose proposals are reaching the right buyers and stalling before they close.

The sector research documents three failure patterns that explain the stall: [C2] (Sourced fact , primary research)

1. Perpetual 10-robot pilots. Ops VP approves the pilot. CFO never sees the ROI deck. Fleet decision requires a different buyer than the pilot.
2. 9-18 month enterprise cycles at 3PLs. The sales cycle burns capital before close. Dexory closed Maersk and GXO at pilot and is stalling at multi-site.
3. RaaS pricing model pressure. RaaS volume-linked pricing collapses when e-commerce demand corrects. Locus’s 2024 layoffs confirm this pattern.

Pattern 1 is the root cause for most PTS-coded companies. Pattern 2 is the root cause for most SCS-coded companies. They share a common source: the vendor’s commercial motion is designed for the wrong buyer.

GreyOrange reported approximately $750 million in annual revenue as of June 2025 and hired both a new COO and a new CRO in late 2025. [C16] (Sourced fact) Leadership restructuring at that revenue scale signals sales execution challenges , the product and the customer base exist, but the commercial motion is not converting them at the pace the valuation requires.

Witron received record EUR 2 billion-plus in incoming orders in FY2025. The Axfood contract (EUR 265 million, December 2025) will not be operational until 2030. [C15] (Sourced fact) A 5-plus year gap from contract to go-live is the implementation capacity constraint version of the same cycle length problem , not a sales failure, but a delivery capacity constraint that limits the company’s ability to take on new contracts.

Where the Market Is Commonly Misread

The standard vendor commercial presentation leads with throughput improvement, pick rate per hour, labor productivity gains, and a payback period calculated at peak system utilization. The CFO’s financial model starts at a different place and reaches a different conclusion.

Misread 1: The CFO evaluates the same ROI model as operations. Operations evaluates whether the system improves throughput and reduces labor chaos. Finance evaluates whether the capital investment will recover within the depreciation window under realistic , not peak , operating conditions. Those are different questions requiring different evidence. [C5] (Sourced fact , operator evaluation insight)

Autonomy Bridge’s buyer-side research documents the Finance position directly: “Finance favors flexibility over fixed capacity, pushing back when client contracts are shorter than the automation depreciation period or when demand volatility threatens utilization assumptions.” [C5] (Sourced fact)

Misread 2: Vendor ROI models use peak utilization. The Robotics ROI Model documents the documented buyer behavior: “Vendor ROI models typically assume utilization based on peak operating periods rather than annualized facility performance.” [C7] (Sourced fact , Autonomy Bridge framework)

Warehouse automation systems are sized to meet peak demand. They operate at significantly lower utilization during average demand periods. If peak daily orders are three times average daily orders, a system installed for peak capacity operates at approximately 33% utilization during average periods. The same fixed capital cost distributes across fewer processed orders, tripling the effective cost per unit relative to the peak assumption. This is the utilization collapse mechanism , and it is the most common failure mode in warehouse automation economics. [C4] (Sourced fact , Phase 1 insight)

Misread 3: Labor savings are the total labor in the automated process. Automation removes travel labor , the time workers spend walking between storage locations. It does not remove station labor, packing, exception handling, supervision, replenishment, or quality control. Those activities remain. In most piece-picking environments, automation removes a defined fraction of total labor hours, not all of them. [C4] (Sourced fact , Phase 1 insight)

Vendor ROI models frequently conflate gross labor productivity improvements with actual removable labor share. The two numbers are materially different. A model that assumes 80% labor displacement when the true removable share is 40-50% overstates achievable savings by approximately 2x. That error does not survive the CFO’s financial model. [C4] (Sourced fact)

Misread 4: Pilot success proves scale viability. The Pilot-to-Scale Failure Framework identifies five structural constraints that cause systems to succeed in controlled pilots and fail under full deployment: fleet congestion, queue formation, orchestration complexity, facility layout, and integration architecture. [C8] (Sourced fact , Autonomy Bridge framework)

Pilots are designed to maximize the probability of a successful demonstration. They operate in limited fleet configurations, simplified workflows, and concentrated engineering oversight. Full deployment introduces interaction effects , between robots, workers, workflows, and the facility’s WMS , that pilots cannot expose. A pilot that proves a robot can execute a task does not prove the system will sustain performance when robot density doubles and the WES is managing 200 concurrent tasks. [C8] (Sourced fact)

Market Structure and Buyer Reality

Warehouse automation decisions involve two structurally different buyers who have different decision criteria, use different models, and rarely sit in the same meeting unless the vendor explicitly creates that meeting. The vendor who designs the commercial motion for one buyer and ignores the other will reach the fleet approval meeting without the evidence it requires.

