A Refresh for Sustainability (Despite the Cuts)
Diversity, Equity, and Inclusion (DEI) and Environment, Social and Governance (ESG) programs are being slashed worldwide. But what if this isn’t the end—what if it’s a reset? A refresh for sustainability? Companies still need to be responsible for how they spend their money. But when it comes to corporate social investments, sustainability initiatives, or ESG programs aimed at complementing government efforts, accountability remains a massive challenge.
Tax Benefits and CSR
Some companies have figured out how to extract marketing and tax benefits from corporate social responsibility (CSR). In South Africa, businesses can write off up to R10,000 in tax deductions by donating to Public Benefit Organizations (PBOs). So, instead of waiting for governments to ensure a thriving economy (aka your future customer base), companies can actively invest in their own long-term market sustainability. Revolutionary? Yes. Understood? Not quite.
Many businesses barely know what metrics to track for their own operations. How, then, can they be expected to set and measure indicators for external businesses or charities they know little about?
One of the main reasons ESG efforts fail is that key performance indicators (KPIs) are confusing or simply too difficult to establish. Plenty of companies don’t even measure their internal operational metrics, never mind their environmental or social impacts. But once they do figure out what to track, real growth often follows.
Imagine an organization seeking funding for sustainability initiatives. They struggle to define clear sustainability KPIs and end up relying on annual PR stories about one individual—a feel-good piece, a write-off, and a pat on the back.
The illusion of control is more persuasive than the reality of uncertainty. So we cling to stories about outcomes being in our control
- Morgan Housel, The Psychology of Money.
Financial reporting is heavily regulated; ESG reporting, however, is often loosely structured and inconsistent. There are hundreds of audits, tax authorities, and oversight bodies that track every cent of a company’s finances. Yet ESG investments, which can be just as complex, often go under-examined.
When we used AI-driven data tracking to correlate a client’s spending, donations, and job creation, we noticed a massive spike in one dataset. At first, we assumed it was just a COVID-related distortion—nothing unusual. But on closer inspection, we saw an NGO jump from R600,000 in monthly income to R66 million in a single month. The cause?
Two extra zeros accidentally added to a report.
Without AI, this mistake would have gone unnoticed, skewing averages, warping grant allocations, and fabricating job creation statistics.
Common Data Pitfalls and Organizational Blind Spots
It’s no surprise that such errors slip through. People wrestle with sprawling spreadsheets containing 40,000 or more rows, untrained staff, and evolving data structures—while each department tries to dodge responsibility. These mistakes aren’t unique to one scenario; they can pop up anywhere, even in your own company, if you spend enough time analyzing a single indicator.
For many businesses, that main indicator is profit: income minus expenses. Yet any finance expert knows the complexities in declaring “profit” over a given period.
So, what happens when the data is just “data”—squares filled with information we never fully interpret? Often, we end up with cluttered cells and mountains of numbers we never use. As the (often-cited) statistic goes, 99% of the world’s data has never been analyzed.
Try starting with just one indicator—one chart—covering the entire history of your company. See what happens. Some clients claim they have five key indicators that all need constant tracking, but that quickly becomes extraordinarily complex. Unless you have teams dedicated to each metric, it’s near impossible to keep tabs on them all—especially when each business, project, or investment has its own environment and set of challenges.
The Inverse ESG Relationship and Poor Reporting
Returning to the aquaponic fishing initiative, we saw grant funding go up while reported production plummeted from three million tons to one million, and then half a million, over three years. No one noticed because the annual report was a jumble of jargon—no clear indicators, no accountability, and no one asking critical questions.
AI, Knowledge Gaps, and the Reality of Unknown Unknowns
Before AI went mainstream, an investment-firm director asked how you can possibly know the efficiency metrics of a shoe factory in Bangladesh. The short answer was: You can’t, and they didn’t even want to. So what was the point? Now, with AI’s emergence, there’s at least a potential for automated, intelligent data scanning. AI might already have those answers—if only we connected it to the right processes and data feeds.
But implementing AI is complex. It involves tokens, workflows, and specialized coding terms—things that create confusion and introduce a new technology layer. Everyone is trying to figure out how to best use it. The promise is big: “Give me your data, and in three months, you’ll have real-time insights.” Still, many remain unconvinced or overwhelmed.
When the “Best Attempt” Is Barely Understandable
Most companies produce some kind of internal or external report loaded with 50+ questions covering everything from water conservation to gender equity. They do their best to cram in impressive numbers, but ultimately, it’s a chaotic mess. Who verifies it? Who validates it? Where are the experts in water-saving, fish farming, capacity building, or job creation? Is anyone collecting raw data, or are we relying on vague statements?
Then along comes an AI interface, which can parse the textual jumble, extract potential indicators, and reorganize them into a coherent timeline—maybe three years’ worth of data. Suddenly, we have a hint of structure, clarity, and a baseline. But that’s only the first step. We still need experts to interpret what’s actually happening—whether it’s an aquaponic fishing project, an education campaign, or any other initiative.
The Core Purpose vs. the Endless Extras
Too often, businesses lose sight of their core purpose—for instance, producing fish in an aquaponic setup. Meanwhile, they get sidetracked by tangential goals, like community workshops or training programs, without first ensuring they can consistently maintain or increase fish production. If production tanks from four tons to 1.5 tons, that’s a glaring red flag.
Yet the data might be buried under a mountain of sustainability agendas. Yes, feeding the world is profitable, and AI can help—but only if you ask the right questions. Coca-Cola, for example, rarely invests time teaching the world about refrigeration; it focuses on selling cold Cokes. Any philanthropic efforts they highlight often double as marketing to boost brand image—sometimes derided as “greenwashing.”
Client Data goes in.
Our core database structure
Indicators are generated and Year on Year data setup
We send AI prompts to return Insights.
And we offer a single chart on your core indicator
Understanding Real Impact
At the end of the day, the best data strategy helps you focus on your core purpose—whether that’s fish farming, building shoes, or selling soda—and measure the metrics that truly matter. That might be tons of fish produced or cases of product sold—not 50 different metrics or superficial PR stories.
If you do want to tackle broader social or environmental objectives, make sure you can measure them effectively. AI can help correlate the data, shine a light on unknown unknowns, and maybe even reveal major opportunities. But it can’t fix poorly defined goals or a total absence of meaningful data collection processes.
In short, success lies in:
1. Knowing your purpose.
2. Tracking a small handful of truly vital indicators.
3. Tapping into AI or other advanced tools to find outliers, errors, or trends early.
Once you’re profitable and stable, you can reinvest in social and environmental causes. Just don’t forget: if your core production has dropped from four tons to one and a half, you might need to pause and fix that before anything else.