There are products that are homogenous because they are the same product. Off-brand paracetamol, for example.
And then there are products that are homogenous because it is really hard to reveal the differentiation behind them. Pepsi and Coke are like this (expecting angry letters from some, but it’s true!).
ESG data is also like this.
We all know ESG data providers don’t deliver the same quality data across different sustainability vectors.
But we also know how hard it is to identify which ESG data provider represents a higher-value proposition, especially given how many thoughtful and intelligent people populate the industry.
We know that the product is not homogenous because the metrics sometimes (frequently? always?) don’t agree.
For products of this nature, the primary purchasing driver is the two “Cs” – convenience and cost.
Convenience comes in many flavours – brand recognition, comprehensiveness of the offer, ease of access, personal relationships and, yes, sometimes the convenience of being able to hide a procurement decision behind a big brand provider.
Cost only has one garment, of course – unfortunately one that sometimes resembles a straitjacket.
Okay, 180 words into this op-ed and I haven’t yet mentioned AI, despite teasing that focus in the title. What’s going on? (Well, don’t rush the artist is what’s going on, I’m painting a picture here!)
At the moment, each investor solves that trade-off between the two “Cs” in their own way. But that equation has now changed dramatically.
AI – bingo! – is radically reducing the costs of data collection and provision. To the point that we are nearing the threshold where some (large) asset managers could in-house a significant part of the data development process.
Scraping and accessing corporate sustainability reports and third-party data platforms (like the Science Based Targets initiative) is becoming trivially easy. And cheap.
It is not just raw ESG data that is exposed to this dynamic.
Why buy one ESG data provider’s view, when you can build algorithms that compute 100 AI-generated market views and let them compete with each other for your attention? (Incidentally, a variant of this idea based on actual asset manager views is the premise of what French fintech ValueCo is doing.)
Want a new metric or a new lens on a new metric? No problem – build it and ship it.
Upstarts vs incumbents
Data-pricing tables will have to compete with this new cost structure. I fear cost pressure will be brutal for the industry, already scarred by previous budget cuts.
In a world where costs drive purchasing, upstarts stand to benefit. Lower overheads, more flexibility and less internal red tape will make competing on price easier.
But AI will also change the convenience equation. After all, the costs are not just the price you pay at the till.
Here, incumbents have the upper hand. One-stop shops, full integration and the lurking fear of AI hallucinations will drive a premium for brand credibility.
Public, respected brands will act as an insurance mechanism for investors fearing job risks from making the wrong procurement decision (or even building a tool in-house).
Traditional brands will also likely benefit from offering the easiest onboarding across the suite of metrics needed to run your sustainability, risk or ESG strategy.
If the transaction costs related to data acquisition are higher than the actual acquisition costs, the Little (start-up) Engine That Could simply isn’t the most attractive offer.
One of the first areas to suffer may be innovation. As budgets are squeezed and the returns on innovation are diminished in a world where building me-too products is becoming trivially easy, the incentive for investing in something new is deteriorating.
The harsh truth is it’s likely that many smaller shops simply won’t make it. M&A will also be harder. Why buy it when I can just rebuild it? Which in turn will make setting up upstarts harder.
When things are cheaper, convenience becomes a more significant determinant of procurement. In that world, big players may see smaller margins, but also more market concentration.
For the past decade, a healthy data ecosystem has driven innovation and scaling across both established and upstart players. Both are needed. But both may not survive.
That is Future #1.
Future #2
But there may be a different future. In this world, ESG data goes the road of financial data – 100 percent commoditised and delivered either at near-marginal costs or even for free as part of financial data packages.
Readers may note that some offers already suspiciously look like this.
At first glance, this will also reward incumbents – even if not their ESG data business itself.
At the same time, “full” commoditisation of “basic” ESG data will open up new procurement opportunities for more bespoke or even “ad-hoc” intelligence.
If big-brand data warehouses stop investing in innovation given the retreat to a low-margin, commoditised and automated data business, upstarts may actually face less competition.
Nobody has invested more in R&D over the past decade than the large data providers. A lot of that R&D budget may disappear (and indeed has already disappeared, some would argue).
In this world, it may be hard to build moats around established products, but easier for truly bespoke, creative and individual service providers to offer something different.
Especially when that data probably demonstrates investment decision-making relevance, which creative sustainable finance data typically has.
Data providers, or even individuals, offering such solutions may find fertile ground. If basic ESG data is cheap or free, in theory that should free up procurement budgets for more creative solutions (assuming the budgets don’t simply get slashed in response).
In this world, size may still win on volume, but lose dramatically both on margins and growth opportunities.
Both futures are plausible in my view. But Future #2 requires something which is currently missing: asset owners willing and interested to design mandates that give room for that kind of creativity, rather than more off-the-shelf solutions.
Without bold asset owners expecting the best, pushing asset managers to the frontier, there is no innovation.
“I think I can,” said the Little Engine That Could. So should data providers! And so should asset owners looking at their mandates. The future demands it.
