What's the best way to price social media advertising or marketing programs?
That's one of those questions that screams for the ever-honest "it depends" answer. Maybe that's because we need a new pricing model. Let's first take a look at some current options and review their strengths and weaknesses in this context:
Cost Per Impression (CPM): It's great for branding, which is the overarching strength of social media. But the value marketers get from social marketing isn't just exposure -- it's something deeper.
Cost Per Click (CPC): The model that made Google what it is won't do the same for Facebook, at least not on the same kind of scale. CPC is tailor-made for direct-response marketers, while social media is centered on building relationships. This is the classic "square peg/round hole" issue.
Cost Per Action (CPA): We should be able to put this to rest easily, as marketers shouldn't be driving a relationship-building vehicle on a conversion track.
Cost Per Engagement (CPE): This gets closer, as some engagement metrics translate well to various social marketing programs, such as time spent with a widget or percentage of a shareable video viewed.
When you want exposure, traffic, conversions, or interactions, each of these four pricing models comes in handy. But social marketing, when done right, achieves something much different: relationships. How does that fit in neatly with these pricing models?
To that end, I'll throw out a fifth model: Cost Per Social Action (CPSA). It's for any action with a distinctly social quality that leads to either new relationships (such as through "viral" referrals or acquiring new followers and fans) or deepening existing relationships (such as through "likes," comments, responses, and ratings).
The main benefit of CPSA is that marketers know they're paying for something social and relationship-oriented. More importantly, marketers know they're not specifically paying for exposure, traffic, conversions, or interactions (though those can all provide additional value). It's an acknowledgement that social media is something else, so it's deserving of a new model, one that stresses relationships above all else.
There are downsides too, and I'll spend even more time on these:
It's a new model. The alphabet soup is too crazy already. There are others too, like CPL (Cost Per Lead) and CPS (Cost Per Sale). So now marketers have to learn another?
Look at the publisher or vendor side too - they now have to sell yet another pricing plan? Then again, some publishers like VideoEgg have used Cost Per Engagement pricing to differentiate themselves. Could some differentiate themselves by promoting CPSA?
There are too many letters. Come on, at least simplify it to three letters so it stands a chance. Though you could argue that to stand out, maybe the four-letter version will differentiate it.
It's vague. A social action can mean so many different things. Just looking at Facebook alone, can you really say that the same pricing structure can be used to cover everything from the number of fans accumulated to the number of virtual gifts sent? Then again, CPM covers everything from an impression on a portal's homepage to an impression that's highly contextually relevant on a blog. Similarly, CPC is used for search and contextual marketing, two very different approaches.
There are already even more specific social metrics, such as CPI -- a Cost Per Install model for widgets. CPSA doesn't account for enough of the intricacies of social media that are already being addressed directly.
What's a social action worth anyway? The further anyone veers from reach and sales, the harder it's going to be to tie this into marketers' traditional metrics.
I'll leave it to you now: is CPSA needed? Are some already using a version of it? Is this latest version of alphabet soup past its expiration date? Share your thoughts in the comments to address CPSA the way it should be: socially.
PRINT
COMMENT
SUBSCRIBE
RSS
RSS FEEDS
Maybe the challenge will be to scale the model?
We are clearly in a new era of defining reach, similar to back in the day when a definition was developed for TV, Print, radio and PR "impressions." I'd love to see an organization or group take a stand and start developing some standards around social media "impressions." Eric, I think you're on the right track at the very least. If you haven't already, you should throw your opinions in the mix here: http://measurementcamp.wikidot.com/start
Kerry: I think you even have to qualify that reach number you suggest in your example - both Facebook and Twitter have pretty high churn rates yet the Fans and Followers numbers in general don't decline (unless you're content is pushing them to), so you end up with a decent number of inactive Fans/Followers in that audience count. I tried to make a very simplistic look at Twitter click-thru rates for Zappos by just reducing the audience number by churn (http://www.voncoelln.com/eric/2009/06/03/after-the-tweet-exploring-twitter-click-through-rate-benchmarking-to-measure-engagement/) and would welcome feedback on the approach or ideas around a standard.
In addition to churn, Kerry mentions that only 48% come to Facebook each day. There is also a glut of info in the Facebook or Twitter stream for a user to peruse and likely only a portion of your fans see your post (I showed that on Facebook that over half the engagements for a post - Likes and Comments - were made in the first 90 minutes - see http://www.voncoelln.com/eric/2009/06/18/how-long-does-it-take-for-facebook-fans-to-react/). I'm not sure how much of the audience number should be reduced to take each of those in and would be interested in the group's thoughts.
This is inherently the issue we're facing - that you can't really measure this directly. I can engage a consumer in an experience, after which they might post to their personal blog, tweet, and post a Facebook status. So that's three actions, but what's the reach of their networks? If they have 1000 twitter followers, 500 FB friends and 2000 blog followers - can you justify that the reach of those three actions is 3500 people? Can we consider those "broadcast impressions" or are they something new entirely? If only 48% of FB users log in on any give day (per FB analytics), did they all actually see the update? And how then do you quantify the comments and conversations that result from those broadcasts?
