I’ve been following an interesting debate over on No Man’s Blog about whether social media monitoring services are all their cracked up to be. In the original post Asi Sharabi launches a visceral assault on social media monitoring services, citing Radian6, BuzzMetrics, BrandWatch and Techrigy, among his targets. His key accusations are:
1. The technology is fairly stupid.
2. The services produce unreliable data.
3. Sentiment analysis is flawed.
4. The don’t offer regional-specific data.
5. Influence analysis is flawed.
6. Visualisation tools are misleading (due to data).
7. The process is time consuming.
8. The services are expensive and not value for money.
The discussion has generally flowed with Asi’s criticisms. Many of the commentors, including staff and founders from several leading SMM service providers, concede that the industry is still relatively new, that it’s grappling with difficult issues: how to gather filter data from the masses of spam and nonsense on the web; how to gauge sentiment accurately without using expensive humans in real-time; working out what “influence” means in different contexts; plus the normal troublesome issues web services have – like getting the usability, price and customer support right.
Having collated the data within this post though, identified the influencers and gauged the overall sentiment – there’s a definite feeling that social media monitoring services to date have been oversold. I’ve certainly seen promises of “80% accuracy for sentiment detection” and “we’ll identify the most influential people in your industry”. Claims which, to anyone lacking a world-weary bullshit detector, must seem fantastic. I can see hard-pressed PR Managers around the globe signing off on $1000/month just to bathe in the warm glow of that promise.
Anyway, Justin Kirstner of WebTrends has just boldly struck back for the social media monitoring services, highlighting the novelty of in industry in which many services are still on version 1.0 and criticising Asi as an “armchair critic” an a Johnny-come-lately to the discussion. I can see this one going into extra time. Well worth a read through the comments if you’ve got a spare 15 mins.
P.s An Invitation: If you’d like to take part in a real-life debate about social media monitoring services, meet the vendors and hear case studies, I’m running a conference in London late in autumn 09. I’ll be publicising this in Sept, but you can email me for more info: luke(at)oursocialtimes.com