Reports to Read from HP’s Social Computing Lab

A few months ago, I highlighted HP Lab’s report on Twitter, and requested that the team contact me for lunch. They did, and a few weeks ago I met with Dr. Bernardo Huberman who is a Senior HP Fellow and Director of the Social Computing Lab in Palo Alto, see photo. While only one of about 20 labs at HP, this one focused on social computing is of particular interest to me. Their studies of social behaviors as it relates to technology, psychology, and sociology is interesting you can see a list of their public reports.

HP’s social computing team has over a dozen scientists to who conduct experiments to see how people will react to a variety of situations –many of these findings lead to product innovation. During lunch in Palo Alto, we talked about the impacts of their social research to HP’s products, Bernado expressed how HP understands how social computing is key to business, one of the first indicators was how they were hiring younger people how they were native to these technologies.

HP Social Computing Reports to Read:
You’ll find these reports to be scientific in nature, which are a different approach than industry analyst reports, I’ve read all of the following:

Report: Crowdsourcing, Attention and Productivity (PDF)
Bernardo uses the term ‘attention’ different than web professionals (that may think of that as time on site) and defines it as “Attention is sharp focusing of your mind on something while ignoring other things.” The report’s key findings indicate that people are more likely to share online when they receive attention from their community –something he pointed out is likely related to my own online behaviors and I’d have to agree. Key takeaway? We’re all social creatures and self-expression in a vacumm won’t provide that social gratification. What should business people do? Focus on giving attention to those that say good things about your brand online by reciprocating with attention.

Report: Predicting the popularity of online content (PDF)
This report highlights how it’s possible to predict how getting people views and votes within the first few hours of posting on YouTube and Digg can predict how popular items will be for 30 days –with considerable accuracy. What should business people do? Ensure that your online web properties are getting the proper attention in immediate launch, to ensure viral spread.

Report: Social networks that matter: Twitter under the microscope I’ve already covered the highlights for the business person for Bernardo’s Twitter report, or you can access the PDF directly.

I’m impressed with the work that the Social Computing Lab is producing over at HP, and am more interested to see how it will be implemented into HP products and services. If you’re a Forrester client, you may see my reference their work in my own future reports.

8 Replies to “Reports to Read from HP’s Social Computing Lab”

  1. Very interesting (also the others on the list).

    Jeremiah, I’m not sure your conclusion of “Ensure that your online web properties are getting the proper attention in immediate launch, to ensure viral spread” is supported by the 2nd paper. It’s a matter of cause and effect, and the conclusion I read there is that you can predict success by looking at initial patterns, but not that boosting that initial pattern will necessarily cause success. They also say themselves “[We] did not consider the semantics of popularity and why some submissions become more popular than others.”

  2. Ofer

    I stand by that actually. Companies should first build good content (a given) then ensure that it gets awareness in the market to become popular. I think both of these things are very, very difficult to do, but I’m hearing of services that help certainly with the first (content) and second with letting it go viral (PR, blogger outreach).

    On a less glamorous front, while I can’t endorse this process, I know of brands hiring social media optimization firms that have a variety of techniques to try to get content to go viral. That’s just a fact of what I see happening.

  3. Oh, I’m not saying it’s not a piece of advice that makes sense, it makes perfect sense, I was just trying to point out that it doesn’t stem directly from tats research.

    BTW – on the SMO issue, that’s precisely what a good service that is based on EGOCENTRIC social graphs prevents. Had Digg prioritized YOUR stories higher based on diggs in YOUR OWN social graph rather than the general collective, such global type of gaming would have become a lot less cost-effective. We’ll probably see those models emerging, in particular with FB connect as an enabler (disclaimer: I work at Delver, where we tried to do the same for web search).

  4. Very interesting (also the others on the list).

    Jeremiah, I'm not sure your conclusion of “Ensure that your online web properties are getting the proper attention in immediate launch, to ensure viral spread” is supported by the 2nd paper. It's a matter of cause and effect, and the conclusion I read there is that you can predict success by looking at initial patterns, but not that boosting that initial pattern will necessarily cause success. They also say themselves “[We] did not consider the semantics of popularity and why some submissions become more popular than others.”

Comments are closed.