Sumit Sharma

Archive for December 2009

Smart Grids and The Web…

leave a comment »

Smart grids will enable for more information about our energy usage patterns. This will unlock more powerful information about our habits and life at home. The new thing here is that data and information will be available about what happens within the 4 walls of our domestic existence – something that has never been available before: inferences can possibly be made that would enable breakthroughs in sales, marketing and customer targeting initiatives. The more information we have about users, the more we can sell. The smart grid will be a huge enabler of this.

Smart Grids providers that learn more about us will have the ability to find the optimal lifestyle of consumption based on our characteristics – this will lead to smart grid operators recommending products and services to us. Imagine an alliance between Amazon.com and the leading SmartGrid operator to enable a smooth delivery of recommended products. The possibilities opened up by the emergence of SmartGrids are actually endless.

Written by Sumit

December 23, 2009 at 6:16 am

Posted in Smart Grids

Facebook needs to get relevance right

leave a comment »

A classic pain point on Facebook is whenever one comments on a friend’s wall post. In doing so we will receive notifications of all other comment and comments related to that post, regardless of their relevance to us or our comment – wouldn’t we be better served if we only received notification of posts relevant to our comments or perhaps posts from friends only? Such an example can be extended to other parralel realms such as in e-commerce when looking for product feedback for example we might only want to hearf feedback from those that have similar characterisitcs as us and so on. Not being able to decipher motive/intention and hence obstructing a valuable experience in such cases will lead to frustration eventually however with more information about us, Facebook can be more intelligent in its delivery of data to us.

Written by Sumit

December 19, 2009 at 6:43 am

Posted in Social Networking

The internet can enable disruptive psychology

leave a comment »

Is it possible that the Internet  will enable a disruption to the study of human psychology? I think it could. Imagine that as we put more of our lives online, and data and information about our habits, behaviors, cultural and social activities gets recorded it could tell us a little more about ourselves. Perhaps, just perhaps it might even tell a story about ourselves that we didn’t even know.

Who are we? What do we like? Have you ever wondered why that really shy person wears such a loud bright yellow jacket, is there anything beyond just, “it’s fashionable” and moreover, how does shyness and bright yellow go together? Why does a mid life crisis occur? Can the Internet avoid a mid life crisis? Is it conceivable that mid life crises occur because of unmet needs being fulfilled? Hence the urge to quit your job, start a new hobby, buy that convertible, travel the world or buy a motorbike? If so, then how can we avoid this? By knowing ourselves.

There are a few layers to what the “self” really is and keeping it a mundane physical level here is how I believe we can make an impact: It is possible that our behaviors, and what we define our personalities to be, aren’t heavily correlated? Take the example of the shy geeky closet gamer sitting in his room on Friday night gaming away – perhaps that’s a very charismatic person waiting to be tapped, or stimulated – perhaps his/her upbringing wasn’t conducive to his/her personality development or perhaps he/she has had some drastic events/situations to deal with in life that has molded them into who they think they are, but really are not. If you believe this to be a plausible argument then perhaps they web of data can enable users to know more about themselves, because of the things they’ve been watching, places they go to, websites they visit or meals they order. The list can go on and on and this data in turn, if sliced and diced appropriately can open up new possibilities in the forms of book recommendations, social/professional networks, products, jobs or places to travel to etc.

Can Facebook be your psychologist? Why not!

Written by Sumit

December 15, 2009 at 6:47 am

My warning to the media industry

leave a comment »

DemandMedia is trying to read our minds, in addition to making money…here should be their next steps: To stop using search engine and advertising inputs as their sole source of information. Get information from social networks- wall feeds, posts and comments, from physical devices and from other user/discussion boards. Collaborate with discussion forums, and other resource groups to figure out ways to read our minds in a much richer way-get into a more direct touch with people. Pandora knows the age, location and music preferences, at a minimum, of all its listeners and based on this, it can provide specific localized advertising real time. This was already evidenced by its initiative where it can target men in their 20s listening to Punk music to 5 different Whole Foods locations in the Bay Area with location specific ads within a 5mile radius of each store. Tivo should be able to recommend programs. TV is here to stay for a vast majority of the population, so to catch up to Youtube and others, how can they retain users – Through understanding users and figuring out which programs they can watch.

This will entail a disruption to scheduled programming – we will migrate from seeing, not what’s available, but what can be watched – or form a push model to more of a pull model. It will be on demand programming all the time. This will disrupt today’s model of only being able to watch what’s available and evolve to watch whatever you want. Instead of us controlling our Tivo’s to record what’s interesting to us, the networks will provide us what is interesting to us based on our characteristics. It will be a convergence of Tivo like capabilities and network programming. Networks will become aggregators, not just providers – this will vastly disrupt Hollywood, and the media and entertainment industries. Instead of forcing us what to choose from (basically we’re given a choice of finite options to consume from Movie Studios, Record Labels and Print Media), we’re beginning to see remnants of change occurring: in the Music Industry, Pandora is providing music based an algorithm that matches the genres of each music title to our characteristics. Spotify is a proprietary peer to peer music streaming service through which one can obtain music titles from the peer to peer network and can specify preferences to narrow down the list of choices. Zoe Keating, a former techie turned independent musician and twitter celebrity, chooses not to sign on to record labels because she doesn’t want to make compromises to her personal style. These are all signs of individualism arising at least in relation to media, and especially with regards to providing and consuming entertainment….this trend will be headlining our relationship with the various media channels in the future and they had better be ready for us by knowing what we want and providing greater relevance to our needs and desires.Media is changing

Written by Sumit

December 12, 2009 at 6:55 am

Making shopping easier for us…

leave a comment »

As we continue to use the internet for purchasing all types of goods, the amount of data being generated in relation to our browsing for goods and services is a huge shift for retailers in getting to know the customer. The first step has been accomplished: Obtaining the data. The next step is to be able to make sense of this data to their advantage.

