Bright Matrices

Writings & musings of Mike Zavarello (a.k.a. brightmatrix), a "red mage" web developer.

Tag: metrics

Scales and Sheen: How Very New Social Media Accounts Can Amass Influence So Quickly

As social influence metrics like Klout continue to gain, well, clout in the industry, instances will arise where folks will cry foul over perceived inconsistencies in how scores are generated. An example I’ll use in this post is how very new Twitter accounts are able to amass very high influence scores in a very short period of time.

Let me preface this discussion with influence as opposed to popularity. Numerous articles were written last year that clearly illustrated that one does not equal the other. Hollywood celebrities, musicians, and other popular figures have a great deal of popularity, but this does not translate well into influence, which is defined as being able to convince others in your social network to take action. Mark Schaefer, author of the {grow} blog, described just how challenging and difficult it can be to get your followers to do something, even if you have the popularity and reach of Alyssa Milano.

So, how can new accounts become so influential so quickly? Let’s study the Twitter accounts of a well-known and controversial actor and well-known, and currently missing, snake.

First, the actor. Charlie Sheen opened his Twitter account on Tuesday, March 1, 2011 to much fanfare. At the time, Sheen had a much-publicized falling out with Chuck Lorre, the director of his CBS show, “Two and a Half Men.” Folks speculated, and were quickly confirmed, that Sheen would use Twitter as a mouthpiece to share his perspective on the situation. Once word got out that his account was up and verified, the followers piled on fast. According to TwitterCounter, Sheen’s account accumulated 519,343 followers on the very first day. March 2 saw a subsequent increase of well over 530,000 additional followers, leading Sheen to reach the 2 million follower mark in well under a week. With over 3.3 million followers as of March 30, @CharlieSheen currently holds an astronomical Klout score of 94.

TwitterCounter chart for @CharlieSheen

Now, on to the reptile. On March 27, 2011, a venomous Egyptian cobra was found missing from the Bronx Zoo’s reptile house in New York City. The zoo promptly closed the enclosure as the search went on for the wayward serpent. It didn’t take long for a clever mind to make light of the situation, and on Monday, March 28, a Twitter account appeared under the handle @BronxZoosCobra that posted updates of the cobra’s adventures around NYC. On the first day the account was opened, TwitterCounter shows 86,140 followers tuned in to follow the cobra’s exploits. By 11 a.m. ET on Wednesday, March 30, the account has nearly 140,000 followers. While not as dramatic as Sheen’s account activity, the trend clearly shows a strong demand from folks on Twitter, resulting in a Klout score of 73 for the cobra (as of March 30).

TwitterCounter chart for @BronxZoosCobra

So, what can these two accounts teach us about social influence and the algorithms that determine their scores relative to others? It’s simple.

Demand, in particular, a sharp increase in your network, is a strong factor in measuring overall social influence. Both @CharlieSheen and @BronxZoosCobra attracted a massive number of followers in a very short period of time, and, in Sheen’s case, the demand has risen over the four weeks since the account’s creation (although it is starting to level off a bit). While their influence scores may not indicate that their network will take any kind of action on their behalf, it does clearly show that they’re producing content that folks want to consume and share with their own networks.

Now, it can be assumed that an accelerated drop in a network’s size (think rats jumping from a sinking ship) would result in a corresponding plummet in influence. To a lesser extent, steady or very slow changes in network size would result in a similarly consistent influence score. I’ve observed this in my own personal account: my network size is growing, albeit grudgingly and gradually, and my Klout score has remained relatively stable for several weeks.

What’s the take-away from all of this?

  • First, the systems themselves aren’t broken or being gamed. Demand is one of many factors these algorithms are using in their calculations, and sharp changes in network size seem to greatly influence (pun intended) the weight this specific metric has on the entire score. It would be informative to study Sheen’s and the cobra’s accounts after several months to see whether their overall influence continues to maintain itself over time. As always, don’t just look at the score itself. Do your homework and dig into the underlying metrics to figure out just what’s driving that high number.
  • Second, don’t look for demand alone to change your fortunes. It’s already been shown how sizeable networks don’t always translate to positive action from their audience. Exercise proactive listening and active engagement to energize your followers and fans. Learn about them and their interests, understand their needs, and give them what they want.
  • Third, be creative and innovative. Whoever created the @BronxZoosCobra account used a captivating news story about the cobra’s release to create something fun and clever for the folks on Twitter. This adaptive and nimble thinking is precisely what Jay Baer and Amber Naslund describe in their book, “The Now Revolution.” Another excellent example is how Aflac turned its fortunes around by creating casting call for their signature mascot after their principle voice actor, Gilbert Gottfried, was fired from the job.

