Record for this area

SaaS acquisition

The path from what a machine is documented doing to the party holding the record that makes the result official, with how well each step is backed. Where the sources settled nothing, the step says so.

This area was assessed before the shared method existed, so its rows answer slightly different questions. Its counts are reported apart and never added into a single figure with the other nine.

What a machine is documented doing

Established by the evidence. Composing and submitting offers; licence and spend optimisation
The record it touches

Established by the evidence. The offer (private offer) and the subscription entitlement record; activation state
What still has to happen first

Established by the evidence. CONFIGURABLE - auto activation is ON by default

Assumed, because nothing says otherwise no document describes a step here; that it stops is inferred from nothing saying otherwise No inspected artifact places a capability at customer acceptance

auto activation is ON by default - a configuration of the rail not an act of a capability
the source's own words, quoted exactly
What makes the change count

Established by the evidence. The marketplace platform sets the outermost grammar; the ISV nests a narrower envelope inside it for the channel partner
Who holds the official record

Established by the evidence. The marketplace platform holds the offer and the entitlement record
What the area relies on

Not answered by the evidence we checked: We looked and the sources we inspected did not settle it.. Not answered by the evidence we checked We looked and the sources we inspected did not settle it.

The action that makes it official Customer acceptance of the offer on the marketplace platform; the platform then invoices and collects Done by The customer; the rail executes activation. It counts because of a record state held by the rail. No machine is documented doing this.

What this area could not settle

What could this area not settle?

What this area recorded against its own findings, in its own words.

Recorded against itself by this area

13 of 13 events from vendor and operator estates; ZERO institutional events; one cutoff ruling turned on a no-page-visible-date limitation

13 dated acts by institutions or vendors

0 searches recorded here - this area filed its searches elsewhere, so a zero here is a filing convention rather than a finding that nothing was looked for

23 capabilities examined

13 systems examined

Can the act be undone, and by whom?

not observable We looked and the sources we inspected did not settle it.

Question by question

What did the evidence settle, dimension by dimension?

Eighteen structural dimensions, each carrying its own evidence state and what it does not establish.

What the evidence settled for this area, question by question
Question asked	What the evidence says	What it does not tell you
consequential-object `D01`	Established by the evidence. The private offer, the subscription entitlement record, and activation state	Frozen by the area contract; SHAPE ONLY across areas.
object-holder `D02`	Established by the evidence. The transaction infrastructure (Microsoft Marketplace) holds the offer and the entitlement record	Not recorded by the method used for this area
grammar-or-form-setter `D03`	Established by the evidence. The transaction infrastructure sets the outermost grammar; the ISV nests a narrower envelope inside it for the channel partner	A double envelope, not a single form-defining body.
composer `D04`	Established by the evidence. Seller or channel partner; the transaction infrastructure itself for express offers	Channel-partner composition is the strongest non-rail composition in the set.
validator `D05`	Established by the evidence. The transaction infrastructure accepts the offer against the authorization scope; the customer accepts	OC-05 records that the document states the authorization constraint but describes no validation step or error path.
binder `D06`	Established by the evidence. Customer acceptance on the transaction infrastructure; the transaction infrastructure then invoices and collects	Not recorded by the method used for this area
deterministic-automation `D07`	Established by the evidence. Rail-side deterministic activation is evidenced: auto activation is ON by default for new plans, and billing begins at purchase	Deterministic transaction infrastructure behaviour on activation and billing, NOT on the binding act, which is customer acceptance.
agentic-capability `D08`	Established by the evidence. Composing and submitting offers; licence and spend optimisation	Composition only. No inspected artifact places a capability at customer acceptance.
furthest-deterministic-action `D09`	Established by the evidence. Activation and billing commencement by the transaction infrastructure, after the customer's acceptance	Downstream of the binding act, not across it.
furthest-agentic-action `D10`	Established by the evidence. Composing and submitting offers	The stop is merely unevidenced, not described.
reversibility `D11`	Not answered by the evidence we checked: We looked and the sources we inspected did not settle it.. Reversibility of the entitlement or activation state is not coded and not observable We looked and the sources we inspected did not settle it.	Never filled by analogy (TA-08).
externality-radius `D12`	Not answered by the evidence we checked: We looked and the sources we inspected did not settle it.. Whether the entitlement state reaches parties beyond the transaction infrastructure, seller and customer is not established We looked and the sources we inspected did not settle it.	13 of 13 events from vendor and operator estates; ZERO institutional events. The estate that would show external reach was not present.
counterparty-assent `D13`	Established by the evidence. Required: the customer accepts the offer, then purchases	Not recorded by the method used for this area
rule-codifiability `D14`	A source states this does not happen: A source addressed it and said no.. The binding act is a counterparty's acceptance, not a rule the transaction infrastructure evaluates	The transaction infrastructure automates activation AFTER acceptance; it does not evaluate acceptance.
institutional-closure `D15`	A source states this does not happen: A source addressed it and said no.. No institution closes this area. The transaction infrastructure is a commercial operator and the corpus records zero institutional events for it	This is the corpus's inverse case and it is a finding, not a gap.
authoritative-state-centralization `D16`	Established by the evidence. CENTRALIZED on the transaction infrastructure: it holds the offer, the entitlement record, invoicing and collection	The most centralized area in the corpus.
consequence-of-error `D17`	Not answered by the evidence we checked: We looked and the sources we inspected did not settle it.. What follows an erroneous activation, entitlement or billing start is not established We looked and the sources we inspected did not settle it.	EV-06's own limitation states the announcement gives no adoption levels and no error path.
accountability `D18`	Not answered by the evidence we checked: We looked and the sources we inspected did not settle it.. No stable answerable party evidenced for the offer or entitlement record We looked and the sources we inspected did not settle it.	TA-07: ten looks, ten nils.

