Retail Salespersons

41-2031.00

Median wage $35,410/yr3,897,860 employed (US)Rank #97 of 923 scored · top 11% by substitution

Sell merchandise, such as furniture, motor vehicles, appliances, or apparel to consumers.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution47

Exposure42

Augmentation52

Substitution — the headline: capability discounted by cost, barriers and adoption.

Exposure — technical capability alone, regardless of whether anyone deploys it.

Augmentation — how much AI assists without replacing. High here + moderate substitution = a changing job, not a disappearing one.

Tasks on the substitution scale

24 rated tasks, binned by substitution score.

Position among all scored occupations

Distribution of 923 occupation scores; the marker is this occupation.

Tasks with substitution ≥ 70

25%

Run 1.0.0-draft.1 · computed 2026-08-05 · rater panel: claude-sonnet-5, claude-haiku-4-5-20251001 · intervals span rater disagreement.

Why this score

The five weighted dimensions of the composite, averaged across this occupation's tasks (importance-weighted, panel mean). Exact weights and formulas: /api/v1/methodology.

Task automatabilityw 35%44

panel mean rating 2.8/5 → substitution pressure 44/100

Technical feasibility todayw 20%38

panel mean rating 2.5/5 → substitution pressure 38/100

Cost vs. human wagew 15%45

panel mean rating 2.8/5 → substitution pressure 45/100

Adoption barriersw 20%inverted — strong barriers lower the score65

panel mean rating 2.4/5 (barrier strength) → substitution pressure 65/100

Sector adoption velocityw 10%40

panel mean rating 2.6/5 → substitution pressure 40/100

Task breakdown (24 tasks)

Substitution pressure per task, weighted by O*NET importance in the composite. Expand a task for the full rater audit trail — every rating, every model, every rationale.

Maintain records related to sales.

CI 89–100 · exposure 92 · augmentation 63 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	5/5	Retail has been among the fastest adopters of point-of-sale and inventory automation for decades. Automated sales record-keeping is now standard practice across e-commerce, chain retail, and SMB retail, with near-universal deployment.
Sector adoption velocity	claude-sonnet-5	4/5	Retail has broadly adopted digital POS and inventory systems for years, though smaller independent retailers may lag slightly behind large chains.
Augmentation potential	claude-haiku-4-5-20251001	4/5	While automation dominates, AI can augment human salespeople by surfacing insights from sales records (trends, customer patterns, recommendations), enabling better real-time decision-making and performance tracking.
Augmentation potential	claude-sonnet-5	3/5	For salespersons who still manually track certain sales notes or customer follow-ups, AI tools can streamline and organize this, but the core task is more automated than augmented.
Task automatability	claude-haiku-4-5-20251001	5/5	Maintaining sales records is a structured, rule-based task involving data entry, transaction logging, and inventory updates—all readily automatable by current AI and RPA systems. Point-of-sale integrations, inventory management software, and automated data pipelines can capture and organize sales data with >50% time savings at equal or better quality than manual record-keeping.
Task automatability	claude-sonnet-5	4/5	Sales record-keeping (logging transactions, updating inventory counts, generating reports) is largely structured data entry and reporting that POS systems and software already automate or can automate with minimal human review.
Adoption barriers	claude-haiku-4-5-20251001	1/5	No licensing, legal, or regulatory barriers prevent automation of sales record-keeping. Organizations face minimal friction; this is purely a business decision with no required human sign-off or contact requirement.
Adoption barriers	claude-sonnet-5	1/5	No licensing, liability, or regulatory requirement mandates a human maintain these records; it's a purely administrative function already commonly automated.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	AI-driven record-keeping (cloud infrastructure, software licensing, minimal human oversight) costs orders of magnitude less than paying a salesperson to manually log and organize sales data, especially at volume.
Cost vs. human wage	claude-sonnet-5	5/5	Automated record-keeping via existing POS/inventory software costs a tiny fraction of dedicating human labor hours to manual record maintenance.
Technical feasibility today	claude-haiku-4-5-20251001	5/5	Mature, deployed products (e.g., Shopify, Square, SAP, Oracle Retail) perform this at scale daily. Automated record-keeping is standard in retail operations, with reliable integration into inventory, financial, and CRM systems in production across millions of transactions.
Technical feasibility today	claude-sonnet-5	5/5	POS systems, ERP software, and inventory management platforms already handle sales record-keeping automatically in virtually all modern retail environments.

Compute sales prices, total purchases, and receive and process cash or credit payment.

CI 92–95 · exposure 100 · augmentation 38 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	5/5	Retail has nearly universally adopted automated POS and payment systems; manual cash-drawer computation without technology is vanishingly rare even in small retailers, indicating extremely fast and deep adoption in this sector.
Sector adoption velocity	claude-sonnet-5	4/5	Retail has adopted self-checkout, mobile payment, and automated POS systems broadly and quickly over the past two decades, though not universally in all store formats.
Augmentation potential	claude-haiku-4-5-20251001	2/5	Once payment processing is automated, there is limited room for AI to further assist; the human's role shifts to customer service rather than computational tasks, so augmentation potential on this specific arithmetic and payment task is minimal.
Augmentation potential	claude-sonnet-5	3/5	Where humans still perform this task, POS software assists with calculations and payment processing, but the assistance is largely already baked into existing systems rather than novel AI augmentation.
Task automatability	claude-haiku-4-5-20251001	5/5	Point-of-sale systems and payment processors already automate price computation, total calculation, and payment processing end-to-end today, delivering well over 50% time savings at equal quality. Current POS systems handle cash and card payments with minimal human intervention.
Task automatability	claude-sonnet-5	5/5	Computing prices, totals, and processing payments is already fully automated by POS systems, self-checkout kiosks, and online checkout flows with no meaningful time savings left to capture from AI specifically.
Adoption barriers	claude-haiku-4-5-20251001	2/5	While PCI-DSS compliance and fraud prevention add some regulatory friction, there are no licensing requirements mandating human involvement in payment processing itself, and automation is already legally standard. Organizational preference for human presence at checkout remains a minor barrier.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement exists for computing totals or processing payments; some friction remains from customer preference for human checkout and loss-prevention concerns in self-checkout.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	Automated payment processing and POS systems cost pennies per transaction while eliminating the need for manual cashier time on these computational tasks, making them orders of magnitude cheaper than human labor for the same output.
Cost vs. human wage	claude-sonnet-5	5/5	Automated checkout and payment processing costs pennies per transaction compared to the loaded wage cost of a cashier performing the same calculation and transaction handling.
Technical feasibility today	claude-haiku-4-5-20251001	5/5	Mature, production-deployed systems (Shopify, Square, Toast, traditional POS terminals) reliably perform these exact functions at scale across millions of transactions daily in retail organizations worldwide.
Technical feasibility today	claude-sonnet-5	5/5	POS terminals, self-checkout, and e-commerce checkout systems reliably perform price computation and payment processing at massive scale in production today.

Prepare sales slips or sales contracts.

CI 70–100 · exposure 87 · augmentation 63 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	Large retail chains and quick-service sectors have rapidly adopted automated receipt and contract generation integrated with POS systems. Adoption is widespread in digitized retail, though smaller independent retailers lag.
Sector adoption velocity	claude-sonnet-5	5/5	Retail has near-universal adoption of automated point-of-sale and checkout systems, representing one of the most complete digitizations of any commercial task.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI substantially assists sales staff by auto-populating forms, ensuring accuracy, and streamlining document generation, allowing salespeople to focus on customer engagement and closing while the system handles compliance and formatting.
Augmentation potential	claude-sonnet-5	3/5	Where human-prepared contracts still occur (e.g., custom sales), software can auto-fill and speed the process, but this is a narrower slice of the overall task now largely automated outright.
Task automatability	claude-haiku-4-5-20251001	4/5	Preparing sales slips and contracts is largely rule-based document generation with structured data entry. Current AI systems can reliably extract purchase details, fill templates, and generate compliant documents end-to-end, saving at least 50% of time versus manual entry and formatting.
Task automatability	claude-sonnet-5	5/5	Preparing sales slips or contracts is a highly structured, templated data-entry task that POS and e-commerce systems already automate fully with equal or better accuracy.
Adoption barriers	claude-haiku-4-5-20251001	3/5	Some friction exists: retail organizations often require human review of contract terms for liability reasons, customer signature requirements may slow automation, and legacy POS integration can create adoption friction. No hard legal barrier prevents AI from performing the task.
Adoption barriers	claude-sonnet-5	1/5	No licensing, liability, or human-contact requirement blocks automated generation of sales slips; this has been standard practice in retail for decades.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	AI-based document generation and automation tools cost significantly less than the fully-loaded wage of a retail salesperson per transaction processed; inference cost is negligible and integration with existing POS systems is straightforward.
Cost vs. human wage	claude-sonnet-5	5/5	Automated POS/checkout systems cost fractions of a cent per transaction compared to manual paperwork time, an order-of-magnitude or greater savings.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Multiple deployed products (document automation platforms, POS integrations, e-signature solutions) reliably generate sales slips and contracts at scale in retail and service environments. Error rates on standard transactions are low, though edge cases and complex terms may require review.
Technical feasibility today	claude-sonnet-5	5/5	Point-of-sale systems, e-commerce checkout, and contract-generation software are mature, widely deployed products that generate sales slips/contracts automatically at massive scale today.