Update (4-1-11): About midday on Thursday, March 31, 2011, it was reported that the Bronx Zoo’s cobra was found, apparently in a “non-public” part of the reptile house. No status update from the @BronxZoosCobra Twitter account as of 8:45 a.m. today, but I suppose the zookeepers took its iPhone away. It will be informative to see whether the owner of the account keeps up the cobra’s commentary, and also whether its followers and influence will fade over time.

The parody account @BPGlobalPR, set up after the Deepwater Horizon disaster in the Gulf Coast last year to poke fun at BP’s series of public relations flubs, is still active but seems to be slowly losing followers as the attention around the incident fades. Its Klout score, however, has stayed relatively steady over the past 30 days, hovering around 64.

Editorial note: I also updated the title of this post to read “Very New Social Media Accounts” vs. its original “Very New Social Media Networks.”

My 2011 Wish List for Klout

This has been quite a year for Klout. Their suite of measurements, specifically their “Klout score,” has risen to become a recognized standard in the realm of user-generated content and social communications. They’ve generated numerous articles on how corporations and industries are considering Klout to award perks, recognize influencers, and even possibly screen for job applicants. Klout has become, in essence, a “credit score” for the social space. Of course, detractors have had their say, and I’ve written my own article on why Klout’s metrics need to be used within their proper context.

I realize Klout is nascent, but the fact that they’ve gained the clout (pun intended) they have today speaks volumes for the need for solid, reliable metrics in social media. With each announcement Klout releases, I admit I find myself craving more. So, in the spirit of the holidays, I’ve cobbled together a “wish list” of features I’ve love to see from the good folks at Klout in 2011.

More Historical Data

Right now, Klout’s various charts show measurements up to 30 days in the past. As an avid student of analytics, the ability to delve further into the past and download that data for further analysis would be valuable for tracking trends and correlating against other communications and campaigns I work on.

I really enjoy the flexibility offered by Google Analytics, where you can show comparisons between two spans of time, send yourself automated reports in various formats, and slice the data in any number of ways. If Klout released a similar suite of data tools, they would make me, and I’m sure many other analytics geeks out there, very, very happy.

Comparison of Metrics

Klout provides a healthy array of charts that show trends and measurements of several supporting metrics, not just the Klout score itself. It would be great to see overlays of various metrics, such as my inbound/outbound message ratio laid atop my amplification score, to see how the various data points interact with and affect each other.

I realize this could stray into “correlation does not equal causation” territory, but we’re not talking “pirates vs. global warming” here. I don’t see the various data points being that off target to create egregiously false assumptions. That is, of course, provided people continue to do their homework.

Data Segmented by Channel

Originally, Klout was focused specifically on Twitter. Starting in October 2010, however, metrics from personal Facebook pages were added to the mix, and a beta for gathering LinkedIn data is in the works. If you had tied your Klout profile to Facebook, historical data was adjusted as of October 21, 2010, which resulted in often dramatic changes to your overall score.

A sound tenet of communications strategy is matching your message to the medium. It may not be effective or possible to broadcast, engage, or interact the same on Twitter as Facebook or LinkedIn. Therefore, I would expect overall influence to differ between channels.

While an overall Klout score (and its supporting metrics) is useful as a broad indicator, it would be immensely helpful to segment the data by channel so I can give more thought and consideration to how I communicate on that channel.

Real-time “Influenced By/Influencer Of” Updates

Klout offers a simple chart of whom you influence and who influences you. Up to five of each is shown in your profile. Currently, this “rogue’s gallery of influence” isn’t updated with any frequency; my set has been the same for many months, and others in my network have reported the same situation. Most the accounts shown in this chart are pretty obvious, but it remains a good insight into the cornerstones of your network. I’d love to see more “drift” here.

Transparent Data on Total Number of Profiles Indexed

Klout doesn’t automatically connect to every single Twitter or personal Facebook profile; in most cases, you need to create a profile to share your data and acquire a score. Why is this important? If you connect your Twitter account to Klout, your score is relative to all other Twitter accounts captured in Klout’s database, not every Twitter account in existence.

As with any study or poll, it’s necessary to know the total size of the data set in order to establish weights or bias on the resulting statistics. Currently, Klout doesn’t share the precise number of accounts they index. I’d like to see more specificity and transparency here.

I feel this is important for measuring and reporting on success in the social web. While I’m sure a healthy amount of influencer heavyweights are already ensconced in Klout’s data sets, knowing the total number of indexed profiles will help put the scores into more accurate and meaningful context.