The four-stage evaluation structure. Autonomy Bridge’s operator evaluation research documents the process: [C5] (Sourced fact)

1. Vendor shortlist formation , based on operational fit and retrofit feasibility. Operations team leads. Criteria: ceiling height, aisle width, WMS compatibility, SKU profile fit. Finance is not involved.
2. Pilot validation , testing integration risk, throughput reliability, and labor substitution limits. Operations and Engineering lead. Success criteria: throughput approaches modeled capacity, labor reduction is realistic, integration complexity is manageable.
3. Internal ROI modeling , built around utilization thresholds and contract duration. Finance enters here for the first time. The vendor’s ROI model is rebuilt using the operator’s own demand data, client portfolio, and contract durations.
4. Cross-functional approval , Operations, Engineering, and Finance negotiate. Finance pushes back on capital rigidity and demand volatility. Engineering resists WMS modification risk. Operations advocates for throughput improvement.

The vendor participates most actively in stages 1 and 2. Stage 3 is where most proposals fail , and the vendor is typically absent from the room when it happens.

The two-buyer problem. The Ops VP who approves the pilot is not the CFO who approves the fleet. They evaluate on different criteria:

(Autonomy Bridge proprietary analysis, 2026) [C5][C7]

The vendor builds a commercial motion for the Operations VP. That motion succeeds at producing pilots. It fails at producing fleet approvals because the CFO’s evidence requirements are never met. [C5] (Sourced fact , operator evaluation research)

The three-scenario CFO test. Autonomy Bridge’s warehouse automation ROI evaluation documents the financial model operators build for capital committee review. It includes three demand scenarios: [C11] (Sourced fact)

1. Base case , current client mix and seasonal demand profile.
2. Downside , loss of the single largest client account.
3. Stress case , two clients simultaneously reduce volume.

Capital recovery must remain viable under the downside scenario for the investment to be defensible in a 3PL environment. A vendor ROI model that passes the base case but fails the downside scenario will be rejected at financial review , even if Operations leadership supports the project. [C11] (Sourced fact)

The three-scenario test is not pessimism. It is the standard capital allocation methodology for any facility that serves multiple clients on contracts shorter than the automation asset life. A 3PL operator whose largest client represents 30% of volume is exposed to a 30% utilization drop if that client exits. The capital committee will model this before approving a fixed-infrastructure investment.

Economics and Competitive Implications

The economics of warehouse automation are governed by two variables that vendor presentations routinely underweight: the utilization threshold (U_min) and the removable labor share ceiling. When those variables are correctly applied, many vendor ROI models that pass Operations approval fail the CFO’s financial review.

The U_min condition. The Robotics ROI Model establishes the economic condition for automation viability: [C7] (Sourced fact , Autonomy Bridge framework)

U ≥ U_min

Automation becomes economically viable only when system utilization remains above the minimum threshold required for capital recovery across the expected asset life. Below that threshold, fixed capital cost distributes across insufficient throughput volume, and labor savings cannot offset the capital burden. [C7]

The practical consequence is that the utilization assumption in the ROI model is the most consequential single input , and it is the input vendors most consistently inflate. If a vendor builds the ROI at 85% utilization (peak quarter performance) but the facility operates at 45% utilization during average demand, the payback period approximately doubles. The investment that appeared to recover in 4 years at peak utilization does not recover in the asset life at average utilization. [C4] (Sourced fact , Phase 1 insight)

The removable labor share limit. Warehouse automation removes travel labor. It does not remove the labor stack. [C4] (Sourced fact)

In a typical piece-picking workflow, workers spend a defined share of their time traveling between storage locations. Goods-to-person automation eliminates that travel by bringing items to workers. The labor that remains , station operation, item verification, packing, labeling, exception handling, supervision, replenishment, inbound processing , is not removed by the goods-to-person system.

If the true removable labor share is 45% of total labor hours, a vendor ROI model that assumes 80% labor displacement overstates savings by approximately 78%. That error compounds across the payback period calculation. A CFO who discovers this error during financial review will rebuild the model with the correct removable share estimate and will arrive at a materially longer payback period than the vendor presented.

