How to Fix Post-Pandemic Supply Chains

Digitization alone won’t guarantee supply chain success

MIT Initiative on the Digital Economy


By Michael Schrage

Courtesy of MIT Sloan Management Review*

That COVID-19 dramatically accelerated digital transformations worldwide has become C-level consensus and a truism. As Microsoft CEO Satya Nadella publicly observed, “We’ve seen two years’ worth of digital transformation in two months. From remote teamwork and learning to sales and customer service, to critical cloud infrastructure and security.”

Even discounting for hyperbole, pundits and practitioners have been echoing Nadella’s sentiment. Organizations understandably seek to wring digital virtue from pandemic necessity. While many business operations can swiftly transform, other key processes defiantly resist digital acceleration. Supply chains are a case in point: Spreadsheet- and ERP-dependent supply chain operations had to radically revisit and revise expectations.

Yesterday’s digital transformation road maps proved largely useless. Organizations understandably seek to wring digital virtue from pandemic necessity.

Why? As heavyweight boxing champion Mike Tyson is said to have declared, “Everybody has a plan until they get punched in the mouth.” COVID-19’s impact revealed that supply chain business continuity plans had both the wrong data and the data wrong. Top management literally couldn’t see what was happening — or needed to happen — to ensure safe and reliable deliveries under duress. This came as a shock. Data, not digitalization, was their immediate problem. Legacy leadership teams need to understand that decisions around data — not digitalization — drive successful supply chain transformation.

Most significantly, targeted transformation investment overwhelmingly emphasizes greater visibility and transparency rather than supply chain optimization. Policies and practices promoting granular, real-time data access at every relevant link in the supply chain assumed primacy and urgency. As organizations confront their global futures in post-pandemic, machine learning, “What about China?” contexts, those decisions take on new urgency, as well as importance. I’ve interviewed — and worked with — more than a dozen supply chain managers and executives since pandemic shutdowns began.

Virtually all of them have fundamentally reset their supply chains’ strategic priorities. “Just in case” now matters more than “just in time.” Assuring worker safety now supersedes improving inventory turns.

“Honestly,” said one supply chain executive from a Latin American food and industrial equipment conglomerate, “we didn’t know how little we knew about our key suppliers until COVID-19 occurred. We had to get all kinds of [new] information from them and — to make [our] factories work — we had to give them information we never had before.”

The takeaway: Digital-first enterprise success demands clarity-first supply chain design. Digitally transforming supply chains requires digitally transforming transparency and visibility. Transformational transparency is what leadership needs to measure; visibility is what leadership needs to assess. Unless they first determine data access, quality, and lineage, digitization can’t deliver agility or reliability.

This is the most important organizational (re)learning from the COVID-19 crisis. (And this learning is gaining mindshare in more recent supply chain discussions.) Digitizing supply chain process improvements without deliberately defining, collecting, and labeling essential data computationally codifies blind spots. Workarounds won’t work without the ability to swiftly see, find, or acquire relevant data. Where is that data coming from? Who owns it? How — and why — is it being made accessible? Supply chains fail when those questions are primarily answered by IT, legal, or a chief digital officer.

The data-driven ability to see — or computationally infer — tacit or explicit supply chain interdependencies may seem obvious. Pre-pandemic systems design, however, indicates that the obvious was misunderstood. Many “go to cloud” supply chain initiatives privileged process models over data models.

Too many managers believed that moving legacy processes, data, and analytics to the cloud would automatically enhance transparency and visibility. They got punched in the mouth.

“Besides safety, the most important investments we’ve been making in our global supply chain focus on improving transparency and visibility worldwide at every step and stage,” said Todd Stohlmeyer, global strategic sourcing and procurement vice president at agricultural giant Cargill. For example, the company quickly deployed digital signatures to handle farm-to-delivery paperwork at every key supply chain checkpoint. Almost all of Cargill’s post-COVID-19 supply chain and sourcing documentation is now remotely enabled, recorded, and made appropriately transparent throughout the enterprise. Instantly tracking authorizations can be as important for delivery as instantly tracking product.

