Teardown

Supply chain · Deep dive

Keychain

New York AI manufacturing platform for packaged goods from Handy founders Oisin Hanrahan and Umang Dua — free for brands and retailers, paid by manufacturers — that indexed 30,000+ North American manufacturers, reached $1B+ in monthly project volume by August 2025, signed 8 of the top 10 US retailers, and is now selling KeychainOS as an AI-era ERP to the plants themselves, with roughly $78M raised through the November 2025 W23 Global round.

emerging

The question that decides it: Keychain's demand side came cheap — 8 of the top 10 US retailers pay nothing, so the logos accumulated fast — but revenue sits on the supply side: mid-tier contract manufacturers paying subscription fees (one Trustpilot account: $5,000 for a three-month trial) for leads, and since August 2025 for KeychainOS as plant software. Does the lead-gen wedge convert into system-of-record software — plants actually running purchasing, compliance and traceability on KeychainOS — before churn from unconverted leads (that same reviewer got ~10 leads, zero deals, and a push into a three-year contract) erodes the only side that pays, and before the retailer-funded Keychain360 pivot turns Keychain into a private-label PLM vendor fighting Trace One instead of a marketplace?

HQ
New York, NY
Founded
2023 (platform launched February 2024)
Ownership
VC-backed private — Lightspeed, BoxGroup, Wellington Management, W23 Global, plus strategic CPG money (General Mills' 301 INC, Schreiber Foods; Tracxn also lists Hershey)
Funding
~$78M disclosed: $18M seed (Lightspeed, Nov 2023), $15M Series A (BoxGroup, Nov 2024), $30M Series B (Wellington Management, Aug 2025), $10M strategic round (W23 Global, Nov 2025); Tracxn pegs total at $80.5M (2026)
Valuation
$260M post-money at the Series A (Fortune, Nov 2024); Series B valuation undisclosed — Hanrahan called it 'a good step up' (TechCrunch, Aug 2025)
Revenue
Undisclosed. The headline metric is $1B+ in monthly manufacturing project volume (August 2025) — demand flowing through the platform, not Keychain revenue; no ARR, take rate or paying-manufacturer count has been published (through July 2026)
Headcount
~70 (August 2025): roughly half in Gurugram, India, ~20 in New York, remainder in Austin; Dublin office announced September 2025 (TechCrunch; Irish Times)
Screen
Early breakout — founded 2023 (within the past 3 years) with $8M+ raised
Published
2026-07-29
Web
www.keychain.com
Elsewhere
LinkedIn · Crunchbase

Founders and leadership

  • Oisin Hanrahan Co-founder & CEO

    Dublin-born serial founder. Co-founded Handy in 2012 out of Harvard Business School — the on-demand home-cleaning marketplace — and sold it to IAC's ANGI Homeservices in October 2018 for roughly $165M, personally netting a reported ~$4M+. Became ANGI's chief product officer, then Angi CEO in February 2021; drove the Angie's List-to-Angi rebrand and a costly push into fixed-price services, and left in October 2022 with the stock down ~82% over his 20-month tenure, replaced by IAC's Joey Levin. Keychain is his argument that the marketplace playbook works better applied to $500B of US packaged-goods production than to home services.

  • Umang Dua Co-founder

    Hanrahan's partner for over a decade: co-founded Handy with him in 2012, ran operations and sales, then served as chief revenue officer of Angi after the acquisition. Left alongside Hanrahan and co-founded Keychain in 2023. The pair's pitch is that they have already built and scaled a two-sided marketplace through hypergrowth and a public-company exit — including the mistakes.

  • Jordan Weitz Co-founder

    The domain hire: came from private equity and venture investing in the CPG sector, supplying the industry knowledge the two marketplace operators lacked when they picked contract manufacturing as the next fragmented market to organize.

Snapshot

Keychain is the Handy founders’ second two-sided marketplace, pointed this time at the $500 billion of packaged goods that 20,000-plus US contract manufacturers produce every year for brands and retailers who still find each other through trade shows and brokers. Launched in February 2024, it indexed manufacturers and their products with AI, gave brands and retailers free search, and charged manufacturers for the leads — reaching over $1 billion in monthly project volume, 8 of the top 10 US retailers and 7 of the top 10 CPG brands by August 2025, on roughly $78 million raised from Lightspeed, BoxGroup, Wellington Management and a Tesco-backed retail fund. The August 2025 launch of KeychainOS — an AI operating system for the plants themselves — and the November 2025 Keychain360 private-label platform for retailers mark the pivot from matchmaking to infrastructure. The open question is whether the side of the market that pays is getting enough value to fund that ambition.

