Case Study 1: TradeLens — The $100M Blockchain That Nobody Used

Case Study 1: TradeLens — The $100M Blockchain That Nobody Used

The Vision

In January 2018, IBM and A.P. Moller-Maersk — the world's largest container shipping company, controlling roughly 17% of global container capacity — announced TradeLens, a blockchain-based platform for digitizing global trade documentation. The vision was ambitious and, on its face, entirely reasonable.

International shipping is an industry drowning in paper. A single shipment of refrigerated goods from East Africa to Europe can generate up to 200 separate communications and involve 30 different organizations: the exporter, the importer, freight forwarders, customs brokers, port authorities, ocean carriers, inland truckers, warehouses, banks (for trade finance), insurance companies, and government agencies. The documentation required — bills of lading, certificates of origin, phytosanitary certificates, customs declarations, commercial invoices, packing lists — is mostly paper-based, prone to error and fraud, and responsible for massive delays. Industry estimates suggested that documentation costs accounted for 15-50% of the total cost of physical transport.

TradeLens proposed to replace this paper labyrinth with a shared digital platform built on Hyperledger Fabric. Each participant in a shipping route would record their data on the blockchain: the shipper would upload the bill of lading, the port authority would record vessel arrival and departure, customs would log clearance status, and the carrier would track container location and condition. All authorized parties could see relevant data in real time, eliminating the faxes, emails, phone calls, and physical document hand-offs that consumed days at every step.

IBM's blockchain division, at its peak employing over 1,500 people dedicated to enterprise blockchain, built the platform. Maersk provided the shipping domain expertise and, critically, access to its massive network of trade partners. The two companies projected that TradeLens would eventually process data for the majority of global container shipments — over 800 million TEUs (twenty-foot equivalent units) per year.

Early demonstrations were impressive. IBM showed that TradeLens could reduce the transit time of a shipment of flowers from Kenya to the Netherlands by 40%, primarily by eliminating documentation delays at borders. In pilot deployments, customs clearance times dropped by as much as 40%. The technology worked.

The Launch and Early Growth

TradeLens launched commercially in late 2018 after a year of pilot testing. IBM and Maersk structured it as a joint venture, with both companies holding equal governance stakes. The platform was built on Hyperledger Fabric, using channels to segregate data between different shipping routes and trade lanes.

Initial adoption was promising, driven largely by Maersk's own supply chain:

2018: Platform launch with Maersk and several port authorities (Rotterdam, Halifax, Algeciras) and customs agencies (Netherlands, Singapore, Australia)
2019: Over 100 organizations enrolled, processing data on 10 million shipping events per week. GTD Solution and Pacific International Lines (PIL), a major Asian carrier, joined.
2020: CMA CGM (world's third-largest shipping line), MSC (world's second-largest), and several additional port authorities joined. Over 175 organizations enrolled.
2021: Claimed over 300 organizations, covering data on approximately half of global container trade by volume.
2022, January: Over 65 million shipping events processed per month. ZIM (Israeli carrier) and Hapag-Lloyd (initially) explored integration.

The numbers looked good on paper. But beneath the surface, critical problems were metastasizing.

The Cracks

The Competitor Problem

TradeLens's original sin was its governance structure. Maersk was not just a co-creator of the platform — it was the world's largest shipping line, a direct competitor to every other carrier that might join. Asking CMA CGM, MSC, and Hapag-Lloyd to put their shipment data on a platform built by and initially controlled by Maersk was like asking Pepsi to record its supply chain data on a platform owned by Coca-Cola.

IBM and Maersk recognized this problem early and took steps to address it. In 2019, they restructured TradeLens as a more independent entity. They created an advisory board with seats for major participants. They offered API access and promised that Maersk would not have preferential access to competitors' data. Fabric's channel architecture ensured that each carrier's data was segregated.

But perception matters as much as technical reality in consortium governance. Competing carriers remained skeptical. Internal surveys and industry interviews conducted by logistics research firms revealed persistent concerns:

Data asymmetry: Even though Maersk couldn't see competitors' raw data through channels, Maersk's privileged position as co-owner of the platform meant it had insight into platform-level analytics — total volume trends, popular trade lanes, growth rates — that competitors didn't have.
Strategic leverage: If TradeLens became the dominant platform, Maersk would have structural influence over the infrastructure that all its competitors depended on.
The governance paradox: Truly neutral governance would require Maersk to cede control of a platform it had invested heavily in building. But without Maersk's commitment, the platform wouldn't exist. There was no stable equilibrium.

