June 22, 2010

The Value Of A Customer's Hat

Yesterday, we talked about my problems with a study called The Value Of A Facebook Fan: An Empirical Review produced by a company called Syncapse. It purported to calculate a value for Facebook fanhood. Today we'll discuss another problem with the study.
 
After 18 pages of charts and graphs and marketing babble that occasionally resembles English, Syncapse concludes that the average value of a Facebook fan is $136.38.

Not $137.12 or $135.29, but $136.38.

They do this through a series of calculations. As they say...
"Understanding long-term Facebook fan value is a complex undertaking with advanced calculations required across multiple variables."
But before doing advanced calculations across multiple variables, you first need a clue about what you're calculating.

The problem with the quantitative part of their study is that the underlying assumption they are using is wrong. It is based on...
"...the measureable (sic) differences between (Facebook) users who have “fanned” a brand and those who have not."
Here's what's wrong with this assumption. If people who wear Red Sox caps spend $250 a year more with the Red Sox than people who don't, does that mean the value of a Red Sox cap is $250?  Of course not. You are taking the value of the customer and attributing it to his hat. But that's essentially the logic of the Syncapse study.

They are taking the added value they've calculated for a Facebook fan and attributing it to fan-hood at the expense of all other possible explanations.

Let's start at the beginning and pretend for a moment that, unlike Syncapse, we are not "global leaders in...digital measurement," we're just a bunch of bozos trying to figure out what a Facebook fan is worth.

We'll start with another example. After doing some calculations, Taco Bell finds that people who have Taco Bell bumper stickers are worth $100 a year more to them than people who don't have bumper stickers. The bumper sticker people buy more stuff at Taco Bell and they recommend it to their friends, etc.

Someone not very good at math or logic might conclude that this proves that a Taco Bell bumper sticker has a value to Taco Bell of $100. 

Someone a little more assiduous in her work, might think differently. She might do another calculation and find that all Taco Bell customers (whether they have bumper stickers or not) are worth $95 a year more than people without Taco Bell bumper stickers. She might conclude that $95 of the $100 attributed to the bumper sticker actually comes from the fact that the person is a Taco Bell customer. She might conclude that the value of the bumper sticker is actually $5 a year.

So the key question for Syncapse is, how much of their imputed value can we attribute to "Facebook fan-ing" and how much to something else -- like the fact that fans are more likely to be customers?

The answer is, I don't know and neither does Syncapse. And until they know, it is impossible for them to calculate a reliable value for a fan, no matter how many charts and graphs they produce.

It is my contention that a substantial part of the value imputed to being a Facebook fan has nothing to do with Facebook, and may be related to being either a customer or an active category participant or something else.

If I am right,  Facebook fanhood has a far smaller value than the one they have calculated. This, of course, is not good for companies like Syncapse who have a vested interest in pumping up social media.

Apparently, when it came time to publish this study, someone at Syncapse with functioning synapses (sorry) must have read the thing and realized the fallacy of the logic. Buried on page 15 is the following disclaimer...
"It is important to note that this audience would still have value without Facebook..."
In other words, what they probably have been measuring is the value of the customer, not the value of his hat.

No comments: