{"api":{"name":"api.sb","description":"Business-as-Code surface for Startups.Studio","home":"https://api.sb","docs":"https://api.sb/docs","version":"1.0.0"},"$context":"https://api.sb/$context","$type":"FoundingHypothesis","$id":"https://api.sb/founding-hypotheses/fh%3Abuyers-and-purchasing-agents-spend-analysis-report%3At-ais-low-tech%3Av1","links":{"self":"https://api.sb/v1/founding-hypotheses/fh%3Abuyers-and-purchasing-agents-spend-analysis-report%3At-ais-low-tech%3Av1","canonical":"https://api.sb/founding-hypotheses/fh%3Abuyers-and-purchasing-agents-spend-analysis-report%3At-ais-low-tech%3Av1","pool":"https://api.sb/v1/founding-hypotheses"},"foundingHypothesis":{"id":"fh:buyers-and-purchasing-agents-spend-analysis-report:t-ais-low-tech:v1","lens":"HeadlessSaaS","type":"founding-hypothesis","click":{"rubricScores":{"C8_lensFit":1,"C7_magicLensFit":1,"C4_competitorHonesty":1,"C6_crossSlotCoherence":1,"C1_customerSpecificity":1,"C2_problemFrictionRealism":1,"C9_killCriteriaAttestability":1,"C3_approachEngineCoverability":1,"C5_differentiationLoservilleEscape":1},"upperRightLoserville":true},"cellRef":{"id":"work-contexts.org.ai/buyers-and-purchasing-agents-spend-analysis-report","stableHash":"wcc:buyers-and-purchasing-agents:spend-analysis-report:document:v1"},"problem":{"slotStatement":"Every category spend-analysis report the startup ships today costs roughly $180–$400 in engineer + analyst time (3–6 hours stitching ERP extracts, UNSPSC/NAICS reclassification, supplier concentration math, and a sourcing-strategy narrative), and half of those hours are wasted re-doing the work when the jurisdiction boundary is wrong — a GPO-covered SKU gets priced as open-market, a state-contract line gets rolled into corporate spend, or a subsidiary's tax-jurisdiction rebate is missed — so the report lands on the purchasing agent's desk and gets kicked back for rework, forcing a wet-signature form to be re-routed for endorsement before the sourcing decision can move."},"approach":{"oneSentence":"Pay-per-report spend-analysis API: POST a CSV of GL+PO lines, get back a JSON category tree with UNSPSC codes, supplier-concentration stats, jurisdiction-tagged sourcing recommendations, and a per-line authority trace — $0.25 per spend line classified or $4 per full category report, free tier of 500 lines/month, paste an API key and run with no implementation call."},"customer":{"icpShape":"VP Engineering or Head of Data at a 40–200-person VC-backed SaaS startup (targetCompanyType: company-type-vc-saas-startup) that sells into procurement/sourcing teams and ships a spend-analysis or category-intelligence feature — where the buyer is the VP Engineering (signs the credit card / controls the dev-tools budget line) and the daily user is a senior backend or data engineer who'd rather curl an endpoint than book a demo.","beachheadShape":"EarlyAdopterJTBD: small VC-backed procurement/spend-intel SaaS startups whose own customers are buyers & purchasing agents, and who need a per-category spend-analysis & sourcing-strategy report generator they can embed under their own brand without building the classification + benchmark logic in-house."},"archetype":"startup-archetypes.org.ai/HeadlessSaaS-UsageMetered","beachhead":"EarlyAdopterJTBD: small VC-backed procurement/spend-intel SaaS startups whose own customers are buyers & purchasing agents, and who need a per-category spend-analysis & sourcing-strategy report generator they can embed under their own brand without building the classification + benchmark logic in-house.","competitors":{"substitutes":[{"name":"The in-house Python + pandas script the data engineer already wrote","category":"informal","uncopyabilityReason":"Not a product — it is the DIY baseline, but it has no UNSPSC/NAICS crosswalk, no GPO/state-contract jurisdiction table, and no supplier-master dedupe, so it silently miscategorizes 15–30% of lines across contracting authorities."},{"name":"Wet-signature sourcing-approval form physically routed for endorsement","category":"manual-bridge"},{"name":"Coupa Spend Analysis (Spend 360)","category":"incumbent","uncopyabilityReason":"Deployment-mode mismatch: Coupa sells a full suite-license seat model with a 60–90 day implementation, SSO/SOC2 review, and taxonomy-workshop onboarding; it cannot re-ship as a $0.25/line paste-an-API-key endpoint without cannibalizing its seat-based contract floor and dismantling its Services org — the entire GTM is priced against implementation revenue."