Hub-and-Spoke Influence Diagram

This is more of a “pie in the sky” request, but it would be sweet to see a hub-and-spoke diagram of influencers. The current “influenced by/influencer of” chart allows you to click on a specific account to jump to their Klout profile, wherein you can see who influences them and who they influence in turn. I’d love to browse through a Flash- or AJAX-based hub-and-spoke diagram that could show me dynamically who connects to whom in the influence realm.

The Obvious Conclusion

The obvious conclusion about the features in this wish list is that Klout could set up a “freemium” model: continue to offer the current suite of metrics and charts at no cost, and then offer an extended array of features to monthly paying subscribers. Hootsuite did much the same recently with their social services. Such a model would allow Klout to continue to add to its user base and secure a source of revenue from dedicated users.

Key Elements for Measuring Your Social Media Campaigns

Now that social media channels like Twitter and Facebook have matured into solid communications platforms, many individuals, organizations, corporations, and agencies have taken to these channels to reach their audiences. What hasn’t quite solidified, however, is how these campaigns are tracked and measured. There’s still a decent amount of confusion and differing opinions of what you should measure and how.

In this post, I’ll share several factors, methods, and steps that I’ve learned this year about how you can effectively and realistically measure your social media campaigns. This isn’t intended to be a complete or exhaustive list of what you could use, but it’s my hope that each topic will assist you in your endeavors.

Define Your Goals and What Constitutes Success

Before you pursue any coordinated effort in your social media channels, think very carefully about your goals and what constitutes success for you. Remember, you cannot measure success (or anything at all, for that matter) if you don’t have defined goals up front or have an unclear vision of what defines a successful venture.

  • Expectations: Set the expectations of the campaign early and often. Make sure the goals and definitions of success can be precisely tracked, measured, and reported on. Build a communications strategy and make sure everyone who will be involved in the campaign has read and understood it completely. If you’re simply testing the waters of a specific social media channel or experimenting with a new approach, that’s perfectly fine, but make certain everyone understands that, especially those in charge.
  • Timeframe: Think about the timeframe of your efforts. What is the life expectancy of the campaign? Is it a short-term promotion, a webinar, conference, or event with specific dates, a brand awareness effort, or a marketing campaign to raise awareness of a cause or resource? Will you announce the effort ahead of time so users are prepared to take action, such as a “save the date” for events? How will you build interest and sustain momentum over the course of the campaign? What are your plans for retiring the campaign once it’s complete? Have you considered follow-up activities like a user satisfaction survey?
  • Transactions: Work out which transactions will define a successful campaign and stick with those decisions throughout the effort. How is the campaign intended to affect your transactions? Are you seeking to drive more traffic to a specific website, blog, or web-based application? Do you want a certain number of users to sign up for a promotion, event, or product trial? Do you want them to buy something? Think about “goal conversions”: the action(s) you want your users to make, and the end result(s) you want from them.
  • Return on Investment (ROI): ROI is a classic business metric, but it’s one that’s generated a lot of heated discussions when it relates to social media. Whenever you think of ROI, it should always boil down to money: hard dollars and cents (or euros, sterling, yen, etc.). Did your campaign generate enough revenue to justify its cost (marketing materials, agency fees, manufacturing costs, etc.)? Did you increase sales of a product or service? Did your transactions balance out the (estimated) hourly rate of the staff spent managing, tracking, and measuring the campaign? You don’t have to always sell something to determine ROI; in essence, you’re figuring out whether the campaign was worth the effort, but that worth has to be a financial measurement.

Look for Active Responses

Among the most valuable success metrics are direct responses from your audience. These can include mentions and direct messages on Twitter and wall posts and comments on Facebook. Be sure to check these regularly, especially if you intend to respond to comments or engage with your audience to keep up the momentum of your campaign.

If one of your transactions is new fans or followers, you may want to activate e-mail notifications to have “hard copies” of these actions. New follower notifications from Twitter, for example, not only show you basic information about the new follower, but also which platform or browser they used, the date and time of their activity, and the size of their audience.

Check How and How Often Users Are Sharing

To track how far your message has reached, look for evidence of sharing within your audience.

On Twitter, keep track of retweets, both the old- and new-style versions. “Old-style” retweets have the prefix “RT” before the original tweet along with the handle of the originating account. Users may choose to add their own comments in the retweet, and may edit or winnow down the original text to fit their comments. “New-style” retweets are simply a reposting of the original tweet in the user’s timeline; they can’t make any changes to the original text. Depending on which tool you’re using to measure retweets, you may need to look in different views. For example, Hootsuite will not show new-style retweets in a “Mentions” view; you need to track them in the “Your Tweets, Retweeted” view.

