Understanding HP Lab’s Twitter Research

Left: The famed HP Labs think tank in Palo Alto.

A few months ago, I spent an entire day with the HP Labs group in Palo Alto, they’re responsible for the R&D and innovation that goes into their thousands of technology products on the market. I was pleased to see this deep dive scientific research on Twitter by Bernardo A. Huberman, Daniel M. Romero and Fang Wu.

You can read the free Social networks that matter: Twitter under the microscope (PDF). Written in an academic style, it’s a bit dense for the casual reader. I write for a business audience and I’ll strip out the most important findings, and add my own insight to what I think matters. As always you’re welcome to chime in the comments.

Understanding HP Lab’s Twitter Research:
If you just need a summary, I wrote this in a way that you can just read the bolded elements to get a sense of the report. I hope this saved you some time.

  • Most users have a smaller inner circle they communicate with: Within a social network, it was found that most only frequently communicate with a small segment of users –even if one has a large community. Makes sense, everyone has an ‘inner circle’. Finding the true network that an individual has (even if they have thousands of “friends”) is what’s really important. Although Scoble solicits imput from thousands of contacts, he leans on a smaller subset of folks to trust above all others.
  • HP Labs Sample Size is @ 6% of the Twittersphere: HP Labs took a random sample set of Twitter users, for a base of number of 309,740 users. According to my social network stats tracking page, Twitter’s total universe is somewhere between 4-5 million (still very small). I’ll value the network on the 5 mil side, so that’s sample size of about 6%, which is pretty healthy.
  • On average, most had 85 followers: They found that the average user has 85 followers in their network, this number seems reasonable when averaged out across the network.
  • On average, most had 80 friends: Most users followed back 80 others, which is close to the actual follower number. Perhaps some weren’t following spam bots, or people that follow everyone. James Governor has been discussing asymmetrical networks, but it appears that on the average, most are symmetrical.
  • Tweet Frequency? About one a day: On average, these users had posted 255 tweets, and since the average users has been around for nearly 7 months, thats about 36 tweets per month, or little bit over one a day.
  • 68%: are active users Social networking stats are almost always flawed, as the vendors don’t disclose how many are truly active. I define active user base as logged in and completed an activity in the last 30 days. Among the 309,740 users only 211,024 posted. It’s unknown if this filtered out spam tweets, although nearly 2/3rds of users have returned (site stickyness. That’s a pretty good return to site rate.
  • Most members have been on twitter nearly 7 months: The research showed that the average person (from first to last post) was active for 206 days. This means that June 2008 (report written in Dec) has become somewhat of a trigger point, perhaps where a growth curve started to point upwards. I noticed an influx of users on April 2008, two months before HPs findings, see comment #579
  • A quarter of tweets (@) are directed at other users: The report showed that Around 25.4% of all posts are directed, by using the “@user” which is responding to others. This could suggest that the other 75% of tweets are updating their network of what users think is interesting or discussing ‘what they are doing’
  • The more followers, the more they tweet –up until a point: Figure 1 indicates frequently in posting the more followers they have, right up until about 500 followers where the frequency starts to level out (if the graph were smoothed). The data around number of friends suggests a similar graph, although there’s no saturation point (see figure 2). I’ll suggest the more connections a user has, the more value they have, and therefore are more active.
  • Despite having large networks, a smaller circle is maintained: For users with a high number of followers, they actually only still communicate with a smaller subset of users. This rule remains constant see figure 4.
  • Where’s the value? within the hidden network: To find out the real value of a twitter user and their network, finding out their true network of folks they communicate with on a regular basis will show their trusted network. Finding out who the Scobles’ communicate with the most will determine will help find out how he is influenced.
  • Business Opportunity for Measurement Vendors
    If you’re a social media measurement company, and can find out the true influence model of who people really trust above all other users by looking at actual “@” behavior and follow behavior, be sure to leave a comment below showing how you can do this. Then, conducting this by topic, will find out the true influencers by market segment within the Twitterpshere.

