There’s been a great deal of chatter in the business community about Twitter. To me, most of it appears to be excited businessmen smelling a hot commodity, but still struggling to understand what makes Twitter so popular. Here is an article where a businessman shares his insights by creating a parallel between Twitter/Facebook and the Id/Ego. Just the fact that he’s trying to force Twitter into an existing and familiar framework not only seems forced, but it’s a reflection of about 90% of the business thinking surrounding new ideas like Twitter. Too often there's a disconnect between a hot technology idea like Twitter, and the people with money to spend on these ideas. There is an inclination to squeeze these ideas into an old-fashioned business model, until profits oozes out.
This is a good summary article of how most business people view twitter. They outline 10 ways that Twitter will change business. I’m going to save you some time by telling you that there’s only about five unique ideas in this piece, the rest is just filler to make the list 10. The core of these ideas fall into one of two business advantages:
- Corporations can gather personal data about people;
- If people follow corporations, the companies have a direct channel to their consumer through tweets.
But when I read the article, I had an out-of-body experience that the year was 1993 and I was reading this exact same article, except it wasn't about Twitter, it was about a new thing called E-mail and the Internet:
- “In utilizing the Internet, we are able to collect personal data from consumers about their preferences!”
- “By collecting hotmail addresses we can use electronic mail to harass the customers about promotions and sales at pennies of the cost as regular mail!”
It’s really all the same stuff. There are definitely some good uses outlined in the article, but there’s no game-changer. Twitter, in-and-of-itself, is a minor connectivity improvement.
Then there’s Howard Lindzon, who has leveraged Twitter to create something called Stock Twits. Essentially, it’s designed to be twitter specifically for day-traders. Follow truth and rumours from the investor gurus in real time! I suppose this is an important step for day-traders, Twitter can help unify hundreds of different investment chat rooms into a single massive investment community. But I'm just not feelin' it.
But then I read this. It blew me away. The idea is that StreamBase Systems wants to collect and interpret tweets to determine if there is a particular sentiment towards a stock or the stock market in general. The challenge for them is to interpret all the tweets and organize that data into forming a collective sentiment.
The article goes as far as citing an old day-trading secret, which serves as a convenient example of how the power of this tool would manifest. Recall last week we discussed how markets are efficient, and that any news about a particular stock, or the general economy as a whole, is quickly incorporated into stock prices as soon as the news is made public. There is an old trick that day-traders use (which might have a slick short name that I am unaware of) called, “Buy on rumour, sell on news.”
For example, let's assume RIM is going to announce its 2nd quarter earnings on July 15th. Now, as day-traders we believe that RIM has a positive sentiment associated with it (by that I mean RIM is a good company, it gives investors a positive feeling because it often exceeds expectations). Because we believe this, we can potentially ride this sentiment to profitability. How? The way we can profit from this is to buy the stock early, and sell the stock right before the news about their earnings is made public, so in this case we would sell it July 14th. At the end of the day, we just have to guess which way the herd is leaning to make money. It doesn't actually matter if RIM has exceeds its estimated profits on July 15th - we're not actually betting on the stock; we're betting on the market's collective perception of the stock.
With StreamBase Systems data, we can potentially tap into a more accurate depiction of that sentiment. We can decide based on their data if the market is relatively positive on RIM, and make an educated guess about whether the price will creep up on anticipation of the Q2 earnings report. Investors will pay StreamBase for this privileged information, because in theory, it provides an information advantage on the market.
There are some potential loopholes here. If the service is truly good at depicting market sentiment about a stock, most or all day-traders will rely on it (heck - institutional investors will be all over this too). If most day-traders rely on the same source, it means they'll receive the same data at the same time, and will transact as a herd which eliminates their profit margin on the buys and sells. The other problem is information being leaked. Currently, I don’t have to read the Wall Street Journal to know what WSJ is writing about because bloggers and tweeters are writing about it for free – I just have to follow the right people and I can get my information for free.
But whether or not one can profit from this service is not why I found the article powerful in the first place. What fascinates me is the potential to interpret tweets and form a general sentiment. The idea of a collective consciousness was always said to be possible with the internet, but will Twitter be the application to make it a reality?
Then there’s Howard Lindzon, who has leveraged Twitter to create something called Stock Twits. Essentially, it’s designed to be twitter specifically for day-traders. Follow truth and rumours from the investor gurus in real time! I suppose this is an important step for day-traders, Twitter can help unify hundreds of different investment chat rooms into a single massive investment community. But I'm just not feelin' it.
But then I read this. It blew me away. The idea is that StreamBase Systems wants to collect and interpret tweets to determine if there is a particular sentiment towards a stock or the stock market in general. The challenge for them is to interpret all the tweets and organize that data into forming a collective sentiment.
The article goes as far as citing an old day-trading secret, which serves as a convenient example of how the power of this tool would manifest. Recall last week we discussed how markets are efficient, and that any news about a particular stock, or the general economy as a whole, is quickly incorporated into stock prices as soon as the news is made public. There is an old trick that day-traders use (which might have a slick short name that I am unaware of) called, “Buy on rumour, sell on news.”
