HP Labs

HP Labs: Attention is more important than money

/ Tom Murphy / 2 Comments

The conclusion from a recent study by HP Labs, Status as a Valued Resource, which confirms that attention is more important than money may seem totally counterintuitive. Outside of totalitarian dictatorships and countries that are totally ramshackle and broken down, we live in economies that could be categorised as capitalist to one degree or another. Without money the governmental and business systems of a State quickly grind to a halt. Our lives, as individuals without money, can become onerous in the extreme. Therefore, on the surface, money can be seen as vital and core to our survival and anything else is a bonus. Claiming that something as seemingly trivial as attention is actually more important is a very challenging idea.

In extremes, the absence of money like the absence of water, food or shelter requires urgent self-management. All one’s focus, unless impeded by injury or illness, automatically turns to satisfying these basic needs. Lack of enough money to function with carries within itself the same imperative to do something to remedy the situation.

However, it has been shown that after one has access to a certain amount of cash the impulse to satisfy financial desires is as a consequence reduced. The following video makes this point in a most delightful manner.

In the paper Status As a Valued Resource authored by Bernardo A. Huberman from Hewlett Packard Laboratories, Christoph H. Loch, INSEAD and Hewlett Packard Laboratories and Ayse Öncüler from INSEAD a key experiment was done that shows that will sacrifice monetary gain in order to gain status and garner increased attention.

“Intrinsic status seeking by individuals has important implications for social and economic systems because it can provide a powerful motivation to perform; it also can lead to unproductive competitions with no obvious social value.”

The experiment consisted of a two part card game. Each subject was given a set of 30 cards and the idea of the game was too deploy them in such a manner so as to proceed to the next round. The cards had been given an arbitrary value so their use when deployed could be described as an expenditure or investment. Once through to the second round the game became a lottery. A winner was picked at random and the game ceased. The more cards that a player had left over from the first round increased their chances of winning the lottery in the second round because they would have had more cards available for the draw.

When the game was played by the rules presented to the subject and other variables were taken into account the subjects deployed their resources in a manner corresponding to what could be perceived as a rational approach. Passing through the first around with sufficient cards left over to increase their chances of being card being chosen in the lottery.

Then the notion of a ‘winner’ was introduced:

“In the first condition (no status), the game was conducted exactly as described above. In the second version (status), we introduced a status condition by stating at the beginning of the game that the winner of Stage 1 would be announced publicly, given a small tag saying “Winner,” and congratulated.”

This altered the subjects game playing strategy. The immediate acknowledgement, attention and status that was perceived to be had from obtaining a winner’s badge and receiving applause proved to be more valuable to the participant than a possible eventual overall victory in a lottery.

We know from our previous article Crowdsourcing: Getting Attention is the Key to Getting the Message Out, that attention is the main driver for content production on Youtube and across other forms of Social Media networks that have similar dynamics and operate in a similar way that the more attention a content provider gets the more content they produce and vice versa.

This research further shows that our prime motivator is attention. In this experiment attention gave the subjects status and it is this sense of status that defines where we are in relations to others. Our essential primate nature has existed long before the use of money came into play. Where we stand in our pack or social group has far more significance then many either suppose or care to admit.

In our prehistoric groupings outside of the inefficient use of violence, being active in increasing one’s status was probably the only way of getting hold of special privileges such as a preferred share of the food and grooming etc.

In the modern world we see this in the office politics and turf wars of the private and public institutions. Most people are earning enough to get by so our inclination to pursue status comes to the fore more easily.

In the online world status may be seem to be suggested by having a high follower account but as explained in a previous article, How to Influence on Twitter: Research from New Algorithm Gives Guidance a high number of followers does not necessarily mean that one has a great deal of influence.

“The experiments reported here imply that people tend to over-invest resources whenever “winning against others” is involved, because winning confers status.”

We now know:

  • People value attention more than money.
  • People will behave differently for the sake of increased status.
  • Increased status brings more influence.
  • More influence means greater leadership and possibily more rewards in online communities.

Knowing what our drives are frees us from being slaves to them. Knowledge of our predilections means we can allow, correct for and take advantage of them. The information we now have from the work done by Bernardo and his colleagues at HP Labs enables the rest of us to think, plan and prepare our online and offline activities with greater precision and accuracy than ever before.

