Really interesting article from Eric Pratum rounding up social ROI opinions from several industry experts. Could have used more concrete examples IMHO, but good stuff nonetheless. For many still, social ROI is little more than voodoo. And I think one of the best ways to demystify it is to share experience with peers. I don’t claim to know more than the next social strategist about ROI, but I’m sure happy to share what we’ve done with ROI at my work place to gauge our SocialService@Scale efforts. At Autodesk we measure against several “center of mass” KPI across the entire support and service division. Social is no different. These KPIs and associated desired outcomes are as follows: 1:1 contacts -> 0 TTA -> 0 NPS -> 100 EE -> 100 These outcomes are of course desired trends rather than hard targets. The idea is to reduce 1:1 contacts by disseminating and amplifying 1:N content via self-help (scaling goal). To increase response and answer time velocity (efficiency and flow). To increase customer satisfaction (brand perception goal). And to increase employees social engagement (cultural goal). We then apply these to social metrics. On the 1:1 contact deflection/self-help, we measure the effectiveness of our publishing efforts, CTRs, reach, post views, etc. via link tracking and Facebook insights (we use Bit.ly and Omniture). The support knowledge content folks also measure solution “quality”, stickiness, and bounce rates. Then we know how effective our publish/promote tactics are for a given set of support content and community links. And based on that week over week we can track 1:N effectiveness against normal and campaign-driven content pushes. On the TTA (time to answers) side, we measure how quickly we respond on social channels. We do this by pulling thread deltas from Radian6 data using their API.  We then compare weekly to our channel-specific SLAs and adjust if and as needed. It’s a well-known fact that velocity in social service directly affects customer satisfaction. Not only that, but consider that average social service resolution (non-technical issues) is measured in minutes versus days for traditional channels. On the sentiment side, the company is aligned around NPS but NPS is neither sufficient nor clearly applicable to social service channels. I prefer to measure efficiency and friction (how easy was it to deal with us), accuracy of information, and resolution time — all of which move the needle on overall satisfaction. Another sentiment index I strongly believe in is what I call the #GI (gratitude index) of a channel. Namely, what percentage of your traffic contains posts expressing any form of “thank you” gratitude? We hit 20% in 2012 at AutodeskCare. Only one other brand surpassed this with 24%. Can you guess? It’s Citibank run by Frank Eliason :) On the EE (employee engagement) side we measure staffing levels and engagement frequencies. Load bearing, and a bunch of other interesting behavioral aspects for each geo. One interesting metric is the size of what we call our “network map” — a dynamic internal system we created to tally social service “contact points” throughout the company as we encounter them. The thicker, wider that network graph gets, the more efficient your social business gets. So at the end of the day we look at these metrics and ask the following questions: Have we made an impact on traditional support pipelines (phones, email, chat, etc)? Have we increased our response velocity? Have we improved customer satisfaction and gratitude? Have we pulled in more colleagues and groups into the corporate social service graph? A positive on all of these justifies our existence and, to the executives, the social and community investment they made. And that’s how we get to stay social my friends! :)
Pop-upView Separately

Really interesting article from Eric Pratum rounding up social ROI opinions from several industry experts. Could have used more concrete examples IMHO, but good stuff nonetheless.

For many still, social ROI is little more than voodoo. And I think one of the best ways to demystify it is to share experience with peers. I don’t claim to know more than the next social strategist about ROI, but I’m sure happy to share what we’ve done with ROI at my work place to gauge our SocialService@Scale efforts.

At Autodesk we measure against several “center of mass” KPI across the entire support and service division. Social is no different. These KPIs and associated desired outcomes are as follows:

  • 1:1 contacts -> 0
  • TTA -> 0
  • NPS -> 100
  • EE -> 100

These outcomes are of course desired trends rather than hard targets. The idea is to reduce 1:1 contacts by disseminating and amplifying 1:N content via self-help (scaling goal). To increase response and answer time velocity (efficiency and flow). To increase customer satisfaction (brand perception goal). And to increase employees social engagement (cultural goal).

We then apply these to social metrics. On the 1:1 contact deflection/self-help, we measure the effectiveness of our publishing efforts, CTRs, reach, post views, etc. via link tracking and Facebook insights (we use Bit.ly and Omniture).