Founding story

Hanrahan and Dua are running the same play a second time, with scar tissue. They co-founded Handy in 2012 from Harvard Business School, scaled the on-demand home-services marketplace through the gig-economy wars, and sold it to IAC’s ANGI Homeservices in October 2018 for roughly $165 million — a real exit, but a modest one against Handy’s ~$110 million raised. Inside ANGI, Hanrahan rose to chief product officer and then, in February 2021, CEO of Angi; Dua became chief revenue officer. The Angi chapter is the uncomfortable part of the resume: Hanrahan bet the company on the Angie’s List-to-Angi rebrand and a capital-hungry push into fixed-price services, and the stock fell roughly 82% over his 20 months before IAC’s Joey Levin replaced him in October 2022 (The Currency, October 2022). The record is genuinely mixed — a marketplace operator who scaled supply and demand twice, and a public-company CEO whose strategy the owner pulled the plug on.

Within a year both had regrouped. With Jordan Weitz — a CPG-focused private equity and venture investor who supplied the industry knowledge — they founded Keychain in 2023 on a simple observation: finding and vetting a contract manufacturer took brands 12 to 18 months of trade shows, brokers and cold calls, in an industry with no directory, no data layer and no Amazon. The seed round’s tell was the cap table: alongside Lightspeed came more than 20 CPG industry executives, bought in before the product launched.

How it works

The core machine is an index. Keychain crawled and structured the packaged-goods universe — by August 2024 it had indexed 24,027 manufacturers and 763,000+ products, and by the Series B over 30,000 manufacturers and 2.5 million products (VentureBeat, August 2025) — using the products on shelves to infer what each plant can actually make: certifications, allergen handling, pack formats, capacity, co-packing history. A brand or retailer describes what it needs — a high-protein granola bar, organic-certified, 500,000 units, nut-free facility — and the matching model returns vetted candidates in days rather than the 12-18 months the industry default takes. Introductions, RFQs, sampling and project management then run through platform workflows; demand-side users pay nothing, which is why the top of US retail adopted it so fast.

KeychainOS, launched August 19, 2025, moves from matching into the plant. It is pitched as an AI-era replacement or overlay for legacy food-manufacturing ERP: modules for purchasing and procurement (demand-predictive purchase-order recommendations generated from live inventory, production plans and supplier performance), production planning, inventory, and compliance-traceability — where the system turns operator-recorded events like batch mixing into auditor-ready documentation and prompts the required follow-on steps automatically (VentureBeat, August 2025). Keychain360, launched November 2025 with W23 Global’s money, is the retailer-facing equivalent: a supply-chain and product-management layer for private-label programs that Hanrahan claims cuts launch timelines 25-30% (Irish Times, November 2025). Engineering is anchored in Gurugram, India, where roughly half the ~70-person team sits (TechCrunch, August 2025).

Product and business overview

Three named surfaces. The marketplace — free discovery, matching and project workflow for 20,000+ brands and retailers (August 2025 count), spanning food and beverage, supplements, and since 2025 beauty and personal care. KeychainOS — the paid manufacturer-side operating system: purchasing, planning, inventory, compliance, traceability. Keychain360 — the retailer private-label platform, launched November 2025 with Tesco’s venture backer as an investor. The build-out ran US first, then Ireland (September 2025) and the UK (November 2025). The strategic investor roster — General Mills’ 301 INC, Schreiber Foods, W23’s retail LPs, Hershey per Tracxn — doubles as a customer pipeline, and that is the point of it.

Business model and pricing

The model is asymmetric by design: brands and retailers ride free, manufacturers pay based on usage and the tools they take — subscription-style access to be discovered, receive leads and run projects (AlleyWatch, August 2025). Keychain takes no per-transaction commission that anyone has disclosed, and publishes no pricing. The only real price point in the public record comes from an aggrieved manufacturer on Trustpilot: $5,000 for a three-month trial — roughly a $20,000 annualized rate — after which the reviewer says Keychain pushed a three-year contract and invoiced a further $15,000, retracted only under legal threat. KeychainOS and Keychain360 pricing is undisclosed; both are classic enterprise SaaS motions into buyers (plant operators, retail private-label teams) with existing systems and long procurement cycles. The structural tension is worth stating plainly: the side with the famous logos pays nothing, so every dollar of revenue depends on mid-tier manufacturers believing the leads and software are worth it — and the company has published no ARR, no paying-manufacturer count and no retention figure through July 2026. “$1B+ in monthly project volume” measures demand intent flowing through a free product, not money Keychain keeps.