    Brand Opportunity
    As we know, traditional advertising doesn’t work well in social networks, ‘carpet bombing’ isn’t effective. However, conversational marketing is also costly, as you have to spend great resources on labor to communicate with influencers. Therefore brands who want to be effective with their resources should find out who is an influencer in their market and focus their conversational marketing primarily on them.

    Thanks to the HP labs team who did a great report and really helped to further understanding Twitter better, when you have time, invite me over for lunch, I’m in the area.

    • Pingback: Social Networks Site Usage: Visitors, Members, Page Views, and Engagement by the Numbers in 2008()

    • i guess i’m an abnormal user. i started late but i am constantly updating. most of the people i follow and/or am followed by i have never met in real life. none of my real friends would bother with it, i don’t think.

    • This post is really helpful, since I’m a newbie to Twitter and learning its value day by day. I’m also interested in personality types and job industry distributions across Twitter. It seems the extroverts seem to tweet about themselves and introverts about links. But both also tweet ideas, and I like that crucible for ideas. Also seems to attract self-taught techies, Mkt/Adv gurus, and those who consider technology a part of their lives. Anyway, thanks again, and I look forward to learning more.

    • About the branding part.. that’s in a sense WOM and originating from influencers. However, the up ramp labor is not low either to find those influencers and woo them and keep them onboard. One way could be tracking the discussion forums that these influencers attend.. and also follow people who make comments to their posts.

      http://twitter.com/pinakis

    • Jeremiah – thanks for an interesting post, but what does it mean in terms of Twitter activity for business right now? As a PR agency immersing ourselves in social media, it has become very useful to keep in touch with what’s innovative and interesting.
      But recently, when devising client proposals, we’ve put in and subsequently taken out Twitter ideas because the justification for using the platform just wasn’t strong enough.
      Is it too soon for Twitter to deliver real benefits to campaigns?

    • matt

      “Twitter’s total universe is somewhere between 4-5 million (still very small).”

      4 to 5 million is anything but “very small”. (especially to turn right around and call a measly 6% a “pretty healthy … sample size”. perhaps I misunderstood… if not, yowza, that line irks me. but anyway…)

      The fact that Twitter is even being researched in such a fashion is one of many-a testament to it’s impact outside of the grim throng of early-adopters (aka, the “very small” group of Twitter users who think they are the “only” twitter users or that they somehow “made” Twitter popular… you know who you are…) But that won’t prevent the inevitable follow-up by this mob from being rife with elitism, self-importance, and downright hot air.

      Save the aforementioned problem with some early verbiage I think you hit on some really key abstracts. I especially appreciate the remark about: “Brand Opportunity” (which falls in line with “Find the hidden network” … that’s all great advice … even though it draws an example from such an “outlier” as Bob Scoble… 😉 hehehe.

    • Very interesting post Jeremiah, I come to a similar conclusion about the opportunity for a measurement business model in my article here: http://experiencecurve.com/archives/why-mlm-will-kill-twitter-hint-because-they-have-a-business-model

      @Jon Clements read the article I posted because in that I provide a solid measurement model for measuring Twitter’s effectiveness at driving action. I used the Kmart idea and said anyone that retweeted my message would be entered to win a one of ten $25 threadless gift certificates. I sent that out to my 1,800 followers and it got rebroadcast over 500 times, and is still echoing now 🙂 Now take http://twitter.com/kevinrose who has 77,000 followers and do a competition to win an iphone based on RT and you might crash twitter.

    • Twitter is simple, but powerful. “Less is more,” really applies in this case.

      As of right now I have 104 followers, I am following 46, and I have posted 724 tweets!

      If you are wondering how businesses can use Twitter, check out my latest blogpost about CheapTweet.

      Another way that businesses are using Twitter is to see what average people are saying about their products. Then using this information to improve their products or advertising campaigns. Not a bad idea.

    • Not sure I agree with the linearity of this line of thinking regarding “influentials.”

