Measuring social media ROI (But what does it mean?)

 Social media return on investment (ROI) is tough to quantify. Recently a very good friend and VP of finance for a popular local company quipped “if you don’t manage it (social media) it will manage you.” True enough, but how can you measure the success of that management? Social media is about brand sentiment, perceptions and the feeling of connection to a company. I’m not saying that participating in social media doesn’t directly affect sales (because it does) but like many other marketing efforts, often in a way that’s difficult to measure.

Still, there are plenty of tools that claim to measure sentiment across multiple platforms, and I wanted to see if any of them truly performed as advertised.

 Tweeteffect.com

This one uses data from Twitter to measure the effectiveness of your tweets – that is which of your tweets cause people to follow or unfollow you. I’ve tried this one for months and I found some very big irregularities in the measurement. For example, I’ve noticed a pattern where the same four followers were lost, gained, and lost within a few tweets, which probably had more to do with the unreliability of  Twitter’s API than the content of my tweet. So, while interesting, not particularly useful.

 Klout.com

But what does it mean?

Calling themselves the “standard for influence” for social media, this tool held promise for me. For months I regularly measured my Klout influence and tried not to obsess over the score. It wasn’t long before I noticed a big problem. While my Twitter followers were steadily increasing along with the number of lists on which I’m included, those same numbers on my Klout score didn’t change. In fact, the Twitter followers were off by more than 300 and the list by at least 25. While Klout has recently modified their algorithm to measure sentiment across both Facebook and Twitter, the fact that the followers and lists are still reported incorrectly is a big red flag to me.

 Twittergrader.com (and it’s counterpart Facebook.grader.com)

 From Hubspot comes Twittergrader and now, Facebookgrader. While my Klout score remains in the 50s, this tool has me at a 95 out of 100. My ego tempts me to say this one must be accurate! But, like the double rainbow viral video of 2010– “What does it mean?” I experimented further with Facebook Grader by attempting to measure a non-profit page and found that while it ranked 47,827 out of 57,625, it placed it at a 15 out of 100. Hmm.  At least the follower and list numbers are accurate, but without knowing how these tools measure sentiment, how can we use it to strategize?

 Twitalyzer

I used this one to measure several of my twitter accounts. Each one gave me an “impact” score (not sure what it means) and the following message under “network in Twitter”–

Something looks funny to us! Since very few people have “no Network” in Twitter, this list being empty is suspicious.

Undeterred I decided to try out their detailed reports, which require authentication with Google. No problem. When granting access through my Google account it returned this message:

Something has gone completely awry! When you see this page something has gone really, really wrong.

I tried this with several accounts and received the same error. Ok – not particularly useful.

So, since I’m not having much luck with these free analytic services, are there any that you recommend? In the meantime, how are you measuring your effectiveness?