I'd love to hear how any of you are approaching this issue - if at all.
I've been helping to construct scoring models for clients, to help them estimate the ROI of their campaigns that include social media.
The models track what Jim (above) calls "CMA - Cost per Monetized Action." (I don't call it that but may start!)
My models also track that more vague engagement: How many have followed, interacted, etc. -- i.e., the Social Action (the SA) of your CPSA. Although we can't make a direct correlation to ROI now, over time we can fiddle with the scoring to become more predictive of future spending.
In the meantime, it's training my clients to focus on the costs/benefits of these social components to campaigns.
Can I suggest an additional contribution to the metric? How about calculating CPSA using a “depth of relationship” multiplier? Kind of like a CTR applied to a CPM. I propose that we can quantify social action with the formula below:
Value of social action (CPSA) = breadth of reach of the social action x the depth of relationship
The DoR score can be totally arbitrary and different buyers will value it based on the vendor’s specific product offering and the marketer’s goals. For example, posting a sponsored comment on a blog (like this one) has a low DoR score while a sponsored Facebook profile post where someone is endorsing a brand by placing a logo on their profile, shows a very high degree of DoR, and therefore may get a higher CPSA.
That said - I like the fact that you're keeping this dialogue alive.
I like CPSA because (as you said) it gives us something we can share with our client so that they feel like they're paying for something social and relationship-oriented.
On the other side of the pancake.
I don't like CPSA because (to your point) it is vague. The tactics recommended within a Social Media plan are as varied and unique as a fingerprint. To compound this issue CPSA basically seems to me to be a new higher level model that uses key performance indicators that could be applied to Social Media plans right now but are more independent and isolated: - Repeat Rate - Time Spent Rate - Take Rate or Download Rate - Send to A Friend Rates - Comment Rates - Linking Rate
Maybe we should consider looking at the tools available to us that listen to the social ecosystem (BuzzMetrics, Spiral16, Radian6, etc.) and the results our efforts produce. In the social space, what has our current spend created in terms of conversation/relationship sentiment (positive, level, or negative). A possible measure could be around the impact our efforts have on that sentiment score and our ability to sustain a positive level.
Just thoughts - thanks for keeping this discussion alive.
CMMA - Cost per Monetized Mobile Action -social, e-coupons or otherwise.
It seems a reasonable argument to say that a social action such as "Liking" something on Facebook would be less valuable than, say, a comment, a ReTweet, some combination thereof, etc. One action clearly takes more effort and thus one could assume that the person taking the action is more incentivized to spread the word. There has to be some way to value the reach that each social action has on a user-by-user basis. I.e. - How much is a social action worth if taken by someone with 100 friends or followers, versus, say, a social action by Ashton Kutcher on Twitter? Third party validation is only as good as the selling power/ brand equity of that third party! There's a reason that we haven't seen "Maury's Book Club" ;)
Obviously, there is a need for clearly defined tracking and real time valuation based on the tracking results. If this is not closely and carefully monitored, it opens up an array of potential problems from scams and programs to who knows what.
In summary, David has a great concept and I believe it is the right one. (We at CenterMedia Partners have been using a similar type of model for a while now, but have yet to perfect the tracking component). We just all need to be careful to monitor this model as closely as possible so as to not de-legitimize the social media marketing platforms that are becoming so valuable for us as marketers.
It seems a reasonable argument to say that a social action such as "Liking" something on Facebook would be less valuable than, say, a comment, a ReTweet, some combination thereof, etc. One action clearly takes more effort and thus one could assume that the person taking the action is more incentivized to spread the word. There has to be some way to value the reach that each social action has on a user-by-user basis. I.e. - How much is a social action worth if taken by someone with 100 friends or followers, versus, say, a social action by Ashton Kutcher on Twitter? Third party validation is only as good as the selling power/ brand equity of that third party! There's a reason that we haven't seen "Maury's Book Club" ;)
Obviously, there is a need for clearly defined tracking and real time valuation based on the tracking results. If this is not closely and carefully monitored, it opens up an array of potential problems from scams and programs to who knows what.
In summary, David has a great concept and I believe it is the right one. (We at CenterMedia Partners have been using a similar type of model for a while now, but have yet to perfect the tracking component). We just all need to be careful to monitor this model as closely as possible so as to not de-legitimize the social media marketing platforms that are becoming so valuable for us as marketers.
I think this is a great idea as well. What you measure determines how you behave, and if you're measuring social actions (whether they're fans of a page, retweets, or referrals), you'll be behaving in such a way as to drive that behavior. Well done.
Cheers, Kaila
Where many measurement tools are focused just on awareness, it highlights the value of an engaged community and the artifacts of driving sustainable word of mouth through meaningful relationships.
Ryan Rasmussen Associate at Zocalo Group http://zocalogroup.com
Oftentimes we measure social media campaign success by total fans, interactions, time spent, etc. but do not always analyze from an overall cost per social interaction standpoint.
Third party ad servers should start leading the way and tracking/calculating this metric. If industry benchmark data for this metric could be established as well, that would help sell it through as a new standard.
Best regards, Gianluigi Cuccureddu