This is a quick round-up of the key points raised at Monitoring Social Media Boston on 5th Oct, the presentations for which are mostly here. I’ll be following it up with a review of the afternoon sessions:
Katie Delayhaye Paine kicked off the day with some fantastic examples of how social media can provide ROI. Apparently Sodexo cut their entire recruitment budget ($300k) by recruiting via Twitter, and USO – a charity that supports American service men and women – saw revenues rocket after recruiting a Community Manager who engages with supporters via social media. She also reiterated her belief that AVE (ad value equivalency) the traditional method of measuring PR value, has been made redundant by social media. Engagement is the new currency and measuring it require creating an Engagement Index (in Katie’s case, a Kick-ass Engagement Index).
Amber Naslund (@ambercadabra) of Radian6 then took to the stage and said that The American Red Cross listens to it’s users and changes it’s website every day. This is part of a trend for organisations to use monitoring to reflect the actions and interests of their users. Amber likened this, neatly I thought, to humans “mirroring the body language of a suitor”. Ever prepared to provide actionable advice, Amber also said that if you want to outdo your competitors, you should create searches for “really wish [product] had [x or y]”. These expressions appear frequently and are obvious feature requests. Similarly if you’re responding to a complaint online you’ve got around 10 mins to act on Twitter before it’s too late. With a blog post you have 1 hour. Kinda scary really.
Keith-Woods Holder of Glide Technologies, a UK-based monitoring tool provider, pointed out that the key issue with sentiment analysis is perspective. Good news for you is bad news for someone. So accurate sentiment analysis requires you to gather feedback and teach the system about your perspective. Because of this he said that if you are serious about monitoring, you shouldn’t get a “black-box” solution that doesn’t allow you to tinker with it’s sentiment settings.
Stephen DiMarco of Compete then presented a case study showing how AMEX increased their page impressions from 50k per month to peaks of 1.5 million per month by launching their community site Open Forum. He also explained how Best Buy have replicated their website within Facebook. Now 75% of the people visiting their Facebook page also visit their website within 1 month and 4.5% of Best Buy’s traffic comes from Facebook. That’s pretty impressive. Later during an example involving Old Spice he also invented the word “scentiment”. Less impressive 😉
The irrepressible Mark Schmulen of Constant Contact then leapt onto the stage to explain how small businesses should be taking advantage of social media – quoting from Tony Hsieh’s book “Delivering Happiness” about how successful modern companies should becompeting to make their customers happy. (Having recently travelled on RyanAir, that’s a refreshing concept!) He then presented a case study of Dingo (a dog food company) which offered people a $20 to “like” them on Facebook and sign up to their newsletter. They increased their Facebook “likes” from 330 to 6,000 and newsletter sign-ups from 9000 to 14000. Wow! Especially since they did it in 3 days. Double wow! Sales rose 22%. Similarly, Glamour Nails, a local salon in the US publicised a “secret password” on Twitter and Facebook, offering discounts for people who use them in the shop. Another neat trick that increased profits.
Seth Grimes, the New York-based analyst and consultant then gave a detailed review of the pitfalls and opportunities offered by sentiment detection and analysis. He explained how Google is already using automated sentiment analysis in it’s star-ratings (taken from reviews written by web users) and how sentiment analysis couldn’t account for wrong facts – of which the Internet has many. We then finished the morning session with a discussion of monitoring tools – of which the upshot was: the tool you choose is only as good as the person who’s going to use it.