The bottleneck displacement problem. Warehouse throughput is bounded by the slowest node in the workflow: [C4] (Sourced fact , Phase 1 insight)

T_facility = min(T_pick, T_pack, T_sort, T_dock)

Automation that increases pick throughput beyond the facility’s packing capacity increases work-in-process without increasing shipped order volume. Pick rate improves. Shipped orders per day do not. The ROI model that assumes pick throughput improvement translates directly to facility revenue is measuring the wrong output. Finance will model the facility constraint , not the isolated pick rate. [C4] (Sourced fact)

The competitive implication. Vendors who provide the CFO’s model proactively close fleet deals faster. The CFO’s model , U_min at average demand, removable labor share at the workflow level, three-scenario downside test , is not proprietary analysis the operator cannot build independently. The operator’s finance team will build it before any fleet approval. The vendor who builds it first, alongside the Operations team during the pilot, removes the information asymmetry that causes proposals to fail at financial review. (Reasoned inference from buyer decision structure)

The practical result is that vendors who sell to Operations and hope Finance approves are producing the same pipeline stalls documented across the sector’s 21 SCS-coded companies. The commercial motion that converts pilots to fleet deals requires selling to Finance through Operations , which means providing Finance’s evidence package while the pilot is running, not after it concludes.

Formic’s structural advantage. Formic operates a RaaS model for industrial robots with no upfront capital from the buyer. The company reports a 97% renewal rate and 5x deployment growth in 2025. [C18] (Sourced fact) The RaaS model eliminates the CFO’s capital recovery objection by converting the decision from capital committee review to operating expense approval , a lower approval threshold with faster cycle time. The constraint shifts to the vendor side (fleet financing), but the buyer-side commercial cycle shortens materially. This is the pricing model architecture adjustment documented in the Vendor Economics Framework. [C9]

What Decision-Makers Should Conclude

The vendor diagnostic starts with identifying which stage of the four-stage evaluation process the deal is blocked at, and which decision-maker is the blocking approver.

Step 1: Stage identification. Is the deal blocked at shortlist (operational fit not established), pilot (integration or throughput risk not resolved), ROI modeling (CFO has not seen a model that passes the downside scenario), or cross-functional approval (Finance and Operations disagreement is not resolved)?

Most deals reported as “stuck in sales cycle” are blocked at stage 3. The pilot succeeded. The Operations team supports the project. Finance has not been given a model it can approve. This is not a sales problem , it is an evidence package problem. (Reasoned inference from sector pattern)

Step 2: Build the CFO’s model during the pilot , not after. The pilot period is the evidence-gathering window for the financial model. By the time the pilot concludes, the vendor should have:

• The facility’s actual average throughput demand (not peak)
• The facility’s actual removable labor share at the workflow level (travel time as a fraction of total labor hours)
• The facility’s client portfolio and contract durations
• The three-scenario demand model: base, downside (loss of largest client), stress (two clients reduce volume)

These inputs produce the utilization-adjusted ROI model that passes financial review. A vendor who waits until the fleet proposal meeting to introduce this model is asking Finance to approve a capital investment on the same day they receive the financial analysis. That is not how capital committees work. [C5][C11] (Sourced fact , operator evaluation research)

Step 3: Address the bottleneck before the fleet proposal. If the facility’s throughput constraint is at packing, not picking, the fleet proposal cannot justify its ROI on picking improvements alone. Either the scope of the project must include downstream capacity, or the ROI model must be built on the constrained throughput output , not the unconstrained pick throughput. Engineering will identify this. Finance will use it to block approval. The vendor should surface it during the pilot and address it in the proposal scope. [C4][C8] (Sourced fact)

Step 4: For 3PL environments specifically, test against client churn. 3PL facilities serve multiple clients whose contracts may be shorter than the automation depreciation period. The capital committee will run the downside scenario , the vendor should run it first. If the investment does not recover when the largest client exits in year 2, the CFO will not approve the fleet proposal. This is the structural constraint the Warehouse Automation ROI Evaluation documents. [C11] (Sourced fact)

Step 5: Consider pricing model architecture. If the Operations VP supports the project but capital committee approval is consistently blocking fleet deals, the pricing model may be the constraint rather than the ROI calculation. A RaaS model converts the fleet decision from capital committee review to OpEx approval, reducing the approval threshold and shortening the cycle time. The tradeoff is the vendor-side fleet financing requirement , the Vendor Economics Framework provides the structure for evaluating whether that tradeoff is commercially viable. [C9] (Sourced fact , framework)