Answer questions regarding the store and its merchandise.

CI 71–84 · exposure 70 · augmentation 63 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	Retail has been aggressively adopting chatbots, voice assistants, and automated customer service systems for years. Major chains deploy these widely online and increasingly in-store; adoption is fast and measurable across the sector.
Sector adoption velocity	claude-sonnet-5	3/5	Retail is adopting AI chatbots and self-service tools moderately, but many stores still rely primarily on staff, especially smaller and physical-format retailers.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can assist human salespeople by retrieving product details, policies, and inventory status in real time, letting the employee focus on relationship-building and upselling. This is genuinely useful but not transformative, since the salesperson still owns the interaction.
Augmentation potential	claude-sonnet-5	4/5	AI tools like in-store search apps, chat assistants, and inventory lookup systems meaningfully help sales associates answer customer questions faster and more accurately.
Task automatability	claude-haiku-4-5-20251001	4/5	Current AI systems can reliably answer factual questions about store policies, product specifications, inventory, and basic merchandise information by retrieving data from knowledge bases or product databases. This captures most of the routine inquiries; only complex judgment calls or contextual issues requiring deep product expertise would require human intervention, achieving >50% time savings.
Task automatability	claude-sonnet-5	4/5	Chatbots and AI assistants can answer routine product/store questions (hours, location, stock, specs) reasonably well, though complex or nuanced in-person queries still need humans for full equivalence.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Minimal legal or regulatory barriers exist; retailers are free to deploy automated answering systems. Some customer preference for human interaction remains, and integration with inventory/POS systems requires upfront setup, but these are soft friction rather than hard legal blockers.
Adoption barriers	claude-sonnet-5	1/5	No licensing, legal, or regulatory requirement mandates a human answer these questions; retailers can freely deploy self-service kiosks or bots.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	Once deployed, inference cost for answering a customer question is negligible compared to the fully loaded wage of a retail salesperson. A single chatbot can handle hundreds of concurrent inquiries, delivering massive cost advantage.
Cost vs. human wage	claude-sonnet-5	5/5	Automated chat/voice systems answering standard questions cost a fraction of a cent per query versus paying an hourly retail wage for the same interactions.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Deployed conversational AI and chatbot systems already perform this task in production at major retailers (e.g., customer service bots, in-store kiosks, website FAQs powered by LLMs). Performance is reliable for standard product and policy questions, though edge cases and ambiguous queries still occur.
Technical feasibility today	claude-sonnet-5	3/5	Retail chatbots, store apps, and voice assistants are deployed widely for FAQs and product lookup, but accuracy on complex or store-specific nuances remains inconsistent, requiring human backup.

Estimate quantity and cost of merchandise required, such as paint or floor covering.

CI 67–87 · exposure 78 · augmentation 75 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	Large retailers and home-improvement chains have already deployed online estimators and calculators at scale; adoption is accelerating in digital-forward and e-commerce segments, though smaller physical stores lag.
Sector adoption velocity	claude-sonnet-5	3/5	Retail is moderately digitized with self-service kiosks and apps rolling out, but many transactions still occur via in-person staff consultation, especially for complex projects.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI estimation tools significantly augment salesperson productivity by instantly generating accurate quotes and recommendations, allowing sales staff to focus on customer service and upselling rather than manual calculations.
Augmentation potential	claude-sonnet-5	4/5	AI-powered estimator tools significantly speed up and improve accuracy of quantity/cost calculations, letting salespersons focus on customer service and upselling while tools handle the math.
Task automatability	claude-haiku-4-5-20251001	5/5	AI can estimate quantities and costs end-to-end by analyzing room dimensions, material types, and current pricing databases. Current systems (including multimodal tools and e-commerce integrations) reliably perform these calculations with >50% time savings compared to manual measurement and lookup.
Task automatability	claude-sonnet-5	4/5	Estimating quantity/cost from room dimensions and product specs is a well-structured calculation task that AI tools (calculators, chatbots, retailer apps) can already do quickly given inputs like square footage and product coverage rates.'
Adoption barriers	claude-haiku-4-5-20251001	2/5	Minimal legal or licensing barriers exist for automated estimation; however, customer preference for human consultation and organizational inertia in legacy retail systems create moderate friction to adoption.
Adoption barriers	claude-sonnet-5	2/5	No licensing is required for this estimation task, though retailers may prefer human interaction to build trust and upsell, creating mild organizational friction.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	Once integrated into a retail platform, AI-driven estimation costs pennies per calculation versus several minutes of paid staff time, yielding at least 10x cost advantage for high-volume retail operations.
Cost vs. human wage	claude-sonnet-5	4/5	A digital estimator or AI-assisted calculator costs a fraction of a cent per use versus employee time, making it dramatically cheaper once built and deployed.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Mature products exist in paint retailers (e.g., paint calculators, flooring estimators integrated into e-commerce platforms) that perform this task reliably at scale, though some edge cases (irregular room shapes, specialty materials) may require human review.
Technical feasibility today	claude-sonnet-5	3/5	Home improvement retailers (Home Depot, Lowe's) deploy online estimator tools and apps, but many in-store interactions still rely on staff judgment for irregular spaces, waste factors, and product matching.

Inventory stock and requisition new stock.

CI 72–75 · exposure 75 · augmentation 75 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	Retail is a digitizing sector with widespread POS and inventory system deployment; chain retailers and growing e-commerce players have largely automated stock tracking and reordering.
Sector adoption velocity	claude-sonnet-5	3/5	Retail as a sector has adopted inventory automation unevenly—large chains have sophisticated systems while small/independent retailers still rely heavily on manual processes, placing this in the middle range.
Augmentation potential	claude-haiku-4-5-20251001	4/5	Automated inventory systems dramatically enhance salesperson productivity by providing real-time stock visibility, alerting on low inventory, and auto-generating orders, enabling focus on sales rather than manual stocktaking.
Augmentation potential	claude-sonnet-5	4/5	AI-based inventory dashboards, demand forecasting, and reorder alerts significantly boost a salesperson's efficiency in tracking stock levels and flagging needs even when humans remain involved in physical counts and decisions.
Task automatability	claude-haiku-4-5-20251001	4/5	Inventory scanning and stock requisitioning can be largely automated with barcode/RFID systems and automated reorder logic based on threshold algorithms; human judgment on seasonal demand or supplier issues remains, but the core cycle is automatable for 50%+ time savings.
Task automatability	claude-sonnet-5	4/5	Inventory tracking and reordering are largely data-driven processes that can be automated via barcode/RFID systems, POS integration, and automated reorder algorithms with minimal human judgment needed for routine SKUs.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Minimal legal or licensing barriers; adoption depends mainly on upfront integration cost and retailer choice to implement, not regulatory or liability constraints.
Adoption barriers	claude-sonnet-5	2/5	No licensing or legal requirement mandates human inventory counting; main friction is integration cost, physical stock verification needs, and occasional exception handling requiring staff judgment.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	Cloud-based inventory systems cost tens to hundreds of dollars monthly per location, vastly cheaper than the labor hours saved on manual counts and reorder management, particularly at scale.
Cost vs. human wage	claude-sonnet-5	4/5	Once integrated, automated inventory and requisition systems process high volumes of SKU data far cheaper per transaction than manual counting and ordering by an hourly retail worker.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Retail inventory management systems (SAP, Oracle Retail, Shopify Plus, Square) perform automated stock tracking and purchase order generation in production at thousands of sites; implementation and integration require setup, but the core task is reliably deployed.
Technical feasibility today	claude-sonnet-5	4/5	Mature inventory management systems (e.g., automated replenishment in retail ERP/POS platforms like NetSuite, Shopify, Oracle Retail) are deployed at scale in production across large and mid-size retailers today.

