Social Media Influence Score - What do you think?

We all are aware that social media is now a very important part of internet marketing. But what about measuring it? I found this on the Forrester blog - http://blogs.forrester.com/marketing/2009/07/why-i-like-the-sim-score.html

1. Calculate the number of positive, negative, and neutral conversations happening for your brand

2. Use the formula (Positive + Neutral - Negative)/ Total brand conversations

3. Calculate the number of positive, negative, and neutral conversations happening for your industry

4. Use the formula (Positive + Neutral - Negative)/ Total industry conversations

5. Divide lines 2/4 and you have your SIM score

What do you think?

The only part I do not agree with is adding the Neutral comments to the positive nor would I want them added to the negative number. I think the Neutrals have to be viewed as a plus/minus swing statistic.

I would suggest the following

Your Brand (Positive - negative) divided by your industry (positive - negative).

then

Your brand Neutral divided by your industry neutral

SIM score is derived from the first equation and the plus/minus swing is derived from the second equation.

I think this would be a little more accurate as neutrals represent people who have not made a firm decision and in time they could break either way. Sort of like when you hear political poll numbers where they say that candidate A has a 47-43 lead over candidate B with 10% undecided. Make sense?

I agree about the neutral ones. Adding the neutral with the positive and negatives may somewhat skew the data. For example, let say I have a blog commentary. I received a total of 10 comments. 3 were good, 3 were bad, and 4 were neutral (like self promoting links, spam, and non relevant comments/links). Thus, using the above metrics, it seems I have lots of positive, but not really since the 4 neutral should not even be calculated.

I agree about the neutral ones. Adding the neutral with the positive and negatives may somewhat skew the data. For example, let say I have a blog commentary. I received a total of 10 comments. 3 were good, 3 were bad, and 4 were neutral (like self promoting links, spam, and non relevant comments/links). Thus, using the above metrics, it seems I have lots of positive, but not really since the 4 neutral should not even be calculated.

Neutrals are the blogospheres version of undecided voters but with less importance to the product they are undecided about.

On one of my blogs, I constantly get the same comment over 100 times (Im not exaggerating) called "Hello World". Thus, if I approved this comment my neutral score will all be influenced by this.

On one of my blogs, I constantly get the same comment over 100 times (Im not exaggerating) called "Hello World". Thus, if I approved this comment my neutral score will all be influenced by this.

What you mention regarding spam is going to be true for positive, negative and neutral. This is where a filter needs to be used to prevent scores from being skewed. And until a solid way to take this into account is designed Social Media Influence Scores will carry the same weight as voting for American Idol or the MLB All-Star game vote.

Well, this sounds interesting to me....I had never thought about something like this...now when I think of it I think everything should be measured like this.

Well, this sounds interesting to me....I had never thought about something like this...now when I think of it I think everything should be measured like this.

You can desire everything to be measured like this as long as you know how easy it is to skew results. I used to work for a market research firm where we gathered shopping info from 4 targets and extrapolated out from their to advice clients as to the sales activity at all Targets. The numbers were not going to be accurate for each store but they allowed my company to fullfil their contract. Basically, ask a question the right way, narrow your target audience down or open it really wide, etc, and you can get any results you want.

You can desire everything to be measured like this as long as you know how easy it is to skew results. I used to work for a market research firm where we gathered shopping info from 4 targets and extrapolated out from their to advice clients as to the sales activity at all Targets. The numbers were not going to be accurate for each store but they allowed my company to fullfil their contract. Basically, ask a question the right way, narrow your target audience down or open it really wide, etc, and you can get any results you want.

Thanks for the nice information....I think this helps in making the good changes which will be in favor of consumers or customers.

Like Marketing Rob, I also come from the market research world and I have an analogy comparing research with journalism, "You gotta ask the right questions to get the right answers."

social media is becoming powerful than ever before. its the growing demand that is exactly making the social networking sites more propelling. especially, dating sites are doing enough good business.

date girls online

Like Marketing Rob, I also come from the market research world and I have an analogy comparing research with journalism, "You gotta ask the right questions to get the right answers."

And the flip side to that is "If you aren't getting the answers you wanted, tweak the question so that you will."

I agree, but you may have to tweet the same question a few times over since the half life of a tweet could be seconds/minutes only.

I agree, but you may have to tweet the same question a few times over since the half life of a tweet could be seconds/minutes only.

What are your feelings on using a survey service online? One of my clients uses the survey option in constant contact.

Oh I have used a few survey tools. I have used survey monkey and survey gizmo. I used of constant contact but I dont think you can white label it which was a factor.

Oh I have used a few survey tools. I have used survey monkey and survey gizmo. I used of constant contact but I dont think you can white label it which was a factor.

How is survey monkey as compared to constant contact? I am looking for alternatives to recommend to one client