Remaining Unknowns

Locus Robotics commercial model trajectory. The research file shows $230 million in revenue, 6 billion picks, and 2024 layoffs , but no confirmed Chapter 11 filing as of the evidence window. [C12] Whether the RaaS model generates sustainable margins at this fleet scale, and whether the path to profitability is achievable without a structural pricing change, is not publicly determinable from current data. This is the sector’s most significant open question about RaaS viability at intralogistics scale. (Open question)

Dexory multi-site conversion rate. Dexory has named enterprise customers at pilot scale. The conversion rate from pilot sites to full warehouse network deployments , the commercial metric that justifies the $165 million Series C , is not publicly disclosed. [C13] (Open question)

SCS-coded company blocking stage. The primary research identifies 21 intralogistics companies as sales cycle stall. The specific blocking stage , whether deals are failing at Finance (ROI modeling), at Engineering (integration risk), or at cross-functional approval (internal misalignment) , is not broken down in the dataset. Knowing the blocking stage would sharpen the diagnostic. (Open question , evidence gap)

RaaS model floor in intralogistics. Formic demonstrates 97% renewal and 5x growth on a RaaS model for industrial robots. [C18] Whether the RaaS model is viable at the larger fleet sizes required for goods-to-person ASRS systems , where per-unit costs are substantially higher , has not been demonstrated at comparable scale. The financing requirement for a 200-robot goods-to-person fleet is materially different from a 20-robot palletizing deployment. (Open question)

Frequently Asked Questions

Why do warehouse automation proposals fail at financial review? The most common cause is that vendors build ROI models using peak utilization assumptions and gross labor displacement figures, while CFOs test proposals against average utilization, actual removable labor share, and downside demand scenarios. The vendor’s model and the CFO’s model start from different inputs and reach different conclusions about payback period. When Finance rebuilds the model with realistic inputs and the payback period extends beyond the depreciation window, the proposal is rejected , not because the technology failed, but because the business case was not built for the financial decision-maker.

What is U_min and why does it matter for warehouse automation vendors? U_min is the minimum system utilization required for automation to recover its capital cost over the expected asset life. The condition is U ≥ U_min. Automation becomes economically viable only when throughput volume keeps the installed system sufficiently active. If a warehouse installs automation sized for peak demand but operates at 40-50% of peak during average periods, the effective cost per processed order is 2-2.5x higher than the peak-based ROI model assumed. The payback period extends accordingly. Vendors who do not calculate U_min at average demand are presenting an ROI model that will fail when Finance rebuilds it.

What is removable labor share and how does it affect the business case? Removable labor share is the subset of total warehouse labor hours that automation actually eliminates from payroll. Automation primarily removes travel labor , the time workers spend walking between storage locations. Labor at pick stations, packing, labeling, exception handling, supervision, replenishment, and quality control remains unchanged. In most piece-picking environments, travel labor represents 40-60% of total labor hours. A vendor ROI model that assumes 80% labor displacement when the true removable share is 45% overstates achievable savings by approximately 78%, producing a payback period that is materially shorter than what Finance will calculate.

What is the three-scenario CFO test for warehouse automation? The three-scenario test models automation economics across base case (current client mix and demand), downside (loss of the single largest client), and stress case (two clients simultaneously reduce volume). Capital recovery must remain viable under the downside scenario for a capital committee to approve the investment in a multi-client 3PL environment. A vendor ROI model that shows positive returns under the base case but negative returns under the downside scenario will be rejected at the capital committee stage , even if Operations leadership supports the project. Vendors should build and present all three scenarios during the pilot phase, not at the fleet proposal meeting.

Why does the vendor need to build the CFO’s model during the pilot? The pilot period is the only time the vendor has access to the facility data needed to build an accurate financial model: actual average throughput demand, actual labor time distribution, actual client portfolio and contract durations. By the time the pilot concludes, the vendor should have the inputs for the three-scenario utilization model and the removable labor share calculation. A vendor who waits until the fleet proposal meeting to introduce this analysis is asking the capital committee to evaluate a major investment on the same day they receive the financial model. Capital committees do not work on that timeline. Proposals introduced this way are deferred, not approved.