Place special orders or call other stores to find desired items.

CI 47–67 · exposure 53 · augmentation 75 · importance 3.7/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	3/5	Retail has adopted e-commerce and order-management systems broadly, but AI-agent automation of the specific task of inter-store lookup and special ordering remains in pilot and early production phases rather than mainstream deployment. Larger chains are moving faster than small retailers.
Sector adoption velocity	claude-sonnet-5	2/5	Retail as a sector has moderate digitization with growing use of inventory management software, but adoption of AI specifically for special-order coordination and inter-store calls remains limited and slow compared to fields like finance or professional services.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI assistants can significantly boost salesperson productivity by instantly checking inventory, suggesting nearby store locations, and auto-drafting orders, allowing staff to focus on customer relationship and upselling. This augmentation is actively being deployed in modern POS and CRM systems.
Augmentation potential	claude-sonnet-5	4/5	AI-powered inventory systems and chat/voice assistants can significantly speed up locating items and drafting special orders, letting sales associates focus on customer interaction while the system handles search and logistics.
Task automatability	claude-haiku-4-5-20251001	4/5	Placing special orders and checking inventory across stores are highly structured tasks involving phone calls or system lookups that current AI agents can largely automate. End-to-end execution (inventory search, order placement, inter-store communication) meets the 50% time-saving threshold, though some edge cases and customer-specific requirements may still require human intervention.
Task automatability	claude-sonnet-5	3/5	Placing an order or checking inventory across stores is a structured, transactional task that AI systems integrated with inventory databases could largely execute, though calling other stores still often requires phone communication or human coordination not fully automated in most retail settings.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Few hard legal or regulatory barriers prevent automation of order placement; however, customer service expectations, human touch preferences, and organizational legacy systems (ERP, inventory databases) create moderate friction to full substitution. No licensing requirement applies.
Adoption barriers	claude-sonnet-5	2/5	No licensing or legal requirement mandates a human for this task, but customer preference for personalized service and the need for judgment in special-order situations create some friction.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	Inference and integration costs for AI agents performing inventory lookups and order placement are substantially lower than the loaded wage of a retail salesperson, especially when deployed at scale across multiple store locations. The labor-per-transaction cost advantage is significant.
Cost vs. human wage	claude-sonnet-5	3/5	Automated inventory lookup systems can be cheap to run per transaction, but integrating and maintaining cross-store systems plus handling exceptions still requires human involvement, keeping costs roughly comparable in many smaller retail contexts.
Technical feasibility today	claude-haiku-4-5-20251001	3/5	Deployed chatbots and order-management systems can handle routine special-order placement and basic inventory queries in production retail environments, but accuracy varies with system integration depth and real-time inventory data quality. Most systems still require some human oversight or escalation for non-standard requests.
Technical feasibility today	claude-sonnet-5	2/5	Some large retailers use inventory-lookup and cross-store fulfillment systems, but the specific act of calling other stores or placing ad hoc special orders is still typically done manually by sales staff, not by deployed AI products at scale.

Maintain knowledge of current sales and promotions, policies regarding payment and exchanges, and security practices.

CI 39–75 · exposure 50 · augmentation 88 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	Retail has rapidly adopted digital POS systems, mobile tools, and knowledge-management apps; major retailers widely deploy these systems for policy and promotion dissemination.
Sector adoption velocity	claude-sonnet-5	2/5	Retail is a moderately digitized sector but frontline staff knowledge management tools are still in early pilot stages rather than widespread production use.
Augmentation potential	claude-haiku-4-5-20251001	5/5	AI systems (search, chatbots, real-time alerts) dramatically assist sales staff by instantly surfacing current promotions, policies, and procedures, transforming their ability to serve customers accurately without replacing their judgment.
Augmentation potential	claude-sonnet-5	4/5	AI-powered internal search tools, chatbots, and training modules can significantly help employees quickly look up current promotions, policies, and procedures, improving speed and accuracy of staying informed.
Task automatability	claude-haiku-4-5-20251001	4/5	Maintaining knowledge of current sales, promotions, policies, and security practices can largely be automated through integrated retail management systems that pull real-time data into accessible formats (receipts, dashboards, chatbots), reducing time spent consulting manuals or searching systems by at least 50%.
Task automatability	claude-sonnet-5	2/5	AI can store and retrieve policy information, but the task requires salespersons to internalize and apply this knowledge live during customer interactions, which is not fully replaceable by AI end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Few regulatory barriers exist to automating internal knowledge distribution; main friction is organizational inertia and employee reliance on direct training, but these are soft barriers.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement, but there is some organizational friction since employees need to demonstrate policy knowledge for compliance and security reasons, and errors in payment/security practices carry liability risk.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	Maintaining a knowledge system (API queries, database hosting, periodic updates) costs far less than paying humans to study and retain all policies and promotions, making AI-driven knowledge infrastructure several times cheaper at scale.
Cost vs. human wage	claude-sonnet-5	3/5	AI-driven knowledge management systems are relatively cheap to run compared to training costs, but human oversight and verification of accuracy still add cost, making it roughly comparable rather than dramatically cheaper.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Mature retail systems (POS, knowledge management platforms, chatbots) reliably serve this function in production at scale across thousands of stores, though some edge cases and policy interpretation still require human oversight.
Technical feasibility today	claude-sonnet-5	2/5	Knowledge-base chatbots and internal AI tools exist to surface policy info, but reliable, comprehensive deployment for keeping staff continuously current on evolving promotions and security practices is uneven across retailers.

Open and close cash registers, performing tasks such as counting money, separating charge slips, coupons, and vouchers, balancing cash drawers, and making deposits.

CI 30–62 · exposure 45 · augmentation 63 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	3/5	Retail is moderately digitized and actively piloting automation (self-checkout, mobile payments, integrated POS), but legacy cash-handling workflows remain common, especially in smaller retailers and franchise locations. Adoption is accelerating but uneven across the sector.
Sector adoption velocity	claude-sonnet-5	2/5	Retail is adopting AI for inventory and customer service faster than for physical cash management, and small-format stores overwhelmingly still rely on manual drawer counts.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI-assisted cash counting, automated reconciliation alerts, and imaging-based verification significantly improve a human cashier's speed and accuracy without removing them from the process. These tools are increasingly deployed to reduce manual counting time and error.
Augmentation potential	claude-sonnet-5	3/5	POS software already assists with automatic balancing, error flagging, and deposit calculations, meaningfully speeding up the reconciliation portion of the task even though physical counting remains manual.
Task automatability	claude-haiku-4-5-20251001	4/5	Most components—counting money, separating charge slips/coupons/vouchers, balancing drawers, and recording deposits—are highly structured, repetitive, and automatable with current computer vision and accounting software. Physical handling of cash and integration with POS systems are straightforward; the task easily exceeds 50% time savings with equal or better accuracy.
Task automatability	claude-sonnet-5	2/5	Physical handling of cash, coupons, and register hardware requires embodied manipulation that current AI cannot perform; POS software already automates the arithmetic but not the physical act.reciprocal reconciliation.rrationale placeholder removed
Adoption barriers	claude-haiku-4-5-20251001	3/5	Regulatory and organizational friction is moderate: cash handling involves audit trails and some liability (discrepancies, fraud detection), and many retailers still prefer human oversight of cash movements for control and security. There is no legal requirement for a human to perform the task, but institutional inertia and risk aversion create friction.
Adoption barriers	claude-sonnet-5	3/5	No licensing requirement, but retailers often require accountable human oversight for cash handling and loss-prevention, creating organizational friction against full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	Cash-handling automation (coin/bill counters, imaging verification, POS integration) has low marginal cost per use and scales across thousands of transactions per location. Over an annual deployment, the per-transaction cost is orders of magnitude lower than paying a retail clerk ($15–18/hour loaded) to manually count and reconcile daily.
Cost vs. human wage	claude-sonnet-5	2/5	Cash-handling automation (e.g., smart safes, automated cash recyclers) exists but requires significant hardware investment, often costing more per-transaction than a low-wage worker performing this task alongside other duties.
Technical feasibility today	claude-haiku-4-5-20251001	3/5	Retail automation systems (self-checkout, mobile payment, integrated POS) handle portions of this task in production, and cash-handling robots exist in banking. However, deployed retail solutions remain narrow (mostly self-checkout) or require significant human oversight, and full end-to-end automation with reliable cash handling is not yet standard practice in most retail environments.
Technical feasibility today	claude-sonnet-5	2/5	Smart registers and POS systems automate balancing calculations, but no deployed AI/robotic system autonomously opens/closes drawers, counts physical cash, and sorts paper slips in production.