Capabilities

What is each capability evidenced doing?

The mode records that a named subject performs an operation. It does not record what kind of performer that subject is, and several rows disclaim an agentic mechanism in their own limitation.

Infrastructure

Whose systems does this area run on?

What part each system plays was decided by what it does to the area's records, never by what kind of company runs it. Where the sources leave that open the entry says so, because an open question is not a kind of system.

Systems examined in this area
System	Who runs it	What part it plays	What it does to the official record
AWS Marketplace `C-012`	Amazon Web Services	authoritative-transaction-rail	creates — the private offer; creates/holds/mutates — the licence entitlement, in AWS License Manager
Microsoft Marketplace private offers `C-013`	Microsoft	authoritative-transaction-rail	creates/holds — the private offer record, in Partner Center; accepts-or-rejects — acceptance gated by billing role and killed by the accept-by date; holds — the private price Microsoft invoices against
Google Cloud Marketplace `C-014`	Google	authoritative-transaction-rail	creates/holds — the entitlement; creates — the order; accepts-or-rejects — approval required; expiry auto-rejects
SoftwareOne Renewal Manager `C-015`	SoftwareOne	not-applicable	observes — renewal agreements; mirrors — SoftwareOne transaction data
Suger cloud-marketplace enablement platform `C-024`	Suger	transaction-enablement-layer	constructs-for — the marketplace private offer; submits-to — AWS, Azure and GCP marketplaces; mirrors — the marketplace entitlement; observes — fulfilment events
Clazar cloud-marketplace automation platform `C-025`	Clazar	transaction-enablement-layer	constructs-for — the private offer, for AWS, Azure and GCP; mirrors — deal and offer state; observes — offer status across marketplaces
ClearSale AI Agents Portal `C-026`	ClearSale	not-applicable	none-established — the only object created is a quote inquiry in ClearSale's own sales pipeline
Microsoft Marketplace SaaS Fulfillment Subscription APIs v2 `C-033`	Microsoft	hybrid	mutates/holds — the Microsoft Marketplace SaaS subscription record; constructs-for — the publisher's own SaaS service state, held by the ISV
CloudBlue Connect `C-034`	CloudBlue	hybrid	creates/holds/mutates — the Connect subscription object and its fulfilment requests; constructs-for — the vendor's own licence activation state
Apideck integrations marketplace `C-035`	Apideck	not-applicable	none-established
Cloud Commerce Consumer Procurement API `C-036`	Google	unresolved	none-established — relative to this area
Zylo SaaS Negotiator `C-017`	Zylo	not-applicable	none-established
commercetools agentic catalog `C-027`	commercetools	not-applicable	none-established

Claims

What may be said from this area, and what may not?

Each one carries how far it is backed and, where it was recorded, the exact stronger sentence that may never be written from it.

CL-01 Observed

Rails write the grammar, not the terms.

The strongest thing against it A party other than the platform also sets grammar: a Microsoft resale authorization is 'tied to a specific offer, one channel partner, and one or more customer markets'.

What may not be said from this Never: 'marketplaces do not control the transaction'. They control grammar, record and settlement.

CL-02 Observed

In nine offer mechanisms across AWS, Google and Microsoft, the transaction infrastructure composed the specific commercial terms in none of them.

The strongest thing against it None was found within the coded set.

What may not be said from this Never: 'transaction infrastructure never compose terms anywhere' or any share-of-market phrasing.

CL-03 Observed

The transaction infrastructure set the transaction grammar in every one of the nine mechanisms.

