Next week we’ll be attending the Council of Institutional Investors (CII) Fall Conference in Washington, D.C., our fourth year going.
For anyone who hasn’t been, CII is one of the few conferences where the people who actually run proxy voting, corporate engagement and stewardship programs—and live in the weeds of corporate governance—end up in the same place.
Which makes it a good moment to step back and reflect on how our business—and the problems governance teams are trying to solve—have changed over the past five years.
When we started this business five years ago, we weren’t experts in the alphabet soup of acronyms in the world we were stepping into—DEI, ESG, HCM, TSR, and the rest. The thought of competing against established, well-funded, and widely respected incumbents like Bloomberg, MSCI, and ISS was intimidating.
What we did bring was deep experience in synthesizing large amounts of unstructured information into structured datasets, deriving insights and signals, and building tools that fit into customers’ workflows—saving them time and helping them make better, more data-informed decisions.
Our early meetings with some of the largest asset managers and advisors in the world were eye-opening. The message was essentially: we are desperate for high-quality data on workforce composition, development, and policies…and the delivery can be in a spreadsheet.
Unlike financial reporting, there was almost no consistency in how or where this information appeared.
Relevant disclosures could show up in SEC filings, sustainability reports, diversity reports, or corporate websites. The information itself might appear in text, tables, graphics, or embedded in PDFs. Even the language used to describe the same underlying concepts varied widely from company to company.
To solve this, we built our own research system in-house—tools that helped us discover the documents we care about, extract relevant information, and route it into analysts’ feeds where it could be vetted, normalized, and transformed into structured datasets.
We differentiated based on depth, timeliness, and quality. While larger vendors focused on covering tens of thousands of companies and hundreds of metrics, our clients valued something simpler: when our data showed something had changed, it was comprehensive, up-to-date, and accurate.
We scaled coverage deliberately over time, adding companies and sources as we went. There was no easy way to automate these processes, even with the best “AI” tools available at the time. Traditional NLP and machine learning helped with classification and discovery, but they didn’t eliminate the need to actually understand and validate the information before delivering it to clients. There was no shortcut that didn’t come at the expense of quality.
Fast forward to today, and the narrative around AI is dramatically different. Headlines suggest we’re in the middle of a technological revolution that will transform every industry overnight. According to the more optimistic forecasts, AI agents will soon be doing half of all white-collar work, though if you ask the pessimists…it might take until 2030.
We’ve been integrating generative AI into our workflows for more than two years now—across research, software development, and data analysis—and there have been real gains. Some tasks that were previously manual and resource-intensive now take a fraction of the time. The ability of modern models to interpret and summarize unstructured information has significantly increased the scale at which we can ingest and analyze new sources.
At the same time, our experience has been more measured than the hype suggests. In many cases, generative AI behaves less like a finished product and more like a team of eager, well-educated interns—curious, capable, and often surprisingly helpful—but still requiring direction, supervision, and constant fact-checking. Sometimes it saves enormous amounts of time. Other times the effort required to design and test a new AI-driven workflow ends up exceeding what we were doing before.
For us, the fundamental constraint remains the same: “good enough” isn’t good enough. Our clients rely on us not just to find information, but to validate it, interpret it, and explain why it matters. Those signals feed directly into engagement strategies, voting decisions, and investment research. That requires a level of accuracy and context that purely automated systems still struggle to deliver.
Over time we realized that the research system we built—document discovery, extraction, validation, and structured analysis—is valuable well beyond the narrow problems we initially focused on.
We started by helping clients track workforce disclosures and governance policies. But over the years the questions expanded: policies on hiring ex-felons, political spending, mentorship programs, religious accommodations, support for private prisons, and dozens of other niche governance or social issues that matter to different investors.
Increasingly, clients aren’t asking us for another dataset. They’re asking questions like: Which companies changed this policy last year? Who stopped disclosing it? What did they replace it with?
The underlying problem is always the same. The information exists, but it’s scattered across thousands of filings, reports, and websites, described in inconsistent language, and constantly evolving.
Answering those questions reliably requires more than just data. It requires the ability to search, interpret, and synthesize information across large collections of unstructured documents.
In other words, the value increasingly lies not in the dataset itself, but in the research system behind it.
This shift is happening at the same time the broader governance landscape is becoming more complicated. Public disclosures around workforce practices and governance commitments are changing quickly, and in many cases becoming less standardized. At the same time, some of the institutions and frameworks that historically provided consistent signals for investors—whether proxy advisors, disclosure standards, or benchmarking systems—are stepping back or narrowing their scope in the U.S.
For governance teams and investors, that creates a real gap. The need to understand what companies are actually doing hasn’t gone away, but the signals are becoming harder to find and interpret. That’s where we see the opportunity for PeopleReturn.
We still produce datasets, and we will continue to invest in expanding them. But increasingly we think of our role less as a traditional data vendor and more as a research and intelligence partner—helping clients ask better questions of non-financial disclosures and synthesize the answers quickly enough to inform real-world decisions.
