5 Lessons on Data and Sustainability

By Salvador Núñez

Berkeley Master of Engineering
8 min readSep 5, 2017

Salvador Núñez has worked at the intersection of strategy, operations, and data science across the life sciences and energy industries for over 8 years. His experience includes projects throughout the United States, Europe, and Latin America. He began deploying CRM software to international affiliates at Alcon Laboratories and now develops data science tools to automate and optimize rooftop solar operations at NRG Energy. His latest venture is a consulting practice called Sustainable Data. Salvador holds a B.S. in Biomedical Engineering from Yale University, as well as a MEng in Industrial Engineering and Operations Research and a Certificate in Engineering and Business for Sustainability from UC Berkeley.

Ibecame interested in data science when I realized how it generalized my training as a scientist to solve problems in any domain. Although my domain knowledge began in biotechnology, I’ve grown increasingly passionate about energy and sustainability throughout my career. The reality of industry work has revealed many things that are often ignored in academia. Although any business can be modeled as an optimization problem with an objective function, a set of constraints, and a set of decision variables, traditional scientists do not work through the challenges of change management and organizational behavior when implementing an algorithm. That’s why both soft skills and analytical ability are critical in solving these problems.

During this era of unprecedented growth, sustainability has emerged as one of the most pressing problems we must now face. From disruptive startups to large established corporations, data can have a big role in driving sustainability. Here are 5 lessons that I’ve learned that can help organizations use data to become more sustainable, efficient, and productive.

1. Reframe the measured objective

Sustainability brings a unique set of challenges to a business by prioritizing long-term value over short-term profits. Too often, sustainability is modeled as a constraint rather than an objective that measures prosperity. If an organization is able to prioritize sustainability, management should change the data it collects, measures, and reports.

Reporting standards can help. Businesses have followed General Accepted Accounting Principles (GAAP) for decades, but have not had standards for sustainability reporting. The Sustainability Accounting Standards Board (SASB) and the Global Reporting Initiative (GRI) are tackling that problem. Companies, like NRG Energy, have joined the GRI Standards Pioneer Program and are adopting best practices on how to measure, report, and manage sustainability.

Other companies may choose to measure a double bottom line through metrics uniquely tailored to their business. For example, Opower’s mission was to motivate everyone on Earth to save energy. The saved kilowatt-hours were reported alongside the dollars in every sale and in every product. Sustainability doesn’t have to be all about the environment either, short term profits can overlook social externalities as well. Beyond a double bottom line, sustainable companies may even pursue a B Corp certification and measure a triple bottom line around economic, environmental, and social measures. Patagonia, for example, chooses to collect and report data on its social and environmental impact in creative ways, such as The Footprint Chronicles.

2. Boldly anticipate opportunity and iterate quickly

Innovative business models may be needed to create new markets and pursue new business opportunities. The advantages that come with being first-to-market can outweigh the risk of failure and many organizations take leaps of faith before calibrating a solution. Using data to anticipate opportunity and iterate quickly can transform these leaps of faith into calculated risks. Seizing sustainable business opportunities, in particular, can depend on influencing and anticipating changes to policy and consumer behavior. For example, at the beginning of the Obama Administration, cap-and-trade seemed like a ripe opportunity for the taking. As a result, C3 Energy began as an emissions management company, which then pivoted to focus on smart-grid applications, and currently provides an Artificial Intelligence & Internet of Things software platform known as C3 IOT.

Today, changes in policy and consumer behavior are disrupting the auto and utility industries. Electric vehicles (EV) sales are expected to increase by 117% in 2017. Countries like Norway, Netherlands, and China plan to put many millions of EVs on the road by 2020. In the United States, Tesla recently made headlines when it became the most valued US automaker by market capitalization. Tesla’s success depends on the fact that it is a tech company as much as an automaker, if not more so. As a tech company, it has a strong data strategy. For example, it has already been capturing video clipsfrom Teslas on the road to train its autopilot algorithms on how to deliver autonomous driving in 2 years.

Utilities are collecting data too. Consolidated Edison’s SmartCharge New York program will pay customers 5 cents per kWh to charge during off-peak hours. While energy curtailment and time-of-use rates can partially support the economics of this program, the real value comes from the data the program will collect. ConEdison will be able to gather data on how the growing popularity of EVs could potentially stress the electric grid’s infrastructure and develop a better informed investment strategy as a result.