Right now, Amazon.com has all this data on what I have navigated in the past 5 years, regardless of whether I followed through with a purchase or not. How useful is this? Basically, Amazon.com only knows I browsed the page for Winnie the Pooh five years ago and that’s it – why I did that – Amazon has no clue whatsoever. Here is the point: Amazon.com may make sense of that search I made 5yrs ago based on plain numbers in which case they’d label me as a Winnie the Pooh fan, or perhaps a Disney fan. Now fast forward to today, I haven’t searched anything close to Winnie the Pooh merchandise in the last 4 years on Amazon.com in which case they might strike me off as a Winnie the Pooh fan – that makes sense if I was only existing in the virtual world. How do they know I haven’t been frequenting the Disney.com website every month for Winnie the pooh stuff? They don’t, and perhaps never will until and unless there is an interface from Disney.com to Amazon.com.

Also, Amazon.com hasn’t got a clue about my physical activities and perhaps I have been visiting Toys R Us for certain specific Winnie the Pooh merchandise every single week because I’m actually a Winnie the Pooh die hard junkie. Perhaps, I only buy Winnie the Pooh merchandise when I’ve visited my baby nephew – how does Amazon know that – they’d know that if they’d read my Facebook wall. Hopefully you’ve realized how in this example, the data Amazon.com had about me was useless.

Had all aspects of my Winnie the Pooh related activity, across both the physical and virtual realms, been able to have been recorded and made available to Amazon.com, that would’ve made a world of a difference to Amazon – it would have told them about how useless their Winnie the Pooh site is, and it would have broken all conceptions about what my customer preferences are in their eyes.

Written by Sumit

December 10, 2009 at 6:44 am

Disruptions to how all data will be structured and processed

leave a comment »

I’m no expert in databases but I’d like to talk about an issue that is quitely sliping into the background, though is very important and crucial to our enjoyment of that quick presentation of search results on our favourite search engine, or why we can easily search through the plethora of posts on our Facebook page….

This whole new approach to contextualizing this massive amount of data will lead to disruptions in data storage and processing, as evidenced by the popularity of the “no SQL” movement today. Google’s MapReduce and Apache’s Hadoop are two popular frameworks for distributed computing and mass data storage and processing that are manifestations of this “no SQL” movement today. Yahoo, Google and Facebook amongst other data powerhouses are increasingly using such frameworks to organize and process queries that span their immense repositories of data in a manner that is more efficient and time consuming than if they were to use traditional relational databases to store data and process information and queries from. This is not to say that this is the end of relational databases, because it is still possible extract efficiencies by intelligent indexing and optimizing how data is stored and processed, but Hadoop and MapReduce are definitely disruptive forces to be taken very seriously.

Written by Sumit

December 6, 2009 at 7:38 am

Google, Facebook, Yahoo, Amazon and all…lets all start sharing

leave a comment »

Aggregators’ value will never truly be realized until and unless there is a breaking of all the highly protected silos of data and information out there –  the baseline set of data used to satisfy the query should be as expansive as possible. For example, in the case of Facebook or other social networking sites, friend recommendations should not just be to friends within certain degrees of freedom, but based on various other sources of data.

If serendipity is to be truly molded, then it is inconceivable to only narrow down aggregation sources to a very finite set of data. Moreover, we’re seeing a breakdown of personal barriers in the workplace and so why should we maintain separate accounts between LinkedIn and Facebook? They need to share APIs. This frustration will be driven by consumers and eventually the providers will need to respond else they might lose business. This is highly disruptive, unless we see a giant consolidation of APIs across the spectrum of providers, the increased silo’ing effect will only frustrate customers more and more.

Over time, Data aggregators, notably Search Engines, will lose power due to user frustration and an argument can be made, that since over the long run it has been proven that switching costs aren’t a significant hindrance to users migrating to new applications and services (i.e.Friendster, MySpace, Facebook, Twitter), the power will shift towards the user as the incumbents gets challenged more and more by new entrants.

Further, as the intent discovery function expands to include other factors based we will begin to see a lack of standardization due to the many search functions acting in silos. This will call for alliances between Google, Amazon, Facebook, Yahoo, Foursquare, Twitter and the list goes on – the core objective for these alliances will be to provide a universal data set from which a meaningful aggregation of data can be done. However, there will still be a lack of standards, integration and this will be a natural consequence of these alliances – a federation of standards.

Written by Sumit

December 1, 2009 at 7:36 am