Why We Need Data Unions to Support the Data Economy
Shiv Malik, CEO of Pool Foundation, on Data Unions and the value they bring to the Data Economy
In the twelfth episode of Voices of the Data Economy, we had a conversation with Shiv Malik, CEO at Pool Foundation and former Head of Growth for Streamr where he evangelized about a new decentralized data economy. During this discussion, he tells us how Data Unions work, its real-life use cases, and legislative regulations supporting the Data Union model. Here are edited excerpts from the podcast.
Note: The episode was recorded on April 9, 2021 — when Shiv was working with Streamr.
What are Data Unions, and how do they work?
As individuals, we all produce information in the form of data, and only a few people are harvesting and profiting from it. There is no way for ordinary people who are producing the value to retain any of that value themselves. Hence, a Data Union is a way of pooling data between various parties and monetizing it as a collective.
However, you still need an organization in the middle because this model involves reaching out to data buyers. You need to pick up a phone and speak to people — and machines can’t do that yet. Lots of people have fleshed this out academically but no one’s really done it practically. When you try and do it in practice, there are some really big stumbling blocks.
A Data Union framework provides a way to bundle a user’s real-time data together with others’ and distribute a share of the revenue when someone pays to access it. On its own, a person’s data does not hold much value, but when combined in a data union, it can aggregate into an attractive product for buyers to extract insights. This is crowdselling, and has the potential to generate unique data sets by incentivizing individuals to trade previously unavailable data.
Data Unions as a revenue source to IoTs
Our smartphones are powerful devices that are collecting information through internal sensors. All you have to do is create an app that collects a specific data set — and then you can imagine there’s a monetary incentive for people to do this.
Why do we only have one map application (Google) or maybe two — and they’re owned by the same company? The reason for this is because they have a monopoly on our location and we create that monopoly by sharing our data freely — or rather we don’t have a choice. “I would like other options — but viable ones. If you take the raw data sets away from them (devices), anyone can build applications on top of the raw datasets. That’s the world we want to live in,” says Shiv.
The problems with data sets and structures
Expanding on the inherent issues with the data collected from mobile devices via sensors, Shiv comments, “From the data buyers’ perspective, this is terrible data too. Why do you want to collect data under the table spying on your users that you’ve buried the consent on page 70 of a contract that no one’s read? You get data just as a byproduct now of our digital lives when actually you really wanted to be the product. That is the problem Data Unions solve — if you want to create good data products, get people to create the data who actually want to be part of this and are happy to do it. It creates better data products — more rich, stable, interesting, and sustainable.”
Another problem is that data brokers go bust all the time. Cambridge Analytica is the most famous. Recently, there was Jumpshot, a subsidiary of the antivirus giant Avast, which went under last year, Vice reports. The company had $30M USD in revenue, projected to reach $70M USD with 400 employees. Vice mentioned, “An antivirus program used by hundreds of millions of people around the world is selling highly sensitive web browsing data to many of the world’s biggest companies, a joint investigation by Motherboard and PCMag has found. Our report relies on leaked user data, contracts, and other company documents that show the sale of this data is both highly sensitive and is in many cases supposed to remain confidential between the company selling the data and the clients purchasing it.”
Which parts of the world are more receptive to Data Unions?
In theory, every government should be pro Data Unions because they provide two benefits. One, it’s clearly giving a monetary value to ordinary people and returning it to ordinary people — as opposed to companies that are usually outside of their own country. If you’re in any way a democratic or populist government — that sounds nice. Second, it also opens up innovation, and it’s actually good for business.
You can also see rights to data portability already starting to take off in Europe. According to a report by Deloitte:
- 76% of respondents are aware of the right to data portability.
- 9% revealed to have already submitted portability requests.
- 23% haven’t heard of the right.
- 24% stated that they have no intention to use it.
Interestingly, 81% of respondents from EU countries are aware of the right to data portability.
- The Netherlands and France lead the way in terms of awareness, both at 83%.
Only 68% of respondents in non‑EU countries were showing awareness of this particular right.
- Australia and Canada recorded the lowest awareness; 60% and 62% respectively. This could be attributed to it being the only new data right that consumers have, and therefore less practiced before now.
“Where Europe leads, I think other countries will follow, especially with respect to GDPR. We are getting good noises from the US, and I believe the UK will try and catch up because of political reasons. India, too, has a natural resonance to this kind of model because of grassroots cooperativism. It’s built into the country’s economic history in the last 100 years and it’s also a democratic country. I don’t know where China will go with them.”
Can Data Unions be acquired? Can we choose to whom we sell our data?
At the moment, we don’t have a technological way of signaling the member preferences of a Data Union. But it’s not that difficult to build that in, according to Shiv. In an ideal situation, people could say at the beginning: these are the kinds of organizations I’m interested in selling to. You could also trust a data cooperative that you believe in to make these decisions for you. A recent EU scheme is based on this fiduciary model. It says if you are a data cooperative, you have to have a legal duty of care to your members. In November last year, the European Commission released the draft of the Data Governance Act which sets out a licensing and regulatory framework for data unions — or data intermediaries as they call them — which will give this nascent sector a huge boost in terms of funding, trust, stability, and assurance and advertising a direction of travel to the world. They have also announced that €2bn in funding will be made available for those seeking out to build enabling software and Data Union projects.
We are not there yet, but at some point, you will say “I only want to sell data to charities and university researchers.” This will put you in a different bucket of data, and only those buckets will be able to be bought by certain types of buyers.
Shiv says that if one day Google decides to buy Data Unions, that is a problem. “Obviously, that haunts a lot of people but certainly haunted me. I would love to see that you can’t force Data Union operators who come along as entrepreneurs to be cooperatives. But I would like a federation of cooperatives. If that doesn’t work, then with the fiduciary aspect of that legal duty of care — it will probably make it really difficult for Google to purchase data sets.”
He adds, “The token model also helps because the whole point of tokens really is that you can disassociate equity and ownership of an underlying capital asset from utility and value. And if you can do that, you can turn revenues to people and then not have to sell out your equity to anyone that also helps with the cooperative stuff. But it is difficult. All of this stuff is stuff that we need to think about. And I know why that keeps me up at night.”
Here is a list of the selected time stamps on the different topics discussed during the podcast:
2:10–5:25: Shiv’s journey from investigative data journalism to Data Privacy and Data Unions evangelist
5:25–10:59: What are Data Unions and how do they work? Examples of Data Unions
10:59–14:20: How can Data Unions be monetized with different use cases?
14:20–22:15: Data Unions as a revenue source to IoTs
22:15–25:20: Which parts of the world are receptive to Data Unions and how have legislative regulations helped?
25:20–29:30: Can Data Unions be acquired? Can we choose to whom we sell our data?
29:30 — End: Why Blockchain and data need to go hand-in-hand: The role of Data DAOs in Data Unions
