This article was originally published as a guest post on socialfresh.com

Here they are! After several weeks of suspense, Facebook has finally rolled out their new page insights for everyone! And, when they claimed this would be a total revamp from the ground up, they weren’t kidding, the old interface looks totally obsolete right now.  It really is a brand new world.

Facebook did a great job improving the user interface and readability for their metrics in many ways, however, there are still many serious flaws in the native insights that only third party tools can help you with.

I’m not going to give you the full tour, others have already done so, such as that great article from Jon Loomer.

Instead, I’m going to focus on the improvements, the main flaws and how to supplement them with third party analytics tools such as Agorapulse.

Related Post: How to Use Facebook Audience Insights for Better Targeting

 

1 – What rocks in the new Facebook insights

    o
  • Better navigation

Page admins complained a lot about the old interface, and they were right to do so.  Where insights used to be clunky and simplistic, the new navigation contains the multiple levels of detail page admins need, and is organized into categories that actually make some sense.  Finally!

1 - Old vs new nav bar

Lots of information that was previously displayed as text only is now a graphical interface (such as the like source or the negative feedback types).

 

    o
  • More graphs and visuals

The old interface had some graphics, but they really looked outdated and there were definitely too few of them. The new version is literally packed with graphics. Where you previously had access to 10 different types of graphs, you now have up to 18- that’s almost twice as much!  And they definitely look as they were designed in 2013, not 2009… Thumbs up on this one!

2 - Old vs new like source

Lots of information that was previously displayed as text only is now a graphical interface (such as the like source or the negative feedback types). These graphics beat the hell out of an Excel export!

 

    o
  • Convenient date selector

In the old interface, it was possible to select a date range for your Facebook metrics, but the setup required you to enter the concerned dates, click apply, then refresh the page to see that data. Frequent naps and aspirin required!  In the new insights, you simply need to scroll left and right to see the dates you’re looking for.  Bonus- all the graphs within the page automagically adapt to the new date range.

This may not sound like a huge improvement, but it’s so fun and much more easy to use!

3 - old vs new date selector

 

    o
  • Comparison versus periods is now clearer

Comparing the performance of one period against the equivalent preceding period (week by week, month by month) used to be a pain. Basically, you had to look at the data for each selected period and conduct a visual check to compare both. Then, to top it all off, it generally wasn’t accurate.

The only place where you actually had a sense of the evolution of your performance over time (one week only) was the overview screen where you had a little sign telling you if you were up or down (and the percentage thereof) for three metrics only (likes, PTAT and weekly total reach).

With the new insights, not only can you benchmark all of your metrics performance evolution over time, but you can also choose the periods you want to compare (not just the current week against the preceding one). This really rocks!

4 - new vs old period comparison

Comparing performance over equivalent periods of time no longer requires aspirin

 

    o
  • More detailed data

The old insights barely scratched the surface of a great deal of very interesting data. For example, you could find your negative feedback data but only on a post per post basis (not over time) and you only had access to the global numbers of negative feedback, not the split of its various types (hide, hide all, report as spam). Now, the negative feedback graph gives you a clearer picture with a detailed view of all of its components.

Same with the likes: the paid likes were buried in the excel download, and determining your net likes gained required doing the math on your own. Fun! Aspirin anyone? Now they are all conveniently displayed in an easy to interpret graph.

Likes, comments and shares numbers that were previously only available through the excel download or by digging through the post insights display then clicking individually on each post (God, I’m tired just thinking about it) are now conveniently displayed in a good looking graph.

Well done, ZuckDawg.

 

2 – What seems like it’s rocking, but actually is not

    o
  • When fans are online

The “When your fans are online” graph is one of those features that made the ‘experts’ a little crazy upon discovering it. That title sure is promising!  Even more promising is the baseline statement: “Knowing when the people who like your Page are on Facebook can help you share posts when they’ll be seen”. Woo-hoo!  No more guessing about the best time or day to post your content!  Call a caterer and the landlord.  There’s about to be a party in this office.

Wait.  Check again.

The truth that most experts missed is that this graph does not tell you when your fans saw your content on Facebook by day of the week or time of the day.  It only tells you “The average number of your fans who saw any posts on Facebook by day of the week / in an hour”.  Read that again.  It’s the “any post on Facebook” part that creates a rift between the promise you thought this would be and what it actually is.

5 - When fans are online bis

Facebook tells you when your fans are online. Looks great on paper! Yet, by looking at the time of the day graph, you’ll learn that you have more fans online during the day than at night, and when looking at the days of the week, you basically have the same number of fans online everyday. Not a big help at the end of the day. Worse, the data seems sometimes buggy as in this screenshot where there seems to be no fan online at all on Monday!

Basically, this tells you when your fans are on Facebook.  It does not tell you when your fans see your posts. Like us, you’ve probably discovered the majority of your fans are on Facebook from about 9AM to 10PM on any given day.  I don’t know about you, but that totally blew our minds.  Golden light…  Angels singing and everything!  Thank you Facebook!   😉

When it comes right down to it, this stat is pretty useless, you don’t need a graph to tell you that your fans will be online (and on Facebook at some point) when the sun is up. What you really need to know is when your fans are actually seeing your content!

Due to edgerank, friends and fans and brands are fiercely competing to appear in your fans’ newsfeeds, not all days of the week or times of day are equal.  The majority of your fans may be on Facebook from 11AM to 7PM, Monday thru Friday, but it’s also likely this is when everyone else is posting their stuff for your fans to see, burying your content in a crowded place.