Describe merchandise and explain use, operation, and care of merchandise to customers.

CI 25–64 · exposure 38 · augmentation 63 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Retail is traditionally a laggard sector for AI adoption; while some large retailers pilot chatbots and digital kiosks, widespread displacement of sales floor staff explaining merchandise remains minimal and adoption is slow relative to information-sector benchmarks.
Sector adoption velocity	claude-sonnet-5	3/5	Retail has adopted AI chatbots and recommendation engines at moderate pace, especially online, but brick-and-mortar in-person sales assistance still lags digitization trends.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can assist salespersons by generating product descriptions, retrieving inventory or care information, and suggesting talking points, moderately raising their efficiency; however, the core task of live customer explanation and persuasion remains largely human-driven.
Augmentation potential	claude-sonnet-5	4/5	AI tools (product databases, chat assistants, translation, tablets in-store) meaningfully help salespeople answer questions faster and access more accurate information, boosting productivity while the human remains customer-facing.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can generate product descriptions and basic information about merchandise, the task requires real-time interaction with diverse customers, understanding nuanced customer needs, and adapting explanations on the fly. Current AI falls short of the ≥50% time-saving bar when including setup and oversight overhead in real retail settings.
Task automatability	claude-sonnet-5	3/5	AI chatbots and product pages can describe merchandise and explain use/care effectively for standardized products, but in-person nuanced questions, upselling, and physical demonstration still require a human for many retail contexts.4
Adoption barriers	claude-haiku-4-5-20251001	4/5	Customer preference for human interaction in retail remains strong; many retail environments require a human presence on the floor for compliance, store liability, and customer trust. Organizational friction and customer expectations create substantial adoption friction even where automation is technically feasible.
Adoption barriers	claude-sonnet-5	1/5	No licensing or legal requirement mandates a human to explain merchandise use; customers already accept online/AI-driven product information routinely.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI deployment for in-store explanation (kiosks, chatbots, agent systems) requires significant infrastructure, training data, and ongoing maintenance; combined with oversight, it remains comparable to or more expensive than a retail salesperson's loaded wage for equivalent quality output.
Cost vs. human wage	claude-sonnet-5	4/5	Automated product information delivery (chat, FAQs, AI assistants) is far cheaper per interaction than paying a salesperson's hourly wage for repetitive explanations.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Chatbots and product information systems exist, but deployed AI in retail still struggles with live customer engagement, handling unexpected questions, and providing the personalized, contextually appropriate explanations that humans deliver. No mature product reliably performs this end-to-end in production retail environments.
Technical feasibility today	claude-sonnet-5	3/5	Deployed chatbots, virtual assistants, and AI-powered product Q&A tools exist widely in e-commerce, but in-store reliance on AI for this task remains limited and error-prone for complex products.

Bag or package purchases and wrap gifts.

CI 26–57 · exposure 41 · augmentation 13 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Most retail remains in the laggard-to-slow adoption category for bagging automation. Self-checkout bagging stations exist but are niche; traditional retail stores continue to employ human baggers. Adoption is concentrated in large e-commerce fulfillment, not mainstream point-of-sale retail.
Sector adoption velocity	claude-sonnet-5	2/5	Retail as a sector has some digitization (POS, self-checkout) but physical fulfillment tasks like bagging/wrapping see minimal AI or robotic adoption in practice.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI/robotic assistance for human baggers is minimal. Systems that support this task (conveyor layouts, better bag dispensers) are mechanical, not AI-driven. There is limited scope for AI to substantially raise human bagger productivity while keeping them in the loop.
Augmentation potential	claude-sonnet-5	1/5	AI offers essentially no meaningful assistance to the physical act of bagging items or wrapping gifts.
Task automatability	claude-haiku-4-5-20251001	4/5	Bagging and basic packaging are highly repetitive, low-variance tasks with clear procedural rules; robotic systems can perform these at scale today. Gift wrapping adds slight complexity due to variations in item size/shape, but the core wrapping motions remain automatable. Time savings of >50% are already achieved in some high-volume retail and warehouse environments.
Task automatability	claude-sonnet-5	2/5	Physically bagging items and gift-wrapping require manual dexterity and physical manipulation of diverse objects, which is not something current AI systems (as opposed to robotics) can do; robotic solutions for this remain narrow and not widely deployed at checkout counters.
Adoption barriers	claude-haiku-4-5-20251001	2/5	No legal licensing or authorization is required to automate bagging. However, organizational friction is moderate: retailers value the customer-facing service aspect of human baggers, and retrofitting existing checkouts is operationally disruptive. No hard regulatory barrier exists.
Adoption barriers	claude-sonnet-5	2/5	No licensing or legal requirement mandates a human for bagging or gift wrapping, but physical dexterity and customer-facing service norms create practical friction against automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Robotic bagging/packaging systems have high upfront capital and integration costs, plus ongoing maintenance. For small-to-medium retailers with lower transaction volumes, the cost per task often exceeds the loaded wage of a cashier or bagger. Only high-volume operations achieve cost advantage.
Cost vs. human wage	claude-sonnet-5	1/5	Robotic packaging systems capable of handling varied retail items and gift wrap are expensive, bespoke, and far costlier than a low-wage retail employee performing this task.
Technical feasibility today	claude-haiku-4-5-20251001	3/5	Automated bagging and packaging systems are deployed in some large retailers and warehouses (e.g., Amazon fulfillment centers, self-checkout stations), but general retail adoption remains limited. Most stores still rely on human baggers; where automation exists, it often requires human oversight or only handles standard package types, limiting reliability across diverse retail scenarios.
Technical feasibility today	claude-sonnet-5	1/5	No mature deployed product performs general-purpose bagging or gift-wrapping in retail settings; self-checkout systems still require customers or employees to physically bag items.

Recommend, select, and help locate or obtain merchandise based on customer needs and desires.

CI 32–41 · exposure 25 · augmentation 75 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	3/5	Large e-commerce and omnichannel retailers are actively deploying recommendation engines and chatbots, but adoption remains pilot-heavy or limited to specific channels (online only). Traditional brick-and-mortar retail and small retailers lag significantly, resulting in middling overall adoption velocity.
Sector adoption velocity	claude-sonnet-5	3/5	E-commerce recommendation AI is widely adopted, but brick-and-mortar retail staffing and in-store assistance show slower, shallower AI adoption typical of physical retail sectors.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI-powered recommendation engines and product search tools already meaningfully assist retail staff by surfacing relevant inventory and customer preferences, boosting their ability to serve customers faster and more comprehensively while the salesperson retains relationship-building and final decision-making roles.
Augmentation potential	claude-sonnet-5	4/5	AI-powered recommendation tools, inventory lookup systems, and product-finder apps meaningfully help salespersons identify and suggest merchandise faster, even though a human remains central to the interaction.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can assist with product recommendations via recommendation engines and basic chatbots, the task requires understanding nuanced customer needs, preferences, and context that AI systems struggle to capture reliably. End-to-end execution with 50% time savings at equal quality is not consistently achievable today without significant human oversight and judgment.
Task automatability	claude-sonnet-5	2/5	Recommending merchandise involves reading customer cues, physical navigation of a store, and handling items, which current AI cannot fully replace end-to-end; chatbot/recommendation engines only handle the informational subset.
Adoption barriers	claude-haiku-4-5-20251001	3/5	Retail has moderate adoption friction: customers often prefer human interaction for complex purchases, retailers value the personal relationship aspect of sales, and liability concerns around incorrect recommendations create organizational hesitance. However, no hard legal barriers prevent automation.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement, but many customers prefer human interaction for purchase decisions and physical assistance, creating moderate organizational and customer-preference friction.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Implementing and maintaining AI recommendation systems, chatbots, and integrations with inventory systems has non-trivial costs. For smaller retailers, the all-in cost (including oversight, error correction, and system maintenance) often exceeds the wage of a part-time salesperson.
Cost vs. human wage	claude-sonnet-5	3/5	Digital recommendation systems are cheap to run online, but the in-person retail floor component still requires human labor, making blended cost roughly comparable rather than dramatically cheaper.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Some retail e-commerce platforms use recommendation algorithms and chatbots that perform parts of this task, but they have material limitations in understanding complex customer desires and handling edge cases. Few deployed systems reliably perform the full task (recommend, select, locate, obtain) at production quality without human intervention.
Technical feasibility today	claude-sonnet-5	2/5	Online recommendation engines and chat-based shopping assistants exist and are deployed, but in-store physical location/help and nuanced needs assessment remain unreliable for AI products today.

