The Open Graph – If you’re in online marketing those three words mean one of two things.
The first set of people see it as a bit of a nerd fad. A new buzz word that doesn’t do a hell of a lot except for make a few things look a little prettier when you share something on Facebook or to get that cute little picture next to your SERPs in Google.
The second set of people see it as a massive feature. The Open Graph (rich snippets, schema mark-up) is a way of categorising content so web properties can use your information in the most effective way. It’s making dumb(ish) content a little bit smarter. The properties you define on your site is what will dictate how a your information gets displayed and treated in various platforms..
I’m one of those people that fit into the second category.
Working as the resident search nerd for a big digital agency, I promote the open graph and rich snippets to almost every client I meet. It’s underutilised by businesses for now (at least here in Australia) and the first ones out of the gate to jump onboard this functionality will benefit the most.
And then a weird thing happened…
My belief is that if you utilise the Open Graph properly, it can only improve your online presence. It makes things stand out on Google, Facebook, Twitter – most of the big online platforms. But then I received an email from a work mate that had the following two screenshots :
One that utilises Open Graph for a Facebook share
And one that didn’t use Open Graph
And a note that said:
“noticing if I don’t put a graphic on Facebook post, its seen by 400%-500% more people”
If I was understanding this correctly, the exact same posts that display Open Graph mark-up were being seen fewer than 10% of the posts without it. WTF!
We gave the posts a little more time to see what would happen, and sure enough they rose to around 3000 views (non OG) and 300 views (w/ OG).
That’s when the heart attack moment started to kick in.
I’d been harping on to businesses almost daily for a while about the benefits of implementing Open Graph and Rich Snippets. Helping them to define strategies that would assist their business, and for the most part (where these snippets were being utilised), we we’re seeing benefits. I was a believer dammit! But this was looking like I was wrong. Very wrong!
Digging a little deeper
Looking at the site (Google) analytics told a different story though. Each link had a similar amount of traffic – surpassing the original lower view counts in some cases.
And then I came across this, and breathed a small (and nerdy) sigh of relief:
“As part of our ongoing investment in Page Insights we recently completed a comprehensive engineering audit of the product. During this audit, we uncovered bugs that impacted impression and reach reporting. We have confirmed that these issues impacted reporting only and not delivery. Ad Insights were not impacted by these bugs.”
You can check out the full announcement in their slick little video with reassuring music overlay here:
So it turns out Facebook reporting had a bug! A bug they assured their engineers were rectifying right now.
“To see the overall impact, if any, on your individual Pages, we recommend looking at your organic, paid and viral reach and impressions for your Page and for your posts over the next few weeks, starting on Monday, February 25. Because these bugs impacted our logging systems we won’t be able to backfill Page Insights with historical data. “
Believing in Data Religiously
Working in search marketing, my job – and success – is completely measured by data. Be it rankings, link profiles, conversions, traffic, you name it, everything about my role is about some kind of number.
I see a lot of discrepancies in numbers sometimes. Some tools show Y amount of links for a site while another shows X amount of links. Even Google properties give different data – anyone who’s tried to link Adwords and Analytics conversion data can attest to that one.
But this once again highlights how trusting we are of the data we are given. Big, expensive decisions are made based on the back of numbers like these.
I guess what this little adventure highlighted to me was the importance of gathering the full story when it comes to translating, reporting and making decisions on the back of data.
We gotta make sure we take care of what we are reporting on. Algorithms sometimes let things slip. Human’s let them slip even more.
On the upside though, all that stuff I’ve been saying about Open Graph the past year or so still stands. Phew…
AND I learnt that if something like this can almost incite a heart attack, then I really need to lay off eating butter from now on.