Looking for Something?
Posts Tagged for

dennis yu

Facebook’s New Virtual Currency is YOUR Reputation


… by Dennis Yu

What do these 3 things below have in common?

If you guessed advertising, you’re only partly right, since one of the three items is an ad. If you guessed personalized results, you’re warmer, but missing the bigger picture. If you ventured that it’s Facebook’s attempt to mine you for all that you know and what you’re worth; you’re exactly right.

Facebook is a free service in the sense that you’re not paying a monthly subscription fee. It’s ad supported. Verizon charges me a few hundred dollars a month for service, which I’d gladly pay over listening to a 30 second ad every time I wanted to make a call. But Facebook chose to use ads, not subscriptions, to make a profit—yet Facebook users expect service as if they were paying customers.

These freeloaders who don’t pay also don’t realize how their activity continues to build value for Facebook’s advertisers. In this article we’ll explore the mechanics behind this system and why Facebook’s revenues will outstrip Google’s in less than 3 years (provided they have no major privacy bumbles).

Let’s start here: Any information you reveal about yourself on Facebook you can pretty much think of as information you’re handing over to advertisers. Sure, there are privacy settings, but do you know of anyone who actually unchecks the pre-checked box to have personalized ads served to yourself and for your image and name to be included in ads? I consider myself more knowledgeable about the platform than the average Facebook user, and even I’m not sure if it’s fully possible to opt out of things like Sponsored Stories.

Sponsored Stories work tremendously well for advertisers. They should, since these ads show you pictures of YOUR friends, creating the ultimate high school cafeteria, enquiring-minds-want-to-know, rubbernecker situation. Who doesn’t want to keep up with what their friends are doing?

Recently, Sarah Bird, the chief legal expert for SEOmoz, and I had a discussion about whether this is okay. If I’m a fan of SEOmoz, and Facebook uses my image to help SEOmoz, a Facebook advertiser, get more fans— doesn’t that mean SEOmoz is benefiting from my endorsement without asking me for permission? What if the advertiser is not even aware that they are achieving stellar results because of this technique? Is this Facebook’s fault or perhaps no one’s fault at all? What if the profile pictures of 13 year olds are being shown to all sorts of people without parental consent in the United States or other countries that have varying privacy laws?

Before you brand me as a privacy shrew or Facebook hater, let me tell you that I don’t think there’s a significant problem here, and there won’t likely be a problem. Here’s why:

  • People routinely sell their information for almost nothing:

    Give up your email address to earn 5 Farm Dollars? Sure! Sell out 10 friends to get 20 coins in some silly game that has no real world value? Absolutely! How about join an airline loyalty program or sign up for a supermarket rewards card? This has been going on for a long time. If you travel a lot like me, you know that many airports have “free” Wi-Fi that activates only upon you watching an ad. Don’t know about you, but I watch the ad versus pay $10 for 45 minutes.

  • We’ve been testing this for FOUR years inside Facebook:

    It was June of 2007 when Facebook released the F8 platform. To the best of my knowledge, it was our team that first came out with ads within Facebook that were personalized with user profile pictures and names. Facebook came out with Sponsored Stories only a few months ago, but arguably they have learned from what we did. If you remember back in 2007, Facebook’s own ad product, not our 3rd party ad-served unit, were those pesky “flyers”—basically static banner ads that had zero personalization and almost no targeting.

  • Advertiser gain doesn’t have to be consumer loss:

    This is not a zero sum game. Scott Richter, one of the pioneers of online advertising, tells me that spam is really just bad advertising. And I agree. Unwanted, irrelevant ads are like the yellow page books being dropped off at your door, killing who knows how many trees. That stuff on your doorknob or windshield when you get back to your car—that’s spam. But a highly targeted, interesting ad that highlights something your friend did—I would appreciate the recommendation, like a personal concierge of sorts.

  • Most users probably don’t know it’s an ad:

    Now this one is devious. How many blog posts or stock recommendation are written by authors with bias (think Michael Arrington of TechCrunch who writes about companies he has ownership in)? How about the free webinars that are put on by Google and Facebook for how to increase your marketing effectiveness—or even those little tips inside AdWords that keep recommending that you spend more money in various ways? Any content anywhere has hidden bias—even what’s in your News Feed, the super positive reviews posted by restaurant employees, or your comments about your friend’s cooking and whether that dress makes her butt look big. As an advertiser, you pay your email company to send emails on your behalf, so how’s that so different than paying Facebook to send messages to your fans?

