Measure what (really) matters

Analytics is all about measuring a business so you can make better business decisions. I’ve done a lot of analytics for a lot of companies. Implementation can be complex and challenging, but the idea is simple: measure business outcomes that you care about, and measure as many of the factors that contribute to those outcomes as possible. Then, use statistics and data modeling methods to understand how things you can change will affect the outcomes you care about. It’s a mature field with a lot of powerful data tools, established methodologies, and best practices.

Analytics has roots in 1950s management theory, and the idea that “what gets measured gets managed”. It grew up alongside computers and relational databases and cloud computing, taking on the (somewhat pretentious) monikers of Business Intelligence, and later, Data Science. By 2012, that combination of business context, computational chops and statistics were so in demand that Forbes called Data Science “sexiest job of the 21st century” (I was still doing biodiversity research in Mexico, so i didn’t have a chance to let it go to my head). In 2018, John Doerr published Measure What Matters, and companies started organizing their data into Objectives and Key Results (OKRs). Software tools sprouted up to convert the growing volumes of data companies collect into metrics, and by borrowing the best practices of software engineering like version control and automated testing, those tools – many of them open-source and free to use – made turning data into insights much faster, cheaper, and more reliable. Many companies found that they could indeed rally people around metrics to drive impressive growth.

But what if you run your business to do more than grow and return a profit? What if those outcomes you care about include some standard business goals, but also goals like biodiversity conservation, reducing the impact on climate change, and creating good jobs and strong communities?

Many businesses have explicit sustainability or social justice missions; they exist to create an impact (or in some sense, less of an impact) as well as a profit. There’s more than 250 B-corps in the greater San Francisco Bay Area, and almost ten thousand worldwide. More than 350 have committed to rigorous climate neutral certification, and there are about 37,000 fair trade certified products worldwide. And those are just some examples of established standards for social and environmental commitments. There are many more companies driven by a mission beyond profit where the measures of impact are less clear.

But how do we measure social and environmental impact? It’s not as straightforward as measuring growth and profit. Money is a concept made to easily quantify a business. It behaves well when you add or average it, even at massive scales. But social and environmental outcomes are more multidimensional and unwieldily, more tricky to quantify, and more prone to the flattening and obscuring effects of measurement. It’s not impossible to measure something like biodiversity or a carbon footprint, but it’s certainly possible to mismeasure them. Mission-driven companies often keep their business metrics in one place and their impact metrics in another, usually a collection of spreadsheets and some proprietary software tools. Answering questions about impact is limited by the constraints of a proprietary carbon accounting tool or supply chain software. If you want to ask questions that involve both business outcomes and impact metrics, you’re manually downloading and lining up spreadsheets every time.

I think we can measure those impacts using the same tools and approaches that we use for business analytics, paired with methodologies from the environmental and social sciences. I think we can use those same analytics tools and best practices: the open-source software, cloud data warehouses, version-controlled code and automated tests – to process data about social and environmental impact. And I think that doing it that way is rigorous and cost effective, and it opens up whole worlds of analysis to understand the interplay of business goals and sustainability objectives.

It’s more rigorous than using spreadsheets and proprietary tools because the data processing is transparent replicable. Anyone with access can see the code, and thus the steps to clean, aggregate, and present data. The methods are not obscured within proprietary software or the specific steps someone took within Excel to process data. It’s more cost effective because the leading analytics tools are largely open-source, and because they automate a lot of the data processing and testing. Business people and sustainability managers can spend less time writing vlookups and if/thens in spreadsheets and more time doing everything that they are great at. By automating data crunching, we can get signals faster, early enough to act on them. And it opens the door to new kinds of questions that touch both impact and business.

And we should be honest: there are ways in which sustainability and social impact goals support and reinforce business goals, and there are ways in which those goals conflict. I do believe that social and environmental impact can be massive business opportunities, especially when measured with rigor and transparency that allows a business to build trust with its customers, find efficiencies that reduce both costs and environmental impact, and strengthen its relationship with employees and communities. But also, there are ways the goals conflict. Navigating and understanding the trade-offs and opportunities multiple goals present, and helping to inform hard decisions in that context, is at the heart of analytics. It really helps to have the data all in one place, in comparable formats, so that you can do analyses across domains.

In Measure What Matters, John Doerr and all the companies that adopted the OKR approach to analytics got it partially right. Measuring your goals and drivers of those goals, rallying people around measures of success, it really did create impressive growth, especially in software and technology companies where data is plentiful. But that approach also narrows the scope of what people work on. By focusing on growth metrics, many of those companies became even more myopic, causing staggering damage to the ecosystems, people, and communities. I know you can think of plenty of examples. In some cases, those harms undermined the trust that companies depended on to grow in the first place.

So let’s measure what really matters. Let’s measure the business stuff, but lets also measure the impact, the goals beyond profit and growth, even though it is hard. I’m not saying that we can perfectly quantify what matters most to us; there is no perfect metric for a resilient ecosystem or a thriving community. But we can get valuable indicators of our impact that help us navigate our ambitious goals to build a business in alignment to what matters most to us.