Estimate and quote trade-in allowances.

CI 30–39 · exposure 25 · augmentation 50 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Adoption is limited to a few sectors (high-volume automotive, major retailers). Most retail environments, especially small to mid-size stores, lack the digitization and data infrastructure to deploy trade-in automation, and cultural attachment to personal negotiation remains high.
Sector adoption velocity	claude-sonnet-5	2/5	Retail is a moderate-to-low digitization sector for this specific task; while some valuation tools are used, broad AI-driven trade-in quoting is not yet common practice.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can usefully assist by pulling comparable sales data, suggesting baseline values, and flagging outliers, allowing the salesperson to focus on condition assessment and customer negotiation. This augmentation raises efficiency without removing human decision-making.
Augmentation potential	claude-sonnet-5	3/5	AI-powered valuation databases and comparison tools can meaningfully assist salespeople in quickly estimating fair trade-in values, speeding up part of the task even if final quoting remains human-led.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can retrieve historical pricing data and apply simple valuation formulas, estimating trade-in allowances typically requires assessing condition, market demand, and negotiation nuance—elements that demand context-aware human judgment. Current systems can assist but cannot reliably handle the full range of product variations and situational factors to meet the 50% time-savings bar consistently.
Task automatability	claude-sonnet-5	2/5	Estimating trade-in value can be aided by data lookups, but final quoting involves negotiation, physical inspection, and judgment calls that current AI cannot fully replace end-to-end in a retail sales context.
Adoption barriers	claude-haiku-4-5-20251001	3/5	Customer preference for face-to-face negotiation and trust in human judgment on valuations creates moderate friction. There are no hard legal barriers, but organizational reluctance to remove the human from sensitive pricing interactions and liability concerns around undervaluation provide some protection.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement typically, but customer trust, negotiation dynamics, and physical inspection needs create moderate organizational friction against full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI infrastructure (data integration, model maintenance, human oversight for disputes) is comparable to or exceeds the cost of a retail employee providing quick estimates, especially given the low complexity of individual quotations and the need for accuracy to avoid margin loss.
Cost vs. human wage	claude-sonnet-5	3/5	Valuation software subscriptions are cheap relative to labor, but human oversight, physical inspection, and negotiation still require paid staff time, making cost savings only partial.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Some automotive and retail platforms have rule-based or machine-learning trade-in estimators (e.g., for used vehicles), but these operate at narrow scope and often require human override or manual adjustment. No mature product reliably automates this end-to-end across diverse retail categories without material error rates.
Technical feasibility today	claude-sonnet-5	2/5	Some dealer/retail systems use valuation databases (e.g., for vehicles or electronics) but these are narrow-domain tools, not general reliable products deployed broadly across retail salesperson trade-in scenarios.

Sell or arrange for delivery, insurance, financing, or service contracts for merchandise.

CI 30–38 · exposure 25 · augmentation 63 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	3/5	Large retailers are piloting AI-assisted sales and chatbot checkout, but displacement remains limited. Adoption is faster in e-commerce than physical retail, and most implementations remain assistive rather than fully autonomous.
Sector adoption velocity	claude-sonnet-5	2/5	Retail is adopting AI in recommendation and checkout systems, but the physical, low-digitization nature of most retail sales floors means adoption of automation for this specific task remains slow.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI readily assists by generating personalized product recommendations, contract templates, financing options, and service bundles while a human completes the sale. Sales associates can use these tools to work faster and close higher-value deals with minimal training overhead.
Augmentation potential	claude-sonnet-5	3/5	AI tools like point-of-sale prompts, CRM suggestions, and product-linked financing recommendations help salespeople more effectively pitch add-ons, though the human still closes the sale.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can draft contract terms, recommendations, and financing options, the task fundamentally requires negotiation, judgment about customer needs, and closing—activities that demand human presence and discretion today. Current systems can assist but cannot autonomously complete the full sales cycle with equivalent quality and time savings.
Task automatability	claude-sonnet-5	2/5	This involves in-person persuasion, negotiation, and cross-selling of add-on services which require situational judgment and rapport; AI chatbots can support parts of this online but cannot fully replace the in-store selling and coordination process.
Adoption barriers	claude-haiku-4-5-20251001	3/5	Customer expectations and trust in human salespeople create moderate friction; some jurisdictions impose liability requirements on financing and insurance agents. However, these are not absolute legal barriers, and retailers are increasingly permitted to automate recommendations and contracts.
Adoption barriers	claude-sonnet-5	2/5	No licensing is typically required for retail sales, but financing and insurance add-ons often require compliance with disclosure regulations and sometimes certified staff, creating some friction.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Integration of AI sales assistance (APIs, training, oversight) remains moderately expensive relative to entry-level retail wages. While chatbots are cheap, end-to-end sales automation with contract generation and financing coordination still requires substantial infrastructure investment.
Cost vs. human wage	claude-sonnet-5	2/5	While automated upsell prompts in e-commerce are cheap, in-store equivalent requires human staff or expensive kiosk/AI integration, and the human wage for retail sales is already low, keeping cost advantage modest.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Deployed chatbots and recommendation engines exist in retail, but they struggle with complex negotiations, flexible financing arrangements, and service customization. No production system reliably handles the full sales-to-contract workflow without human intervention and decision-making.
Technical feasibility today	claude-sonnet-5	2/5	E-commerce checkout flows use recommendation engines to upsell warranties/financing, but in physical retail, deployed AI systems do not reliably arrange delivery, insurance, or service contracts through natural conversation with customers.

Greet customers and ascertain what each customer wants or needs.

CI 25–39 · exposure 25 · augmentation 50 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Retail is digitizing slowly compared to information and finance sectors; most stores still rely on human greeters and salespeople. AI adoption is limited to self-checkout and information kiosks in select high-traffic locations, not mainstream replacement of the greeting and needs-assessment function.
Sector adoption velocity	claude-sonnet-5	2/5	Retail is a physical, moderately digitized sector; self-checkout and chat-based tools are spreading but in-person greeting automation is not a major adoption trend.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can assist by suggesting products, pulling customer history, or prompting sales staff with recommendations, but current systems are narrow in scope and require human validation. Staff augmentation tools exist but are not yet transformative to core greeting and needs discovery.
Augmentation potential	claude-sonnet-5	3/5	AI-powered CRM and recommendation tools can help sales associates anticipate customer needs and tailor interactions, offering moderate productivity support without replacing the interpersonal task.
Task automatability	claude-haiku-4-5-20251001	2/5	Current AI cannot reliably greet customers and ascertain their needs in real-world retail environments, which require context-aware conversation, emotional intelligence, and dynamic adjustment. While chatbots can handle simple requests, they frequently misunderstand customer intent and cannot match the nuanced interaction of human sales staff.
Task automatability	claude-sonnet-5	2/5	Physical greeting and needs assessment in a store requires presence and real-time social perception that current AI cannot replicate end-to-end, though chatbots handle a narrow analog online.dentifying customer needs in person remains largely human-driven.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Strong barriers exist due to customer preference for human interaction in retail, risk of poor customer experience and lost sales, and organizational friction in replacing frontline staff who handle exceptions and build relationships. Liability concerns for customer dissatisfaction and brand damage also deter substitution.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement, but customer preference for human interaction and the interpersonal nature of greeting create moderate friction against pure automation.
Cost vs. human wage	claude-haiku-4-5-20251001	3/5	Chatbot infrastructure and AI service costs are now comparable to lower-wage retail positions, but accounting for integration, customization, and oversight still roughly matches the loaded cost of a part-time retail employee.
Cost vs. human wage	claude-sonnet-5	2/5	In-store greeting still requires human staff presence for most retail formats; AI kiosks or apps have deployment and hardware costs that don't clearly undercut low-wage retail labor for this specific task.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Some limited chatbot and kiosk systems exist in retail, but they have high error rates in understanding customer needs and are typically deployed only for basic FAQ responses, not full greeting and need-assessment. No mature deployed system reliably performs this task at scale in traditional retail.
Technical feasibility today	claude-sonnet-5	2/5	Some retailers deploy chatbots or kiosks for online/initial inquiries, but no deployed product reliably greets and diagnoses in-person customer needs at scale in physical stores.