The strongest thing against it Grammar-setting is layered: the ISV sets a nested envelope at Microsoft.

What may not be said from this Never: 'transaction infrastructure control the market' or 'transaction infrastructure set prices'.

CL-04 Observed

No non-rail party binds the transaction independently in any evidenced mechanism.

The strongest thing against it None was found.

What may not be said from this Never: 'only marketplaces can bind SaaS transactions'.

CL-05 Observed

Microsoft hands private-offer creation to channel partners while keeping invoicing and collection.

The strongest thing against it None within Microsoft's own documentation.

What may not be said from this Never generalise to 'marketplaces are giving away the offer'.

CL-06 Observed

AWS automates offer generation for standard deals from seller-defined rate cards and routes custom requests to the seller's sales team.

The strongest thing against it None.

What may not be said from this Never: 'AWS sets the price'. The pricing is the seller's.

CL-07 Observed

The two evidenced enablement layers compose no terms of their own, hold no official record, obligation or other binding object and bind nothing.

The strongest thing against it Clazar's 'automated checks that gate publishing until all criteria are met' is a control of its own.

What may not be said from this Never: 'enablement layers add no value' or 'enablement layers will disappear'.

CL-08 Observed

Activation moved onto the transaction infrastructure at Microsoft and remained with the vendor on the Google path.

The strongest thing against it This claim IS the counterexample to any one-way absorption reading.

What may not be said from this Never: 'activation is centralising' as a general trend.

CL-09 Pattern

Enterprise SaaS marketplaces are constitutionalising the offer: they define which structures may be expressed, which parties may transact and in which markets, and they alone hold the record and collect.

The strongest thing against it Grammar-setting is layered; activation is not uniformly rail-held.

What may not be said from this Never as 'marketplaces have taken over enterprise SaaS buying'.

CL-10 Pattern

The enablement-layer position is compressed between sellers composing terms and transaction infrastructure distributing creation rights and automating generation.

The strongest thing against it Clazar's pre-publication gate; and Suger's and Clazar's reach across three transaction infrastructure, which no single transaction infrastructure offers.

What may not be said from this Never: 'enablement layers are being disintermediated' or any survival forecast.

CL-11 Observed

Grammar-setting is layered rather than exclusive to transaction infrastructure: a software company's resale authorization defines a nested envelope for a channel partner's composition.

The strongest thing against it This claim IS the strongest counterexample to CL-01 and CL-03.

What may not be said from this Never used to argue that transaction infrastructure do not hold the outermost grammar.

CL-12 Not established

Participant-side formation cannot be characterised from this evidence.

The strongest thing against it n/a

What may not be said from this Never replaced by a softer phrase such as 'participants are moving more slowly'.

CL-13 Not established

Whether this grammar-versus-terms separation holds outside cloud-marketplace-mediated procurement is untested.

The strongest thing against it n/a

What may not be said from this Never implied portable to procurement generally.

Dated acts

What dated acts touch this area's records?

A list, in date order. It is not a timeline and carries no rate: the collection capped each publisher, so this is a floor on what was retrievable rather than a count of what occurred.

Dated acts touching this area
Date	Kind of act	Consequence recorded	Source
2024-05-27 `day precision`	workflow expansion	none	source
2024-06-17 `day precision`	commercial-model change	none	source
2025-04-21 `day precision`	commercial-model change	none	source
2024-11-21 `day precision`	workflow expansion	none	source
2025-11-30 `day precision`	commercial-model change	none	source
2026-05-18 `day precision`	workflow expansion	none	source
2026-05-26 `day precision`	transaction infrastructure access	none	source
2026-05-27 `day precision`	transaction infrastructure access	none	source
2025-08-04 `day precision`	transaction infrastructure access	none	source
2025-07-10 `day precision`	workflow expansion	none	source
2026-01-07 `day precision`	transaction infrastructure access	none	source
2025-09-11 `day precision`	named deployment	none	source
2026-04-30 `day precision`	workflow expansion	none	source

