Currently showing: Other

04 Aug 15 12:55

We are seeing an explosion of data that people are creating about every aspect of themselves. Health data, behavior data, purchase behavior data, etc. Not only are people creating this data but also collecting it for themselves. This, essentially, is the concept of "Quantified Self". The centralization and digitization of your data and behavior. It is an undeniable trend and I think an interesting one. Yet, so many people are so scared of putting this data to full use (which partially includes me) for their own benefit and convenience.

I find there are some totally divergent scenarios around personal data that we might see in the future. These are mainly emerging to address concerns around privacy, security and accuracy. One extreme that we could see happen is where there is no more "public" or "corporately-held" data on individuals. A main driver of this would be the developments in the area of personal data vaults. What is this personal data vault? As you create this personal data you won't actually save the data on some other company's cloud server or system anymore but rather your own personal cloud server. This could even be an actual physical PC standing in your house that collects all your data. In short, you own and host the data. This gives users a new level of control over their data and if they wanted that data or info to be removed then they could do it without needing to apply for it to be removed or provide any big explanation.

Now that you have control over your personal data, you get to choose what services and analytics tools you would want to have "analyze" your personal data for your own insights and benefits. You also get to choose the level of protection and encryption you want to have around your data. So far so good. Who better to be in charge of your personal data than yourself.

Where it gets really interesting and somewhat "out there" is around what happens next. We now have people who collect and encrypt their own data and then go one step further and host it themselves on a platform that is out of reach for companies and corporates. What stops individuals now charging companies to get access to their data? People can now make some money off their own data rather than other companies making money off their data. People could allow companies "access" to certain fields but not an actual copy of your data.

Imagine signing up to a newsletter. Instead of giving a company your email address you allow them access to your email address field on in your data vault. As they send out the newsletter email, your vault assess whether you have actually provided them access to the field and then fills in the email address and it is on the way to your inbox. The company would have no actual record of your actual email address. When you want to unsubscribe you just revoke access or permission for that provider to that field. They still will not have a copy of your email address. Talk about control over your data, right?

It does create a scary scenario for companies of the future if this trend would develop. What I am not sure about is whether the commercial detriment from this development will outweigh the data protection and control advantages and thus not happen ... OR ... are there actually such massive possibilities here in getting people onboard in producing data that it will be worth it and supported by regulators and adopted by companies.

What do you think? Craziness or exciting concepts?

Category: Other

Location: Zurich, Switzerland


Bill Tyson - 4 Aug 2015, 9:44 p.m.

Oliver - a really relevant, exciting topic. As for "where it gets really interesting - when consumers take control over their data and then sell it to companies" is a concept that has been around since John Hagel, III's book The Coming Battle for Consumer Information was published back in 1997. A few start-up companies have tried, albeit unsuccessfully, to develop this "infomediary" model. Back around the .com era I developed a facility for this purpose but it never got the necessary funding so we all moved on. I agree that this concept does represent a huge market opportunity and one that will materialize as the quantity and proportionate value of consumer data increases. After all, why shouldn't the consumer make a buck on their data (including their medical data and history)? The key is to have a vehicle that empowers the consumer to wrestle their data away from the big data companies - not an easy task.

Oliver Werneyer - 5 Aug 2015, 9:32 a.m.

Wow, back in 1997 already. I am not really surprised but I think as always it is about timing. I feel the concept of "quantified self", which I think is the key driver here, was not a concept (or as comprehensive as now) back in 1997. So maybe the need or perceived value not as big as now. Why not dust off what you built back then and see if it still applies? :)

I think that one of the tricks will also be to build something that is easily accessible and integrated. It needs to be built to be technology and platform agnostic and that will in itself not be that easy.

As for the "wrestle the data away form big data companies", I have always wondered if that is so important. If you look at data developments, it's all about live or current data. If these companies stop getting new information then after 9 months companies will not see too much value in this "old" data. Or am I too harsh on the time value of data?

Christian Schmitz - 5 Aug 2015, 7:30 p.m.