Ticket, arrange, and display merchandise to promote sales.

CI 26–35 · exposure 20 · augmentation 50 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Retail has been slow to adopt physical automation for merchandising tasks; most adoption remains in back-office (inventory systems) rather than floor-level arrangement and display. Pilots exist but production deployment at scale is limited.
Sector adoption velocity	claude-sonnet-5	2/5	Retail is a mixed-digitization sector; while e-commerce and inventory systems adopt AI quickly, in-store physical merchandising tasks see slow robotic or AI adoption due to unstructured environments.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can assist with planogram generation, inventory visibility, and sales analytics to guide where and what to display, meaningfully improving a human salesperson's efficiency in deciding what to highlight and how to arrange stock.
Augmentation potential	claude-sonnet-5	3/5	AI can assist with planogram design, sales data analysis to suggest optimal displays, and price-tagging automation via software, improving efficiency of planning even though physical execution remains manual.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI could assist with arrangement suggestions and inventory tracking, the task fundamentally requires physical manipulation of merchandise in three-dimensional retail space, which current robotics cannot reliably perform at scale. Some digital planning (layout optimization) could save time, but the physical arrangement and in-store display setup remains beyond current automation.
Task automatability	claude-sonnet-5	2/5	This task involves substantial physical manipulation of merchandise, spatial arrangement, and in-store aesthetic judgment that current AI cannot perform end-to-end; AI can assist with planning but not execute the physical work.4/5 remains human labor.'
Adoption barriers	claude-haiku-4-5-20251001	2/5	While not legally licensed, retail merchandising requires human judgment, real-time customer interaction, and flexibility that organizational inertia and customer preferences for human service maintain. However, there are no hard regulatory bars to automation.
Adoption barriers	claude-sonnet-5	2/5	No licensing or legal requirement mandates a human for this task, but physical presence and manual dexterity requirements create a natural barrier to full automation with current robotics.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	The capital cost of robotic systems capable of physical merchandise handling, combined with integration and maintenance overhead, significantly exceeds the loaded wage of a retail associate, especially in lower-margin retail environments.
Cost vs. human wage	claude-sonnet-5	1/5	Since AI cannot perform the physical placement and arrangement, a human must still be paid to do the actual work, making AI substitution costs largely irrelevant or additive rather than cheaper.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed AI or robotic system reliably handles the full end-to-end task of ticketing, arranging, and displaying merchandise in real retail environments. Computer vision systems can identify placement, but integration with physical manipulation and real-time adaptation to store conditions is not in production use.
Technical feasibility today	claude-sonnet-5	1/5	There are no deployed products that physically ticket, arrange, and display merchandise in retail stores; this remains manual work performed by employees.

Demonstrate use or operation of merchandise.

CI 25–35 · exposure 25 · augmentation 50 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Retail remains fragmented and low-digitization in many segments; adoption of robotic or AI-driven product demonstration is still pilot-stage and limited to high-end or technology-forward retailers. Most retail environments continue to rely on human floor staff.
Sector adoption velocity	claude-sonnet-5	2/5	Retail is a mixed-digitization sector; e-commerce has adopted AR/video demos steadily but in-person retail floor demonstration remains largely untouched by AI, showing slow penetration into this specific task.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can assist by providing real-time product specs, inventory lookup, or customer preference recommendations that salespeople display or reference, moderately improving their efficiency in answering questions and identifying fits. However, the human typically remains the active demonstrator.
Augmentation potential	claude-sonnet-5	3/5	AI can supplement human demonstrators with product knowledge lookups, AR overlays, or scripted talking points, improving consistency and speed while the salesperson still performs the interactive demonstration.
Task automatability	claude-haiku-4-5-20251001	2/5	Demonstrating merchandise use requires physical manipulation, spatial reasoning, and real-time customer engagement. Current AI can handle some scripted product information and simple virtual demos, but most retail contexts demand human presence and adaptive physical demonstration that AI cannot reliably perform end-to-end today.
Task automatability	claude-sonnet-5	2/5	Physically demonstrating products to customers in-store requires physical presence, dexterity, and real-time responsiveness that current AI cannot replicate end-to-end; some elements (video demos, chatbot explanations) can be automated but the core in-person task cannot.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Retail sales depend heavily on human customer engagement, trust-building, and legal liability for demonstrating potentially dangerous products. Customer expectations for human interaction and regulatory concerns about product liability create friction against full automation.
Adoption barriers	claude-sonnet-5	2/5	No licensing or legal requirement mandates a human demonstrate merchandise, but customer preference for human interaction and trust in physical handling creates moderate organizational and consumer-behavior friction.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Building and maintaining AI demonstration systems (robots, AR/VR infrastructure, integration costs, oversight) is expensive relative to a salesperson's hourly wage for most retail environments, and does not yet achieve the same quality or customer trust at lower total cost.
Cost vs. human wage	claude-sonnet-5	2/5	Digital demo tools (video, AR) are cheap to run once built, but building/maintaining product-specific interactive demo content plus the still-needed human staff for physical interaction keeps overall costs comparable to human labor for many retail contexts.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Virtual product demonstrations and AI-powered chatbots exist, but they are narrow in scope and cannot reliably perform physical demonstrations in-store. Deployed systems lack the embodied capability and contextual flexibility real salespeople provide, and customer preference for human interaction remains strong.
Technical feasibility today	claude-sonnet-5	2/5	Some retailers deploy video demos, AR try-on tools, or chatbots that explain product features, but no product reliably performs full physical demonstration and interactive customer engagement in-store today.

Exchange merchandise for customers and accept returns.

CI 25–35 · exposure 25 · augmentation 38 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Retail remains fragmented and low-digitization at the task execution level; while large retailers experiment with self-checkout, human-assisted returns remain dominant, and deployment is limited to pilot environments.
Sector adoption velocity	claude-sonnet-5	2/5	Retail is a lower-digitization, high-turnover sector where automated return kiosks and self-checkout are spreading slowly and unevenly, mostly in big-box and grocery chains.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can assist with return eligibility checking, return reason classification, and transaction logging, moderately raising efficiency; however, augmentation is narrower than the full task scope.
Augmentation potential	claude-sonnet-5	2/5	POS software and inventory systems assist by looking up transactions and processing refunds faster, but the core interpersonal and physical inspection work remains largely unaided by AI specifically.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can handle some aspects like receipt processing and return authorization logic, the task requires physical handling of merchandise, customer interaction judgment, and exception handling that current systems cannot perform end-to-end at 50% time savings without significant human oversight.
Task automatability	claude-sonnet-5	2/5	Involves physical inspection of returned merchandise, register/POS handling, and judgment calls on store policy exceptions that require a human present, though self-service kiosks handle simple cases.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Customer preference for human interaction, liability concerns for merchandise damage assessment, and potential regulatory requirements for refund documentation create meaningful friction against full automation; many retailers legally or contractually require human sign-off on returns.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement, but customer preference for face-to-face resolution of complaints, fraud/theft prevention needs, and physical handling of goods create moderate friction.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Current AI solutions require substantial infrastructure (vision systems, robotic handling, integration with POS) that would cost more than the loaded wage of a retail employee performing this task.
Cost vs. human wage	claude-sonnet-5	2/5	Kiosk/automated return systems have upfront hardware and integration costs that are only cheaper than a human at very high volume; for typical retail stores a cashier remains cost-comparable or cheaper.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed product reliably performs the full exchange/return workflow autonomously. Point-of-sale systems assist with transactions, but customer interaction, merchandise inspection, and dispute resolution remain human-dependent in production retail environments.
Technical feasibility today	claude-sonnet-5	2/5	Self-checkout return kiosks and online return portals exist but in-store exchange/return handling with physical merchandise inspection is still predominantly done by human associates.

