With the numerous modifications to news feeds, I think it's high time Facebook dropped the number of page likes as it's top metric for Facebook Business pages. But what should it replace it with?
Just a quick gripe first: too many numbers not enough clarity
There are essentially 3 sources of data on the performance of your Facebook business page:
- The Insights interface - This is intended to give you an overview of the performance of your Facebook page. It has data on the fan base, the demographics, the individual posts etc.
Recently this has been upgraded and is now clearer and more helpful than the previous version, but in my opinion it was a missed opportunity to provide a clear visual representation of your page's performance.
- The exportable data - These are the excel style data sheets which you can export. Essentially raw data which you can analyse on your own. There is currently an option for an 'old' or a 'new' version of the data. These two differ by the type of metrics which are provided. Facebook has phased out some of the analytical indicators and introduced new ones. This is what many Facebook page managers or social media analytic companies make use of in their client reports.
3. API data - The Application Program Interface dataset is the bank of information available to developers who may want to dynamically produce their own reports on a Facebook page. This is typically used by bigger social media analytic companies, as well as online services which in effect provide you with their own version of Insights. As with the export data, some of the indicators are being phased out and new ones being introduced.
Mark Twain apparently said, "Facts are stubborn, but statistics are more pliable". In Facebook's case they are particularly pliable.
The problem is that there is a lack of transparency and clarity from Facebook on its Insights metrics. The definitions are just not specific enough.
I don't think this lack of clarity is intentional; I simply think Facebook has not grasped two important concepts when reporting performance. One, the user interface should be incredibly simple and user friendly, and the export data should be extremely nerdy and specific! Even if this means writing a paragraph on each measure! These raw datasets are going to be used by number crunchers, so it's important that they are very clear about what they are crunching!
To be fair, dozens of data columns are provided by Facebook at the page and post level, but their definitions and how they relate to each other are not specific enough.
Just read a great blog post by Ben Donkor on the difference between engagement vs consumption, which opens up a can of worms on a number of Facebook definitions. Or see my post on how the new prominent engagement rate stat from Facebook can be wildly skewed and manipulated.
On top of that, there are of course occasional bugs in Facebook's performance reporting, which while understandable, can push you from mild confusion to complete frustration.
The highlighted indicators are not always worthy
Clients naturally gravitate towards the measures which are more prominent on Facebook's interface, but often these are not the best indicators of how a page is performing. The reason for this is that:
What are the indicators many clients are fixated on? I'll refer to them as 'the unsuccessful candidates'. Some may provide one piece of the performance jigsaw puzzle, but none on their own are that revealing and worthy of most of your client's attention.
Some of the unsuccessful candidates
Number of page likes
There was a time when the total number of fans a page had was the ultimate stat. These were the 'loyal' fans, the ones who had an 'affinity' with the brand.
For some pages this may still be true, but how many times have you liked a page not because you care about that page or brand, but simply because of a promotional offer, or because your friend liked the page, or you felt like being nice to a 'Like our Facebook page' request. There is of course the matter of the many who have liked Facebook pages they don't even know about, because of click jacking scripts.
At one time you couldn't even get to see a voucher on a Facebook page without going through a 'like' wall first. Then there are the multitude of likes generated by advertising campaigns.
What is the point of having hundreds of thousands of fans, if only a few hundred are exposed to your post (exposed does not even mean they see it, they had the potential to see it) and a few dozen interact with it. Who cares how many people are on your party invite list, you should care about how many actually come to your party.
If it was up to me, I would now take this statistic off the homepage and relegate it to the insights panel!
With the removal of the Edgerank algorithm, Facebook embarked upon a much more sophisticated method for selecting and displaying stories in news feeds. This meant many clients were now not so concerned with their fan base but with their reach instead. But the reach figure in itself can be misleading and should be used with caution.
Firstly, x number of users reached does not mean x number of people saw your post. It simply means it appeared in their news feeds. To be fair, this is the case for the 'reach' figure in all media.
It can be boosted with paid advertising. So many clients demand to know the 'organic reach'. But 'organic reach' may be under reported by Facebook as it does not include the organic reach of a link to your post from another Facebook business page.
Nevertheless, organic reach is a useful parameter for judging the success of your posts. Emeric Ernoult has a passionate post defending the 'reach' measure.
