The Graveyard: Three Waves of Failed Chemical Marketplaces
TPC Special Edition — Why Chemical Commerce Resists Digitization — Part 1 of 4
The chemical industry has spent twenty-five years and hundreds of millions of dollars trying to build digital marketplaces. The results are in.
ChemConnect. Omnexus. CheMatch. Chembid. CheMondis. Each one launched with a reasonable thesis: the global chemical market is worth over $5 trillion, fragmented across thousands of producers and tens of thousands of buyers, and still runs largely on phone calls, WhatsApp messages, and personal relationships. Connect the two sides digitally, reduce friction, and capture a sliver of the transaction value.
It sounds right. It has sounded right for a quarter century. And the ventures that pursued this thesis — across three distinct waves — have either shut down, pivoted away from the marketplace model, or raised hundreds of millions of dollars while struggling to convert website visitors into actual transactions.
This is not an execution story. When smart founders backed by Sequoia, Coatue, BASF, Dow, and DuPont all hit the same wall, the wall is structural.
Wave 1: The Dot-Com Exchanges (1995–2007)
The first generation emerged during the B2B marketplace boom. ChemConnect, founded in 1995 and one of the earliest chemical e-commerce platforms, built a spot-trading exchange modeled on commodity markets. The pitch: chemicals should trade like oil and metals. List your product, find a counterparty, execute.
ChemConnect was not alone. CheMatch and several others launched in the same window. ChemConnect itself raised $114 million in venture capital and operated a trading platform called the World Chemical Exchange.
The most ambitious attempt was Omnexus — a joint venture launched in 2000 by five of the world's largest chemical producers: BASF, Bayer, Dow Chemical, DuPont, and Ticona (Celanese), each contributing $10 million. If anyone could make a chemical marketplace work, it should have been the producers themselves. They had the supply, the technical data, the customer relationships, and the capital.
Omnexus shut down in 2003. ChemConnect survived the dot-com crash but never achieved the liquidity its exchange model required. CheMatch was absorbed by ChemConnect in 2002, and the combined entity's commodity trading business was eventually acquired by ICE in 2007. The other spot exchanges of the era disappeared in the same period.
The lesson the industry took from Wave 1: chemicals are not oil. An exchange model assumes product fungibility — that one seller's HDPE is interchangeable with another's. In practice, polymer grades require qualification on a buyer's specific production line, and a buyer who needs weeks of trial runs before switching suppliers has no reason to find them on a spot platform.
The exchanges tried to capture value at the transaction — the moment of matching buyer to seller. But in an industry where the hard work happens before the transaction (qualifying the grade, evaluating the producer, assessing the risk), the transaction itself was the wrong place to look.
Wave 2: "This Time It's Different" (2017–2023)
A decade later, a second generation emerged with a more nuanced approach. Better UX. Mobile-first design. Regulatory integration. Request-for-quote workflows instead of spot trading. The new founders had studied why Wave 1 failed and believed better technology could overcome the structural barriers.
CheMondis, founded in 2018 in Cologne, launched with corporate backing from Covestro, LANXESS, and Evonik — major European chemical producers who invested directly in the platform. The logic: if the producers are shareholders, they'll list their products and drive supply-side adoption. CheMondis built RFQ systems, integrated European REACH compliance tools, and targeted the fragmented European chemical distribution landscape.
Chembid in Germany took a different angle — a search engine aggregating chemical listings from across the web, trying to be Google for chemicals rather than Amazon for chemicals.
CheMondis held on longer than most, but in the second half of 2023, Covestro divested its entire stake. Covestro was itself undergoing strategic changes at the time, but when a founding corporate backer exits a platform investment, the signal is hard to ignore.
The same structural wall held. Buyers used the platforms for product discovery — browsing specifications, comparing grades, identifying potential suppliers — and then picked up the phone to transact through their existing relationships. The platforms digitized the easy part (finding out who sells what) while the hard part (qualifying the product, negotiating credit terms, managing logistics, building trust) remained stubbornly offline.
Wave 2 moved upstream from the transaction to the connection — but the connection was not where the value was either. The value was in the evaluation: knowing which grade would actually work, which producer was reliable, and whether the price justified the switching cost. No platform addressed that layer.
Wave 3: The Well-Funded Current Generation
The third wave brought Silicon Valley capital and ambition. Knowde, founded in 2017, raised approximately $175 million from investors including Sequoia Capital and Coatue Management, reaching a reported $500 million valuation by 2021. Its founder, Ali Amin-Javaheri, was a veteran of more than a decade at ChemPoint, a specialty chemical distribution company later acquired by Univar. He understood the industry from the inside. By the time of the Series C in August 2024, the company had 8,000 suppliers, 230,000 products listed, and 2.9 million unique visitors annually.
Those are impressive numbers. But one metric tells a different story: in late 2023 — the most recent publicly available data — Contrary Research reported that the platform generated approximately 1,200 quotes in a month. Across 8,000 suppliers and 290,000 monthly product searches, that works out to roughly 0.4% of searches converting to a quote request — and a quote is still several steps away from a transaction.
The Series C was a down round — valued below the $500 million valuation from 2021, per TechCrunch. And the company's own messaging has shifted. Where Knowde once positioned itself as a chemical marketplace, the 2024 fundraise emphasized its "AI-powered Master Data Management platform." The website now leads with data management and AI capabilities, not marketplace transactions.
