Failed sustainability projects and the lessons learned

Enterprise sustainability initiatives increasingly depend on technology, from carbon accounting platforms and digital twins to cloud infrastructure and data systems.

But the right technology purchases don't guarantee an organization will achieve its sustainability goals. Poor data, siloed teams, unrealistic expectations, unclear ownership and sustainability goals disconnected from business priorities can undermine an initiative before the technology has a chance to deliver.

Cost-efficient regulatory compliance remains the most common goal of sustainability initiatives examined by global technology research and advisory firm ISG, according to Matt Warburton, digital sustainability lead at ISG. Decarbonization follows closely as organizations recognize potential cost, efficiency and resilience benefits from reducing natural resource consumption.

Understanding how failed sustainability projects unravel can help IT and sustainability leaders recognize warning signs earlier and avoid repeating the same enterprise sustainability mistakes.

Carbon accounting software couldn't compensate for missing supplier data

In one failed project, a midsize manufacturer bought a carbon accounting platform to automate its Scope 3 emissions reporting, according to Stanislav Kazanov, head of GRC, cybersecurity, sustainability and data at Innowise. The organization expected the software to calculate emissions using procurement and logistics data. But it had not determined whether its approximately 400 suppliers could provide the data the system required.

Three months into the project, the implementation partner requested API access to supplier systems and discovered that many suppliers didn't have APIs. Some still relied on paper records. Eighteen months later, the sustainability team continued chasing suppliers for spreadsheets and PDFs. The platform worked, but it didn't have the data needed to deliver the expected automation.

Leadership continued the contract for another year because of the money already invested, Kazanov said. Eventually, the organization essentially turned the platform into a repository for manually uploaded information.

"Most problems with sustainability tech aren't about the tech itself. They're about data," Kazanov said.

Warburton sees a related problem across enterprise sustainability initiatives. Organizations can underestimate the risks and costs of integrating environmental, social and governance (ESG) data across the business. Compliance-oriented SaaS applications and spreadsheets can create more silos rather than address data lineage, workflow and organizational issues.

The IT sustainability lesson is to evaluate the data before evaluating the technology. Kazanov recommends auditing data before approaching vendors and proving the entire pipeline on a limited scale, such as one site, emissions scope or building system, before expanding an implementation.

A data center digital twin ran into organizational silos

Another organization that Kazanov advised attempted to use a digital twin to improve data center cooling. The technology required real-time information from facilities systems and IT infrastructure. But facilities engineering controlled the building systems, while IT controlled network and server data. The groups had not standardized their data formats before the project began.

It took approximately nine months to calibrate sensors across three sites, Kazanov said. By the 14th month, finance leaders were questioning why the organization continued to pay for a smart platform whose data still required manual correction. More significantly, no one could name a concrete decision that the digital twin had improved. The organization ultimately canceled the project.

The problem wasn't that digital twin technology couldn't optimize cooling, Kazanov said. The organization implemented it without first resolving the organizational and data problems on which it depended.

Warburton identified unclear ownership as a recurring contributor to the failure of sustainability programs. Sustainability initiatives require collaboration among IT, finance, procurement, operations and sustainability teams, making clearly defined roles and workflows critical.

For IT leaders, the lesson is to address organizational integration before technology integration. Kazanov recommends assigning a data owner with sufficient authority to work across IT, facilities and operations.

Cloud migrations can move emissions instead of reducing them

Cloud migrations can appear to be a sustainability success without delivering the environmental improvement that an organization assumes it has achieved. Srinivas Chippagiri, a senior member of the technical staff at Salesforce who works in cloud-native engineering, distributed systems and platform infrastructure, sees problems when organizations fail to reconcile IT efficiency metrics with sustainability metrics.

A genuine reduction can occur when an underutilized workload moves to a consolidated infrastructure requiring fewer physical machines, Chippagiri said. A cloud facility might also achieve better power usage effectiveness (PUE) than a company's server room, while the destination region could use electricity from a lower-carbon grid. Rightsizing or re-architecting workloads can produce more gains.

