The Marginal Value of Your Data
Famously, Gregory Barber sold his personal data for a whooping 0.3 US cents. Today we look into the marginal value of data, and how we as consumers can capture the value our data creates.
The global data economy is worth something like 3 trillion USD (yes, trillion with a T), and we as consumers, see very little to none of it directly. With or without or knowledge, every bit of data about us is bought and sold on the open market – from our bank transactions to our cell-phone numbers; if someone has access to it, they’re selling it.
Gregory Barber took the blockchain and self-sovereign data revolution into his own hands, and decided to sell his personal data directly to data-hungry corporations. By doing this, he can control exactly uses his personal data, and profit from the sale. From this experiment he made a whooping 0.3 US cents.
You can read his original account of the experience (which sounds exhausting, by the way) over here, but today we want to dive into why there’s such a huge discrepancy between the profits Gregory Barber realised, and what we might expect the value to be, based on the size of the world wide data economy.
Why Does Data Have Value?
As a quick aside, the value for data comes from corporates creating competitive advantage against one another. Figuring out what people want, their habits, and how they choose to spend money means they’re more readily able to target advertising, develop products, or create pricing schemes to meet the needs of their consumers (and therefore, make more profit.)
Marginal Value vs Average Value
Here’s something to understand – we’re not yet at a stage where we can realise truly personalised advertising, at least not profitably. For now, that belongs to the futuristic worlds (and debatably dystopian and/or Orwellian) of Bladerunner.
The sorts of advertising we do see are formed by creating a profile of people with similar characteristics to us – things like age, geography, and maybe interests or spending behaviours, and then learning about that market segment at a whole. We receive advertising based on these data, not exactly personalised data as such.
Knowing that, we can suggest that having a vast database of consumer information to mine for trends and habits is hugely valuable. Tens or hundreds of thousands of data points about a huge number of potential customers can tell as a lot. Compared to that, how much value is there in one more data point?
In fact, if we assume that the value of a dataset is proportional to its predictive power, then it is going to be proportional to the square root of the number of data points – which is the same thing we learn from statistical power analysis. Every additional data point adds less predictive power than the one before. Double the number of data points only increases the predictive power by around 40%. Tripling the number of data points only increases the predictive power by around 70%.
As an individual selling data, without a broker selling thousands of data sets at a time, that tiny increase in power/predictability/profit is all we can capture.
How Can We Still Realise Value?
If this sounds dismal, never fear, there are still ways we as consumers might capture value for our data.
The first that comes to mind is selling our data to organisations that care deeply about our personal data. Think insurance companies. Insurance payouts are a heavily right-tailed distribution, so being able to establish which users are specifically more at risk is going to be a huge competitive advantage. Although health insurance is the most relevant to us here at Juri, this is also applicable to life insurance, car insurance, and even travel insurance or professional indemnity insurance! (Read about how we at Juri approach this in our post, Staking on Your Health.)
Secondly is cartel-like behaviour. In this case, we’d knowingly agree to have some representative sell our data on our behalf, along with thousands of other people’s, and then receive a share of the profit from selling this large pool of data. This is much the same as the current model, except in this case we’d opt-in, and the agreement would be reciprocal, instead of being snuck into the fine print of a credit card agreement or terms and conditions document we never read.
Brave web browser is doing something similar to the above right now, where users can opt-in to receive advertising, and in return, receive a share of the advertising revenue (I haven’t personally used Brave yet.)
The increase in consumer awareness of their data, and the self-sovereign data movement means we’re closer than ever to being able to capture the value from our own data. One day, we’ll probably be able to profit directly off the data we create every day, but there’s still some work to be done to get to that stage.