The Future Is Here: Passive Monitoring
I remember as a kid being very excited about the future. One concept that was extremely enticing was the coming age of robots. I was assured that they were only around the corner by my favorite cartoon at the time, The Jetsons, and their robot Rosie. I knew in the future I would want my very own robot to take care of the laundry, vacuum the house, and cook all my meals.
Fast forward to today and unfortunately the closest thing we have to commercialized personal robots is the iRobot Roomba that can vacuum your carpets. That's sadly about it.
Yet the underlying beauty of having a personal robot, something that was always taking care of things for you while you were busy with the rest of your life, has found it's way into many interesting software solutions.
The concept has materialized into passive monitoring solutions that are constantly monitoring some set of activities that are relevant to you and providing you with insight in the form of summarized and aggregated views, relevant alerts, and more. While large scale passive monitoring solutions have existed in B2B industries in a variety of verticals, only recently have these solutions started to become commonplace in consumer internet. Much of this has to do with a set of trends that are coming together. For one, all of our communications, transactions, and activities continue to be digitized and move online, making it easier for software to monitor these activities. In addition, open APIs are starting to pervade many verticals, making the task of extracting the interesting data from across disjoint data sets even easier. Finally, cloud computing has also played a significant role is decreasing the cost of constantly monitoring a set of activities with the flexibility to scale up and down resources with your user needs.
Four such passive monitoring solutions have found their way into my own day-to-day life.
Mint does an amazing job of aggregating all of my financial accounts, across credit cards, saving and checking accounts, and investment and retirement accounts, to provide an aggregated summary of my spending and overall net worth. While I can log in to the dashboard whenever I like, I love that Mint sends me a weekly summary of my financial data. Mint not only provides a great overview of your financial assets, but also provides deep drill down capability into specific transactions categories or transactions themselves. I've also found it extremely helpful that Mint keeps my transaction history forever, versus many of the credit card and bank online interfaces limit my view into my own transactions to sometimes as little as 90 days.
The key enabling technology behind Mint is a company called Yodlee which developed the interfaces to connect to your bank transactions. Mint simply licensed their service and created a compelling experience on top of it.
RescueTime provides a view into how I spend my time each week in front of my computer, breaking down all of my activities into specific application and web site usage as well as summarized category usage. Categories include communication, social networking, news, software development, and more.
By looking at my weekly RescueTime summaries, I can get a great sense of how productive I was that week as well as an understanding of how much time was wasted on non-productive activities. It's a great way for myself to help realize where all my time is going and then make course corrections based on them. RescueTime is able to get this detailed information about my computer activities through a simple monitoring tool they install on your desktop to collect your data.
Tripit has become an invaluable aspect of my travel experience. Everytime I get an itinerary confirmation from an airline, I simply email it to firstname.lastname@example.org and Tripit lets me rest assured that the rest of my travel experience will be extremely simplified. Once Tripit receives your e-mail, it parses it to understand exactly what flights you are taking as well as the time for that flight. It then creates a calendar that you can subscribe to that shows you exact departure and arrival times. I used to do this manually, but now it's all automated. In addition, their iPhone app provides me access to my itinerary on the way to the airport. I don't have to hunt for that e-mail with the confirmation in it. For premium subscribers, it even provides text message alerts on delays on your flight through regular monitoring of your flight status.
Feedera provides a daily summary of the latest buzz amongst my friends on Twitter. It makes it much easier for me to keep up with Twitter without a significant time investment in constantly watching my feed.
Feedera provides this summary by constantly monitoring your feed, scoring all the content that comes across it by popularity and friend relevance, and delivering a daily digest of the highest ranked content. Feedera basically reads my feed to find the good stuff so I don't have to. It's significantly decreased the time I spend on Twitter and that is a great thing.
While these solutions may seem like no-brainers, there are a set of key criteria that these types of applications need to satisfy in order to ensure successful adoption.
Maximize data accuracy.
When a user decides to adopt one of these solutions, they are placing their trust in the application to monitor the given activity for them so that they don't have to. This is a high trust level required for a piece of software. The issue with this is that trust is quickly rescinded if the application fails to uphold it's end of the bargain. And the most common breach of trust is the result of data accuracy issues.
A friend of mine told me that the reason he abandoned Mint in the early days was that their categorization of transactions was always off. This forced him to go through and look at every transaction to ensure it was categorized correctly. He therefore saw little value in the automation and ultimately abandoned the solution.
In addition to a user expecting to trust the output of such an application, a user needs to place even more trust on the system ensuring the security of their credentials used to get access to the data as well as the security of the data itself.
One of the things Mint had to deal with in the early days was getting users to put their financial account credentials into their service. They spent a lot of effort building this trust through education, security reviews, compliance tests, etc. Similarly, RescueTime collected potentially personal data about your application usage and web site history. User's need to trust that this data will be kept private. It's thus important for applications to invest heavily in maintaining and building this trust. Blippy, a startup that displays your credit card transactions to your friends, unfortunately suffered from a serious breach of trust recently that will be sure to effect at least their short term growth rate.
Regularly re-engage your user.
Given that these solutions are constantly monitoring a set of activities, oftentimes there isn't a natural point at which the user should re-engage with the application. This may result in high expense for the application to monitor the customer's data, but little customer utility.
Therefore it's important for the solution to provide meaningful re-engagement points with their user. The most effective ones I have seen are daily or weekly e-mail summaries as well as real-time notifications. Feedera sends out a daily summary, while Mint and RescueTime send out weekly summaries. Tripit, on the other hand, sends out real-time notifications before and during your travel plans.
Overall I'm very impressed with how far passive monitoring solutions have come and expect to see many more pervade our daily lives. While it's not the robots that I dreamed of as a kid, it's definitely an analogous trend towards personal assistants in the software world.
Enjoyed this essay?
Get my weekly essays on product management & entrepreneurship delivered to your inbox.
May 17, 2010