Predictive Analytics Can Deliver More Sustainable Outcomes

What savvy investors need to know to deliver outsized impact this decade

All investors make decisions without knowing what the future holds — uncertainty is baked into the job description. But savvy investors use data and insights to make better decisions, increasing the likelihood of successful outcomes. That’s why it’s common practice for investors to analyze the financial fundamentals of a company by examining market insights and forecasts before allocating capital.

Impact investments, now with more than $500 billion in assets worldwide, have a dual purpose: making money and having a measurable positive impact. Financial outcomes are forecasted pre- investment, yet most impact outcomes are measured and managed post-investment. We’d make better investment decisions if we forecasted impact before checks are cut.

There’s a growing movement within the impact investing community to use predictive impact analysis in forecasting outcomes. New methods rooted in social return on investment principles, such as Y Analytics’ impact multiplier of money (IMM), IFC’s Anticipated Impact and Monitoring (AIMM), and CRANE’s emission reduction potential, offer rigor in what would otherwise be a subjective process.

There is a similar movement in climate venture capital and venture philanthropy. The Autodesk Foundation invests in and supports a growing portfolio of low carbon technologies. With our peers at Prime Coalition, Clean Energy Trust, Project Drawdown, and Breakthrough Energy Ventures, we’re quantifying potential climate impact before allocating capital. We do this even when startups have little-to-no commercial traction to speak of.

Predictive impact analysis methods are new, and more collaborative action is needed to bring them into the mainstream. But we have found that by collecting more information up front, not only do we better understand the investment opportunity – we better understand the entrepreneurs and their full potential.

Bringing predictive impact analysis, not just financial fundamentals, into mainstream investing is the opportunity of this decisive decade. If successful, the reward — which we believe includes prosperity and well-being for people and the planet — can’t be understated.

In this post, we share insights gleaned, and address shortcomings we’ve found, in using these tools in our due diligence process across our portfolio.

Forecasting impact in climate mitigation startups with ERP

Low carbon tech investors are at the forefront of predictive impact analysis. One method gaining traction in the community is Emission Reduction Potential (ERP), which assigns emission abatement to commercial sales. For every product sold, or kWh deployed, there’s a corresponding impact on reducing greenhouse gas emissions — and as a company grows, its direct impact on climate grows in lockstep.

ERP tells you the net impact a company could achieve over the status quo, not what it will achieve. Knowing the externalities of commercial scale can reinforce a conviction that, if successful, a startup will have a meaningful positive result.

Predictive impact analysis in practice

The Autodesk Foundation uses predictive impact analysis, like ERP, in our growing impact investing portfolio. We do so to inform due diligence in service of our mission. At the highest level we aim to:

Invest catalytic capital in entrepreneurs using design and engineering to address the world’s most pressing problems — namely climate change and inequality. We focus on early stage (pre-Seed to Series A) ventures that uniquely benefit from Autodesk’s technology and talent. Our work covers three areas: low carbon innovation, resilient communities, and future of work.

ERP is most helpful when assessing the positive social and environmental impact of investments in low carbon innovation. We’re working to track progress against the following three goals:

• Reduce global greenhouse emissions on the potential scale of 500 MMT CO2e or more
• Lead our architecture, engineering, and construction (AEC) and design and manufacturing (D&M) industries to increased energy and materials productivity
• Increase our impact investment allocation and recycle financial returns to reinvest and support more innovations

Our current portfolio consists of 12 low carbon technologies – ranging from a startup mitigating emissions at the microbial level to a startup drastically reducing energy in food warehousing. Together the portfolio has a net emissions reduction potential of 24 gigatons of carbon dioxide equivalents (Gt CO2e) by 2050. That’s enough to cut greenhouse gas emissions from today’s forecasts by 1.5-2% if every company in our portfolio reached its full market potential.

Of course, we understand it’s unlikely that the companies in our portfolio will follow the growth we’ve modeled. Entrepreneurs face infinite twists and turns at the earliest stages and our assumptions won’t all be accurate. Some companies may grow faster than others. Some companies will face new competitors. Others will fail altogether.

The absolute ERP numbers aren’t the point. But they do help build confidence that if our companies are commercially successful it’ll be a homerun for the planet. And homeruns in this field are essential if we’re going to avoid the most catastrophic effects of climate change.