3. Develop data stewardship before data science

Regardless of whether a business has sustainability goals, today’s business growth increasingly depends on sustainable data growth. On the one hand, the go-to-market strategy and the cash flow for a business may require the plane to be built while it’s flying. On the other, short-sighted system design and data management can prove costly and prevent the business from scaling later. Successfully informing business operations requires agile and farsighted collaboration between IT, data, and business professionals. In fact, many organizations may benefit more from better data stewardship than from cutting-edge data science. Management can help foster a sustainable data culture by measuring data quality, proactively handling exceptions in the data pipeline, incentivizing key stewards across the organization, and measuring long-term improvements in organizational behavior and data governance.

Information systems should prevent faulty data from ever being recorded. If errors occur due to changing business requirements or the shortcomings of an initial design, exception reporting can be a safety net that allows these issues to be addressed early on. The difficulty and effort involved in rectifying these issues will only compound with time. Fixing these errors should be done while the context is fresh and before the downstream risk of inaccurate metrics, reports, and insights weaken business operations. Furthermore, adopting software development techniques such as automated exception handling can help minimize manual intervention that can distract from other pressing business needs.

From salespeople to data analysts, management can identify and incentivize key data stewards to prevent or immediately fix these errors. I’ve seen sales organizations where compensation is directly tied to the data quality entered in Customer Relationship Management (CRM) software. On the other hand, IT and analytics teams should also be staffed with junior analysts that focus on maintaining tidy data sets that maximize the impact of more senior data scientists. Instead, many data scientists often spend 50 to 80% of their time simply collecting and cleaning data.

While spreadsheets may continue to be the lingua franca in business, a comprehensive Business Intelligence (BI) strategy may be necessary to break down business silos and improve system integration. Upgrading to more sophisticated information systems and procuring the latest BI platform may not solve the issue without changes to organizational behavior and data governance. To guide this change management, it’s important to define goals and measure improvements. For example, organizations may choose to keep track of how many separate information systems need to be queried to answer a simple business question, how quickly the data behind these answers goes stale, or how many teams manage information systems feeding the same key performance indicator (KPI).

4. Nurture and empower data scientists with soft skills

Although there is a huge demand for data scientists, data insights do not drive value on their own. Actions do. Delivering results requires more than mastering machine learning algorithms and NoSQL frameworks. In fact, tech employers say soft skills are top hiring factors for data analytics jobs.

The most impactful data scientists bridge the gap between opaque statistical models and actionable business insights, work across teams to implement changes within the organization, and care deeply about the outcomes. Motivating data scientists to push the business in following through with their insights all the way to implementation can transform them into a force multiplier. With data science being such a new field, organizations can empower young talent by leveraging their longing for passion, authenticity, and purpose. Sustainable businesses can leverage their mission-oriented culture, for example, to motivate data scientists to take their data insights to the next level.

5. Shape consumer behavior through personalized data insights and behavioral science

Sustainability can often involve deep changes in human behavior. Beyond managing organizational behavior, it is critical to guide sustainable consumer behavior as well. Online advertising has long used personalized data insights and behavioral science techniques to shape consumer behavior. Its success has been largely driven by the ability to use data to present highly relevant and personalized content. A study from Venture Beat, for example, shows that email personalization has 2.5 times higher click-through rates (CTR) than static emails. These data also inform when and how to apply the psychology of persuasion to make consumers more likely to buy your product. How many of us buy that last ticket on kayak or last item on amazon because we were nudged by the illusion of scarcity?

Fortunately, psychological jiu jitsu informed by personalized data insights also has the potential to transform consumers into “prosumers” that drive sustainability in the marketplace. A simple cue carrying social pressure, like a normative comparison, can be more effective in changing consumer behavior than economic incentives. Opower built a successful business engaging consumers on a traditionally boring subject, and helped save over 11 terawatt-hours of energy, simply by stacking up a household’s energy usage with its neighbors and tapping into our innate competitive spirit. That’s enough energy to power over 1 million U.S. homes for a full year — and that’s just one example. From educating teenagers about the risk of HIV in South Africa to conserving water in Costa Rica, organizations like ideas42 are showing us that data and psychology can also be positive forces in driving sustainable behavior around the world.

According to the former US Chief Data Scientist, DJ Patil, wisdom is attained by reflecting on experience. I’m grateful for the lessons I’ve learned and think we should be optimistic about the promise of data science. While sustainably managing our scarce resources is an increasingly pressing problem, data is one resource that continues to grow. But, like all resources, it’s still up to us to use it wisely.



Berkeley Master of Engineering

Master of Engineering at UC Berkeley with a focus on leadership. Learn more about the program through our publication.