A lift-and-shift migration presents a different scenario. Moving an inefficient workload unchanged can shift emissions to the cloud provider without addressing the underlying inefficiency. An organization's own reporting might improve because it no longer directly measures the energy consumption, even though the workload still consumes resources.

"The tell is simple," Chippagiri said. "If the only thing that changed is whose books the emissions sit on, and not the amount of work or the energy per unit of it, you relocated emissions rather than reduced them."

Chippagiri recommends measuring energy consumption or carbon emissions per unit of useful work before and after migration. Depending on the workload, that unit could be a transaction, request or completed computing job. Organizations can also compare hardware use, the grid carbon intensity where and when the workload runs, and PUE at the source and destination facilities.

"If you cannot state your carbon per unit of work before and after and show it went down, you have not demonstrated a reduction; you have demonstrated a change of address," Chippagiri said.

A lower monthly cloud bill isn't sufficient evidence of sustainability progress because costs can fall while emissions remain unchanged or increase, he added.

Sustainability dashboards can measure problems without fixing them

Sustainability goals might belong to a chief sustainability officer, infrastructure to IT, and individual workloads to product or engineering teams. Each group can optimize its own responsibilities without anyone owning the overall sustainability outcome, Chippagiri said. The result can be sophisticated carbon or energy dashboards that don't produce operational changes.

"Instrumentation is mistaken for change," Chippagiri said.

Warburton considers disconnected metrics a warning sign. When organizations report sustainability measurements separately from normal operational and financial metrics, those initiatives face a greater risk of being deprioritized.

Chippagiri recommends selecting a metric that connects cost and carbon, such as energy consumption or emissions per unit of useful computing work, and assigning an engineering team to work on and improve it.

Unrealistic ROI expectations can tank workable sustainability projects

Some sustainability initiatives fail because organizations evaluate long-term investments against short-term financial expectations.

Kazanov has seen renewable energy and smart building initiatives run into this problem when organizations expect returns within a year from investments designed to pay off over five to 10 years. In some cases, solar and battery projects have been terminated at the 18-month review because the savings didn't appear when measured against financial models developed for shorter-term investments, he said.

Chippagiri has seen a similar pattern in data center and infrastructure sustainability efforts. Benefits from consolidation, rightsizing and re-architecture can appear later in a project's lifecycle, meaning an initiative built around aggressive short-term ROI expectations can lose funding before those benefits materialize.

Sustainability, IT and finance teams need to establish success metrics, measurement periods and expected returns before they seek approval. Kazanov said teams that make it through the early stages tend to agree with finance on the measurement horizon at the beginning, rather than renegotiating expectations after initial results disappoint.

Sustainability projects become vulnerable when separated from the business

One warning sign cuts across many failed sustainability projects: Sustainability metrics exist independently from the operational and financial metrics executives use to run the business.

Organizations are increasingly moving away from treating sustainability as a separate goal and instead using it as another lens through which they examine existing operations, Warburton said. That can mean finding waste in the value chain, figuring out whether reducing a product's environmental impact increases its value to customers or assessing how climate-related risks could affect the balance sheet.

Warburton recommends that IT leaders identify where sustainability, financial and organizational goals align; integrate metrics used to evaluate them; invest in enterprise-scale ESG data foundations; and establish cross-functional ownership. The failures offer a consistent warning: Technology cannot compensate for poor data, unclear accountability, unrealistic financial expectations or organizational silos.

Avoiding those enterprise sustainability mistakes starts before an organization buys a platform or migrates a workload. Leaders need to establish what success means, decide how to measure it, identify who owns the outcome and ensure the initiative supports business goals that can survive changing priorities and budget cycles.

Christine Campbell is a freelance writer specializing in business and B2B technology.