I say this all the time to my clients. Your end goal should not be more likes. They are worthless. The goal should be more actual interactions with your app or website or product.
The same goes for people focusing on getting people to click the big green button on their homepage that says "learn more" which takes them to another page (?!?). The buttons should encourage interaction and progressing that user to use, experience, and depend on your product.
I don't use Facebook much these days except as a way to log in to other web sites. I didn't have a day when I decided that "I hate Facebook", I just drifted away.
My profile is full of all the anime I was watching two years ago. It's pretty thin, and when you consider that it intersects with a population of friends and family I knew 20 years ago, I don't know if anything useful is going to come out if it, at least for me.
Some people might be in a more thickly connected part of the graph and get better results.
Another issue with Facebook is that content is created for consumption by all of your connections. It's social preening. It's what you aspire to be, but wholly incomplete as a medium for search.
A few of the missing items:
-All the dirty stuff you won't show your grandmother. "Watch what you post" is part of raising kids today.
-What you fear. Do you post that you need an abortion or that you might have cancer?
-Gross stuff. Nobody posts that they have hemorrhoids. Or any number of things that you don't want girls you might date to see.
-Things we're ashamed of.
-Things that make us different. If you're a gun-toting neanderthal in SF or a gay man in the bible belt, you might not want everyone on facebook knowing that.
-Things the government will read. Reading revolutionary ideas, means of destruction, might be interesting and educational. However, posting on these topics or having a traceable record of them is a wholly different level. Think about prosecutorial misconduct.
I think FB will have a reasonable degree of success in search. However, it's badly gimped.
How many people actually care whether their friends like (whatever "like" means) Coke or Pepsi? The most useful graph searches would most likely be job or travel-related queries like "friends who live near San Francisco" or "friends who work for Google". Such truly personal data will never be polluted by advertising campaigns.
And if "friends who like Restaurant X" doesn't turn out to be a reliable way of getting a good restaurant review, people will quickly stop using it. Or they'll send a message to their friend to confirm whether they really recommend the restaurant or they just liked it for a coupon. Eventually they'll figure out which of their friends are reliable reviewers and which just click on every "like" button.
> The truth however is that the link between query intent and your social interactions for interests and places is much weaker than FB wants you to believe.
i must diasgree with this line. facebook is not the place that you could search for a specialized query (for example algorithms or scientific data). instead it is a place to search for people, restaurants and other day to day subjects that people tend to talk about in daily chats, and they may share links or write status about them in social networks. yes, maybe number of likes is a bad measure to rank the results, but it definitely is a measure.
I have been thinking about this a little and a lot of this is to do with weighting. THe simplest starting point I can imagine is the value of a like in the overall search should be weighted by the total number of likes given by that person.
So a like from Robert Scoble will be 0.0000001% of the value of a like from my maiden aunt who has only liked three restaurants in her home town since she was given a iPhone for Xmas.
What will really force this to take off is the ability for a like from Robert Scoble / Aunty Mable to be tracked to me liking the same restaurant -> and so there being a like-rank
value
I think it is this:
L = ((1/TotalUserLikes) / totalWeightedLikesOfRestaurant) * numberOfSubsequentLikes
Even an ineffective Graph Search is still probably a net win for Facebook, however. The ineffectiveness itself could be a kind of lure for less-popular businesses who want more exposure (and perhaps have some interests in appearing more popular than they are)... it doesn't fit every business, but it is certainly a non-empty market.
But it is a good bet - people I know are more likely to
Use services I will also want to use.
There is dirty data in there, but I would suggest the people who like coke to get a token are the kind of people I will happily ignore recommendations from anyway
But do you know why they chose to like something? With Coke, it's possibly a fair bet that it was because of a competition. But for a local restaurant, even if they've gone there and liked it, that could simply be because they saw a "get a 10% discount on your bill if you like us on Facebook",
Its a simple catch-22 - a "like" has no downside cost to me, so I will happily trade it in for 10% of my bill (Hell I wish more UK restaurants did this!)
However in the new graph search world (and indeed before this really) the downside cost was that I am seen as a like-whore trading my affections.
The more facebook becomes real the more the cost of like will increase.
There will be a virtual rate of exchange - as indeed there is today. Some of my friends tell everyone everything, some play their cards close to their chest. I judge their recommendations accordingly. Somehow social graphs need to do the same - and frankly that is pretty easy weighting really.
the point is that the real intent is somewhat independent and almost orthogonal to pressing the "like" button. that's the point the article is making about noisy and dirty data.
The same goes for people focusing on getting people to click the big green button on their homepage that says "learn more" which takes them to another page (?!?). The buttons should encourage interaction and progressing that user to use, experience, and depend on your product.