      (a) I am not sure twitter network analysis is superior to real world testing that measures business results, e.g., who gets the most traffic. In a sense, the many gift card giveaways that are all over twitter (e.g., Walmart engaging 11 bloggers who also tweet) provides comparative data as to which of them is more productive.

      (b) Influencers and platforms are two different entry points; this should be a study of how influencers build social media influence, not just how they build it on twitter.

      Business wants to identify influencers. Twitter is a component of building influence for some people. I think the approach should be more comprehensive. Not every influencer will be on twitter (e.g., Seth Godin) or even a heavy user (e.g., Anita Campbell).

      Also, important to distinguish “household word” v. “niche” influencers. The question depends on who is asking the question and who they are trying to influence.

    • Derek

      This is a great write-up, thanks Jeremiah. The one statistic which I think may be the hardest to accurately measure, is the following/follower ratio and what type of priorities that implies for the Twitter community. First you have the proliferation of ‘auto-follow’ tools. Add to that the fact that many people try going the auto-follow route, then decide it’s not for them, and stop doing it.

      Then there’s differing philosophies about whether or not you should follow everyone who follows you as a courtesy or whether it is desirable to achieve a certain ratio of follows to followers. I’ve come across a number of Twitter users who believe that ideally you should be following far fewer people than follow you in order to create some sort of reputation. Some even use a 2:1 ratio as their goal.

      Seems like that particular metric is the target which moves the most for this study.

    • Social relationships are built on trust, manners and common interests – and so are twitter relationships.

      It’s not a tool of marketing, it *may* be a tool of discussion marketing for some – any excessive marketing oriented abuse will be spotted and naturally put aside.

      More comments in my post – Twitter and manners that matter: ttp://italiaotoko.blogspot.com/2008/12/thoughts-on-twitter-manners-that-matter.html

    • @Susan Kuhn Frost

      You are absolutely on the point.. it has to be an integrated approach. However, Twitter can be cross- platform integrated with one’s profile signatures whether Facebook, LinkedIn, blogs, or whoever.

      Twitter being a micro-blog tool, it may not be descriptive by itself.. however it can be referenced through objective conversations that happen across all other profile signatures. That’s where API pool will come into play. Just the way how other tech evolves on the growth curve.. next will be when developers are all crazy and developing hundreds of apps that will connect twitter to emails, vlogs, and what not.

      I will keep my eyes open for how Twitter apps come to the surface as more eyeballs gather around it.

      For the 2.0 PR community, it will be mission critical to track ways that Twitter as a service is finding its place in the daily communication channels of the influencers. Also depends how tech savvy these influencers are to absorb these open services into their existing routine.

    • matt

      I’m not sure I follow your comments. Twittersphere of 5 mil is small compared to Facebook and MySpace who have over 100 mil.

      Regardless, a 6% sample size of nearly any social network is still a good sample, make sense?

      Scoble is an outlier, that’s my point, would you mind trying to read the post again, then let me know if you still have questions. Let me know if I need to clarify something more.

      Thanks.

    • Jonathan

      A clarification question on your strategy of identifying influentials, Jeremy:

      Is your point that just broadcasting micro-posts — even pithy, personal, relevant, frequent and well-written ones — isn’t sufficient to build a brand, and that a business needs to ‘reply to’ users also? But that, since replying to lots of people is expensive, they should focus on those most influential users?

      Two questions:

      1. Are we sure that broadcasting doesn’t work, at least, somewhat? Blogs where the blogger doesn’t respond in the comments are quite common, and many are very popular.
      2. Can a business afford not to reply to customers and/or potential customers, anyway?

      Sorry if these are obvious…

    • Liz

      This is very timely information for me. Thank you for posting about this study.

    • One thing is clear, big business is asking questions and we need to be prepared with answers. Companies are still trying to figure out what a website is for, but they know they need one.

      If you are consulting in social media you better have a recommendation on why your clients should or shouldn’t use Twitter.

