Paul M. Banas on Consumer Insights, Marketing Research, and the Digital Media Landscape
Random header image... Refresh for more!

A Better Way To Measure Social Media Marketing?

While Social Media Marketing has seen explosive growth, the tools and metrics necessary to measure the effectiveness of this new medium haven’t seen quite as much innovation.

In fact, the tools I’ve seen discussed are a lot of the same tools that marketers in Search have had success with.

Unfortunately, these tools generally lack the ability to capture the viral nature of programs and campaigns in Social Media.

The chart below is from a study by the Society for New Communications, entitled “New Media, New Influencers, and Implications for the Public Relations Profession” which asked a couple hundred industry professionals a series of questions about Social Media.

This is in response to a question about what Social Media metrics they found most important:

What struck me is that things like search rankings and visitor tracking are the only tools that are seen as being very effective. And I don’t think it is because they truly reflect the nuances of Social Media, it’s just that they are what people are used to measuring when it comes to digital media.

Jeremiah Owyang pointed out a potential measure on his blog that may be more appropriate, something he calls Velocity.

“Velocity, when applied to Social Media, is the measurement of how fast an idea, embed, widget, or other like media spreads over web properties. Benchmarked over time, acceleration and deceleration indicate relevancy.”

He also gives the following example of Velocity in use:

  • Week One: A widget was installed on 5,000 Facebook profiles within 7 days, resulting in a weekly velocity of 714.
  • Week Two: A widget was installed on 15,000 Facebook profiles within 7 days, resulting in a weekly velocity of 2142.
  • You can then look at this over time and benchmark, and then look for accelerations and decelerations, in this case, week two accelerated from week one by 300%.

While the measure itself is very interesting, the challenge for most market researchers trying to work with this new measure will be capturing all the data necessary for analysis in an automated way.

However, at least it is a step towards measuring the concept of relevancy with a measure that is itself relevant.

10 comments

1 Nick { 04.15.08 at 1:08 am }

Velocity is a good measurement for some channels in social media. It seems though that it wont work for all content. I’ve seen time and time again where content peaks, disappears and peaks again. This might only occurs when the initial peak isn’t so high. Perhaps instead of only measuring the sprint, we should also be measuring the marathon.

2 Tad Chef { 04.15.08 at 2:43 am }

Yeah! That’s it. Velocity is on point. The only drawback is that it’s dependable on chance and the users who submit stuff. So a low velocity does not mean a certain product is bad. Velocity needs to be compared with contentiousness (like in “viral”) of the submission. A post submitted by a weak user which still has a velocity of, say, 100 would have to be divided by the contentiousness of the submitter/s.
So if top digger Mr Babyman submits it to Digg and it gets on the frontpage with 300 diggs as he has a very high c it would be low end result: v/c = low whereas as noob user who gets it to the frontpage with 3000 diggs would result in a high measured virality.

3 Paul M. Banas { 04.15.08 at 3:36 am }

@Nick
I like your point about measuring the sprint versus the marathon. This is why overall velocities from sites such as Digg may be different than those from sites like Stumbleupon, if you look at the first few days of traffic. Over time however, Stumbleupon will still probably be showing some velocity, although smaller than the initial surge, while Digg will be zero.

@Tad Chef
Interesting you should say that, since that is exactly how I was thinking velocity should be used. I’ve been following some of your posts on the diminishing returns of Stumbleupon traffic and I was wondering what metric would help provide insight in that area.

Another way to look at your question would be to hold time constant (let’s say 7 days) and then calculate an initial velocity per post for Stumbleupon: unique visits over 7 days/# of stumblers. You can then analyze the initial velocity of posts submitted to Stumbleupon by new users versus top stumblers. Or how does velocity compare when a post is submitted by yourself or someone who has submitted you before, versus the velocity of posts submitted by someone outside of your inner circle of readers.

4 Alex Nesbitt { 04.15.08 at 2:15 pm }

The problem I see with measures being relevant is the lack of context for why any particular data point is tracked. Velocity might be important for some objectives, but not others. Search engine ranking might be good, but if it’s for the wrong keywords then maybe it’s not so good.

We are swimming in data, but lack insight.

We need to start with a framework that starts with top level objectives and drives down to specific measures. Without such a framework, it is impossible to evaluate the usefulness of any specific measure. With a strong framework, you can see what measures matter and which ones have the most leverage.

5 Paul M. Banas { 04.16.08 at 6:05 am }

@Alex
I couldn’t agree more with starting with specific objectives. I’ve always thought that doing research or marketing without a clear objective is like driving in a foreign country without a map or directions. In those cases, it really doesn’t matter what kind of car your driving.

Thanks for the good points.

6 Don't Think Traffic – Think Demographics | Social Media Trader { 04.17.08 at 6:02 am }

[...] could already be vast amounts of market research to use depending on the topic of your site. Don’t rely on just one or two studies for your [...]

7 AndersonAnalytics { 05.04.08 at 8:46 am }

There will soon be more powerful tools for market research on Facebook and LI, see: http://www.tomhcanderson.com/2008/05/04/facebook-and-now-linked-enter-the-%e2%80%9cmarket-research%e2%80%9d-game/

8 Paul M. Banas { 05.05.08 at 8:27 pm }

@Anderson Analytics
Tom,
Thanks for visiting. Like several of the commenters on your post, I’m also intrigued by the BtoB research possibilities of LinkedIn.

Depending on the question and the sample base, Facebook may also be interesting as well. But I think LinkedIn probably has the most immediate utility.

9 links for 2008-05-24 { 05.24.08 at 5:32 pm }

[...] A Better Way To Measure Social Media Marketing? | Insight Buzz A view on Social Media measurement. (tags: Marketing metrics socialmedia) [...]

10 Topics covered in 6/30/10 workshop: How to know your demographic « INDIE Marketing Genius { 06.28.10 at 3:48 pm }

[...] could already be vast amounts of market research to use depending on the topic of your site. Don’t rely on just one or two studies for your [...]

Leave a Comment