Why I like the SIM Score

Riley, Emily If you are like most marketers I talk to, you are probably trying to find a reliable way to measure the effects of your social marketing efforts. One of the tactics I typically recommend is to use a listening platform to understand and benchmark the scope and quality of conversations happening around your brand. In Razorfish's recent publication "Fluent"they along with TNS Cymfony introduce a formula that really captures the value of listening platforms for measuring social media. They call it a "SocialInfluence Measurement" or SIM score, and it works like this:

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

A few people have compared it to a Net Promoter Score, but I think it goes one step further in that it compares your brand's performance to that of the industry.

Not only that, but elements of the formula give you other cool insights. For example, the result from line 2 can be divided by your total brand conversations to give you a "net sentiment" score. You can do the same thing with line 4 and divide that number by the total industry conversations. If your net sentiment score is lower than the industry average, your competitors have better positive sentiment across social media than you do.

You can recalculate this score over and over as you try new forms of social marketing. While the score will certainly reflect larger efforts, such as big TV campaigns or news stories, you can typically correlate the effects by mapping your scores on a time line and plotting all of the different campaigns and news stories to see which things affected your score the most.

If you don't have a listening platform, ask your interactive agency about it, and ask them to develop a SIM score for you and your industry.


re: Why I like the SIM Score

Interesting post, it's something i've been thinking about a lot and the sim score seems like a good way to measure progress. When your talking about the 'listening platform' do you have any examples,Great post here about reputation management dashboard.But I can't help feeling there's a better way to do it. Any ideas?

re: Why I like the SIM Score

Hey Emily'ask your interactive agency about it'?Surely that should read 'ask your social media agency about it'?;)

re: Why I like the SIM Score

We'd caution marketers to be wary of overly obsessing on "sentiment scoring" when it comes to your brand's social marketing strategy. It's important, but with deep and abiding caveats...Within our social marketing listening and activation programs we've worked with almost all available techniques and methodologies for assigning sentiment to collected and analyzed conversational data. It is tricky and elusive and, well, massively inexact.Most social data aggregators (like TNS, who is great, along with many of their stellar competitors like Visible Technologies, Radian6 et al) usually apply sentiment scores using a mix of people and programs. Unless a live analyst (fluent btw in both the local language and vernacular) is making the decision about whether a post to a blog or a tweet or blip is positive, negative or neither/neutral, you should be very suspicious of the accuracy - and more importantly, actionability - of any "score".And it gets expensive quickly asking a live analyst to manually read, assess and score hundreds of thousands of such posts or tweets. So we all rely on technology to help, and it's a great but not perfect science.You can do like we do and create language cartridges specific to the product domain or category and that helps. Or have our social media analysts scan samplings of the machine scored outputs of our data aggregator partners and fine-tune the "sentiment scores" from this.But here's a little exercise to illustrate the vagaries of sentiment scoring social conversational media... try it yourself and see.Your brand is Alpha Man hair color and your biggest competitor GrayOut is gaining market share. You stumble across this gem on a Gawker comment to a relevant article on celebrity male vanity.“I gotta tell you, I’m almost 40 and have been using GrayOut for over 10 years. I’ve loved it, and no one knows I’m cheating , or so I thought. Last week some creep in the gym sees my bottle and makes a stupid crack about how GrayOut sucks and he’s an Alpha Man man. I look at him like he’s an ignorant neanderthal and then go pick up a bottle of (stupidly expensive btw) Alpha Man. I used it for a week, and am totally sold. Dude, I thought GrayOut was awesome til this dweeb turned me on to Alpha Man. According to my wife it looks pretty freaking natural. Christ, I sound like some shill. But it works, so there.”Ok – what’s the score? Who wins this battle of the sentiments? See how certain hot words seem to be modifying one brand in shifting sentiments across the narrative of one man’s tale of hair coloring salvation?So, sure, baseline sentiment, using the most robust methods you can afford and stick with that method so that, at the very least, there is some directional value in measuring and assessing the trends over time.But if you really want to invest in a metric that means something when it comes to analyzing your social media, focus on “actions from social”, e.g. sales, downloads, shares, trials, activations, usage, etc. The place where your brand’s social meets your commerce is where you need to be scoring - and optimizing - your social marketing programs accordingly.

re: Why I like the SIM Score

I love the idea of measuring Social Media Influence, but there are 3 questions I have:1. Wouldn't a true measure of social media influence include a variable governing the influence of the conversation? (i.e. if CNN posts something positive on Nissan, and I tweet something negative should that really = 0?)2. Does the score care about the number of people saying good or bad things, or just whether they're saying good or bad things as compared to your industry?(i.e. If no one is talking about Nissan except 5 people, and they're all saying positive or neutral things, the result from line 2 would = 1.0; and if there are 1000 people talking about cars, and 600 are saying positive or neutral things, the result from line 4 would = 0.2 and SIM = 5)3. What happens if the industry is unpopular, but your brand is popular? (say the banking industry)(i.e. result from line 2= 0.5 and the result from line 4= -0.25; would the result be SIM = -2?)If anyone has an explanation, I just want to be able to speak to them when talking about Social Media Influence.

re: Why I like the SIM Score

I am curious; if it is Social "Media Influence" why is the acronym SIM why not SMI?

re: Why I like the SIM Score

Interesting measure for social sentiment. This may, however, be a poor measure for brands with a passionate group of believers and non-believers. If brand X has a score of 50 positives (many of whom are rabid brand loyalists) and 50 negatives (many of whom hate it, perhaps.) Versus brand Y with 10 positive, 50 neutral and 40 negative. Brand Y would "win" in the SIM game, but in the end wouldn't you rather be in brand X's position?Obviously, measurement of brand value, sentiment, and ROI in social media is extremely difficult to address and this is a start, albeit perhaps a flawed one.

re: Why I like the SIM Score

I find it an interesting -but- incomplete measurement.SIM stands for Social Influence Measurement. But social influence or influence isn't measured only by the sentiment of the content.Influence is more, it's also who says it, what their reach is, what their influence is etc.The total picture consists of communication (SIM) and the sender and receiver(s).