For example, let's assume RIM is going to announce its 2nd quarter earnings on July 15th. Now, as day-traders we believe that RIM has a positive sentiment associated with it (by that I mean RIM is a good company, it gives investors a positive feeling because it often exceeds expectations). Because we believe this, we can potentially ride this sentiment to profitability. How? The way we can profit from this is to buy the stock early, and sell the stock right before the news about their earnings is made public, so in this case we would sell it July 14th. At the end of the day, we just have to guess which way the herd is leaning to make money. It doesn't actually matter if RIM has exceeds its estimated profits on July 15th - we're not actually betting on the stock; we're betting on the market's collective perception of the stock.
With StreamBase Systems data, we can potentially tap into a more accurate depiction of that sentiment. We can decide based on their data if the market is relatively positive on RIM, and make an educated guess about whether the price will creep up on anticipation of the Q2 earnings report. Investors will pay StreamBase for this privileged information, because in theory, it provides an information advantage on the market.
There are some potential loopholes here. If the service is truly good at depicting market sentiment about a stock, most or all day-traders will rely on it (heck - institutional investors will be all over this too). If most day-traders rely on the same source, it means they'll receive the same data at the same time, and will transact as a herd which eliminates their profit margin on the buys and sells. The other problem is information being leaked. Currently, I don’t have to read the Wall Street Journal to know what WSJ is writing about because bloggers and tweeters are writing about it for free – I just have to follow the right people and I can get my information for free.
But whether or not one can profit from this service is not why I found the article powerful in the first place. What fascinates me is the potential to interpret tweets and form a general sentiment. The idea of a collective consciousness was always said to be possible with the internet, but will Twitter be the application to make it a reality?
Posts
6 comments:
hey umar..interesting insights as usual...i like how you broke down the practicality of twitter but you didn't really address what annoys me so much about twitter (and blogging to a certain extent) that it creates a venue for all rank amateurs to spout their "expert" opinions on whatever..and twitter is the worst..cuz it seems like its all about ppl obsessing over the 140 character comments of celeb's
i didnt realize the corporate use of it..sorta cool...but also sorta overwhelming in the sense that the media/entermainent/big brother/corporations are all over us throught any outlet they can find.
honestly, i'm interested now that i'm looking at it differently, since i read that article. it's not about one particular opinion or comment, it's a body of information.
if we can find a way to tap the collective comments to draw sentiment, won't that be a powerful tool? it's like instant polls (but where everyone is unknowingly participating).
The trick really depends how well trended the information is supplied to twitter. Simply aggregating mentions and re-twits gives you only part of the picture. But when hashing is taken into account - prefacing a term with a # - which is generally only used by people more familiar with Twitter/advanced users i think you can get an accurate picture of how users are feeling about a particular subject. But even then, how actionable the information is is questionable. Being real-time the information is more or less transient in peoples minds. Its not necessarily something they really have thought deeply enough about. I agree that "investor guru's" generally won't idly post, my point mainly regarding StreamBase's seeming model.
Some of the assertions from the 247wallst.com article are frankly offbase. The only benefits, as far as I can see, is building relationships and providing the impression of responsiveness.
For example my company simply tweets crap out there, interesting crap, but crap nonetheless, simply for creating the a fun impression of the company. That really serves little in the way of one-to-one sales or the like, simply creating an image of the company.
Also, in the tech world twitter really does give added benefits in terms of informing customers of ongoing issues. Just last week when Rackspace - a large colocation facility - went down, they reported on each stage of the problem and solution via Twitter. This was really helpful in terms of quickly disseminating information to their customer base and did help mitigate any ill will (as the companies whose websites went down were losing money each passing minute).
Though, perhaps as its not my world, I am not thinking in the same vein as you are and not seeing the correlations you see.
We're on the same wavelength.
Maybe Twitter will add details or fields to each tweet, like something like emoticons, to give some general "feel" of the message (or the sender of the message, whatever). Maybe something like that will help filter through all the crap?
And as for that 24/7 article, the original draft of this piece was heavily focused on how stupid they sound. Thankfully I found a better article (it pays to wait before posting).
The thing you said about Twitter acting as a public service announcement is definitely a benefit. You can imagine the City of Toronto using it, to say, update us on how that whole garbage strike thing is going. Look at how San Fran is using it:
http://www.techcrunch.com/2009/06/02/the-city-of-san-francisco-now-lets-you-submit-complaints-via-twitter/
I'm really just brainstorming. I'm wondering if there's a way to organize information in Twitter into something meaningful.
Like, how about the next time there's a shooting at Osgoode station people can tweet which way the shooter is headed and what he's wearing and the cops can follow this and catch the bad guy?
I'll stop now.
There are a few companies out there that filter through Twitter data. Check out
http://www.radian6.com/cms/home
Working at a marketing agency there is a huge interest in this kind of technology. Focus groups cost a lot of money for the opinions of up to 20 people? Then the next day a rumor (or truth) hits the paper on your company and your focus group findings are pretty much garbage, the sentiment has changed.
Scares me with stcoks... but if you are Ford and put out a new car, pretty cool to get opinions from consumers so quickly. Think how quickly you could change your next email communication to not talk about the "man step" because people in general think it's dumb, instead talk about the all digital dashboard.
I do agree it's a lot of hype, like email, messenger and facebook all got. But given the above I think Twitter will be big for the marketing world, like email has been. Email gave a way to talk to the consumer... now the consumer can talk back with twitter.
~AK
Hey AK, thanks for the link. I like.
And I totally totally hear you on the marketing angle, if companies can somehow tap the sentiment, even just knowing that they are being tweeted about a lot is nice to know (for example it's good feedback on a new ad that aired on tv, etc).
Post a Comment