      Thanks J for the summary and highlighting the key points, instead of stripping them out 😉

    • Thx for the time to make a summary AND link to the full post.

      It’s nice to see some more “academic” research popping up about social media. Not much to find about it, but always provides very interesting perspectives.

      http://twitter.com/Adgenius

    • Pingback: Measuring Twitter: Marketing, Conversations and Individuals | Real Estate Internet Marketing by Union Street Media()

    • Pingback: Printer Tweet 5: Reply to This Tweet | Magicomm Blog()

    • This reminds me of a study that I read about (I think it was HP Labs) that tracked emails for internal projects and identified who was actually leading them. Couldn’t find a link to the study (extra points to anyone who finds it), but very interesting stuff.

      Michael B.
      http://twitter.com/michaelbiven

    • Facebook now claims “more than 130 million active users,” so yes, Twitter’s universe in terms of social networks is “very small.” The remark about healthy sample size refers to statistical significance, not social networking. Twitter’s 3 million probably reach 30 million, since so many have extended networks, are bloggers and are also reposting links elsewhere.

    • Although HP Labs does equate 2-way Twitter connections with value — an unproven notion from a marketing or PR perspective — that’s not exactly what the data showed. For one thing, the study showed a correlation between posting volume and number of friends. HP seems to say “value” is based on the assumption that 2-way communication implies passalong. But how do they know ideas actually pass? And are they underappreciating the value of Twitter broadcasting?

      The quote is:

      This implies the existence of two different
      networks: a very dense one made up of followers and followees, and a sparser and simpler network of actual friends. The latter proves to be a more influential network in driving Twitter usage since users with many actual friends tend to post more updates than users with few actual friends…

      Finding someone with a lot of “real” friends might cut the effort of IDing and influencing real influencers or connectors.

    • Anne

      Jeremiah

      Thanks for this. I help run a services company and am trying to get understand how Twitter can help us (or just be another distraction). You’ve clarified some key things for us with minimal time investment. Appreciate it!

    • Derek

      @ David Card

      Twitter will soon offer the notion of friends and groups http://is.gd/abzQ , so identifying these real influencers may be easier.

    • Most of my inner circle isn’t on Twitter though!

    • Wow ! I would never have thought this was possible – “75% of tweets are updating their network of what users think is interesting or discussing ‘what they are doing’”

      Thanks for the insight into the HP study.

      Shashi

    • Oak

      I find it very interesting that a prestigious outfit as HP will spend the time and expense on doing such an n in-depth study on twitter. Might be an indicator of what is to come….I started using twitter in August or July after I attended the New Media Expo. For me the other things like Youtube etc seemed to time consuming for me. I am certainly a below average user but from what I see this type of application could be a tremendous tool for businesses in terms of market research and as many do just blatant self promotion. I am surprised that none of the big boys have copied twitter yet. Imagine next election hearing an announcement “Log on to twitter to cast your vote” unofficially of course 🙂

    • Ed

      I think the data showing that asymmetrical networks are not the norm is the most fascinating and shows the difference in usage between the majority of users and the social core.

    • Hmmm

      So, most people talk more to their core group than they do to people they barely know. The way to use this as a marketing platform is to communicate with the Twitter Elite who have the largest circles of influence.

      And they spent a lot of time and money on this research.

      Just about every social media “guru” in the business has already explained this (all of it being pretty logical and obvious) … if it is too costly to market directly on social networks because of the time/money ratio…how cost effective is doing an exhaustive think-tank study of the same network?

      Not trying to be sarcastic, but man…unless there is a lot more to it than this, they could have hired a few high school grads to tweet for a couple of weeks and gotten the gist of all this.

      DNW

    • David

      Yes, I agree, we knew a lot of this intuitively, but now it’s backed by research and data.

      Brands, marketers, companies, base their investments off logic, and research helps them to be more confident.

      I assure you, having research on your boss’s desk makes up a more powerful ar argument guement when funding is getting slashed during an economic downturn.

