Confessions of a Facebook Feed Spammer. (Facebook Notes = Blogs. People Tagging = Feed Spam.)

  People Tagging = Feed Visibility 
  uploaded by davemc500hats

Recently i've been playing around with Facebook 'Notes' & people tagging.

altho i guess it's obvious, i had sort of overlooked the fact that 'Notes' are pretty much the equivalent of 'Blogs' for Facebook (see my earlier brief post on Status = Twitter, Share = Del.icio.us). 

However, there are a few important differences i'm still discovering... one diff is that it's very easy to 'people tag' a Note, and provide a very visible way for your blog post Facebook Note to get discovered by your friends & contacts.

i think this mirrors some points i made in another previous post on Marketing Facebook Apps... namely that it's REALLY IMPORTANT to people tag content in Facebook so it has association to specific individuals, which thereby connects to the social graph & provides relevance.

I don't know if Facebook has anti-spam algorithms built to prevent excessive "people tag spam" in Facebook, but some of my experiments tagging people on Facebook Videos and Facebook Notes are showing up in other folks' feed rather obnoxiously (sorry atish), similar to Dare Obasanjo's post about Scoble hijacking his feed (note: Scoble claims to have over 4,000 "friends", whereas right now i only have ~600, still pretty large but not extreme, and about the same amount i have in LinkedIn). 

i'm pretty sure the reason i'm N Yr Feed so much (even when the "volume" is dialed down) is due to the fact that i'm "over-clocking" the normal Facebook feed prioritization algorithm by adding lots of people tags to selected content, and then magnifying that effect even further due to the slightly larger-than-average friend network i maintain (guessing average is ~100-200?).  also possible Notes & Videos aren't used as often as other content, so i'm "dominating" less frequently-used feed story categories more than most folks (except Scoble).  not sure about that, but seems plausible.

my guess is eventually some alpha-geek Facebook engineers will dial-down the magnification on the high-end, and i'll only be slightly more visible than average.  also, other folks will discover these tricks, and the market will 'settle' a bit further.  still, i'm pretty sure this technique will work well for some time to provide above-average feed visibility. (note: you can also use cocoa krispies for eyeballs, and that will also attract attention ;)

in any case, i think it bears repeating:

* If you want your content / apps to get distributed virally (ie, show up prominently in the feed), you don't have to go all Super-Spammer on the invites... the feed works quite well to promote item visibility via the social graph. people discover items via the feed and/or tagging, and they adopt. 

simply include people tags so that the item has reference points to the social graph.  this provides visibility both to the people you tag directly AND ALSO indirectly to other people connected to the people you tag.

having performed this experiment manually with Notes & Videos & Share, i'm more convinced than ever this can be simulated in Facebook Apps by choosing the right frequency & content for app notifications / requests with embedded people tag references.

(again, sorry if my tinkering is jamming your feed.  it's only because i care ;)

Comments

[Eric K.]

Dave, this tactic was popular among college students for awhile to sell their football tickets.

Before Facebook limited the number of people tags, we would tag a picture of a football ticket with every single one of my friends and that would spread not to just my network but theirs also. Quick way to get the word out that you want to sell a ticket.

[David Berkowitz]

Thanks for the good read, Dave. Can't wait for the tag spam epidemic.

[Ed Vielmetti]

dave - if you syndicate your blog into your notes, you can tag those imported notes with people too.

i don't quite know the semantics - is it "to the attention of"? "mentioned within is"? because you might want to tag someone who's not mentioned because you want to tell them about it.

it reminds me a tiny bit of delicious's "for:username" mechanism, in that it lets you push something into someone else's attention bucket without interrupting them. i'm a big fan also of the delicious "via:" tag, which they never mechanized but which shows up tidily in searches so I can tell who's rebookmarking my stuff.

[dave mcclure]

@Scott:

My people? I LOVE my people... PULL!

[shoots peasant flung into air]

;)

[Scott Ruthfield]

And all this time, Dave, I thought you were tagging me because of pure love, not facebook-testing shenanigans. :)

-Scott, a tiny bit heartbroken

[Jeremy Toeman]

Funny you should ask that - I look at "user votes" as the most important metric for apprate. It doesn't matter as much to me if people install the app as it does if they *use* it. Either to contribute back to the community, or to find apps they want. I feel the current "popularity contest" metric is terrible.

My favorite feature in the FB app we built is the page that shows Top Apps - Friends Votes. This one shows me what MY friends actually like, in one aggregated view. There's no other way for me to get that info within FB.

But, I know I'm an outlier when it comes to these topics. The reality is people seem to be numbers-obsessed these days. 11 million people use Top Friends (up from 4 million one month ago), an application I personally despise (more out of principle than anything - do I need to tell the world I'm better friends with XX than YY?)...

[dave mcclure]

hey Jeremy -

thanks for the comment... and i'm sure you do have the metrics to back that up.

still i wonder: ***are you measuring the right metric?***

app installs is probably a lousy way to measure active users, certainly not a good way to measure retention or monetization (tho no one is doing much of that just yet).

if all you're measuring is the high-water mark of app installs (minus uninstalls), i think what you get is a false metric for success.

again, this would be the equivalent of measuring retail store success based on the # of unique people who had ever crossed the doorstep just once -- rather than repeat visits overall, or better yet actual sales.

while i'm not saying that # users isn't a proxy for other growth, i think it's easy to look at successful apps and justify everything that they're doing, instead of the one or two *unique* things that caused success. we had similar stories at PayPal, where the primary viral engine of growth was so successful, many other questionable actions & decisions were masked.

ultimately, i still strongly believe most successful apps will be measured more based on retention or revenue metrics than unique visitors or app installs. i could be wrong on that point, but i'd love to see a deeper analysis & correlation of initial invite-happy apps vs others using more engaged metrics.

my .02,

- dave mc

[Jeremy Toeman]

Great post Dave. One other thing I've noticed with apprate.com so far is that the virality of an application that does NOT have built-in ponzilike features (eg "bite a friend") is significantly slower than one that does have them. For us, the app is starting to spread faster than before (from 2-3 new users a day to 5-10 now), and our outreach was limited to informing friends&family about the existence of the app...