On Facebook, look for both “likes” and “shares”. Whenever a user likes a post, it increments a “thumbs up” value associated with the post; all fans of the page see this value. If the user chooses to share your post, the original post will appear in the user’s timeline. They can also choose to add a comment to prefix the shared post.

Likes and new-style retweets are what I consider “lazy successes”. It takes a user only one click to perform these actions, so it may not indicate that they read the complete post or whether that post was informative, helpful, or useful to them. Facebook comments and shares, along with old-style Twitter retweets with additional text, are more valuable for quantitative feedback.

Use Query Variables or URL Shorteners

Whether you’re sending out links in your tweets and Facebook posts, including website URLs in direct mailings, or displaying QR codes in your promotional materials, go the extra mile to insure that you can properly measure user activity.

Several URL shortening services, such as bit.ly or Hootsuite’s ow.ly, have back-end reporting tools you can use to track and analyze click-throughs. Hootsuite can also overlay tweet click-throughs with your Google Analytics reports to show possible relationships between your posts and website traffic.

Query variables, which are additional pieces of information you can include in a website URL, can provide you with richer metrics. Web analytics applications and services such as WebTrends, Google Analytics, and Radian6 can collect information from query variables whenever a user visits the URL. What’s nice about these variables is that you define them: add whatever information you’d like to collect, such as the campaign’s name, a specific marketing code, or which channel you’re sending the URL out to. You can then work within your chosen analytics platform to track, monitor, and report on the variables.

Never Rely on a Single Metric

When it comes time to gather, analyze, and report on the results of your campaign, never base your success or failure on a single metric, such as followers or fans. Always use multiple measurements and track trends in their activity over time. This allows you to form the complete picture of your campaign, and makes you better informed about what contributed to the results.

Crossing the Finish Line

Building realistic expectations, setting your timeframe, defining your transactions, and understanding how ROI will relate to your campaign will all help immensely when it comes time to measure your efforts. Be diligent in checking (and responding to) active responses from your audience and how they’re spreading the word about your campaign. Give your measurements a boost by considering URL shorteners and query variables to supply additional information about your users. And, finally, don’t hang the success (or failure) of your efforts on a single value or metric.

As I mentioned at the start of this post, I don’t consider the thoughts I’ve gathered here as the “be-all-and-end-all” compendium of social media measurements. If there are specific methods or insights that have worked well for you, please feel free to share them in the comments.

I want to thank the good folks at Marketwire and the #smmeasure chats for helping to inspire this post. Their weekly social media measurement Twitter chats have been quite valuable in my own social media efforts, especially since metrics in this space are still not clearly defined or universal. The #smmeasure chats take place each Thursday at 12 noon Eastern Standard Time. Follow the @smmeasure Twitter account or the #smmeasure hashtag to participate.

Backing Away From Milestones

One of the basic key metrics of Twitter is the number of followers you have. It’s generally interpreted to indicate how much reach you have in your messages, and folks often like to announce when they’ve reached a certain milestone (“We now have over 1000 followers! Thanks for your support!”). I’ve been on the verge of several “milestones” with my account (100, then 150, now 200), but I keep backing away from them each time. Why? Because I value quality much more than quantity.

In your travels on Twitter, you’ll snag followers for several reasons. You have great content to share. You make a connection, and they pass you along to their followers. You get recommended. You get followed because of specific keywords in your tweets. Spammers really like you. Oh, and don’t forget the porn. The numbers start to add up. Before long, if you’re lucky, you’ll reach a “milestone”.

But what does that really mean to you? Is each and every one of those accounts actually listening to you? How many do you regularly engage with? Once you take a more critical eye to your roster, you get a picture of how your online relationships are shaped. And a good percentage of them are garbage.

I regularly prune my followers. Irrelevant keyword matches? Blocked. Obvious product promotions? Blocked. Offers to make me money or grant me more followers? Blocked. Dead or inactive accounts? Blocked.

It’s not that I don’t want these accounts to follow me; I could care less about which of my posts they read. It’s simply this: I want to know that I’m being followed for a reason. That my words and thoughts are being heard by folks who actively, purposely decided to follow me. This way, when I look over the numbers, I can truly see how much of an effect I’m having in my small corner of the Twitterverse. Call it narcissism or whatever you want; I want quality, and I’m going to get it.

© 2020 Bright Matrices

Theme by Anders NorenUp ↑