The support knowledge content folks also measure solution “quality”, stickiness, and bounce rates. Then we know how effective our publish/promote tactics are for a given set of support content and community links. And based on that week over week we can track 1:N effectiveness against normal and campaign-driven content pushes.

On the TTA (time to answers) side, we measure how quickly we respond on social channels. We do this by pulling thread deltas from Radian6 data using their API. 

We then compare weekly to our channel-specific SLAs and adjust if and as needed. It’s a well-known fact that velocity in social service directly affects customer satisfaction. Not only that, but consider that average social service resolution (non-technical issues) is measured in minutes versus days for traditional channels.

On the sentiment side, the company is aligned around NPS but NPS is neither sufficient nor clearly applicable to social service channels. I prefer to measure efficiency and friction (how easy was it to deal with us), accuracy of information, and resolution time — all of which move the needle on overall satisfaction.

Another sentiment index I strongly believe in is what I call the #GI (gratitude index) of a channel. Namely, what percentage of your traffic contains posts expressing any form of “thank you” gratitude? We hit 20% in 2012 at AutodeskCare. Only one other brand surpassed this with 24%. Can you guess? It’s Citibank run by Frank Eliason :)

On the EE (employee engagement) side we measure staffing levels and engagement frequencies. Load bearing, and a bunch of other interesting behavioral aspects for each geo. One interesting metric is the size of what we call our “network map” — a dynamic internal system we created to tally social service “contact points” throughout the company as we encounter them. The thicker, wider that network graph gets, the more efficient your social business gets.

So at the end of the day we look at these metrics and ask the following questions: Have we made an impact on traditional support pipelines (phones, email, chat, etc)? Have we increased our response velocity? Have we improved customer satisfaction and gratitude? Have we pulled in more colleagues and groups into the corporate social service graph? A positive on all of these justifies our existence and, to the executives, the social and community investment they made.

And that’s how we get to stay social my friends! :)

Really interesting social service post I happened to catch recently. And roughly in line with what I’ve seen at work for the past 18 months.  Almost 50% of customers seek assistance via social media channels? No shock there. It’s probably lower in B2B spheres. Much higher among students and consumers. Our social service crews cater to vast and diverse constituencies including enterprises (B2B), SMBs, Students, and consumers. So not unlike Autodesk as a global business bellwether company, our social services also taps into various swaths of social populations and behaviors. And we draw appropriate lessons - and experience - from addressing such varied market segments. Frequency of social service requests is an interesting figure. Does it mean you have significant product or service issues? Or does it mean you’ve done a really good jobs building a “humanized” conversational front-end to your brand? Or all of the above? We do get a small set of rather frequent “engagers” on Twitter. We don’t see this on other channels as much. The most persistent ones are either professionals or students. Bless their hearts - we’ve actually made real “friends” in both communities. And truth be told, some portions of our online engagements are indeed purely conversational. Sometimes even emotional. I like that. Response “effectiveness” has been a real focus point for us because we realize that “engagement velocity” alone is not enough. It’s nice to acknowledge customers within minutes of a problem. But it’s even nicer to actually resolve their issue and “close” the deal, as we used to say in sales.  And in social service, there are a lot of deal makers but few “closers”. This is likely why people rate social service as “ineffective” - a placebo of sorts. In the “old days” customer service would close cases from an internal perspective. In other words, a support rep would decide that the case was resolved and close it. Pretty normal since auditing usually revolves around case closure rates (a questionable metric if you ask me). But I find that approach less than ideal in social support. Why? Because on social channels it’s easy to get a “confirmed hit” by asking the customer straight-up in real time. So not doing so is lazy at best, IMO. Why assume when you can be sure? It’s true that doing it outside-in affects deflection metrics - you “deflect” less cases because people don’t always get back to you - or they “hit and run” with a question, and you never know what the outcome was. But intellectually, I feel better using “resolved” if, and only if, the customer confirmed it was or inferred resolution is unambiguous to any reasonable observer.  I’m convinced this approach greatly improves overall customer experience (instant response time or not) and corresponding NPS scores. I also know it makes “deflection” scores look worse internally - at least initially - but we should be in this business to improve customer experience. Not beef up pretty internal dashboards. Social service is a tough gig. It punishes shortcuts.  Stay social my friends :)
Pop-upView Separately

Really interesting social service post I happened to catch recently. And roughly in line with what I’ve seen at work for the past 18 months. 

