What Should Twitter Analytics Look Like?

It was revealed this week that a choice group has been invited to test Twitter Analytics. Expected to debut at the end of this year, it should provide tons of insight, such as which tweets are most successful, Mashable reports.

We asked a few PR measurement gurus which features they think Twitter Analytics should include and how it will contribute to the ongoing efforts to measure social media initiatives. After the jump, their responses, received via e-mail.

Katie Paine, CEO of KDPaine & Partners: Like any system that only measures your own site – i.e. Google Analytics – it’s a great step in the right direction. With luck it will show that people who use Twitter to spam lose followers and don’t get retweeted.  What it needs is the best features of Google Analytics and Twitalyzer (the ability to benchmark against your peers or competition), the ability to produce ongoing reports automatically, and the ability to track the outcomes – in other words how many people actually click on the links you post.

I look at Twitter engagement as a continuum. At the most basic level, you may be following someone but never responding. A second level of engagement is when you actually start responding and possibly retweeting. The third level of engagement is direct messaging and using your hashtags and clicking thru to your links.  Ultimately you want to be able to connect all those different levels to actual actions.

Sean Gelles, research director, digital and social media measurement, Ketchum: Twitter Analytics should provide (in order of importance from most to least important):

1.     Analytics covering historical data spanning the entire life of the handle
2.     Aggregate retweets per tweet – this is a no-brainer but far from easy to obtain from third-parties
3.     Aggregate impressions per tweet based on actual impressions and not just followers
4.     An influencer score that is based on more than just how many followers a handle has – one that takes into account variables such as: How many followers the handle’s followers have; How often the handle’s tweets are retweeted (retweets per tweet); And impressions per tweet (based on actual impressions and not just followers)
5.     Aggregate hashtag mentions for all hashtags generated by the handle
6.     Click thrus for hyperlinks shared on the handle
7.     Visualization of handle’s Twitter network

Mark Weiner, CEO North America, Prime Research: The element most often missing from twitter monitoring and evaluation services is the “human element.” Most services rely exclusively on automation which is great for speed but not good for accuracy, context and assistance in decision-making.

The assumption by many is that “tools” are “solutions:” if this were true, everyone with Microsoft Word would be Tolstoy. What we need to retain (and, in some cases, reintroduce) is the notion that humans offer a unique form of brilliance by which organizations shape business results.

Jonny Bentwood, one of the creators of Edelman’s Twitter measurement tool TweetLevel, spoke with PRNewser: There’s quite a few misconceptions within Twitter analytics that popularity equals influence. This couldn’t be further from the truth. Influence is gauged more importantly by how engaged individuals are with their communities.

Any analytics tool needs to understand engagement and trust and the type of content posted, whether its relevant, informative, and interesting. But perhaps most significantly, people shouldn’t just focus on one channel. Our objective is to create tools to help identify those people that are important within any microtopic, within any communications channel, within any location, on mobile, online, wherever.