LightSource vs. Fairmarkit

Where Fairmarkit is strong: automating the tail spend nobody has time to source

Fairmarkit built its product around a real and expensive problem: a large share of company spend never gets competitively sourced (Fairmarkit puts it at more than 30%), because no buyer has time to run events on $2,000 purchases. Its answer is automation at volume. The platform auto-constructs quoting events from requisitions, suggests suppliers per line item from a 2.7M+ supplier marketplace spanning 195 countries, and collects structured bids through a supplier portal that is free forever and averages about two minutes per quote. The company said it was on pace for 200,000 sourcing events in 2022, claims buyers run 300% more events on the platform, and cites Boeing eliminating 115,000 manual hours per year in its April 2026 launch announcement.

The intake side is a real strength: KIT, its conversational GenAI intake with automatic SOW generation and policy routing, shipped in September 2023, early for the category. Pre-built connectors for SAP Ariba, S/4HANA, Coupa, Oracle, and ServiceNow (an investor since its Series C) let it embed in an existing procure-to-pay stack rather than replace it. Reviews are positive on ease of use and rollout speed: 4.6/5 on G2, though across only 17 reviews.

Where LightSource is strong: sourcing parts that have drawings, revisions, and a BOM

LightSource is purpose-built for direct materials, the parts that go into the product. The foundation is a revision-managed item master and nested bill of materials: AI ingestion turns Excel files, PDFs, and drawings into structured items, and BOMs carry snapshots, revision diffs, and cost roll-ups. RFQs go out at the part level with drawings attached, suppliers quote in custom bid sheets or plain Excel, and AI normalizes and compares quotes in any format. Award scenarios, split awards, and per-part price history turn each sourcing cycle into cost knowledge that survives buyer turnover. See the product overview for the full workflow.

None of this exists in Fairmarkit's model. There is no item master, no BOM concept, no revision handling, and no PLM integration; its line items are description and part-number strings, and reviewers describe its bid evaluation as price-centric. Supplier management is also deeper on the LightSource side: scorecards, data rooms with AI document analysis, and certification collection are product modules, not qualification flags. Manufacturers like BRP, Canada Goose, Shure, and Infinite Machine run production sourcing on it (see customer stories); Infinite Machine reports 25% lower BOM cost and $10M saved in nine months.

Both say autonomous sourcing, but the unit of work is different

Both companies use the phrase, and both are credibly AI-native. Fairmarkit rebuilt its platform around generative AI in 2023 and launched a full agentic architecture, Total Agentic Sourcing, in April 2026 (in private beta at launch). The difference is what the AI automates. Fairmarkit's unit of work is a requisition line: a description, a part number, a quantity. For tail spend that is the right abstraction, because the goal is three competitive bids on thousands of small purchases with near-zero buyer effort, and the winning design removes humans from the loop.

An engineered part is the opposite case. The sourcing object is a revision-controlled item with a drawing, tolerances, tooling, and qualification requirements, and the decision involves cost breakdowns, capacity, and engineering judgment. LightSource automates the labor around that decision (ingestion, event setup, quote normalization, drafted award scenarios) while the buyer owns the award. Fairmarkit's 2026 launch extends its coverage to strategic events, auctions, and TCO scenarios, but none of that adds an engineering data model: there is still no BOM to roll costs into and no PLM connection for parts to come from. Its own manufacturing page centers on MRO sourcing, which is consistent with how its manufacturing customers describe using it.

Running both: Fairmarkit on the requisition tail, LightSource on the product

This pairing is more coherent than most in the category, because the two products barely touch. Fairmarkit sits on top of a Coupa, Ariba, Oracle, or ServiceNow stack and auto-competes the requisition-driven purchases that would otherwise be rubber-stamped or auto-renewed: MRO consumables, shipping supplies, services SOWs, office equipment. LightSource is the source-to-contract system of record for the parts on the BOM, from RFQ through award, with awarded pricing exported to the ERP through its GraphQL API or flat files.

A manufacturer running both would route indirect requisitions through Fairmarkit's intake and let its agents compete the tail, while strategic sourcing and engineering run BOM-driven RFQs, quote analysis, award scenarios, and supplier scorecards in LightSource. The overlap is narrow: a simple build-to-print part with no revision history could technically be quoted in either. The tiebreaker is where the line item lives. If it comes from engineering data and will be re-sourced when the drawing changes, it belongs in LightSource. If it comes from a requisition and the goal is three bids with no buyer touch, it belongs in Fairmarkit. Neither product replaces the other's core.

What adding LightSource looks like next to an existing stack

LightSource does not require displacing anything Fairmarkit touches, and it does not require an integration project to start. Teams go live in days on flat files: Item Sync uploads and AI Item Ingestion structure existing part data (the ingestion feature processed more than 3,300 items in its first month, per LightSource product documentation), and quote formats come from maintained templates, so first sourcing events typically run within about 30 days. Implementation is included rather than sold as a separate integrator engagement, and suppliers onboard free in under 30 minutes with Excel round-trip quoting, which matters because supplier adoption is where sourcing tools usually stall.

Deeper integration comes later, at whatever pace IT prefers: a GraphQL API mirrors the platform for PLM pulls and awarded-pricing pushes to ERP systems. There is no packaged connector catalog, so plan on API work rather than a plug-in marketplace; details are on the integrations page. A common starting shape is a one-program or one-commodity proof of concept before wider rollout. LightSource is venture-backed with $33M co-led by Lightspeed and Bain Capital Ventures, and was named a Gartner Cool Vendor 2025 (Sourcing and Procurement Technology).

What public reviewers say

On G2, Fairmarkit holds 4.6/5 across 17 reviews and LightSource holds 4.9/5 across a small early base of about 11 reviews, both as of September 2026. Both bases are small, so individual reviews carry more information than the star averages. The quotes below were picked for specificity rather than sentiment, and both links go to the full, unfiltered review pages.

Fairmarkit's reviews are genuinely positive, and the most concrete praise is about the thing the product is designed to do: sit inside the P2P stack so buyers stop re-keying data.

"...tightly integrated into the Procure to Pay system and requires ZERO re-entry of data."

-- G2 reviewer (2025)

The criticisms reviewers list are short, and they cluster on sourcing depth rather than usability:

"Hard integration with SAP."

-- G2 reviewer (Enterprise, 2024)

"Lack of rounds of negotiation in RFQ events."

-- G2 reviewer (2024)

Neither reads as a pan; Fairmarkit has no G2 review below four stars. But both complaints fit the pattern this page describes: a platform optimized to run many events with near-zero touch leaves less room for deep, multi-round negotiation on complex purchases.

LightSource's early reviewers describe the direct-materials workflow this comparison is about, from BOM ingestion through bid analysis:

"The platform allows for the seamless addition of RFP items/SKUs and BOM details through a simple drop/drag feature... The platform also has a robust set of analysis tools to quickly compare the different supplier responses, calculate forecasted savings and identify best of the best pricing for the RFP... We typically see a company deploy LightSource within a day or two."

-- G2 reviewer (Mid-Market, 2024)

"Its AI capabilities are seriously helpful, especially for our supplier management and contract work... Having the cost and quoting information in a single location helps with rapid product launches and better visibility for our leadership group."

-- G2 reviewer (VP of Supply Chain, Mid-Market, 2024)

For balance, the recurring dislike in LightSource reviews is occasional minor bugs, and the most candid one is about adoption rather than the product: "The biggest challenge of LightSource is just getting the suppliers to try it. Once they try it --- they're also hooked," per a G2 reviewer (Mid-Market, 2024).