Globality: agentic services sourcing still needs governed award authority

External-source research synthesis, not an OfferSignal client result. Based on the retained October 2026 research and source-specific dates.

A complex services request rarely arrives as a clean catalogue line. The buyer must describe an outcome, find capable providers, compare proposals, negotiate terms and move the decision into purchasing systems. Globality is a useful case because it connects those steps without treating a recommendation as commercial authority.

From structured brief to agentic sourcing

Globality began with AI-assisted services procurement: a business need became a structured “Smart Brief,” matched to supplier “Smart Profiles,” with proposal comparison, negotiation, selection and statement-of-work creation. Its retained analyst material describes a continuously modelled “Service Graph” intended to improve matching and scoping. The product expanded from business-user self-service into procurement-led, procurement-owned and sole-source workflows, broader categories and verticals, internationalisation and SAP integrations. 5 9

The current platform describes Glo as reasoning over spend history, market information and supplier performance; deciding routing, event goals and negotiation strategy; and acting by messaging suppliers, countering offers and negotiating price and terms. Glo 2.0 offers collaborative execution with people in the loop or autonomous progression that makes decisions, negotiates and presents award scenarios. The advertised flow runs from natural-language intake and request-for-quotation generation through supplier invitation, multi-round negotiation across price, quality, delivery, service and risk, scenario analysis and an award recommendation. 1 2 10

That is stronger evidence of agentic action than a generic copilot claim. It does not establish that every customer permits autonomous award, or that contract signature, purchase-order approval, delivery, invoice matching and supplier-performance management occur without human or enterprise-system controls. The defensible boundary is award recommendation or award workflow plus downstream handoff.

Reported outcomes, with the denominator visible

Globality’s customer page attributes several material results to named customers. BMS reportedly ran 10 times more sourcing events and cut RFP cycle time by 6–10 times, without additional headcount. Fidelity reportedly reduced time to contract by about 50% and achieved a 20% savings rate in professional services. UCB reports a 50% shorter contract cycle time and more than 400 projects. LSE reports tenders processed twice as quickly and 15% savings in the first three months. The homepage also claims that 80% of customers launch within four weeks and that average satisfaction is 9.2/10. 1 3

These remain provider-published customer or vendor aggregate claims. The record does not provide the relevant baseline counts, sample sizes, measurement method, control group, pilot-versus-production split or independent audit. The LSE period is stated; most other periods are not. Treat the figures as attributed outcomes, not audited or universal return on investment.

Integration is not approval

Globality documents connections to SAP Ariba, SAP Fieldglass, Coupa, middleware, application programming interfaces and webhooks. Guided-buying intake can create a project and prepopulate data; contract workspaces and purchase requisitions can be created or updated. A 2020 SAP announcement described demand creation, supplier identification, tendering, proposal evaluation and statement-of-work creation followed by contract or requisition handoff. 6 7 8

For an industrial buyer, that is a credible route from RFQ to procurement-system control. It is not evidence that the platform can safely configure custom metalwork, machinery or robotics from drawings, revisions, materials, tolerances, tooling, capacity and binding quotations. Those fields require version controls and human engineering and commercial review.

An illustrative controlled application

For a manufacturer buying engineering or IT services, an illustrative pilot could provide historical briefs, supplier master, approved panels, rate cards, category rules, contract templates and prior spend outcomes. Compare Glo’s scope, shortlist, negotiation and award recommendation with a human baseline, then verify the contract or purchase-requisition handoff. Technical owners remain authoritative for requirements and acceptance; procurement for supplier eligibility and award approval; budget owners for spend and exceptions.

Log source revisions, units, currency, effective dates, supplier identity, permissions and every agent action. Test stale prices, duplicate suppliers, incompatible capabilities, malicious retrieved instructions, unauthorised awards and attempts to alter ranking or authority. Expand only if cycle time, savings, exception rate and approval compliance beat baseline.

Funding explains capacity, not adoption

Globality’s January 2021 announcement records a $138.3 million Series E from Sienna Capital and the SoftBank Vision Fund, with $310.3 million in cumulative funding since 2015. An earlier January 2019 announcement records a separate $100 million SoftBank investment and $172 million cumulative funding at that point. These figures help explain product and go-to-market capacity; they are not evidence of customer adoption, savings or return on investment. 4 5

The practical lesson is narrow but useful: Globality is a credible active platform for complex services sourcing, with configurable negotiation and enterprise handoffs. It is a strong candidate for a governed services-procurement experiment, not evidence of a safe autonomous industrial custom-manufacturing quoting system.

Sources and next reading

Globality homepage · Globality platform · Globality integrations · Globality customer stories

Explore OfferSignal · Focused Business & AI Workshop · All research

References