In practice, that means building tools and systems that allow governance teams to explore large bodies of corporate filings, identify meaningful changes in policies, commitments, and board composition, and surface the signals that matter during proxy season.
The governance landscape is shifting quickly, and the tools investors rely on to understand companies are evolving along with it.
If you’re wrestling with questions buried in thousands of filings and disclosures, that’s exactly the type of problem we’ve been building our system to solve. That’s the conversation we’re hoping to have at CII next week.
What we’re reading:
- The State of Organizations 2026: Three tectonic forces that are reshaping organizations These are challenging times for organizations everywhere. Forces ranging from artificial intelligence, economic uncertainty, and geopolitical fragmentation to evolving workforce expectations, increasing customer demands, and tougher competitive dynamics are redefining how leaders create value and sustain performance. (McKinsey)
- Billionaire chipmaker CEO Lisa Su holds meetings on weekends and sends feedback after midnight because leaders aren’t born: ‘They’re trained’ As businesses around the world embrace a four-day week and “the right to disconnect”, Lisa Su, the CEO of chipmaker Advanced Micro Devices (AMD), goes against the grain and asks some staff to show up on Saturdays. (Fortune)
- Federal Focus on DEI Initiatives Expands in Corporate Hiring Practices and Regulated Industries Recent remarks from Trump Administration senior officials signal an expanding federal focus on diversity, equity, and inclusion (DEI) initiatives—particularly where they intersect with corporate hiring, promotion, and compensation decisions. (Ballard Spahr) and DEI Task Force Update On February 26, U.S. Equal Employment Opportunity Commission (“EEOC”) Chair Andrea Lucas issued a letter to chief executive officers, general counsel, and board chairs of the 500 largest employers in the United States to educate them about her views on the potential for race- and sex-based discrimination in corporate DEI programs. The letter asserts that the “bedrock American principle[]” of equality has been “under attack” by “movements and ideologies” that “demand equal outcomes over equal treatment” and “promote discrimination against certain races or groups.” (Gibson Dunn)
- The End of DEI In 2023, I (Alex Edmans) published The End of ESG. The title was not anti-ESG; instead, it highlighted how the practice of ESG had become niche and partisan—championed mainly by ESG specialists and opposed by those without ESG titles or on the political right—and involved ticking ESG boxes rather than creating long-term value. Later that year, Applying Economics – Not Gut Feel – To ESG showed how ESG issues can be analyzed with rigor rather than ideology. The trilogy concluded with Rational Sustainability, which proposed a new approach to ESG—one that fully harnesses its benefits while addressing its criticisms. In a new article, The End of DEI, I similarly propose a way forward for DEI that seeks to harness its benefits while addressing its challenges. (Oxford Law Blog)
- Reframing Board Diversity Disclosure in 2026 Proxy Statements Board diversity disclosure is undergoing a meaningful recalibration. After years of increasing pressure by shareholders and other stakeholders to increase the number of women and underrepresented minorities on boards and provide robust disclosure of board demographic information, the framework is now shifting. (Cleary Gottlieb)
- A New Lost Generation: Why Gen Z Is Unprepared for the Workplace The workplace can be a tricky place to navigate. Almost everything we do at work—identifying the experts, managing tough feedback from a boss, figuring out how to work in teams made up of different personalities—comes down to our ability to manage relationships. And to do so, we need savvy social skills. Most employees acquire those skills over time—by learning from their nonwork relationships, watching how colleagues behave in the office, and by seeing what happens when they stumble in their own workplace interactions. But the newest workplace generation—Gen Z—is unlike anything we’ve seen. (Wall Street Journal)
- Colgate to Defend DEI Criteria for Board Seats as Goldman, Amex Retreat Colgate-Palmolive Co. told a conservative shareholder group that it intends to defend using race, gender and sexual orientation as criteria for identifying future board members, even as Goldman Sachs Group, American Express Co. and others signal they will drop the practice. (Bloomberg Law)
- 2026 Global Human Capital Trends In our 2026 Global Human Capital Trends survey, 7 in 10 business leaders say their primary competitive strategy over the next three years is to be fast and nimble—to quickly adapt to and capitalize on changing business, customer or market needs. Leaders also report that the two most important drivers of success are accelerating how people and resources are orchestrated to perform work and increasing their organization’s and workforce’s ability to adapt to change and speed. (Deloitte)
- The Financial Implications of Having a Relationship With the Boss Most people know someone who’s dated their boss. In fact, office romances, especially those with power imbalances, are so common that they barely register as scandal—more than 25% of U.S. workers have engaged in a romantic relationship with a colleague, 18% of which were with a superior, according to SHRM, formerly the Society for Human Resource Management. But there has been little evidence on the financial impact of these relationships on the people involved. Now, new research offers just such evidence, and finds that the financial benefits and costs are substantial—especially for women. (Wall Street Journal)