Traction over time

MarkerNov 2023 (seed)Aug 2024Nov 2024 (Series A)Aug 2025 (Series B)Nov 2025
Capital raised (cum.)$18M$18M$33M$68M~$78M
Manufacturers indexed~10,000 claimed24,02730,000-37,000
Project volume$500M fulfilled since Feb launch$500M+/month$1B+/month
Demand sidePre-launchBig brands + retailers activeGeneral Mills, Schreiber invest8 of top 10 US retailers; 7 of top 10 CPG brands; 7-Eleven, Whole Foods; 20,000+ brands/retailersTesco-linked W23 invests; UK launch
ProductMarketplaceMarketplace + workflowsKeychainOS launchKeychain360 launch
HeadcountSmall NY team~70 (half Gurugram)Dublin + UK offices

Eighteen months from launch to $68 million raised and $1 billion in monthly project volume is fast by any standard. The gaps are equally visible: no revenue series, no disclosed paying-supplier count, valuation disclosed only once ($260M post, November 2024), and the flagship volume metric measures the free side of the marketplace. In May 2026 the company shifted to publishing industry survey reports (52% of manufacturers planning new lines) — thought-leadership output, not traction disclosure.

Market analysis

The wedge market is large and genuinely under-digitized. Global food contract manufacturing alone was estimated at roughly $174-187 billion in 2025, growing 8-14% annually depending on the source (The Business Research Company; SkyQuest, 2025); US packaged-goods production runs about $500 billion a year across 20,000+ manufacturers, and CPG overall exceeds $2 trillion. Three structural forces favor Keychain: outsourcing is rising as brands shed capex and chase speed-to-market; private label keeps taking US shelf share, which is precisely the Keychain360 and W23 thesis — retailers need sourcing infrastructure even more than brands do; and the manufacturer base is aging into a software generation gap that AI-native tooling can jump. The countervailing force: the buyers of Keychain’s paid products are thin-margin plants in a cost-pressured cycle — Keychain’s own May 2026 report describes manufacturers pulling back from expansion — and selling discretionary software into that cohort is harder than signing free retailers.

Competitive intel

No competitor combines both of Keychain’s fronts, which cuts both ways — it also means Keychain fights on two. In matching, PartnerSlate ($4M raised, ~6,000 co-mans) is years older but a fraction of the scale; Novi ($51M+) owns the beauty and ingredient-transparency lane Keychain is entering; Wonnda and GrowinCo hold regional ground in Europe and LatAm. Keychain’s edge is capital, index breadth and — decisively — the retailer demand side, which none of the marketplace rivals signed. In plant software, the opponents change entirely: Aptean, Plex, Wherefour and SAP already run food plants’ purchasing and traceability, and Trace One and Specright already run retailer private-label spec management. Those incumbents own installed bases and switching costs; Keychain’s counter is that it arrives holding the demand relationships manufacturers want and an AI-native product surface the legacy vendors can’t rebuild quickly. The risk is the classic two-front problem: a marketplace with better software rivals and a software vendor with better-entrenched rivals.

History and evolution

What people say

The case for. The demand side votes with adoption: 8 of the top 10 US retailers and 7 of the top 10 CPG brands on the platform within 18 months of launch (company, August 2025), with 7-Eleven, Whole Foods and General Mills named. General Mills’ venture arm didn’t just use it, it invested, saying Keychain would change how the CPG industry works (AgFunderNews, November 2024) — and Wellington Management, a crossover investor that typically buys pre-IPO quality, led the B. FeaturedCustomers references call it a genuine marketplace innovation and one client described Keychain as among the biggest partners it has had. Inc. (2025) profiled Hanrahan’s unusually analog go-to-market — roughly $10K a month on in-person industry events — as a deliberate trust-building machine in a handshake industry.

The complaints. The sharpest criticism comes from the side that pays. A manufacturer’s detailed Trustpilot review describes paying $5,000 for a three-month trial that produced about ten leads — startups, unverifiable foreign companies claiming 500K-5M unit volumes, and sample-seekers who wanted free product — none of which converted; the reviewer says Keychain then pushed a three-year contract, invoiced an additional $15,000, and relented only under threat of legal action. One review is one review, but it maps exactly onto the model’s structural weakness: lead quality is the entire value proposition for the paying side, and a free demand side inflates volume metrics with low-intent inquiries. Glassdoor sits at 3.0 across 24 reviews (2025-26) — soft for a hot startup — with praise for talent and camaraderie against recurring complaints of abrupt, undocumented terminations and a culture of uncertainty. And the disclosure pattern invites skepticism: the headline metric measures other people’s demand, not Keychain’s revenue, and hasn’t been accompanied by a single monetization number in three years of press releases.