Crowdsourcing: Getting Attention is the Key to getting the message out

/ Tom Murphy / 1 Comment

It’s either ironic or admirably humble that an article in Wikipedia on crowdsourcing contains no reference to Wikipedia itself being one of the largest and most significant crowdsourcing project going on today. The entry itself claims that the word crowdsourcing is derived from ‘crowd’ and ‘outsourcing’ and first coined by Jeff Howe in this Wired article from June, 2006. Neither of them mention the term ‘wisdom of the crowds’ which was used by Francis Galton, a nineteenth century English polymath. He attended a local fair and noted that when other fair goers joined a competition to guess the weight of an ox the mean of the guesses was surprisingly accurate although no individuals were precisely right and some of the experts were really quite wrong.

But this conventional idea that a sample average is taken from a group or of people and assessed by others to formulate public opinion or design marketing programs may no longer have the relevance it once had.

Legacy media, the old style news and broadcasting mediums which are still in existence and still influential, no longer dominate the landscape of public thought in anything like the reach or depth they once did. However, the idea that there is such a thing as public opinion is still there along with the notion that it matters.

“How does the public agenda get set when most of the conversation is bottom up instead of top down?” is the question that Bernardo A. Huberman, Fang Wu both of the Social Computing Lab, HP Laboratories in Palo Alto ( Fang Wu has moved on since the report was written) and Daniel M. Romero from the Center for Applied Mathematics at Cornell University attempted to answer in their paper Crowdsourcing, Attention and Productivity.

Figuring out how and what bubbles to the top of the public agenda is a very important problem. Bernardo says, “Should we decide x,y or z about something? That about something is now getting very confused because grass roots movements though technology can have a huge sway.

“In the United States we have tea party movements and so on that are getting an immense amount of attention internally and externally. But it’s truly something we don’t understand; how is it out of all these chaotic conversations in Twitter, in Facebook, in blogs, in email and so on stuff bubbles all the way to the top? That to me is something that is going to be profound because eventually it’s going to start setting up the public agenda which is what society as a group, a community and an organization pays attention to.”

If you are going to examine the behaviour of crowds and mine their activities and opinions for data then it makes sense to go where the crowds are. Wikipedia, Digg, flickr are obvious contenders for examination as virtually all the material is provided voluntarily by contributors as and when they please with minimal restrictions. Huberman, Wu and Romero eventually settled on Youtube. There is a great deal of activity on the site. In May 2010 Youtube exceeded 2 billion views a day. At certain points it has accounted for about 10% of all traffic on the internet.

In April, 2008 The HP Labs team put together a dataset of almost 10 million videos which had been submitted by nearly 600,000 users. These were weighted accordingly to account for such factors as older videos having a greater amount of views and so on.

For topics to grow or even exist in the first place then content has to be provided. So what would be a major motivating factor to engage the time and effort to create a video? Since at the very basic levels of distribution where there is no financial reward or incentive then we can turn to considering attention as a possible primary driving force. After all why put anything up anywhere on Social Media if you don’t want people to see it?

In studying “the dynamic interplay between productivity and attention” it can be confirmed through experiment and evidence what most of us thought might have been true. “For those contributors who were active for a minimum number of periods, the more views they received in one period, the more videos they uploaded during the following period.” A virtuous cycle was created where the more attention a video received the more inclined the producer of the video was to produce more videos. It wasn’t down to what the contributor thought about their own video. They were motivated to produce more content because other people were giving their previous videos attention.

They also found the opposite to be true. People stopped uploading videos if attention declined relative to previous uploads. Without attention being paid to their product contributors ceased to upload them.

In the conclusion to the paper Bernardo Huberman, Daniel M. Romero and Fang Wu say the following “By analysing a massive data set from YouTube we have shown that the productivity exhibited in crowdsourcing exhibits a strong positive dependence on attention. Conversely, a lack of attention leads to a decrease in the number of videos uploaded and the consequent drop in productivity, which in many cases asymptotes to no uploads whatsoever.”

All this suggests a mechanism for ideas to bubble up through oceans of data and set the public agenda begins with contributors being rewarded by attention being given to their work, the subject of which could be shared beliefs of a political, financial or of any other nature. As more attention becomes centred on these ideas the more motivation the contributors have to create more product along the same lines and thus create more attention and so on.