Watch for and recognize security risks and thefts and know how to prevent or handle these situations.

CI 25–35 · exposure 30 · augmentation 50 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Retail remains fragmented and risk-averse on loss prevention; while major chains are piloting AI cameras, meaningful deployment of AI-only security oversight is rare. Most retailers still rely on human floor staff and dedicated security personnel rather than delegating threat recognition wholly to machines.
Sector adoption velocity	claude-sonnet-5	2/5	Retail loss-prevention tech adoption is growing but concentrated in large chains; most small and mid-size retailers still rely on employee vigilance and basic EAS tags.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI-powered monitoring systems (heat maps, dwell-time analytics, movement alerts) can assist floor staff by highlighting areas or behaviors to investigate, improving their efficiency and coverage. However, the augmentation is partial—human judgment and physical intervention remain essential for handling suspected theft or security incidents.
Augmentation potential	claude-sonnet-5	3/5	AI-powered camera alerts and analytics can help notify staff of suspicious behavior patterns, improving their ability to watch for theft, though the human still handles recognition nuance and response.
Task automatability	claude-haiku-4-5-20251001	2/5	Watching for and recognizing security risks requires visual surveillance and contextual judgment about suspicious behavior, which current AI systems struggle with reliably in the varied, dynamic environment of a retail store. While AI can flag some obvious triggers (movement patterns, abandoned items), the nuanced human judgment needed to distinguish genuine threats from innocent behavior and handle escalation remains difficult to fully automate.
Task automatability	claude-sonnet-5	2/5	AI-based video analytics and EAS systems can flag some theft patterns, but real-time judgment, de-escalation, and physical intervention still require a human presence on the floor.5-9%time savings at best from surveillance-assist tools.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Retail security and loss prevention are heavily regulated, require trained personnel in many jurisdictions, and involve liability for wrongful accusation or improper handling of suspects. Most retailers maintain human security staff or floor associates for legal and reputational protection, creating strong organizational and regulatory friction against full automation.
Adoption barriers	claude-sonnet-5	3/5	No licensing requirement specifically, but liability for wrongful accusations, physical confrontation risks, and store policy typically require trained human staff to make judgment calls and intervene.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Integrating AI security systems (cameras, analytics, monitoring) into a retail environment is capital-intensive, and the ongoing false-positive rates require significant human oversight. The all-in cost per theft prevented or risk mitigated typically exceeds the cost of a floor associate watching for security issues.
Cost vs. human wage	claude-sonnet-5	2/5	Surveillance AI systems require significant camera infrastructure, monitoring staff, and integration costs, and still need human responders on the floor, so net cost savings versus a salesperson's vigilance are modest.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Security cameras with AI-powered anomaly detection exist in some deployments, but they produce high false-positive rates and cannot reliably distinguish between theft, genuine confusion, and normal shopping behavior. Current systems are research-adjacent to production; they assist but do not reliably replace human floor staff for this task.
Technical feasibility today	claude-sonnet-5	3/5	Loss-prevention camera analytics and AI-based anomaly detection are deployed in some large retailers, but they supplement rather than replace the salesperson's situational awareness and response.

Estimate cost of repair or alteration of merchandise.

CI 25–35 · exposure 20 · augmentation 38 · importance 3.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Retail adoption of AI is growing but remains concentrated in customer-service chatbots and inventory; specialized cost-estimation automation for repairs/alterations is not yet common in production retail operations, reflecting the task's complexity and low volume per location.
Sector adoption velocity	claude-sonnet-5	2/5	Retail sales as a sector shows moderate AI adoption in areas like chatbots and inventory, but this specific physical-inspection task sees little to no AI deployment currently.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI could assist by suggesting comparable repair costs, historical pricing data, or labor-hour estimates from a database, helping a salesperson make faster, more consistent estimates while they retain judgment on condition assessment and final pricing.
Augmentation potential	claude-sonnet-5	2/5	AI could assist with reference pricing lookups or standard alteration cost guides, but it offers limited help with the core judgment of assessing physical damage or fit.
Task automatability	claude-haiku-4-5-20251001	2/5	Estimating repair/alteration costs requires judgment about material condition, labor intensity, and market rates. While AI could assist with templates or lookups for standard items, end-to-end estimation without human inspection and expertise is not reliable enough to meet the 50% time-saving bar today.
Task automatability	claude-sonnet-5	2/5	Estimating repair/alteration costs requires physical inspection of items and tacit knowledge of local labor/material pricing, which current general AI cannot reliably perform without human input at the point of interaction.“},
Adoption barriers	claude-haiku-4-5-20251001	3/5	Some friction exists: customers expect a knowledgeable human to estimate costs, and liability for incorrect estimates creates incentive to have a person accountable. However, there is no hard regulatory requirement that a licensed professional must sign off.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement exists, but the task requires physical handling and visual/tactile assessment of goods, creating practical friction against remote automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	The labor cost of a retail salesperson performing this task is relatively low (entry-level wages), and the setup/oversight cost for an AI system would be comparable or higher given the need for human verification and frequent adjustments for accuracy.
Cost vs. human wage	claude-sonnet-5	2/5	Since AI cannot independently perform the physical inspection, any AI assistance would supplement rather than replace the human, so all-in cost including human oversight remains comparable to or higher than just having a person do it.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed retail product reliably performs independent cost estimation for repairs/alterations at scale; this task typically requires human expertise, item inspection, and contextual knowledge that current AI systems cannot consistently provide in production environments.
Technical feasibility today	claude-sonnet-5	1/5	No deployed retail product autonomously inspects merchandise and generates repair/alteration cost estimates in production; this remains a manual, judgment-based task performed by store staff or tailors.

Prepare merchandise for purchase or rental.

CI 24–26 · exposure 16 · augmentation 25 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Retail remains a low-digitization, high-human-contact sector with limited capital investment in task automation; physical preparation work is among the least-automated retail functions.
Sector adoption velocity	claude-sonnet-5	2/5	Retail is adopting AI for scheduling, inventory, and customer service, but physical merchandise prep remains largely untouched by automation outside of some self-checkout kiosks.
Augmentation potential	claude-haiku-4-5-20251001	2/5	Digital tools (inventory management, pricing systems) can assist with decision-making around what to prepare and how to price it, but provide minimal assistance with the core physical preparation work itself.
Augmentation potential	claude-sonnet-5	2/5	AI can assist with inventory tracking or printing labels/tags to support this task, but it doesn't materially transform the physical preparation work itself.
Task automatability	claude-haiku-4-5-20251001	2/5	Most of the task involves physical handling—tagging, arranging, inspecting condition—which requires dexterity and spatial reasoning in unstructured retail environments. Current robotics can handle this only in highly controlled settings; general retail preparation remains primarily manual work.
Task automatability	claude-sonnet-5	2/5	This involves physical handling like tagging, folding, packaging, or activating items which requires manipulation of physical goods in-store, not something current AI can execute end-to-end without robotics.'
Adoption barriers	claude-haiku-4-5-20251001	2/5	Physical retail work has low regulatory barriers and no licensing requirement, but organizational friction around replacing manual work and customer preference for human-touched merchandise create modest adoption friction.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement, but the physical nature of the work (handling, security tagging, fitting for rentals) creates practical barriers to non-human automation.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The capital cost of robotics capable of handling diverse merchandise types, plus integration and maintenance, far exceeds the loaded wage of retail associates who perform this task efficiently and flexibly.
Cost vs. human wage	claude-sonnet-5	1/5	There is no AI system substituting for this physical task, so cost comparison favors the human worker entirely.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably prepares merchandise end-to-end in real retail environments today. Some inventory and labeling software exist, but the core physical tasks (wrapping, tagging, arranging displays) remain outside practical AI/robotic deployment at retail scale.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs the physical preparation of merchandise for sale or rental; this remains a manual, in-person retail activity.