This is not failure. Knowde may build a successful enterprise SaaS business helping chemical companies organize their product data. But it is a concession that the marketplace thesis — connect chemical buyers and sellers on a platform, capture transaction value at scale — did not achieve escape velocity. Even with $175 million, Sequoia's backing, and a founder who understood the industry from the inside.
The 0.4% conversion rate is not just a low number. It is evidence of a structural mismatch. Buyers searched because they had a decision to make — which grade, from which producer, at what price. But 99.6% of them did not need the platform to execute the transaction. They needed it to inform their judgment. Then they transacted through their existing relationships, where trust, credit, and technical support already lived. The platform generated the insight; someone else collected the revenue.
The Survivors — And What They Have in Common
Not every chemical technology venture failed. But the ones that survived all share a revealing characteristic: they abandoned the marketplace model.
Elemica, organized in 1999 and launched in 2000 as a joint venture by BASF, Bayer, Celanese, Dow, DuPont, and Shell, survives today — in part because it never tried to disintermediate trading relationships. Instead, it built a supply chain connectivity platform — enabling standardized data exchange between trading partners' systems. Order management, logistics visibility, invoice reconciliation. Thoma Bravo acquired it in 2021, which tells you it was generating real revenue. But that revenue comes from workflow automation, not from matching buyers with sellers.
SpecialChem acquired the Omnexus brand after the original joint venture shut down and rebuilt it as a content and lead-generation platform. Engineers search for materials; suppliers pay for visibility and qualified leads. No transactions happen on the platform. The value is upstream of the purchase decision — discovery and specification, not commerce.
ICIS, S&P Global Platts, ChemOrbis — the durable businesses in chemical commerce technology are all intelligence providers. They sell information: pricing data, market analysis, trade flow analytics. They never touch a transaction.
Matmatch — a materials data and discovery platform — was acquired by Dassault Systemes in 2023. Not as a marketplace, but as a dataset to feed engineering workflows within Dassault's 3DEXPERIENCE platform. The transaction value was in the structured data, not in the marketplace model.
Horizontal B2B platforms like Alibaba handle some chemical transactions, particularly within China, but have not displaced relationship-based procurement for cross-border industrial trade — the trust, qualification, and compliance layers that polymer buyers depend on remain outside the platform.
The pattern is consistent across twenty-five years: ventures that tried to intermediate chemical transactions failed or pivoted. Ventures that sold information, automated workflows, or organized data survived. Every survivor moved away from the marketplace model.
What the Graveyard Reveals
The chemical marketplace graveyard is not a story about bad founders or bad timing — and it is not a story about digital chemical commerce being impossible. It is a story about capturing value at the wrong point in the procurement process.
Every wave tried to own a different piece of the transaction: Wave 1 tried to own the match. Wave 2 tried to own the connection. Wave 3 tried to own the catalog. And every survivor — Elemica, SpecialChem, ICIS, Matmatch — found its business in a different place entirely: in the intelligence, the data, and the workflow support that inform the buyer's decision.
The chemical industry does not have a transaction problem. Buyers know how to buy. They have traders, freight forwarders, customs brokers, and banks. The mechanics of executing a purchase — however complex — are operational and understood.
What the industry has is a decision problem. Which grade, from which producer, at what price, on what terms, at what time? These are the questions that determine whether a procurement outcome is good or bad — and in chemical trade, unlike e-commerce, you cannot reverse a bad decision. A container of off-spec polymer cannot be returned. Under standard Incoterms, the risk transfers to the buyer at loading. If the grade does not perform in the end application, the buyer negotiates, files a claim, or writes off the loss.
The stakes of the decision are high. And the judgment needed to make it well does not exist on any platform. No platform offers verified producer track records built over years of actual deliveries. No platform tells you how a grade performs in the end application, not just on the spec sheet. No platform evaluates whether a supplier's REACH registration actually covers your intended use. The practical knowledge of which producers deliver what they promise — and which do not — lives in the heads of experienced traders and procurement professionals who have built it through decades of transactions. It has never been systematically organized in a way that a buyer at fifty tons per month can draw on.
Every marketplace assumed the bottleneck was connecting buyer to seller. The bottleneck is the quality of the decision the buyer makes before, during, and after that connection. The ventures that understood this survived. The ones that did not are in the graveyard.
Part 2 of this series will examine why that decision problem is so acute for mid-tier buyers specifically — and why the cost of solving it has been, until now, structurally out of reach.
Next in this series: The cost floor — why mid-tier chemical buyers get worse service than they pay for.
Related
The fragmentation problem these marketplaces failed to solve persists in structural form today. The Chinese polymer market sourcing audit shows that even with 1,600+ producers and 600+ active merchants in China, most mid-tier buyers work through 2-4 contacts — seeing less than 3% of available supply, with $20-40/MT price spreads on the same grade the same day.
Sources referenced: TechCrunch (Knowde Series C, Aug 2024), Contrary Research (Knowde business breakdown, Dec 2023), Plastics News (Omnexus launch, CheMatch acquisition), Chemical & Engineering News (Omnexus shutdown, 2003), Canadian Plastics (Omnexus closure), Covestro corporate announcements (CheMondis divestiture, H2 2023), ICE press release (ChemConnect acquisition, Jul 2007), CBInsights (ChemConnect, CheMatch funding data), Knowde corporate blog (funding announcements, annual results), Dassault Systemes investor relations (Matmatch acquisition, Sep 2023), Digital Commerce 360 (Knowde Series C coverage), Thoma Bravo (Elemica acquisition, 2021).
Free market intelligence for polymer distributors. Twice weekly.