Assessing ROP (Return on Participation) and the DemoGOM

Every social proposal that I write is tied to, usually, just one business metric. When you walk into a CMO or a PR Manager's office at a large brand, the first thing that they're going to tell you is, "A lot of our people have serious doubts about the business value of social media."

If they're speaking about the cultural values of their organization, they're right, no matter what you think. That's their reality. And if you want to sell them on social media, you've got to live in that reality.

What they're looking for, to justify any social media spend in the C-suite, is ROI. And you've got to tell them, right up front, that you can't show ROI, yet. Social media, for better or for worse, is likely in the second year of a ten year growth curve that should probably shake itself into a more standardized business practice model (a la direct marketing in the '60s or public relations in the '70s). But these businesses need help now. The way that they're communicating with their network of stakeholders (prospects, customers, media) is not working.

What you can show them is ROP. (That link from Brian Solis's PR 2.0 gives a pretty healthy rundown of the minute differences between ROP and ROI.) I'm going to try to delineate the big picture differences between the two here, and give a concrete example.

ROP (return on participation) is what your brand gets in exchange for participating in social media. To put this in a more realistic setting, let's examine a brand whose social media approach goes beyond blogging alone. Let's look at  my second-favorite home-products vendor, Target. (Sorry, but Costco takes first prize. You know, living wages and all...)

Revolutionary, Target is not.  I remember reading something  a couple of years ago about Target where they said that blogging was not in line with their company's values. Whatever. They seem to have picked up the ball since then, beginning with a Facebook initiative (led by AKQA). Their first steps into social media looked like this:

• A marketing campaign that existed in-network only (within Facebook) that leveraged the  existing platform
• The campaign allowed students to tell a story and personalize the content
• There was no pre-existing entity within the network (Facebook) that boasted the same premise for social interaction (a dorm survival guide) AND delivered utilitarian content (recipes based on dorm-fridge food, furniture-preference-based personality quizzes and other seriously engaging stuff for 18-year-olds)

I don't have the figures on what the AKQA engagement cost Target, but if I had to guess, I'd ballpark it at $100k-500k. For the sake of argument, let's call that number  $300,000.

According to AdWeek,  the micro-site attracted "7176 members, 409 photos, 483 posts  and 37 discussion groups." To reduce that data in a "ROI" way, you could say that for every $41 spent on this campaign, Target acquired one "touch," and that "touch" had about a 1-in-20 chance of really interacting with the campaign, based on the amount of photos and posts. But this type of analysis is freakin' wrong. And here's why:

• You can't boil a social media engagement down to a one-time, one-touch interaction. This is part of an ongoing conversation.
• It's totally nuts for a brand to pay $41 to engage with a prospective customer that is likely already a customer. These aren't customers. These are real people, and at this stage in the game, they're influencers to you, buddy.

To really calculate the ROP in this situation, you've got to figure out a few things:

1. Does the fact that these people are engaging with Target report out to their social graph in any way? Are they telling their friends? Does the network allow for this kind of reporting out? (Myspace doesn't.)
   
2. If so, how many people are in the aggregate social graph of these influencers, collectively? (BTW, that's where your $300k plus the opportunity cost of this initiative  went.) I'd guess that number  is somewhere around 538,200 people. This is based on the stat that the average Facebook user has 150 friends. And yes you've got to dilute this number to account for social graph overlap. We'll do that below - that's called the DemoGOM.
    
3. Of the people in this person's social graph, how many will be influenced and will engage with your brand as a result of seeing the influencers' activity? (Is your application viral?)
4. If you're going to say that because your application is viral that it engages more people, then  you must simultaneously account for the people in the social graph that you alienate with your brand-related intrusion. For example, the shitty McDonalds campaign in my newsfeed made me not only avoid McDonalds like I always do, but I also emailed a few dozen people to tell them how much McDonalds sucked because they targeted me with their "chatter". Let's call this figure the SGAM, the social graph alienation multiplier. As in,"The creative was good, but the SGAM killed the WalMart engagement." Better luck next time, Bentonville.

Well, if you can answer those three questions, and you know the total number of man-hours (opportunity cost) and dollars that go into this kind of engagement, then you can calculate ROP.

And you can't calculate that without knowing the DemoGOM. 

The DemoGOM is a demographically stratified Graph Overlap Multiplier. (You're thinking - WTF - just stay with me here, okay?)

My wife Alie has about 100 friends on Facebook, and so do I. The odds that one of her friends is one of my friends is about .5 (50%). However, people in different age groups(14-18, 18-22, 22-25, 25-30, etc.) who live in different environments (small rural high school, large urban dorm, etc.) will have different DemoGOMs.

For example, two high school freshmen in the same class in Northampton, Massachusetts will likely have a DemoGOM of close to 1.0. Two social media hyperconnectors like Chris Heuer and Shel Israel will likely have a DemoGOM of something more like 0.1 (10%), because they're both connected to a lot of people, and there's a slight disparity in their ages.

If you can't calculate the DemoGOM for the network in which your social media engagement takes place in, then you're going to have a lot of trouble accounting for how many people are engaged.

So, to sum things up, here are the critical differences between ROI and ROP:

1. ROP accounts for opportunity cost of using social media versus other methods of outreach. ROI simply measures outreach spend against sales.
2. ROP attempts to account for the SGAM (social graph alienation multiplier) to figure out the cost of your targeting (or lack thereof, since it's a cost).
3. The end goal in monitoring ROP is (1) awareness and the (2) creation of new conversational streams and viral memes. If those two things don't lead to the sales that the C-suite desires, then your brand may have a problem that social media just can't solve.

Comments

[Alie Cohen]

The math behind this is super interesting. If you assume that each person in the target group joined because they had one or more friends already in there, it makes the calculation of the Demo GOM possible. Stay tuned for some interesting algorithms.