Whether the delivery channel requires payment or not, is independent of whether that content is relevant and appreciated by the recipient. Some would argue that if the CPC advertisers pay is dependent upon Quality Score or CTR, that there is a built-in mechanism for good behavior. But we know of a plethora of big advertisers who spent millions on untargeted nonsense.

There is definitely a move towards more personalized, daresay more enjoyable, ads on Facebook. In the research study that we put out in January 2011, we observed that the average CTRs on Facebook were 0.051% in 2010. Some folks took this to mean that Facebook traffic was bad, as opposed to Facebook advertisers being unskilled in their methods.

Even today, the average marketer trying to create social ads is like shoving a bus driver into the cockpit of a fighter jet. They might have driven that bus competently for years, but they are going to need some training for this new vehicle that has so many more knobs to twist. And it’s no wonder that most Facebook advertisers crash and burn.

Last we checked, Facebook made $1.7 billion in 2010—let’s round that up to $2 billion for sake of simple algebra. And Google does about $30 billion a year. So right there, Facebook earns less than 1/15th of Google’s revenue.

Facebook also earns about 30 cents per 1000 impressions served (eCPM). It varies by country, device, and other factors, but we can use this number for now. And Google earns about $80—not 80 cents—per thousand queries. That means Google is 260 times better at extracting value.

Some search die-hards will jump in and say that search queries are more valuable because people were actively looking versus just goofing off when the boss isn’t looking. They will argue that search intent is more valuable than display because the conversion figures prove it. Several PhDs in statistics on staff will calculate it down to the 99%th percentile and 5 significant digits. But they might have overlooked the difference between demand generation and demand collection—and then the distortions that occur in last click attribution.

When marketers are able to give credit to clicks that are non-branded and not navigational, they’ll appreciate the value of word of mouth; which is what Facebook is.

The reality of advertising for us online marketers is that only a fraction of what influences purchase behavior is actually trackable via links and cookies. Marshall Sponder calls this the “ultraviolet” layer. It includes testimonials, peer pressure, TV advertising, and other things in the real world that don’t have pixels attached to them.

When advertisers realize this, and are able to assign value to these “assists”, and have tools that allow them to properly fly their jet aircraft, (punch in the destination, and the autopilot takes over,) then it’s possible that Facebook’s earnings could catch up to Google’s. Further, I believe that Google’s eCPMs will slide over time, as we have witnessed in any channel that matures, since advertisers will not be willing to count conversions that are actually navigational versus search. Maybe Google and Facebook could both meet in the $30 eCPM range.

The price of traffic MUST decrease from a pure economics standpoint. The growth rate in available traffic is outstripping the value of goods being sold, so mechanically this must be true. Unless you double the money supply overnight, a doubling of traffic globally because of more information across more devices on a more connected population gives each ad impression fractionally less value.

Watch as the waters equalize; with Facebook increasing their earnings with traffic being more valuable, and traditional search losing power, as advertisers are smarter about assignment of credit influence and demand generation.

Dennis Yu is Chief Executive Officer of BlitzLocal, a Webtrends partner that builds social media dashboards to measure brand engagement and ROI, specializing in the intersection of Facebook and local advertising. You can reach him on Facebook, Twitter, LinkedIn, his blog, or good old-fashioned email at dennis@blitzlocal.com. BlitzLocal is a leader in social and local advertising and analytics, creating mass micro-targeted campaigns. Mr. Yu is an internationally sought-after speaker and author on all things Facebook, and has been featured in National Public Radio, TechCrunch, Entrepreneur Magazine, CBS Evening News, and other venues.

Facebook EdgeRank Insights: Why You DON’T Want a Large Fan Base and Who Your Quality Fans Are


… by Dennis Yu

From the lessons of our BlogWorld panel came a number of questions we didn’t have time to answer.

Is acquiring a lot of “junk” fans bad for your page? It depends.

If you are a software company, B2B company or have a super niche audience, then fans for the heck of it are bad, acquired via contests or whatever. When your “real” audience comes, they’ll not feel among peers.

If you are a fast food company, or a brand that has a diffuse base (Walmart, for example), you can target pretty much anyone. Major retail brands that get power via mass media will get power via untargeted fans for the same reasons that Super Bowl commercials work. If 60% of the population uses your shampoos and deodorants, you might as well grab as many fans as you can, so long as you’re not using creative that is off-brand (example: Spirit Airlines’ recent Weiner jokes about fares that are “hard” to resist and such).

If you are an agency, it can go either way. The pros of junk fans are that clients may be impressed by a large fan count, in the same way that they’re often deterred if they see your agency has only 58 fans. If you are posting great content before a junk audience base, however, you’re not only throwing pearls before swine, but collecting responses that clearly reveal your audience is not on-target. Puzzling responses or lack of responses are the tell-tale signs.

The hidden benefit of having fans is that you can sometimes leverage connection targeting in a big way.  Perhaps the fan himself isn’t part of your target, but one of their 310 friends is. In areas where personal relationships and word of mouth are important for conversion, (local restaurants, insurance agents, and multi-level marketing,) then you have to weigh in this effect

In other words, the value of a fan = That fan’s direct conversion value + the influence value they have on their friends.

And you could supposedly chain this value all the way down such that:

A particular fan, Jenny, might be worth $3; (what percentage of lifetime value you can trace back to Facebook marketing efforts).

Jenny has 100 friends of which 20 friends are targets. (Let’s say this is Lane Bryant, seller of plus-size women’s clothing.)

The influence value on each of these 10 friends is 5 cents (if that fan buys because Jenny did).

Jenny’s value is then $3 directly, 50 cents of influence, and then 3* 50/300 *50/300 for the third level. If you run this equation down to the 10th level, you’re at $3.60 in total value.

If you increase the number of friends Jenny has (her influence) to 300, increase the percentage of her friends that are in-target (dependent upon your industry) to 30%, and increase the value of friends-of-friends (a function of your product) to 20 cents, then the equation works out to $5.15.

Now the value of friends of fans goes from 60 cents to $2.15, which is almost as much as the fan is worth directly.

The beauty of this math is we can actually start to measure these values versus talk about them hypothetically.  Facebook has given us tools to determine who our top fans are.  Lane Bryant might have only 300,000 fans compared to the 1.8 million of JC Penney.  But their fans are 4 times more engaging, based on how actively they interact on the page. You can see the same thing in Adidas vs. Nike, Monster Energy vs. Red Bull, and so forth.

You can pull this data via their graph API for your own page or set of pages to determine if you have junk fans, how engaged your fans are, and whether fans are influencing their friends.  We happen to sell these dashboards, but you can certainly do it yourself with some clever programming.  Regardless, it’s not available in the web-based Insights, for those familiar with the tool.

Facebook has told us they are considering releasing influence tracking that will allow brands to determine who the “first click” influencer is—who initiated a purchase, for example, that is then later seen by friends who buy, and then friends of friends who buy. When that’s coming is anyone’s guess. But we do know it’s a long ways out, since they still haven’t figured out how to even share the data with us.

EdgeRank as a theoretical concept can best be thought of as influence; how many fans does your page have, which of those fans are influential, and what measurable conversion events occur because of it.  In the real world there are loyalty programs for airlines, supermarkets, and other companies.  I travel a LOT, so I get special treatment with certain travel providers—they recognize their best customers.

Are you measuring your best fans—the ones who have the most influence, are spending the most money, and are most loyal to you? Have you started to integrate your marketing efforts between email, Facebook, PPC, and your point of sale to identify Jenny when you see her?

Are you able to tell if that person of high engagement is also one of high influence?  Most brands aren’t yet able to measure engagement vs. influence vs. sentiment.

Dennis Yu is Chief Executive Officer of BlitzLocal, a Webtrends partner that builds social media dashboards to measure brand engagement and ROI, specializing in the intersection of Facebook and local advertising.   BlitzLocal is a leader in social and local advertising and analytics, creating mass micro-targeted campaigns. Mr. Yu has been featured in National Public Radio, TechCrunch, Entrepreneur Magazine, CBS Evening News, and other venues. He is an internationally sought-after speaker and author on all things Facebook. BlitzLocal serves both national brands and local service businesses.


Learn About NMX


Recent Comments