Instead of an old style news editor sitting in their office deciding what hundreds of thousands of their readers are going to read about or what millions of viewers are going to watch on their televisions we now have random individuals coagulating around an idea and creating content simply because other people are willing to pay attention to it. Instead of the world being presented to us through the filters and ‘judgment’ of a relatively tiny amount of editors and their editorial teams we now have the world being shown to us by content creators who have managed, by whatever means, to bring attention to their work.

Predicting Box Office Returns using Twitter: Bernardo Huberman from HP Labs

/ Tom Murphy / 2 Comments

Last week, Technology Voice spoke to Bernardo Huberman, Senior HP Fellow and Director of the Social Computing Lab at HP Labs. For the last two years, he and his colleagues have been focusing on the phenomenon of, in his words, “social attention”. Our world has exploded in terms of the amount of information that we have available to us. But like so many plentiful things, all this information has very little value. As Bernardo puts it, “ The only things in life that have value are the things that are scarce.”

One commodity that is very valuable because it is limited is our ability to give attention to the myriad potential distractions we face each day. Only so much time and energy is available to us to focus on anything. It is the competitive vying for our attention that has led to the constant bombardment from providers of goods and ideas with messages such as “look at me, consume me, buy me, watch me.” Spam is a great example.

Social attention is not measured from a perceptual or psychological point of view, as fascinating as that is, but from how certain things draw our attention as measured by the number of downloads of a given video, the amount of times an article is read, the number of things people discuss, and so on.

Bernardo says, “One interesting follow up of this was my conjecture that where attention goes usually it portends something and the money follows. So that’s why providers want a lot of attention to their content.

“So one thing we decided to do was to see whether or not we could actually predict something about the future by looking at how attention is allocated to certain things that are about to happen… How do people discuss forthcoming movies that are about to open and the rate which they tweet about them on Twitter? Could we come up with predictions about how well those movies were going to do at the box office?

“Now the reason I chose movies is because that is a very clean-cut example of something that you can measure. If I say to you I will predict certain things will happen and they are diffuse, then it is very hard. But [we can] predict the box office revenue at the end of the day or the end of the weekend… We know a number, that number’s public, how much that movie made.

“So, what we did, basically, we chose about 24 movies that were about to open. We followed the rate at which people were tweeting and we were able to – by calibrating to movies that had already opened and [we] saw the attention they were already getting on Twitter – we were able to come up with a fairly accurate prediction about how much money the movie was going to make.”

Bernardo and his team then compared their findings with the Hollywood Stock Exchange (HSX). This is a site where people can buy stock on actors, movies, directors and other aspects of the movie industry. The HSX has a reputation for being a good predictor of box office revenue. The predictions of the HP Labs team worked so well that patents have been applied for and a business mechanism has been created.

The HP Labs team were also able to improve the quality of their forecast for the next week because now that people saw the movie, they were expressing feelings about whether they liked or disliked it, etc. With this feedback, they were able to fine tune and calibrate their predictive methods.

“It is obvious you can use this for other things. We know that it works for movies, you could do that for new products, new trends, and so on. So if you have a way of tapping a social medium, and that’s our idea, you [can] look at the way that attention is allocated to certain things – sometimes you can predict fairly accurately how certain things will happen.”

This predictive technology has very important implications for political as well as all sorts of other marketing campaigns. The work of Bernardo and his colleagues combined with the techniques of sentiment analysis offer a solution to the Provider’s Dilemma.

But there is still important research to be done using these techniques to inquire into aspects of other modern-day social phenomena. The public agenda is no longer being set by the ‘legacy media’. The power of old-school editors lay in dictating the news agenda by their choices as to what featured on the front pages of their newspapers or what the lead story of the main nightly newscast would be. But that ability to influence public debate is fading. The power for setting the terms of what constitutes public discussion is now dispersing from the few to the many.

Increasingly, our main topics of discussion that concern us as a society are what rise up from and through the various social mediums, but as yet no one has identified the mechanism by which this works.

As Bernardo says, “It’s truly something we don’t understand. How is it out of all these chaotic conversations in Twitter, in Facebook, in blogs, in e-mail, and so on, [does] stuff bubble all the way to the top.”

