Lately we’ve all seen the ads on Facebook ‘Promote this post and have it seen by a wider audience!’ And many of us pish-posh it, because the people who liked us will see our posts, and who needs it. Right? Wrong. Not even 25% of the people who like you, or your page, will ever see your posts if you don’t promote them. What they call ‘organic’ reach is highly limited, especially with their new timelines. The way in which they filter the new timeline is going to make this even harder.
A lot of people I know had no idea that Facebook limits who can see your posts. When you start combining this with Google’s encrypting of search terms, the easy valuation of your SEO is creeping swiftly into overly complicated. A lot of A/B testing relies on this information and now that it’s being taken away, we’re back to the age old metrics of grabbing people off the street and asking them which sock is whiter.(Of course with on-line prompts to fill out those Q&As, we’ve already hurt ourselves. We’ve all trained ourselves to ignore those ad-like things that get between us and content. But that’s another post.)
Twitter has had ‘promoted‘ tweets for a while, which is probably why they get so tetchy about the other Twitter apps. The concept is that you pay and more people see your tweets. Not a bad model, really, though most of us just roll our eyes and ignore them. Still, all your tweets are just as popular and shared with people who follow you.
Not so, Facebook. For a long time, Facebook did something kind of similar. Your sidebar has “sponsored” posts, which are just plain ads. If you have a Page, you can pay to sponsor your posts, similar to Twitter, and push your brand. But here’s where Facebook’s a shit-bird: Edgerank.
The concept is like Pagerank from Google. The more popular, and active, your page/post are, the more they’re worth. But that doesn’t make sense for people, since my brother’s edgerank may be low, but I still want to see all his posts. Supposedly Edgerank doesn’t affect this, but I’m not so sure, given how many ‘important’ Facebook messages I miss. This probably stands out more to be since I don’t visit the site with any regularity. If I have a blog post (like this one) that I push to Facebook via Wordpress, I may go back to see if people commented there. That’s really the only time I notice what’s in my timeline, and while I do quick scan it, it’s filled with cruft.
If you’re on Facebook all the time, you’d never notice. If you got Facebook emails, you’d never notice. I don’t meet either of those requirements. If I use Facebook, I use it. If I don’t, I don’t want a hundred emails cluttering up my space, and this is a problem. See, if I don’t participate actively, by clicking like (something I rarely do), then I cause your edgerank to drop and fewer people to see your posts, so fewer of us click like and thus it sucks. They aren’t wrong with their algorithm, as Facebook is a work based on connections. A likes B likes C who shares A with D. That’s how things get around.(If you’re interested in the math, Dan Zarrella did the math and Harvard reblogged it.)
With that in mind, if you’re trying to improve your ranking for your product, Ari Herzog has a suggestion: Concentrate on interactions. If you have some regular people who leave comments, talk with them. They’ll be more inclined to share and retweet your posts, which gets you better ranking. Remember, we’re working under the assumption that everyone wants to get noticed more for business, and while SEO is not a zero-sum game, there are winners and losers. Concentrating too much on the media aspect of social media is a quick way to lose.
For the rest of us who just want to communicate with our friends and family, you’re better off getting a blog that emails them directly. At least then the only fear you have is the spam filters. But of course, that falls into the argument of why you should own your own content anyway, and is a post for another day.
2 replies on “Social Throttling”
Technically, Dan Zarrella did the math.
Zarrella isn’t affiliated with Harvard — he did not attend the University nor is he a faculty member. The Harvard Business Review blog syndicating his article.
Dan hasn’t revealed the foundation of his formula. So we don’t know how it was derived (intuition, speculation, data modeling); how (or if) it has been validated; what use cases it fits; and what edge cases it doesn’t fit.
Edited to say Dan wrote and Harvard reblogged. I really do hope Harvard actually looks into this stuff before reblogging. They have a higher standard.