You could argue, that given that Facebook uses engagement as an important factor in determining whether posts are shown to fans, organic reach in itself is the most important Facebook parameter, as it gives you information on how many people have potentially been exposed to your message, and in theory this should comprise mostly of the fans who have shown active interest.
But for one thing, this figure may be vastly deflated as it currently doesn't seem to include the reach generated by someone linking to your post from another Facebook page. And of course it will also reach many people who have not engaged with your page and may not even look at your post.
This was a useful sub measure of the overall reach provided by Facebook, but which has now been phased out in the new datasets. You can still see it in the old dataset for export and API, but it seems Facebook wishes to get rid of it. It does seem a shame, as it is now lumped with organic reach.
Although a useful stat is showing the reach generated by stories from friends, in isolation it doesn't tell you that much. You need to compare it to the non-viral reach to get some idea of how your post performed. In short, I don't understand why Facebook is not reporting it anymore, but it could never be the all revealing figure.
People Talking About This (PTA)
This used to be one of the most revered measures for Facebook managers. Even Facebook used to show this on the business Facebook Page next to the number of fans, in order to give some sort of idea of engagement. But it was seriously flawed:
- It included new likes of the page, and for me this was the biggest flaw as clients interpreted PTA as the indicator of 'engagement'. New likes for the page or lack of them would completely skew the figures month to month.
- It didn't include post clicks or video plays etc., so it didn't capture much of the engagement with a page's content.
- It was being used by marketing companies as a base measure to report the % of engaged fans. But this is seriously flawed as PTA includes likes, comments and shares from non-fans. It is also a measure of the 'interactions' and not 'people', so dividing it by the number of fans is logically absurd. In theory you could end up with a PTA rate of over 100%!
IPM (Interactions per mille)
Many agencies use this measure. They simply add up the likes, shares comments per post (some may also include clicks) and divide this by the fan base and multiply it by 1000. Hence interactions per 1000 fans. Again it suffers from the same problems as the PTA. Interactions can come from your non-fan base. This figure tells you nothing about how engaged your fans are, or how engaging your posts are! Too many non-engaged fans can make your IPM look mediocre even if you have engaging posts. As clients insist on the IPM, many agencies produce this figure.
This is the relatively new metric to feature in Facebook's insights. It reports likes, comments, shares, clicks and even negative interactions on a post as a % of the reach.
This is inline with what many social media marketers are now doing. i.e. not reporting engagement as a % of the fan base but of the actual reach. The exact engagement measure differs between companies. Many just use likes, comments and shares, whereas others use clicks, and with the introduction of the Facebook engagement rate, agencies may now just report this figure.
The problem with this figure is that it doesn't fairly represent likes as a percentage of the reach! It under estimates the reach, and also underestimates some of the interactions with the post, such as replies to comments. But there is no other reach figure available from Facebook, so any engagement rate as a % of the reach will have to based on this flawed measure.
So what should Facebook be reporting as the headline figure?
I think it should be the following:
The Buzz: A sum total of likes, comments, shares, clicks, opening photos, video plays etc. for each post for the past 30 days.
But hang on a sec, shouldn't we total the unique users rather than the interactions themselves. No, I think as long as we don't include impressions, and just include actual interactions with a post, the Buzz figure would be a very good indicator of the 'practical worth' of a Facebook page. Yes there will be plenty of duplicate users when we total this figure, especially as we are doing the sum for the past 30 days, but this measure is not about counting people, it's about giving you a sense of how engaging your page is with the people who see its content.
What is good about this measure is that it provides an instantaneous representation of a Facebook page's activity and influence. It's taking into account the amount of content you are posting and also engagement (in its broadest definition) with those posts. Pages which are the most active, with the most engaging content, with the most users interacting, will deservedly have the greatest Buzz rating. And this is surely the quality we are after from the headline metric.
You can make up for your smaller fan base with increasing activity and engagement.
The Buzz will give a better sense of how sticky and influential your Facebook presence is. A client can look at this figure and immediately get a sense of how they did in relation to last month and how they compare to their competitors.
Don't get me wrong the number of fans, engagement based on reach and all those other measures have their place. But not as the number one headline figure which the public can see on the main page.
The Buzz doesn't take into account views of the page itself, or views of tabs within the page, or interaction with the page. But I think most social media marketers would agree that it's all about the posts. We need to somehow aggregate the data relating to the posts, and this would be a simple understandable way of doing it by Facebook. What do you think?