Hi Oliver,

I discovered your posting through linked-in. You write that "I think that one of the tricks will also be to build something that is easily accessible and integrated. It needs to be built to be technology and platform agnostic and that will in itself not be that easy."
We set up a task force this year, including members such as CERN or ETH Zürich, to adress exactly this problem in an open source based standard for interoperability between private clouds, called OpenCloudMesh.
It aims at allowing data pods, pods, however you call them, to exchange data if the sovereign of the pods allows it, on any scale. So this works as exchange between individual private clouds, say on the home nas between family members, or even between large global enterprises, say in a large project in a supply chain.
We would be excited to have your vector on the currently existing drafts, I would appreciate your feedback. I have pinged you on Linked-in.



Dan Teague - 6 Aug 2015, 9:09 a.m.

I understand the theory - but I question whether this will ever take off at scale beyond maybe a minimal subset of data.

Unless legally compelled, consumers of data will always want to build their own vaults (i.e. the status quo) with an increasing number of data points because of the value and flexibility that exists in them.

Also, to be of use your data has has to leave your personal vault in some way, at some point - and then it can get stored by someone else regardless of local legislation (just ask the NSA!).

Oliver Werneyer - 6 Aug 2015, 12:36 p.m.

Hi Christian

That sounds very interesting what you guys are working on. I will definitely read up on that. Thanks for sharing that info.

Sounds fascinating. I would love to cast my eye over that piece of work. I think if we find a way of working on that and finding a simple viable way to develop an environment of more control for users then we will see an explosion in production of data as well as peoples' willingness to make use of it at a grand scale.

Oliver Werneyer - 6 Aug 2015, 12:44 p.m.

Hello Dan

The thing is your skepticism is shared by many and I find the opinions quite divided on this trend in data centralization. One of the reasons I wrote this. I want to hear why people feel this will or will not take off.

I would have to agree with you that consumers of data will want to build their own vaults and monetize that. But I think that view is leading to serious limitations being placed on all levels of the data. I believe that if people would feel in control of their data and are the primary beneficiary of commercial value from their data then you will see much more people participating in the data economy. We see many people not participating yet (most people). Once you have voluntary participation I also believe you will see much more detailed and valuable data being produced. Because the person who produces the data also collects it, you would have ALL their data in one place and will be able to make much better suggestions/predictions/offers/customization/etc. Currently, and company would only ever have a small subset of a user's data. Even the Googles and Amazons of the world only have access to a small subset of data.

I also think that from a regulatory perspective there is some value to be had. Currently regulation is hyper complex to cover all the various scenarios in terms of responsibility and fall-out as organisations are so diverse and their intent for the use of data even more so. If people were responsible for their own data, it would infinitely simplify regulation and the regulator can focus on organisations' use of data (ethically, but only once given express permission by the customer owner) rather than all the other things.

Also, from a cost perspective it could be cost beneficial if organisations no longer had to maintain and protect big data repositories.

I believe companies might, in the scenario described above, might no longer "own" or "host" the data but would have access to more data than ever before.

Am I dreaming?

Melissa Leitner - 19 Aug 2015, 8:34 a.m.

This is an interesting and very relevant topic. I see another consequence of an increased usage of personal data vaults: it will force data companies to be more conscious of the value of the data they want to use. At the moment, there's a whole lot of "free" data we're generating which can be used to do all sorts of exploratory analyses by whichever companies are collecting it... But if companies have to pay individuals to use there data, they suddenly have to know BEFORE doing these analyses, what value they will get from it, so that they can calculate a price they're willing to pay the individual to access his/her data vault (or a selection of fields from it).
This consequence has both pros and cons...
E.g. A pro - stronger commercial cases for big data projects will have to be built earlier in the process, and hence more stable return on investment in this space.
A con - it could slow down R&D and innovation on big data applications.

Oliver Werneyer - 21 Aug 2015, 2:23 p.m.

That is a cool way of looking at it. once something starts costing money then people will be more selective about what they want to use. It could be also good for people in these jobs to not get a scenario where they get a whole bunch of data and get a request like "See what you can find.".

I am not sure it will slow down R&D too much because more detailed data might be available than before. It might cost more but the insights could be much better.

If you would like to leave a comment, please, log in.