CMA CGM and MSC joined, but their participation was described by industry analysts as "cautious" — they connected some routes but did not fully migrate their operations to TradeLens. Hapag-Lloyd, the world's fifth-largest carrier, joined the rival GSBN (Global Shipping Business Network) platform instead, founded by COSCO Shipping Lines and other primarily Asian carriers. The shipping industry had bifurcated rather than unifying on a single platform.

The Small Player Problem

TradeLens's value proposition depended on network effects. A shared shipping platform is useful only if most of the parties in a given shipping route are on it. Consider a shipment from Shanghai to Rotterdam involving a Chinese exporter, a customs broker, the Port of Shanghai, CMA CGM (ocean carrier), the Port of Rotterdam, a Dutch customs authority, a freight forwarder, an inland trucker, and the importer. If the Port of Shanghai, CMA CGM, and the Port of Rotterdam are on TradeLens but the customs broker, freight forwarder, and trucker are not, the platform provides only partial visibility. The user still needs to make phone calls and track paper for the missing links.

This fragmentation was TradeLens's daily reality. While over 300 organizations were technically enrolled, many were small players whose integration was shallow — they could view data on TradeLens but hadn't integrated the platform into their operational workflows. A freight forwarder in Lagos or a customs broker in Vietnam might have signed up for TradeLens but continued using their existing systems for day-to-day operations, manually uploading data to TradeLens as an afterthought.

The integration challenge was particularly acute for small and medium enterprises (SMEs), which constitute the vast majority of actors in global shipping. A large port authority with a dedicated IT department could integrate with TradeLens's APIs. A small freight forwarding company in Mombasa, operating on a thin margin with legacy software (or paper-based systems), could not. TradeLens offered simplified integration tools, but the effort and cost remained prohibitive for many smaller participants.

The Revenue Problem

TradeLens needed a sustainable business model, and it never found one. The platform's costs were substantial: IBM's development team, Hyperledger Fabric infrastructure, customer support, integration assistance, and ongoing development. These costs needed to be covered by revenue from participants.

But who should pay? Carriers were reluctant to pay significant fees for a platform that their competitors co-controlled. Port authorities, many of which are government entities, had limited budgets and competing priorities. Small freight forwarders operated on razor-thin margins. And the ultimate beneficiaries — importers and exporters who would benefit from faster, cheaper shipping — were the most removed from the platform.

TradeLens experimented with various revenue models: per-transaction fees, subscription tiers, premium analytics. None generated sufficient revenue. Maersk and IBM subsidized the platform for years, absorbing operating losses while hoping that scale would eventually create a viable business. Scale never came.

The Technology Wasn't the Problem

This is the crucial insight. TradeLens's Hyperledger Fabric infrastructure worked as designed. The platform could process millions of shipping events. The channel architecture provided appropriate data segregation. The smart contracts correctly enforced business rules. IBM's blockchain engineers delivered a technically sound product.

The failure was not technological — it was organizational, economic, and strategic. The consortium coordination problem (competing carriers), the network effects problem (insufficient adoption for the platform to be useful), the integration problem (too costly for small players), and the business model problem (no viable revenue path) combined to create a spiral of insufficient adoption that ultimately made the platform commercially unviable.

The Shutdown

On November 29, 2022, IBM and Maersk issued a joint statement: TradeLens would wind down operations by the end of Q1 2023. The statement cited the need for "full global industry collaboration" to achieve "commercial viability" — bureaucratic language for "not enough people used it to pay the bills."

The industry reaction was telling. Many logistics professionals expressed disappointment but not surprise. In interviews and industry forums, the most common sentiment was not "the technology didn't work" but "we always knew getting the industry to agree would be the hard part."

IBM began dismantling its blockchain division, which had once been touted as a growth engine. Over the following months, the division was substantially downsized, with many of the 1,500+ blockchain employees reassigned or laid off. IBM's broader strategic pivot away from blockchain services was accelerated by TradeLens's failure.

Maersk, for its part, continued pursuing digitization of shipping — just without blockchain. The underlying problem that TradeLens tried to solve (paper-based shipping documentation causing delays and costs) remained real. But Maersk's subsequent initiatives focused on conventional digital platforms, APIs, and partnerships rather than distributed ledger technology.