The answer to that question may have more profound implications than being able to create a more accurately-targeted advertising campaign.

How To Influence On Twitter: Research Results from New Algorithm Give Guidance

/ Tom Murphy / 3 Comments

Recent work done at HP Labs, the exploratory and research group for Hewlett Packard, shows what most of us suspected as being true all all along; that just because a person has a lot of followers, it doesn’t mean they have a lot of influence.

In September 2009, using an algorithm they devised called the IP (Influence/Passivity) algorithm, a team of researchers from HP Labs continuously queried the Twitter Search API for 300 straight hours for all tweets containing the string of letters ‘http’. Finding this string in a tweet would indicate the presence of a URL, and demonstrate that a web page was being shared or retweeted by means of a link.

In that time period, they acquired 22 million tweets with URLs present. This accounted for 1/15th of the entire activity of Twitter at the time. The URLs were checked for validity, and by revisiting the Twitter API they could determine who the user for each URL was, and in particular who their followers and followees were as well. From that information, a complete social graph was constructed from the dataset generated by the users sampled.

The research team worked on the following assumptions which are taken from their report “Influence and Passivity in Social Media“:

  • A user’s influence score depends on the number of people she influences as well as their passivity.
  • A user’s influence score depends on how dedicated the people she influences are. Dedication is measured by the amount of attention a user pays to a given user as compared to everyone else.
  • A user’s passivity score depends on the influence of those who she’s exposed to but not influenced by.
  • A user’s passivity score depends on how much she rejects other user’s influence compared to everyone else.

A whole industry has grown up around Twitter with the aim of developing various tools that enable Twitter users to increase their number of followers. But now all these efforts seem to have been in vain. An average Twitter user retweets only one in 318 URLs. It seems most users are passive information consumers, and do not forward the content to the network at any kind of rate that could be described as ‘little more than partially engaged’. Consequently, having a large follower count is not a lot of use from a message propagation perspective if most of the followers are made up of these passive users.

If you want to be a person of influence on Twitter, then the way to do it is to acquire engaged followers who are themselves active on Twitter. That would at the very least mean being active and engaged yourself.

Of course, this makes life difficult for marketers and others engaged in viral activity who want to take advantage of the enormous reach that Twitter has. They can no longer rely on a single dubious metric, follower count, as a guide to how far their message gets out.

It also means that to find active and engaged people, they will have to become active and engaged themselves. Fun, maybe. Time-consuming, certainly. But it is only through interacting with highly connected people that they will be able to propagate themselves and their message through the social network.

Through the process of finding the most influential people on Twitter, the team also managed to turn up the most passive users on the service as well. The majority of these users tended to be spammers and robot users.

It is as important to identify the highly-passive Twitter user because “they provide a barrier to propagation that is often hard to overcome.” It’s good to know where the Twitter dead ends are as it gives us a useful benchmark to contrast with someone who is influential. This information aids navigation through the vastness of the Twitter network, and knowing where not to go can be every bit as useful as knowing where to go.

The HP Lab report finishes with the following conclusion:

“This study shows that the correlation between popularity and influence is weaker than it might be expected. This is a reflection of the fact that for information to propagate in a network, individuals need to forward it to the other members, thus having to actively engage rather than passively read it and cease to act on it. Moreover, since our measure of influence is not specific to Twitter it is applicable to many other social networks. This opens the possibility of discovering influential individuals within a network which can on average have a further reach than others in the same medium regardless of their popularity.”

In a way, it is understandable that the findings from this research on Twitter would be applicable to other social networks as well. Influential people tend to be influential wherever they are. The IP algorithm that has been developed by the HP Labs team is going to be very useful across all of the social media domains, wherever people gather and exchange ideas and news via links.

The key metric to determining how effective any given person is in propagating information is to measure how often a URL that they tweet or retweet is clicked. And it is important to allow for the fact that not everyone is very adept at giving credit for their shared links, and some shared links go beyond the Twitterverse on to other services.

But unless you have access to a given individual’s bit.ly account or other some search service which keeps tabs on retweets, the only sure way to know if they are a person of influence is to get to know them.

Happy twittering.

BlogTalk is returning to Ireland on 26-27 August; check out the speaker list.