While using digital signatures may look and sound like a digitization initiative, focus on the underlying problem and use case: The process failure’s root cause wasn’t manual paperwork — it was the absence of visibility. Stakeholders literally couldn’t see the authorizations in a timely manner. Transformation didn’t drive transparency; the need for transparency drove transformation.

Transparency describes a supply chain’s capability and commitment to data sharing; visibility defines what the enterprise explicitly and demonstrably knows.

Improving Supply Chain Transparency

Championing transparency guarantees more challenging — and more strategic — debate about what post-pandemic optimization means for suppliers and customers. Those optimization discussions and disagreements, in turn, lead to more rigorous definitions of transparency. Meaningful optimization — the ability to weigh measurable trade-offs — requires clearly delineated parameters. Greater clarity invites greater insight into options and opportunities. Greater insight ensures better and more timely decision-making. For example, increased visibility makes clearer whether better managing complexity or supply chain simplification is the better investment.

Cargill’s transparency and visibility imperatives shape its supply chain digitalization strategy. They make everything — whether worker safety, supplier capability, or real-time customer delivery — measurably clearer, more accessible, and more accountable. That openness, said Stohlmeyer, creates options and opportunity.

This epiphany helps explain why accelerating supply chain transformation proved less tractable and more troublesome than most legacy companies — including Cargill — expected. Much of the challenge, noted MIT’s Alexis Bateman in a pre-COVID-19 interview, is that, “as we all know, supply chains are not designed to be transparent. As you move upstream with suppliers, they don’t want to disclose information they think of as a competitive secret.” Post-pandemic supply chains literally renegotiate the terms of transparency and visibility throughout their value chains.

What’s more, transparency initiatives typically favor adherence over agility, especially when rooted in sustainability, compliance, and corporate social responsibility (CSR) aspirations. Intriguingly, none of the managers I spoke with cited compliance-based data as integral to adaptiveness, responsiveness, or recovery. Compliance and CSR-oriented transparency typically doesn’t enable the anticipation or flexibility that today’s supply chains — and their customers — need. Contextually accurate data is not inherently actionable.

To be sure, public disclosures promise benefits as well as costs. As a recent research paper from MIT Sloan professor Y. Karen Zheng, Sloan visiting assistant professor Tim Kraft, and University of Pittsburgh assistant professor Leon Valdes concluded, “Increasing supply chain visibility always strengthens consumer trust. Furthermore, opportunities exist for a trust-driven revenue benefit due to greater visibility.”

Optimizing transparency to ensure effective compliance and disclosure, however, is a purposefully and profoundly different task than optimizing transparency to facilitate flexibility and response. To be sure, these competing transparencies are not inherently at odds, but — as COVID-19 made clear — they’ve proved neither complementary nor congruent.

Perhaps the most disconcerting supply chain revelation around pre-COVID transparency/visibility investment has been its inward focus. Organizations might scrupulously monitor regulatory compliance and track disrupted stock-keeping units (SKUs), stores, categories, distribution centers, or logistics on their networks. But observing how other actors — suppliers and partners — improvised or failed to cope with crisis frequently proved impossible.

Actionable data and analytics for evaluating downstream impact simply weren’t there.

“If you think of data analytics, all the models, tools that have been put together for forecasting sales or forecasting any of the things just stopped working once COVID hit. They didn’t make sense,” Anil Kaul, CEO of AI business platform Absolutdata declared in a recent article. “There was nothing historically in the data that would be able to inform you what is going to be happening.”

For Stanley Black & Decker, the world’s largest toolmaker, Kaul’s sentiment translated into tighter cooperation between the company’s data scientists and supply chain managers. With over 100,000 SKUs coordinated by complex procurement and production processes, better demand management forecasts could measurably improve prioritization, said Aleksandar Lazarevic, advanced analytics and data engineering vice president at Stanley Black & Decker in a recent email exchange. Supply chain teams could better rethink and re-rank to-do lists with more visible customer and channel signals.

This analytic challenge went beyond quantitative technique and customer data sets, Lazarevic noted during a DataRobot AI webinar. Meeting the transparency/visibility challenge required translating the demand forecasts into formats that described what was needed to take meaningful actions immediately. The forecasts were worthless unless their users had visibility into the actionable next steps that they should take. In other words, the deliverable needed to be more of a map than a checklist.

