I just finished hearing Gary Angel, CEO and CTO of Semphonic and Scott Wilder, GM – Online Communities / Social Media at Small Business Division Intuit give a webinar entitled How do you Measure Social Media ROI? If you didn’t hear/see it, you
will be able to can find it at Slideshare and talk about it on Gary Angel’s blog.
As I am in the process of putting the finishing touches on my next book, Social Media Metrics: How to Measure and Optimize Your Marketing Investment, I was very interested to see what they had to say.
First of all, it was great to have that balance of consultant and practitioner. Gary can tell you what happens in most companies and Scott was able to provide some hands-on experience and examples.
They spent a good chunk of time on how to prepare your organization for social media / social media metrics. It’s all about governance.
. . Who owns social media?
. . How do you set the proper level of expectations?
. . How do you identify metrics?
. . How do you identify goals?
They talked about Reach and Conversion of course but here are the tidbits that stuck with me…
For one large client, Gary fond conclusively that website visits from social communities were less successful than the average visitor. Success in terms of how much product information they looked at, how many inquiries they made and, of course, conversion. HOWEVER – this was only true for the visit, not the visitor. Visitors from social sources were more successful over time.
Once you start tracking your visitors on a longer term basis, you find that those who have engaged with you via social communities are more interested, more responsive and more likely to buy. Seems sort of intuitive doesn’t it? But we’re so focused on immediate results per campaign that we forget to watch the long-term effects.
Gary and Scott also spent time on using social media for customer service – tracking success:
They threw in a great rule of thumb re: expectations. Don’t expect more than 5% of your customers to sign up for a community that you create yourself. To that, I add, don’t expect more than 5% of those to participate beyond lurking. Scott was quick to point out that if you have a small customer base, your own community just doesn’t make sense.
Gary hit me with something that hadn’t occurred to me. He suggested that you find a community that you are NOT targeting in order to use it as a control group. I typed in a question about how that’s done, but we ran out of time. I’m hoping to prod him into answering here….
Quote of the Day: Don’t send a machine to do a man’s job.
This is not database marketing. This is something you can easily over-automate and run into trouble. Autoresponders, auto-following and friending and re-tweeting will bite you eventually. Be careful.
Second Quote of the Day (or would have been had either of them said it just this way): Understanding trumps counting
What they said was – Don’t forget to Listen! It’s more important to understand what people are they actually saying than to count up the number of times they say something.
That lead Scott to express my deeply felt opinion that sentiment analysis just ain’t there yet. If you’re promoting a sentiment analysis capability – that provides trustworthy and actionable information, I’d love to hear from you.
Third Quote of the Day: Measure as much as you can but use as few metrics as possible to convey your message,
Finally, this purloined Model of Engagement that they got from Ross Mayfield and I got from them:
Great stuff. Thanks Gary and Scott!