The Aftermath and Alternative Approaches

TradeLens's shutdown did not end efforts to digitize shipping documentation. GSBN, the rival platform founded by COSCO and other Asian carriers, continued to operate, though with limited scope. DCSA (Digital Container Shipping Association), a non-profit founded by nine major container lines, pursued digitization through industry standards rather than a shared platform — a fundamentally different approach that sidesteps the consortium governance problem by focusing on interoperability standards rather than shared infrastructure.

The most significant post-TradeLens development may be the eBL (electronic bill of lading) movement, enabled by changes to the UK's Electronic Trade Documents Act 2023 and similar legislation in other jurisdictions. By giving legal recognition to electronic trade documents, these laws reduce the need for a blockchain-based solution — digital documents with appropriate legal status can be transmitted through conventional means.

Lessons Learned

Lesson 1: Governance First, Technology Second

TradeLens's leaders acknowledged in retrospective interviews that they underestimated the governance challenge. IBM's former head of blockchain acknowledged that "the technology was the easy part." The hard part was convincing competitors to collaborate on shared infrastructure, agreeing on who controls the platform, and establishing a governance structure that all parties perceived as fair.

Takeaway: Any enterprise blockchain initiative must solve the governance problem before investing heavily in technology. If you cannot get credible commitments from a critical mass of participants under a governance structure they find acceptable, the project should not proceed.

Lesson 2: One Dominant Player Is Not Enough

Maersk controlled 17% of global container capacity — significant, but far from dominant. Compare this to Walmart, which controls a substantial portion of U.S. grocery retail and could unilaterally mandate supplier participation. Maersk could not compel CMA CGM or MSC to do anything. In a consortium model, market power sufficient to compel participation is necessary for success, and that threshold is much higher than many organizations assume.

Takeaway: Assess honestly whether any participant (or small coalition) has sufficient market power to drive adoption. If not, the consortium coordination problem is likely insurmountable.

Lesson 3: Network Effects Are Binary

TradeLens suffered from the "last mile" problem of network effects. A shipping platform that covers 60% of a trade lane is only marginally more useful than one that covers 0%, because users still need alternative systems for the remaining 40%. This makes adoption a chicken-and-egg problem: the platform isn't useful enough until critical mass is achieved, and critical mass can't be achieved until the platform is useful enough.

Takeaway: Model the network effects carefully. Identify the minimum viable network — the smallest set of participants needed for the platform to be useful in a specific trade lane or use case. If that minimum set is not achievable, scale back the scope or reconsider the approach.

Lesson 4: Consider Whether Blockchain Adds Genuine Value

In retrospect, many of TradeLens's benefits — digitized documentation, shared visibility, reduced paperwork — could have been achieved through conventional digital platforms, APIs, and data standards without a blockchain. The blockchain provided immutability and tamper-proof records, but it was unclear how much those specific features mattered to most participants. The real value was digitization and data sharing, not distributed consensus.

Takeaway: Always apply the decision framework. Ask whether the specific benefits of blockchain (immutability, decentralized trust, tamper-proof audit trail) are genuinely necessary for the use case, or whether the same outcome can be achieved through simpler means.

Lesson 5: The Business Model Cannot Be an Afterthought

TradeLens invested years and over $100 million before establishing a sustainable revenue model — and never actually found one. The assumption was "build it, reach critical mass, and the business model will follow." It didn't.

Takeaway: Define the business model — who pays, how much, and for what value — before building. If no viable business model exists, the project will fail regardless of the technology's merits.

Discussion Questions

If you had been advising IBM and Maersk in 2017, before TradeLens launched, what governance structure would you have recommended to address the competitor problem? Could any governance structure have overcome the fundamental tension of a competitor-led platform?
TradeLens enrolled over 300 organizations and processed data on half of global container trade by volume. By many metrics, it was a successful platform. Yet it was commercially unviable. What does this tell us about the relationship between adoption metrics and commercial viability in platform businesses?
The DCSA's standards-based approach (defining interoperability standards rather than building a shared platform) is arguably a response to TradeLens's failure. Compare the platform approach and the standards approach. Under what circumstances is each more appropriate?
Maersk subsequently pursued digitization through conventional technology (APIs, digital platforms) rather than blockchain. Does this suggest that TradeLens's core problem was the blockchain architecture, or that the blockchain was incidental to a deeper organizational challenge?
If a national government mandated that all shipping documentation pass through a blockchain-based system, would TradeLens's problems have been solved? What new problems might such a mandate create?