    • Pingback: Cloudy Thinking » Blog Archive » Jeremiah Owyang on HP Lab’s Twitter Research()

    • Pingback: Canadian Business Blog » Blog Archive » More on Twitter()

    • michael

      taximike new to twitter and new to new media.My perception is that nothing is new.big tweets eating small tweets

    • Pingback: Five in the Morning 120908 « StickyFigure()

    • Pingback: links for 2008-12-08 « Donghai Ma()

    • Hi Jeremiah, another great post. But I don’t agree with “Most users have a smaller inner circle they communicate with.”

      I’ve been on Twitter for a while now and yes, as you create your network, you have a small inner circle. But as soon as you add a Twitter app on your phone, a client on your computer and play with Twitter Search, users find different circles that in many cases will never overlap.

      I follow 1,200 people but I have at least 10 different “inner circles.” By city (Miami, Bend, San Francisco, New York, Colombia) by industry (technology, newspaper, previous job, current job) etc.

      Thanks, @SocialJulio

    • Pingback: Between the Lines mobile edition()

    • Great post/insights, many thanks!

      I am one of those described above (or worst), as I created the account in April07, and started re-using it Nov/Dec08!

      Personal view: I think Twitter will not be mainstream until:

      1.) It provide better privacy protection

      (not everyone want to ‘blog’ nor share the innermost secrets (within reasons), so, I have to use a ‘public’ profile and friends will have to follow me on my private & protected feed)

      2.) groups

      At present, even with help of TweekDeck, it is still too many tweets and no easy way of managing all the information

      3.) better computer/mobile client

      it will need more ‘integrated’ approach, as yes, it has 6+million users, but they have to rely on various ‘hacks’ or 3rd party clients to view/track/comment/re-tweet.. its too geeky (not that I am not.. but I am not joe blog the mass market..

      and only when there is a more seamless client that the brand can really leverage/track/invest in using the channel to approach the market.

      Lets face it, the brand/media/FMCG leaders are just starting to understand online (limited mobile)..

      BR
      Gareth

      twitter: @garethwong

    • Pingback: links for 2008-12-09 « The social media revolution (in 15 minutes)()

    • Thank you so much for the insights and alerting us to the article. “Dense” is good. The loose tie to your post created by some search tools allows time to read the article. I’m not sure that a tweet would have achieved the same results.
      Aloha,
      Dan

    • Pingback: Networks that matter on Twitter: the @-Crowd < life under electronic conditions()

    • Pingback: renaissance chambara | Ged Carroll - Links of the day()

    • Certainly

      I favor a mash-up of platforms and messages. Using a blend of social networking tools with informal messages to new and old friends, sprinkled with a judicious bit of business development that adds value works just fine. After all, isn’t that what people do at parties?

      Give the economic times; combining messages, people, offerings and even industries can be a super way to prosper. My friend Jon wrote a cool series of pieces on corporate mash-ups positing Exxon bailout GM. A worthwhile read at http://www.touchmarketingblog.com/.

    • The “hidden network” of people you are replying to is only one part of the story. There’s also another, maybe even more hidden network of those people who are replying to you most often. I hacked together a tool to show both sides: http://metaroll.de/relevantnet.php

      I don’t quite agree with the notion of a “sparse” hidden network. I calculated my hidden network and it doesn’t look like the sparse network in the paper at all. Actually it’s a quite dense social network.

    • Pingback: Twitter stats and starting your own corporate Twitter account — Archive — RD2 Blog()

    • Pingback: @ the friends within followers | brants()

    • Pingback: 博客士 » Blog Archive » 每日关注 [2008-12-12]:ahgua @ del.icio.us()

    • Pingback: UID Teatime Blog » Facebook status updates: an informal study()

    • Pingback: All of a Twitter » pr-media-blog.co.uk()

    • Pingback: Why I’m Qwitting You On Twitter | OnlineMarketerBlog.com()