Evidence Base

Sources used in this article:

1. Problem_Proof_Matrix , Intralogistics Filter , 69 companies: SCS(21), CC(22), PTS(16), QF(8), PM(2). Autonomy Bridge primary research, 2024-2026. [C1]
2. Sector Research , Intralogistics Section , three failure patterns; named companies Locus, 6 River Systems, Covariant, Dexory. Autonomy Bridge primary research, 2026. [C2]
3. Intralogistics Company Research , Dexory, Exotec, GreyOrange, Locus Robotics, Seegrid, Vecna Robotics, Witron, Attabotics, Formic entries. Autonomy Bridge primary research, 2026. [C3]
4. How Warehouse Robotics Economics Actually Works , U_min, utilization collapse, removable labor share, bottleneck displacement mechanisms. Autonomy Bridge, 2026. [C4]
5. How Warehouse Operators Evaluate Robotics Vendors , four-stage evaluation, two-buyer structure, Finance vs. Operations criteria, cross-functional conflict. Autonomy Bridge, 2026. [C5]
6. Why Robotics Pilots Fail to Scale , pilot success vs. scale viability distinction, six scaling failure patterns. Autonomy Bridge, 2026. [C6]
7. Robotics ROI Model , U_min condition, ROI formula, utilization economics. Autonomy Bridge, 2026. [C7]
8. Pilot-to-Scale Failure Framework , five structural scaling constraints: fleet congestion, queue formation, orchestration complexity, facility layout, integration architecture. Autonomy Bridge, 2026. [C8]
9. Vendor Economics Framework , capex/RaaS/hybrid pricing model architecture and utilization economics. Autonomy Bridge, 2026. [C9]
10. Vendor Evaluation Framework , purchase friction mapping, approval paths. Autonomy Bridge, 2026. [C10]
11. Warehouse Automation ROI Evaluation , three-scenario CFO test: base, downside, stress. Autonomy Bridge, 2026. [C11]
12. Locus Robotics revenue, layoff, operational data , $230M revenue, 6B+ picks, 2024 layoffs, no confirmed Chapter 11. Public disclosures, 2024-2026. [C12]
13. Dexory Series C and customer network , $165M Series C Oct 2025; GXO, Maersk, DHL, Stellantis; pilot-to-multi-site stall documented. Public disclosures, 2025-2026. [C13]
14. Exotec scale and valuation , 10,000+ robots, 200+ sites, $1B+ cumulative sales, $2B+ valuation, 12-18 month cycles. Public disclosures, 2025-2026. [C14]
15. Witron order book and timeline , EUR 2B+ FY2025 orders; Axfood EUR 265M; operational by 2030. Public disclosures. [C15]
16. GreyOrange revenue and leadership restructuring , ~$750M revenue June 2025; new COO December 2025, new CRO November 2025. Public disclosures, 2025. [C16]
17. Attabotics bankruptcy and acquisition , Chapter 11 July 2025; LaFayette acquisition September 2025; integrator partnership April 2026. Public record. [C17]
18. Formic RaaS deployment and renewal data , 97% renewal, 5x growth 2025, $59.1M funding, Mitsubishi HC Capital agreement. Public disclosures, 2025-2026. [C18]

Highest-confidence conclusions (sourced fact):

• 21 intralogistics companies stalling at sales cycle length , primary research
• “Ops VP approves pilot, CFO never sees the ROI deck” , sector research direct finding
• Vendor ROI models use peak utilization; operators rebuild at average demand , Phase 1 research
• Removable labor share = travel labor only; station/exception/supervision labor remains , Phase 1 research
• Three-scenario CFO test: base, downside, stress , documented buyer behavior
• Dexory: pilot-to-multi-site stall documented at GXO, Maersk, DHL
• Exotec: 12-18 month sales cycles at $2B+ valuation
• Attabotics: Chapter 11 July 2025, hardware unit economics mismatch

Moderate-confidence conclusions (reasoned inference):

• Vendors who build the CFO’s model during the pilot close fleet deals faster , inferred from buyer decision structure
• Most SCS-coded deals are blocked at Finance ROI stage , inferred from sector pattern
• Bottleneck displacement is a common hidden factor in failed ROI proposals , inferred from Phase 1 framework

Known evidence gaps:

• Locus Robotics: no confirmed Chapter 11 in research file; sector_research.md states it occurred April 2024; discrepancy not resolved
• Dexory multi-site conversion rate , not publicly disclosed
• SCS-coded company specific blocking stage , not broken down in dataset
• RaaS model viability at large AS/RS fleet scale , not demonstrated at comparable scale