Outlook: the open question

Keychain works if, by roughly the end of 2027, the paying side proves durable: a disclosed base of manufacturers renewing subscriptions because leads actually convert, KeychainOS running as the operating layer in real plants (named deployments, not launch press), Keychain360 winning at least one flagship retailer private-label program against Trace One-class incumbents, and a Series C that discloses a valuation meaningfully above $260 million — evidence that free-demand-paid-supply generates software economics, not just volume metrics. The assets are real: the best demand-side logo wall in the category, an index no rival matches, strategic capital from the exact companies that are the customers, and founders who have operated a marketplace at scale before. It fails if the Trustpilot pattern is the norm rather than the outlier — manufacturers churning when $20K a year of leads doesn’t convert, leaving the marketplace hollow on the only side that pays — while the ERP and PLM incumbents’ switching costs blunt KeychainOS and Keychain360, and the company keeps substituting index size and project volume for revenue disclosure; in that world the pivot to plant software reads as an admission that matchmaking alone couldn’t carry the valuation, echoing the Angi years, when top-line reinvention outran unit economics. Tells to watch: any published paying-manufacturer or ARR figure; whether the next round is led by a new outside investor at a disclosed step-up or quietly filled by insiders and strategics; named KeychainOS plant deployments; and whether Tesco converts from LP-adjacent investor to disclosed Keychain360 customer.

Sources and further reading

Capital history

DateRoundAmountValuationLead(s)
Nov 14, 2023 Seed $18M Undisclosed Lightspeed Venture Partners; BoxGroup, Afore Capital, SV Angel and 20+ CPG industry executives participated
Nov 19, 2024 Series A $15M $260M post-money (Fortune) BoxGroup ($10M of the round); General Mills' 301 INC and Schreiber Foods joined as strategics; Lightspeed and SV Angel returned
Aug 19, 2025 Series B $30M Undisclosed ('a good step up' — Hanrahan to TechCrunch) Wellington Management, with existing investor BoxGroup; total raised $68M; KeychainOS launched same day
Nov 17, 2025 Strategic $10M (£7.5M) Undisclosed W23 Global — the retail venture fund backed by Tesco, Woolworths, Ahold Delhaize, Shoprite and Tata's Trent; funded UK entry and the Keychain360 private-label platform launch

Investors / owners: Lightspeed Venture Partners, BoxGroup, Wellington Management, W23 Global, 301 INC (General Mills), Schreiber Foods, Afore Capital, SV Angel

Competitive set

  • PartnerSlate — The direct marketplace predecessor: San Francisco, ~$4M raised (Supply Change Capital, June 2023), ~6,000 co-manufacturers in its network, food and beverage only. Attacks Keychain from below on price and co-packer trust built over years; Keychain outguns it ~20x on capital, ~5x on indexed manufacturers, and owns the retailer demand side PartnerSlate never signed.
  • Novi (Novi Connect) — $51M+ raised (Greylock-led rounds through 2022). Ingredient-level sourcing and transparency marketplace for beauty, personal care and home goods with retailer compliance programs (notably Target). As Keychain expands from food into beauty and personal care (2025), Novi is the incumbent in exactly that vertical, with deeper ingredient data; it lacks Keychain's food scale and ERP ambition.
  • Wonnda — Berlin-based private-label and contract-manufacturing marketplace for European consumer brands (seed-stage, single-digit millions raised). Small, but it is the local incumbent in the European private-label market Keychain entered via Ireland (Sep 2025) and the UK (Nov 2025).
  • GrowinCo. — Brazil-based co-manufacturing and ingredients network (founded 2018) covering LatAm CPG. Limited US overlap today; matters if Keychain's global-sourcing pitch to multinationals extends south.
  • Food ERP / plant-software incumbents (Aptean, Plex, Wherefour, SAP) — The real fight KeychainOS picked in August 2025: purchasing, inventory, compliance and traceability are already sold to food plants by entrenched ERP vendors — Aptean alone has rolled up dozens of food-specific systems. They own the installed base and the switching costs; Keychain's counter is AI-native workflows and the demand-side relationships manufacturers want access to.
  • Trace One / Specright (private-label PLM and spec management) — Keychain360 (Nov 2025) walks into retailer private-label product lifecycle management, where Trace One has served grocers for two decades and Specright owns spec data management. These are entrenched, retailer-embedded systems — a different, harder sale than free marketplace access.