Selling shovels to prospectors: Social Media Marketing's epic fail

Ok, so the headline is probably a bit extreme. Sorry, I don't usually like to trade in sensationalism. But as I was discussing yet another pitch by a social media metrics company with a colleague, the metric of 'friend counts' came up.

The pitch goes something like this: "You know, our research shows that 80% of soc-net users have 0-100 friends and 20% have more than 100...the 80-20 rule. So we use this key metric to identify influencers for special marketing treatment...because if they have more than 100 friends and a bad experience, well, that's alot of risk".

It sounds logical. But counting friends is meaningless. Here's a few reasons:

• Are they really friends? It is a well known phenomenon—Trophy Friends--that some people collect and keep score with their friend counts (or linked in connections, or Twitter follows or friendfeeds, etc.)

• Are influencers always influential? Identifying influencers based on gross generalizations about individuals (e.g., friend counts) is something marketers have always tried to do…Is someone you friended after a 15 minute conversation really influenced by your opinion on detergent? There are literally millions of us, though, who get to be influential once in a while among small groups of people.

• What does 100 friends mean in creating a definition of influencer? Why not 50? Would a social media metrics company feel good about saying it’s discovered a new rule: 95-10, where 95% of users have less than 5 friends? Probably not...because it doesn't promise easy. If you really wanted to understand influence, versus the willingness to click ‘add’, then you’d want to track a recommendation or comment that made it’s way through the ‘friends’ network of a supposed influencer.

Here’s a somewhat relevant example of that using twitter. The list filters popular content against a list of Twitter users who were first to Tweet about it…it’s a proxy, but one could make the case that, at least in this example, some of the first movers are influencing those who follow (i.e., friend) them. The trouble is that once you get below the top 100, you have thousands of people who get to be first…once. Classic Long Tail statistics


So what?

This gets at the issue afflicting social network marketing…the point of social networks is NOT marketing…that’s why Facebook has millions of members and no revenue. (see here for a post on this point).

Finding ways to appeal to millions of people around shared interests will require thousands of approaches…none of which will make the marketers job easy…pitches referencing 80-20 rules or other hokum feel right because it conforms to the marketers pre-existing belief system (which, at least in the client data we’ve worked with, isn’t generally true--or useful).

So getting to the point of this post: If there’s a gold rush on, the real money is made is selling picks and shovels to the prospectors.

MyAds: Banner ads reborn?

MySpace, the oft-maligned, drab older cousin to Facebook's freaky fashionability is taking the growing up seriously. Ever since their acquisition by Fox Interactive Media, they've pursued a revenue generation strategy around advertising innovation. 

And now, one month from it's launch, MySpace is generating estimated revenue of $140,000-$180,000 per day using a pay-per-click display advertising model (according to TechCrunch).

So what?

Display advertising (also referred to as banner ads) has known issues (banner blindness and CPM deflation being two among many). But the MySpace model is intriguing for what it enables:

1. Do it yourself ad creation

2. Pay per click pricing

Combined with MySpace's long tail, tribal approach to community (i.e., you associate with those whose interests are relevant to your own--like music, pet ownership or tatoos!), one might expect many community-generated banner ads to actually reflect the community's values rather than an ad agency or marketer's interpretation of those values. 

Combined with pay-per-click pricing, one might expect that these potentially more authentic ads might outperform their less relevant, intrusive messaging foils--and therefore attract more spending. In fact, IAB reports for the 3rd quarter of 2008 show that CPM-based approaches to online advertising are already showing flattening spend levels, while performance-based models continue to rise (see prior post here). MySpace would appear to be on the right road there.

But what about the 'quality' of the ads? That argument, like many subjective arguments over quality, will have to have performance data to back it up or it will be an argument of interest only to those making it. An ad created by someone within the community has alot of intrinsic advantages over an outsider with an art degree.

Online, good design is design that works. If a person with inexpensive, off-the-shelf tools  (e.g., Flash, Photoshop) can create display ads that get measurable results, those who have made a living on self-evident value judgements may have to rethink their approach...or focus on the communities that they are part of.