This insight had an enormous impact on the data engineering and analytic function: Answers, recommendations, and solutions that did not allow — or empower — users to see the essential data they needed to access that day were not helpful. Strategic approaches required operational visibility to get organizational buy-in.

This visibly actionable imperative was consistent across several COVID-19 supply chain planning sessions I participated in:

  • A consumer packaged goods company sought to identify its least profitable SKUs and determine what portion of its materials could be easily repurposed for more profitable/higher-volume lines. This company quickly discovered data gaps between how procurement described and labeled materials and how production actually processed them.
  • A regional grocer positioned as a champion of locally sourced/organic produce feared for the future solvency of several suppliers. The company struggled to figure out what kind of support it could provide to simultaneously ensure predictable supply levels while preserving supplier safety and viability.
  • A foods company (not Cargill) sought to quickly model how the processing-plant order/delivery tempo should shift if staffing levels dropped to 75% and extra time was allotted during each shift for cleaning. Preliminary findings indicated that overall productivity would fall so precipitously that the production mix would need to be recalculated. The knock-on supply chain impact would be even more complex. (The plant was later shut down for over a week.)
  • An industrial logistics company debated the cost and convenience of switching to cellphone-based QR codes to coordinate pickups and delivery. A nascent regional pilot delivered valuable insights within two weeks from both delivery personnel and several clients.

Four Requirements

Although generalizations are dangerous, four transcendent issues dominated discussions:

  • Outside data. Supply chain teams immediately needed data outside their remit to appropriately define, not just solve, their apparent problems. Without exception, they needed visibility into data and/or analytics owned and operated by other functions that they had not accessed before. Existing data governance processes and protocols had not anticipated these circumstances. Legacy supply chain leadership teams undervalue data governance.
  • Internal partners. Existing technical and analytic competencies of the supply chain function were not up to converting existing or accessible data into actionable insight in the face of a pandemic disruption. (My favorite anecdote is that the most valuable data from one supplier’s sustainability disclosure was the name of an external counsel; he proved surprisingly helpful in renegotiating supplier contract terms.) Partnering with legal, compliance, finance, data engineering, and sales is unavoidable.
  • Direct communications. The supply chain data gaps and discrepancies described as most difficult or challenging required direct outreach to suppliers, external partners, and even customers. Supply chain teams realized that procurement and sourcing agreements did not facilitate reasonable data access or exchange in a timely manner. Negotiated delivery terms consistently prized price over openness. You get what you pay for.
  • Improved metrics. Existing supply chain transparency and visibility policies and practices were explicitly aligned to internal key performance indicators and metrics. That is, supply chain managers invested in transparency and visibility as a function of how well it allowed them to measurably improve supply chain efficiencies. Organizations incorporating external customer metrics for supply chain assessment — such as on-time delivery, customer satisfaction, and Net Promoter Score — enjoy greater trackability/traceability capabilities.

In almost every case, meeting supply chain challenges relied less on digital maturity or orchestrating incompatible systems than on finding — or creating — relevant data fast. Even organizations with data catalogs and/or master data management quickly determined that their existing data would not effectively support key use cases. Essentially, and unsurprisingly, supply chain management discovered that poorly defined data adds little value to well-defined use cases.


Key lessons can be learned by observing how seriously supply chain leaders struggling to turn COVID-19 disruption into digital transformation addressed the following five questions. Each question reflects and respects the importance of accessible and actionable data as the organizing principle for supply chain transformation. Data first, not digital first, is the key that unlocks anticipatory, agile, and resilient supply chains.

1. Do data governance policies and processes make visibility simpler, easier, and more accessible throughout the enterprise? Ironically, many data governance programs digitally replicate the very supply chain inefficiencies they seek to eliminate. Data is warehoused until needed and labeled in ways that prioritize legacy categories over actual “use cases” Self-service is difficult; the data is often decoupled from the contexts from which it’s drawn — for example, the order is in the queue, but the warehouse from where it will actually ship is not visible.

In other words, the use cases support the job to be done rather than the data essential to doing it. Former Google engineer Steve Yegge’s infamous 2011 rant detailing how his company must architect data to be easily accessed through externalized interfaces offers a template for data governance design. People pay attention when the CEO/founder mandates data accessibility — or else.

2. (How) are you measuring visibility? Declaring visibility as an aspiration or complying with disclosure requirements is not enough. Supply chain managers need to be able to instantly assess whether they can see the data they need in order to make informed decisions or perform further analysis. Does the organization actually have the data? If so, how accessible is it? Is it in an appropriate format? What additional cleaning — or processing — is necessary?

If supply chain managers can’t instantly see the data they need, where can that missing data be found? Who is authorized to get it? Does that authorization map to accountability? Is this a one-off instance, or should this data type become part of standard use cases and/or processes?

At least one of the observed supply chain operations assigned data wranglers/stewards to create metadata for specific data sets, ingredients, and suppliers. Leadership recognized that better context promoted greater visibility. This organization pushed visibility to accelerate categorization and resolve supply chain problems faster.

3. Are you as transparent as your best customers want you to be? Customers now expect to be able to track an Amazon shipment or an Uber Eats delivery. People can literally see where their orders are and when they expect to be delivered. In Australia, Domino’s Pizza sends customers actual photos of the bespoke pizzas that are on the way.

Transparency both sets and manages customer expectations. While the underlying logistics processes may be digital, customers can exercise their power to see what’s going on. Customer-centric organizations must ask themselves how transparent — that is, how visible — they want to be for their customers. Mercedes, for example, allows customers to see their cars being built. Should grocery shoppers be able to see where their food is processed? Should fashion shoppers see the ateliers where their clothes are designed and the factories where they’re made?

Enabling visibility and transparency from the customer perspective creates greater demand for internal visibility and transparency to both anticipate and resolve customer concerns. For one industrial equipment supplier, for example, sending a digital product photo from the factory floor built trust with a potentially important customer.

4. (How) are you championing visibility culture inside and outside the organization? Visibility can be a capability without being a value. Without exception, supply chain managers I spoke with asserted that they wanted to encourage and embrace greater visibility as a value. The costs of opacity and inaccessibility had been too high. Their biggest challenges didn’t involve internal silo-ization or data governance issues; the issues were outside the organization.

As observed above, supply chain managers acknowledged that relationships with key suppliers needed to become more transparent. Contracts would have to be renegotiated; specific levels of transparency/visibility would need to become part of service-level agreements. Procurement requests for proposals would have to better balance price, quality, and visibility. Most important, visibility that measurably improved internal efficiencies and customer outcomes would be recognized and rewarded.

These conversations — internal and external alike — have just begun. Will greater trust lead to greater transparency? Or will greater transparency lead to greater trust? Or will lean trust and lean transparency suffice in a post-pandemic era?

5. How will our visibility/transparency commitment facilitate AI/machine learning/automation adoption? In the first and final analysis, better visibility and better transparency ensure better data. Similarly, better data ensures more effective and reliable training sets for machine learning and AI systems. Better visibility and transparency allow organizations to better manage the inherent risks of automating supply chain processes in warehouses and distribution centers worldwide. This makes digital twins for supply chains as meaningful and manageable as digital twins for production. Without exception, supply chain managers I spoke with recognized that investments they made in visibility and transparency would directly impact the efficacy of automation and AI investments. The COVID-19 crisis both intensifies and elevates the conversations supply chain leaders and the C-suite need to have about the future of digital process automation and optimization in an era of smart machines.

Trying to predict the supply chain future when relationships with China are in flux, global regulations around sustainability are in transition, and regional economies struggle to overcome the impact of the COVID-19 pandemic is likely a fool’s errand. The smarter bet recognizes that the odds favor supply chains becoming supply networks that, increasingly, will learn to optimize not just internal coordination but customer outcomes. The leadership principles making successful transformation possible will be transparent and visible to anyone who looks for them.

Michael Schrage is a research fellow at the MIT Sloan School of Management’s Initiative on the Digital Economy, where he does research and advisory work on how digital media transforms agency, human capital, and innovation.

*This blog was originally published at on July 29, 2020.



MIT Initiative on the Digital Economy

Addressing one of the most critical issues of our time: the impact of digital technology on businesses, the economy, and society.