Over the weekend, something very exciting happened at Recommend.ly
Conversation Score (http://csco.re) moved out of the excel sheets and white boards, to become a web prototype for the first time on Friday. Raj Dsouza thought it would be nice to take Mari Smith’s views on the alpha version and asked her if she could get more testers. She obliged and posted her CScore card on that hugely popular Facebook Page of hers.
Lo and behold! Here is a screenshot of Google Analytics for the 5 day old CScore tool:
Goes without saying that we are overwhelmed with the immense response and words like ‘nifty’, ‘cool’, ‘accurate’, ‘interesting’ being used by Social Media enthusiasts like Mari Smith, Pam Moore, Michael Q Todd, Kimberley Castleberry etc. Initially, we tried telling everyone that we are still in Alpha, so don’t expect too much right now. But it went viral already even before we made a feature freeze! So we doubled down on the development schedule and are trying to put in all the features by next week and make it look like Beta at least. Why let a good thing fizzle out?
Among the various queries we got, including from Scott Ayres (who wrote a cool review comparing CScore to Edgerankchecker), the biggest one was – ‘How do we interpret CScore?’ Without further ado, here it goes:
THE GUIDING PHILOSOPHY OF CScore
Popularity isn’t the same as success on Facebook. Quality interaction is what separates a great Facebook Page from a popular one. Recommend.ly believes that a Page’s success can be measured on four fundamental parameters:
1. How often does the Page start conversations by making updates, posting media or asking questions? (“Broadcasting“)
2. How many self-posts and fan-posts does the Page participate in, adding value to the conversation? (“Response to Fans“)
3. How viral is the Page’s content? This measures not just quantity but also quality of fan response (“Virality“)
4. How popular is the Page with Facebook users, in terms of fan count and active fan ratio? (“Fan Love“) Yes, fan count is aslo taken into account for CScore, but gets a lower weight.
We tried to keep it simple and make CScore cards self-explanatory. Also, we didn’t want to keep the algorithm shrouded, like some of our (hugely popular) competitors do. The only thing we don’t reveal is the weights. That’s a trade secret! We believe that any scoring system should simplify the measurement of your Page’s strengths and weaknesses, and help you correct course quickly. CScore isn’t a confounding metric that complicates a Page owner’s life. It is probably the easiest way to know if you are on the right track.
What does CScore study?
We study the recent activity on a Page including posting habits, likes, comments, shares etc – as much data as publicly and legally accessible. Facebook limits the quantity of data you can publicly access, so the time period of study varies based on Page activity.
What’s with the name tags?
That brings us to those popular name tags (Viral Wonder, Conversationalist etc) we are giving along with the CScore. There’s a lot of excitement around the tags and aren’t we glad! You earn a cool tag, based on your overall score and the combination of the four parameters. We envisaged 81 different grades for Pages based on the levels of performance on all the four parameters. For the Alpha version, we consolidated them all into just 5 name tags. Rest of the tags will be active once all the grades go live next week.
What do the name tags mean?
The name tags describe your Page’s behaviour. There is no ‘good’ or ‘bad’ here. Even a Page with a 0 score gets a ‘Starter’ tag. Not a ‘Geez-you-are-bad-at-this’ tag. Also, there is no hierarchy in names. You can get a score of 90 and be a Broadcaster if you are big on posting and your friend who scored 90 too, is a Fan Favorite with 160% PTAT and a gazillion fans. Different Pages have different objectives and behave differently. Here’s an example for how to interpret the name tags:
News networks tend to be Broadcasters more than anything else. That’s because they target real time dissemination of news on social media. Entertainers like Eminem target fan loyalty and probably virality. They don’t tend to make 15 posts a day.
If NYTimes were a Fan Favorite and Eminem were a Broadcaster… umm, they need help managing their Pages! Likewise, most Social Media experts are Conversationalists and Viral Wonders. If you are a marketing consultant and you end up being a Broadcaster, may be you are trying too hard? You should hold back on posting and try to generate interesting content and participate more in conversations.
Once we have the full library of 81 grades and the associated name tags, the difference between a low scoring Broadcaster and a high scoring one becomes much more stark. That’s what CScore attempts to do – measure precise performance on all parameters and help you align your goals with activity.