Almost 50% of customers seek assistance via social media channels? No shock there. It’s probably lower in B2B spheres. Much higher among students and consumers.

Our social service crews cater to vast and diverse constituencies including enterprises (B2B), SMBs, Students, and consumers. So not unlike Autodesk as a global business bellwether company, our social services also taps into various swaths of social populations and behaviors. And we draw appropriate lessons - and experience - from addressing such varied market segments.

Frequency of social service requests is an interesting figure. Does it mean you have significant product or service issues? Or does it mean you’ve done a really good jobs building a “humanized” conversational front-end to your brand? Or all of the above?

We do get a small set of rather frequent “engagers” on Twitter. We don’t see this on other channels as much. The most persistent ones are either professionals or students. Bless their hearts - we’ve actually made real “friends” in both communities. And truth be told, some portions of our online engagements are indeed purely conversational. Sometimes even emotional. I like that.

Response “effectiveness” has been a real focus point for us because we realize that “engagement velocity” alone is not enough. It’s nice to acknowledge customers within minutes of a problem. But it’s even nicer to actually resolve their issue and “close” the deal, as we used to say in sales. 

And in social service, there are a lot of deal makers but few “closers”. This is likely why people rate social service as “ineffective” - a placebo of sorts. In the “old days” customer service would close cases from an internal perspective. In other words, a support rep would decide that the case was resolved and close it. Pretty normal since auditing usually revolves around case closure rates (a questionable metric if you ask me).

But I find that approach less than ideal in social support. Why? Because on social channels it’s easy to get a “confirmed hit” by asking the customer straight-up in real time. So not doing so is lazy at best, IMO. Why assume when you can be sure?

It’s true that doing it outside-in affects deflection metrics - you “deflect” less cases because people don’t always get back to you - or they “hit and run” with a question, and you never know what the outcome was. But intellectually, I feel better using “resolved” if, and only if, the customer confirmed it was or inferred resolution is unambiguous to any reasonable observer. 

I’m convinced this approach greatly improves overall customer experience (instant response time or not) and corresponding NPS scores. I also know it makes “deflection” scores look worse internally - at least initially - but we should be in this business to improve customer experience. Not beef up pretty internal dashboards. Social service is a tough gig. It punishes shortcuts. 

Stay social my friends :)

According to this Luxury Daily post, we have breaking news. Making customers happy is good for business! “Brands must put consumer needs, ease of transaction and joy from the purchase at the center of their business model to compete in today’s market and to raise profits, according to an executive at Forrester’s Customer Experience Forum 2012.” You’re kidding me. You need an official “Customer Experience Forum” to figure that one out? Okay, okay, snarky = bad, but I mean come on :) Because every so often you hear about research and analysis having produced such groundbreaking discoveries: The product or service must meet the needs that the consumer has The product or service needs to be easy to obtain and use The experience needs to be enjoyable for the consumers In other words, don’t sell people junk and treat them like crap in the process. Humm. What a concept. Frank Eliason once put it best when describing Apple’s success: “People like their product, they like their customer experience. It’s as simple as that.” Social media won’t magically improve your product or your customers’ experience. It’s just a bunch of tools with creative, passionate people at the helm. And if you don’t have the basics, you just don’t got ‘em. Social media doesn’t replace common sense and the old “Don’t do unto others…” adage.
Pop-upView Separately

According to this Luxury Daily post, we have breaking news. Making customers happy is good for business!

“Brands must put consumer needs, ease of transaction and joy from the purchase at the center of their business model to compete in today’s market and to raise profits, according to an executive at Forrester’s Customer Experience Forum 2012.”

You’re kidding me. You need an official “Customer Experience Forum” to figure that one out? Okay, okay, snarky = bad, but I mean come on :) Because every so often you hear about research and analysis having produced such groundbreaking discoveries:

  1. The product or service must meet the needs that the consumer has
  2. The product or service needs to be easy to obtain and use
  3. The experience needs to be enjoyable for the consumers

In other words, don’t sell people junk and treat them like crap in the process. Humm. What a concept.

Frank Eliason once put it best when describing Apple’s success: “People like their product, they like their customer experience. It’s as simple as that.”

Social media won’t magically improve your product or your customers’ experience. It’s just a bunch of tools with creative, passionate people at the helm. And if you don’t have the basics, you just don’t got ‘em.

Social media doesn’t replace common sense and the old “Don’t do unto others…” adage.