Capability	How far it goes	What it does	Human work still needed	What this does not tell you
CloudEagle.ai Renewal Agents `C-001`	prepares the work, stops before it counts	Not recorded by the method used for this area	Not recorded by the method used for this area	Qualifying artifact E-06 is dated 2026-08-12, AFTER the cutoff; E-04 and E-05 are undated.
Zylo Clarity AI `C-002`	advises only	Not recorded by the method used for this area	Not recorded by the method used for this area	Resolved by one artifact dated 2026-02-20, before the cutoff, which establishes the NEGATIVE and therefore also settles the cutoff.
Tropic Proposal Agent and Purchase Prep `C-003`	advises only	Not recorded by the method used for this area	Not recorded by the method used for this area	Excluded on substance; dating immaterial.
Vertice Ana `C-004`	the sources do not settle whether it acts or only prepares	Not recorded by the method used for this area	Not recorded by the method used for this area	E-18 is dated 2026-06-01, BEFORE the cutoff, and dates Ana's capability — but it names '250,000 negotiated contracts, ranging from SOFTWARE TO SERVICES', not enterprise SaaS specifically. The SaaS-bearing artifact E-11 is undated. Same defect as AWS and Clazar, ruled the same way.
Spendflo Flo AI `C-005`	performs the operation itself	Not recorded by the method used for this area	Not recorded by the method used for this area	E-21 is dated 2026-05-15, BEFORE the cutoff, and dates the Flo AI launch and its purchase-order capability — but it contains no SaaS, subscription, seat or licence statement. The SaaS-bearing artifact E-14 is undated.
Torii Eko `C-006`	advises only	Not recorded by the method used for this area	Not recorded by the method used for this area	Fails Gate 3 and Gate 4 independently. Area execution exists but is non-agentic.
Vendr Pricing Agent / Negotiation Agent / Ruth `C-007`	the sources do not settle whether this belongs in scope	Not recorded by the method used for this area	Not recorded by the method used for this area	A search summary asserted the agent's outputs are 'all managed by Vendr's team of human expert negotiators who review all outputs' — NOT USED AS EVIDENCE, and nothing of the sort appears on the retrieved page. Ownership and continuity rulings from Batch 2 are preserved unchanged: Vertice-owned at the cutoff, identifiable and available, not a duplicate.
Sastrify SaaS management platform `C-008`	advises only	Not recorded by the method used for this area	Not recorded by the method used for this area	Deel-owned at cutoff; absorption stated as intent, not completed.
Cledara software subscription management `C-009`	advises only	Not recorded by the method used for this area	Not recorded by the method used for this area	No named agentic capability on the inspected estate.
Flexera One SaaS Management `C-010`	advises only	Not recorded by the method used for this area	Not recorded by the method used for this area	Licence-reclamation documentation blocked twice; exclusion survives because those pages describe a workflow, not a named agent.
Zip Renewals `C-011`	advises only	Not recorded by the method used for this area	Not recorded by the method used for this area	Fails Gate 1 and Gates 3/4 independently; the renewal-management URL resolves to Intake-to-Procure.
Calero SaaS management `C-016`	advises only	Not recorded by the method used for this area	Not recorded by the method used for this area	'Calero Assistant' is not established to exist on any inspected artifact.
Zluri SaaS management `C-019`	advises only	Not recorded by the method used for this area	Not recorded by the method used for this area	Governs third-party agent identities; names no agent of its own.
BetterCloud IT Agent `C-020`	advises only	Not recorded by the method used for this area	Not recorded by the method used for this area	Fails Gate 3 and Gate 4 independently; execution-grade on non-records in scope for the area.
License Logic contract-first SaaS management `C-021`	prepares the work, stops before it counts	Not recorded by the method used for this area	Not recorded by the method used for this area	E-57 is dated 2026-07-02, AFTER the cutoff; E-47 is undated.
Productiv SaaS management `C-018`	does not apply here	Not recorded by the method used for this area	Not recorded by the method used for this area	STANDING BLOCKER — census unresolved. The only retrievable artifact is a cessation-of-operations notice with no capability text. NO RETRIEVAL PERFORMED, as instructed. No active claim depends on this actor and no further work is proportionate.
Suplari SaaS procurement `C-022`	does not apply here	Not recorded by the method used for this area	Not recorded by the method used for this area	census exclusion — generic procurement only; area-scope ruling, not an evidence-quality ruling
Varisource SaaS procurement `C-023`	does not apply here	Not recorded by the method used for this area	Not recorded by the method used for this area	census exclusion — out-of-scope object, multi-category, not SaaS-specific
Oracle Fusion Agentic Applications for Procurement `C-028`	does not apply here	Not recorded by the method used for this area	Not recorded by the method used for this area	census exclusion — generic procurement only
Workday Contract Negotiation Agent `C-029`	does not apply here	Not recorded by the method used for this area	Not recorded by the method used for this area	census exclusion — generic procurement only
Levelpath `C-030`	does not apply here	Not recorded by the method used for this area	Not recorded by the method used for this area	census exclusion — generic procurement only
Icertis Contract Intelligence `C-031`	does not apply here	Not recorded by the method used for this area	Not recorded by the method used for this area	census exclusion — general contract management
Spendesk procure-to-pay `C-032`	does not apply here	Not recorded by the method used for this area	Not recorded by the method used for this area	census exclusion — generic procurement only