},{"name":"SAP Ariba Spend Analysis","category":"incumbent","uncopyabilityReason":"Integration depth: Ariba's classification engine is wired into the Ariba Network supplier graph and SAP MDG master-data, and its jurisdiction logic lives inside the SAP tax/contract modules. Exposing a clean per-event API that a 40-person startup could embed would require peeling classification out of a graph that is the reason CFOs buy Ariba in the first place."},{"name":"SpendHQ / GEP SMART spend analytics","category":"incumbent","uncopyabilityReason":"Vertical-data-corpus lock-in: SpendHQ and GEP's moat is their curated taxonomy + benchmark corpus collected under NDA from enterprise spend-cube engagements, licensed for use only inside their hosted UI. They cannot publish a per-call API against that corpus without violating the data-sharing terms that built it."},{"name":"ChatGPT / Claude with a pasted CSV","category":"AI-native horizontal","uncopyabilityReason":"Liability posture + vertical-data-corpus gap: a horizontal LLM has no authoritative UNSPSC↔NAICS crosswalk, no GPO/state-master contract registry, and no stable per-line trace — it will confidently cross jurisdictions and hallucinate supplier parents, and neither OpenAI nor Anthropic will attach the authority-boundary guarantee that a purchasing agent can defend to their controller."}]},"studioThesis":"T-AIS-LOW-TECH","killThreshold":{"K":11,"M":45,"N":8,"rubricItemSet":["C1_customerSpecificity","C2_problemFrictionRealism","C3_approachEngineCoverability","C4_competitorHonesty","C5_differentiationLoservilleEscape","C6_crossSlotCoherence","C7_magicLensFit","C8_lensFit","C9_killCriteriaAttestability"],"verdictPolicy":"all-load-bearing-pass-and-overall-ge-X","loadBearingItemSet":["C1_customerSpecificity","C2_problemFrictionRealism","C3_approachEngineCoverability","C4_competitorHonesty","C5_differentiationLoservilleEscape","C6_crossSlotCoherence","C9_killCriteriaAttestability"],"verdictPolicyVerbatim":"KILL unless every load-bearing rubric item passes per workbook AND overall pass-rate ≥ 8/9 (CASCADE.md §4 Stage 9 commit threshold)."},"lifecycleState":"Active","differentiation":{"twoByTwo":{"xAxis":"Per-report cost shape (seat-license + implementation fee ↔ pay-per-spend-line API, free tier, no minimum)","yAxis":"Jurisdiction-authority resolution (single flat category tree ↔ per-line GPO / state-master / corporate / subsidiary authority tag with reconcile trace)","winningQuadrant":"Pay-per-line API with per-line jurisdiction-authority tagging — a 40-person spend-intel startup pipes GL+PO lines in, gets back a category tree where every line carries the contracting-authority it falls under and why, and pays $0.25/line with no implementation call.","loservilleEscape":true,"loservilleQuadrant":"Seat-licensed suite (Coupa/Ariba/SpendHQ) with a single flat category tree and no jurisdiction-authority field — the documented failure mode where GPO-covered SKUs get priced as open-market and the report gets kicked back by the purchasing agent for rework."}},"unmetRequirements":[],"pricingArchitecture":"usage-meter"},"actions":{},"options":{},"relationships":{"runtimeUnit":"https://api.sb/v1/runtime-units?startupRef=startup%3Afh%3Abuyers-and-purchasing-agents-spend-analysis-report%3At-ais-low-tech%3Av1","brand":"https://api.sb/v1/brands?startupId=startup%3Afh%3Abuyers-and-purchasing-agents-spend-analysis-report%3At-ais-low-tech%3Av1","listing":"https://api.services/listings?foundingHypothesisRef=fh%3Abuyers-and-purchasing-agents-spend-analysis-report%3At-ais-low-tech%3Av1","cell":"https://api.sb/v1/cells/work-contexts.org.ai/buyers-and-purchasing-agents-spend-analysis-report","thesis":"https://api.sb/v1/theses/T-AIS-LOW-TECH"},"meta":{"level":"L0","scopes":[]},"user":{"requestId":"a0575ba6a9e6a8fe","edgeLocation":"a0575ba6a9e6a8fe","geo":{"country":"US"},"ua":{"browser":"Claude"}},"references":{"total":0,"limit":25,"page":1,"links":{"self":"https://api.sb/v1/founding-hypotheses/fh%3Abuyers-and-purchasing-agents-spend-analysis-report%3At-ais-low-tech%3Av1/references"},"items":[]}}