Clean shelves, counters, and tables.

CI 19–24 · exposure 16 · augmentation 0 · importance 3.7/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Retail remains a laggard sector for automation due to physical dexterity demands, low per-unit profitability, and customer-facing constraints. Actual displacement via robotics cleaning is negligible in production retail environments.
Sector adoption velocity	claude-sonnet-5	1/5	Retail cleaning tasks are physical and low-digitization; adoption of robotics for this specific task is minimal, unlike white-collar sectors seeing rapid AI uptake.
Augmentation potential	claude-haiku-4-5-20251001	1/5	Current AI offers no meaningful assistance to human retail cleaners performing this task; cleaning tools remain simple manual implements with no augmentative AI component.
Augmentation potential	claude-sonnet-5	1/5	AI offers essentially no assistance to a human performing manual cleaning of shelves, counters, and tables.
Task automatability	claude-haiku-4-5-20251001	2/5	Physical manipulation of retail shelves, counters, and tables requires dexterous robotics and spatial reasoning that current general-purpose systems cannot reliably execute. While narrow specialized robots exist in controlled settings, no scalable off-the-shelf AI system today achieves 50% time savings over human cleaners on this task.
Task automatability	claude-sonnet-5	2/5	Physical cleaning of shelves, counters, and tables requires manipulation in unstructured retail environments, which current AI-driven robotics cannot reliably perform end-to-end at equal quality with major time savings. Software AI has no role here since it's a physical labor task.
Adoption barriers	claude-haiku-4-5-20251001	3/5	While cleaning is not licensed, organizational friction around robotics deployment in customer-facing retail spaces, liability concerns for equipment damage, and preference for human staff create moderate adoption friction.
Adoption barriers	claude-sonnet-5	2/5	No licensing or legal requirement mandates a human clean shelves, but physical space constraints, liability for damaging merchandise, and the need for judgment about clutter create moderate practical friction against automation.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Current robotic cleaning systems require significant upfront capital investment, specialized infrastructure, maintenance, and oversight, making them substantially more expensive than deploying a human retail worker per unit of cleaning completed.
Cost vs. human wage	claude-sonnet-5	1/5	Purchasing, deploying, and maintaining robotic cleaning systems capable of handling varied shelf/counter/table cleaning is far costlier than paying a low-wage retail worker to do it manually.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed commercial product reliably performs end-to-end shelf and counter cleaning in production retail environments. General-purpose robotics in retail remain in pilot phase with high error rates and limited scope.
Technical feasibility today	claude-sonnet-5	1/5	No mature, widely deployed commercial robotic products perform general retail cleaning of shelves/counters/tables in occupied stores today; existing floor-cleaning robots address only a narrow subset of surface types.

Rent merchandise to customers.

CI 11–30 · exposure 13 · augmentation 50 · importance 3.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Retail rental automation has been slow outside niche sectors (e.g., car rental, streaming services). Most small-to-mid retail rental operations still rely heavily on manual processes, and adoption of AI-driven rental systems remains limited compared to other retail functions like sales or inventory management.
Sector adoption velocity	claude-sonnet-5	2/5	Retail rental businesses are generally low-digitization, physical-goods-focused sectors where AI adoption for such tasks remains limited to booking/scheduling software rather than the core rental task.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can assist retail staff by automating eligibility checks, suggesting pricing, generating rental agreements, and flagging high-risk customers, meaningfully improving salesperson efficiency on administrative aspects while the human retains control over final approval and customer relationship management.
Augmentation potential	claude-sonnet-5	3/5	AI can assist with online booking, inventory tracking, and automated invoicing, improving efficiency around the task even though the core physical handoff isn't automated.
Task automatability	claude-haiku-4-5-20251001	2/5	A significant part of the rental process—inventory lookup, pricing calculation, contract generation—can be automated, but the task requires nuanced customer interaction, payment processing oversight, and judgment about customer reliability that current AI struggles with end-to-end. Time savings are unlikely to reach 50% at equal quality without substantial human involvement.
Task automatability	claude-sonnet-5	1/5	Renting merchandise involves physical handoff of goods, contract execution, inspection/condition checks, and in-person customer interaction that current AI cannot perform end-to-end.'
Adoption barriers	claude-haiku-4-5-20251001	4/5	Rental transactions carry significant liability for damage, loss, and fraud; many jurisdictions require human accountability for contract formation and customer verification. Additionally, customers often expect human interaction for high-value or complex rentals, creating both legal and organizational friction against full automation.
Adoption barriers	claude-sonnet-5	3/5	No licensing requirement typically exists, but liability for damaged/lost merchandise, customer trust, and physical custody of goods create meaningful organizational friction against full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Integration of rental software, payment systems, and fraud detection, combined with ongoing human oversight for exceptions and disputes, keeps all-in costs comparable to or potentially higher than direct human staff for many retail contexts, especially small merchants.
Cost vs. human wage	claude-sonnet-5	2/5	While self-service kiosks and online booking can reduce some labor costs, the physical handling and inspection steps still require human labor, keeping overall costs comparable to human-staffed operations.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	While some e-commerce platforms offer basic rental functionality with automated checkouts, production systems handling the full lifecycle (eligibility verification, damage assessment, customer communication, dispute resolution) remain limited and typically require significant human oversight. Reliable end-to-end rental management by AI alone is not yet demonstrated at scale.
Technical feasibility today	claude-sonnet-5	1/5	There are no deployed AI products that physically hand over rental merchandise, inspect items, or manage the full rental transaction autonomously in production.

Help customers try on or fit merchandise.

CI 5–13 · exposure 0 · augmentation 25 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Retail sectors have shown minimal adoption of automation for hands-on fitting tasks; this remains almost entirely human-performed in physical stores because of the physical and interpersonal demands.
Sector adoption velocity	claude-sonnet-5	2/5	Retail is adopting AI for recommendations, chatbots, and inventory, but hands-on fitting assistance sees essentially no AI deployment.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AR/virtual fitting tools provide limited assistance by showing how items might look, but they do not meaningfully augment the human salesperson's ability to physically fit and adjust merchandise on a customer in real time.
Augmentation potential	claude-sonnet-5	2/5	AI can offer some assistance via virtual try-on apps or size recommendation tools, but it doesn't meaningfully augment the physical act of helping a customer try on merchandise in-store.
Task automatability	claude-haiku-4-5-20251001	1/5	Helping customers try on or fit merchandise requires physical manipulation of garments, real-time adjustment to body shape and comfort feedback, and nuanced interpersonal judgment. Current AI cannot perform the physical aspects of fitting or the embodied assessment of how items feel and look on a person.
Task automatability	claude-sonnet-5	1/5	Physically assisting a customer with trying on clothing, shoes, or other merchandise requires physical presence, dexterity, and interpersonal interaction that current AI cannot perform.'
Adoption barriers	claude-haiku-4-5-20251001	4/5	Strong barriers exist: customer preference for human interaction during fitting, liability concerns if a robotic or autonomous system damages merchandise or injures a customer, and the intimate personal nature of garment fitting that customers expect to be handled by a human.
Adoption barriers	claude-sonnet-5	3/5	No licensing barrier exists, but physical presence and customer service expectations create strong practical friction against automation.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Setting aside the automation barrier, any robotic system capable of physical fitting would be vastly more expensive to deploy, maintain, and insure than paying a retail associate's loaded wage for this service.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI substitute performing this physical task, so any comparison of cost favors the human by default.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably performs the core physical task of helping a customer try on and fit merchandise in a real retail environment. Virtual try-on systems exist but do not replace the hands-on fitting assistance required by this task.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product exists that physically fits or assists customers with merchandise; this remains purely a research/robotics frontier problem, not a commercial reality.

Related occupations — Sales & Related

How to read this

A high substitution score does not mean this job disappears — it means a large share of its current tasks face replacement pressure, so the mix of tasks is likely to change. High augmentation alongside substitution typically means the occupation reorganizes around the protected tasks. Wide confidence intervals mean the rater panel disagreed: treat those scores as open questions, not verdicts.

What would change this score

New model capabilities (automatability, feasibility), falling inference costs (cost ratio), regulation and licensing shifts (barriers), and measured sector adoption (velocity) all re-enter at every index release. Each release is recomputed, versioned and kept queryable — scores are claims with a date on them, not permanent labels.