Vivint’s deal with Cloudera offers a small insight into big data in the smart home


One of the tropes of the internet of things is that it will produce a lot of little data that needs aggregation and analysis. The rush is on to both collect that data (witness Google’s(s goog) $3.2 billion deal for Nest) as well as profit from selling gear to help companies parse that data (see Intel’s(s intc) investment into Cloudera).

I get dozens of pitches for what I think of as the back-end of the internet of things — the services and boxes that help companies track APIs, monitor up-time and yes, store and analyze data. But it’s rare that in those pitches I also get the larger truth behind why the internet of things and connected devices has everyone salivating.

Yet, in a release about how home security provider Vivint’s smart home services is using Cloudera’s Hadoop platform to aggregate data, Brandon Bunker, senior director, customer analytics and insights at Vivint puts it plainly:

My what a lot of data you are collecting, Vivint! The better to sell you services and create a competitive barrier to entry, my dear! Yes, there is a very real potential to eliminate waste — such as turning off the A/C when someone isn’t home — or to enhance security, but when you know what’s in someone’s home and the status of those things, it becomes much easier to sell them other things. The level of knowledge a company might have could also translate into influence — whether it’s benign like encouraging exercise or questionable like punishing you with higher insurance premiums if you eat poorly.

In Vivint’s case, its customers have an average of 20 to 30 sensors installed in their homes, but it couldn’t aggregate all that data to make larger inferences. The news release is about how Cloudera helps with that, but really it’s about the economics. Listen to Cloudera’s VP of Marketing Alan Saldich [emphasis added]:

“When we consider all the net new data generated by devices embedded with sensors and geo-location services that hasn’t been touched before, we open ourselves to new possibilities – to realize not just new revenue streams for business but to solve much bigger problems because we can dive into those points of convergence in our world.”

Frankly, if we were really after solving those bigger problems, we would have open data standards, a clear statement of consumer rights to their data and incentives to share select data widely among the consumers who generate that data and those with an interest in making changes. Yes, the insights derived from that data can and should be packaged up as services, but it can’t stop there.

For example, fine-grained data about home energy use could help change home design if researchers had access to enough data from enough homes to figure out the most optimal designs for an area. On the flip side, a customer’s video camera data or motion sensors might say more about her life than she really wants to share. Either way, right now, these connected products hold a lot of promise for adding convenience to our lives, but we’re woefully ignorant of the potential costs and why providers are leaping into the smart home. Let’s start having that conversation.

Update: This article’s headline was changed at 3:20pm PST to fix the spelling of Vivint.


Blair MacGregor

“Either way, right now, these connected products hold a lot of promise for adding convenience to our lives, but we’re woefully ignorant of the potential costs and why providers are leaping into the smart home. Let’s start having that conversation.”

I think for me anyway, the conversation begins and ends when data starts being used against me in a transactional manner. In other words, encouraging (or even discouraging) behavior through the data accumulated should be fine so long as I have the choice to act or not act on that data. If I start getting blasted with ads for treadmills because I cut out my walking routine for 2 weeks or if, as in the scenario re: the insurance companies, I’m actively having money taken out of my pocket based on my not walking/exercising, then I think it crosses the line.

We probably all have different thresholds for this kind of stuff. But two common points of agreement that Stacey alluded to that we all ought to be on board with: 1. the end user should have the final say on whether or not this data is used by companies in a non-anonymous way and 2. that a set of open standards governing the use of these devices has to be put in place. If you start from those two building blocks and move from there, I think you can start to advance the conversation in a meaningful way.


It’s spelled “Vivint” in the article, but “Vivent” in the title. The former is correct.

Comments are closed.