The value of predictive impact analysis

The Autodesk Foundation made its first direct investment in 2018 in Treau, an HVAC-innovation startup developing air conditioners that promised to be twice as efficient as traditional window units. We used predictive impact analysis to understand the positive and negative externalities the company faced when scaling.

This work led us to add three impact fundamentals to our decision-making process:

1. Understand the people and culture driving the business.

Forecasting impact outcomes affords a unique view of the most important driver of startup success: the entrepreneur.

In working with Vince Romanin, Treau’s founder, on our ERP model, we came to better understand the level of depth of his commitment to Treau’s vision. All of Vince’s decisions stemmed from an unwavering commitment to mitigate emissions and revolutionize heating and cooling globally. Two years into the investment, that shared vision underpins the culture of Treau’s growing team.

2. Understand impact assumptions underpinning decisions.

Forecasting impact brings forward all assumptions made by VC investors, including explicit assumptions around impact.

With Treau, it became clear that some impact assumptions were backed by quality data — like the likely global phase-out of refrigerants in response to the Kigali Amendment to the Montreal Protocol. But some assumptions were not — like the energy efficiency improvements of competitive products over time. Exposing the fallacies in our judgement allows us to analyze our own assumptions about what “impact” really means and how it’s realized.

3. Understand impact directionality of the investment.

Treau isn’t the only startup hoping to disrupt the legacy heating and cooling industry. But it is one startup combining tech innovation, a scalable business model, and a go-to-market strategy that could yield potentially meaningful emissions reduction. By keeping assumptions constant across investment opportunities, predictive impact analysis provides directionality. We can choose to focus on the deals that mitigate 900 MMT CO2e, not 9 MMT CO2e.

Shortcomings of forecasting impact outcomes

Forecasting impact doesn’t replace the need for thorough diligence or impact management post- investment. It’s an added task. Forecasting can be challenging and lead to shortcomings worth highlighting:

1. Time-intensive research and varying data quality

Changing existing diligence processes requires deep thinking on futuristic outcomes. Questioning underlying impact assumptions and building a new model requires time and research, especially at the onset.

Startups don’t come with randomized control trails. Data quality varies by technology and by sector.

Luckily, new market-building tools are emerging to facilitate the widespread adoption of impact forecasting. One example is Carbon Reduction Assessment: New Enterprise (CRANE). By gathering ERP assumptions and data across 200+ technology solutions, CRANE will help investors bring climate externalities into their decisions. CRANE is a collaboration between Prime Coalition, RhoAI, Geenometry, Clean Energy Trust, and Project Drawdown and will be released in April 2020.

2. Market impacts and system-level interventions

Forecasting impact outcomes assumes a one-to-one correlation between business success and impact outcomes. But sometimes this can be a problematic assumption. Take energy storage, for example. Low-cost, long-duration storage solutions are needed to make wind and solar the norm, but the ultimate impact of storage solutions is in enabling other technologies to proliferate — an indirect impact that can be challenging to forecast.

Companies may have primary and secondary impacts on the market. By creating consumer demand for electric vehicles, Tesla put pressure on other car companies to create fully electric vehicles. The result is a positive externality (more electric vehicles) and negative externality (demand on extractive material supply chains) that’s more art than science to forecast.

System-level interventions — including those capable of reshaping the fabric of a complex system — may fall through the cracks of a model rooted in the realities of today’s world. Frameworks championed by the Impact Management Project are already bringing together investors of all types to better understand these nuances, but much more work is needed.

Bringing a movement to mainstream investing

Bringing impact fundamentals into mainstream investing is the opportunity of a generation. If successful, the reward — which we believe includes prosperity and well-being for people and the planet — can’t be understated. That’s why predictive impact analysis, or any tool or method that has the potential to improve outcomes, is so important.

At Autodesk Foundation we believe the only way to realize the potential of impact investing is to do so together. Today’s impact forecasting methodologies and tools aren’t perfect – further

collaborative development with investors across asset classes and risk/reward thresholds is needed. There’s a lot of co-development to do.

If you invest in low-carbon technologies, sign up to beta test CRANE prior to its public launch in April. If you don’t invest with a climate lens, consider building and sharing similar open source tools, so other investors can stand on your shoulders.

Regardless of your investment philosophy, we challenge you to engage and we welcome your open input and feedback. Only as a collective investment community will we be able to rewire commercial markets to work for the planet and all its people.

- Lynelle Cameron, CEO, Autodesk Foundation

- Ryan Macpherson, Impact Investing Lead Autodesk Foundation