Maybe the best time for your content to surface in your fan’s newsfeed is actually during the weekend. Or, between 7:00 to 10 PM.  Facebook’s native insights won’t tell you this.

The only way to know the best time for your content to reach the greatest percentage of your fans is through a third party analytics tool which creates the same type of graph based on YOUR content’s average performance

6 - Users reached agorapulse

The main difference between Facebook native insights and a third party Facebook statistics application, such as Agorapulse, is that the statistics will not tell you when your fans are online but when you have the best chance to actually reach them and get engagement from them. For example, if you look at days of the week in Facebook, the number of fans online is the same every day, not a very actionable data. If you look at Agorapulse statistics, you can see that on Sundays, this page was able to reach almost twice as many fans as on Thursday (the best weekday) and three times as many fans as on Tuesday (the weakest weekday. That’s data you can act on. Post more on Sundays than on Tuesdays.

 
 

    o
  • Best post type

Here’s another stat everyone got excited about. Now that you know when to post (when the majority of people are not sleeping ;-), Facebook will now tell you what to post! Be ye thankful!

Yeah, unfortunately, this stat doesn’t tell you much, nor does it give you actionable data you can make any use of. What will you learn? Probably the same thing for every page: status updates have less engagement and no clicks, photos have more engagement and tons of clicks, videos and links have usually less reach than status updates and less engagement than photos.

7 - best post type Facebook

The best post types stat is somewhat interesting, although it lacks really actionable data such as how many times you’ve posted each type. If you’ve only been posting one great link or 25 photos or status, in this example, the data you get is not enough, you may have been lucky with that one link but it does not mean that you need to publish only links now. In addition to that, you have no information about the most important information: what is the type of content that resonates the most with the audience. You know what format works, but not what content.

What would actually be useful would be to know which “themes” are most effective, not which “format”.  Format performance won’t help you develop your content strategy. Are you going to post more status updates just because Facebook tells you they generate the greatest reach? Or only publish photos because they get more clicks?

You need to know how your content is performing with your audience, and what type of message gets your fans interested.  You need to know that promotional content only gets you half the reach or engagement as “how- to’s” or product reviews.  And here again, only a third party analytics tool can help you out.

8 - Reach by tag

With a third party Facebook statistics application such as Agorapulse , you will not only know what format of content reaches the most fans, but you will also know what subject matter (and type of content) gets the most attention. In this example, news, politics and promotions are reaching much more fans than posts about events or non profits. That defines more the direction your editorial strategy should take.

 

    o
  • Average comparison

This is probably the worst looks-good-but-provides-no-real-value new idea. The principle is simple, Facebook will show you how your page’s fans demographics compare with the overall Facebook users’ demographic.  How useful and actionable is that?

9 - Fan demographic compared

Facebook will tell you how your fans’ age & gender compare with the overall Facebook audience. Not very actionable.

In a nutshell, if you are a motorbike brand like the example shown in the screenshot above, Facebook tells you that your fan base contains 30% men aged 25 to 34, where the average population of men between 25 and 34 has 14% of men. If you previously had no clue that your motorbike brand appealed to males between 25 and 35 rather than women in their 50s, this is great stuff. But I assume you were probably aware of that already, weren’t you?

It’s a nice try on Facebook’s part to provide some context to benchmark your data, but… yeah.

More on “context” below.

 

#3 – What sucks in the new Facebook Insights

    o
  • No context

Context is what gives sense to a metric. For example, when you look at a conversion rate or a bounce rate in your usual social media analytics tools, you have a rough idea whether it’s good or not, because this kind of data has been around for a long time and you have some context to put it in. If you know that, in your industry, 1% is an OK conversion rate, then putting your own 0.5% or 2% conversion rate in that context will allow you to understand how you’re doing.

Remove that element of context and your 0.5% or 2% conversion rate become a meaningless metric. Is that good? Lame? Beyond lame? No clue.

Context is everything when it comes to understanding metrics.

That context can be your usual average. For example, landing page A/B testing usually takes your own average metrics into account. If your landing pages convert 1% of visitors on average, then a new test landing page that gets 1.5% is a good one.

Context can be your competition, your industry, companies of the same size, from the same location… whatever.   It’s always a benchmark that makes sense for your business.

Apart from the failed attempt concerning your fan demographic mentioned above, Facebook doesn’t give you a single element of context. As a consequence, you’re left wondering if all the stats Facebook throws at you are good news you want to share with the world, or something you should keep to yourself.

Third party analytics tools are the only way to get some context and put some light on your Facebook metrics.

  • No reporting tool

I manage Facebook pages too. I also have people working for me who manage Facebook pages for clients. And at the end of the month, the moment of truth comes. Have we done a good job? Did we improve? Are we going the right direction? Have we spent too much time moderating the page?

Your client or your boss wants to know what you’ve been doing for the last month and if the money he’s paying you is worth it!

It can be very basic, or very detailed, but a report is mandatory.

Is Facebook helping you build that reporting? Not really. Yes, the metrics are there, but you’ll have to take the time to adjust the period, make the necessary screenshot, filter through the noise (what the client/boss wants to see and what they don’t care about) and create the report in PowerPoint.

For some metrics, it’s even worse. If you are moderating a page for example, the number of posts and comments posted during a given period of time, or the number of fan’s content reviewed is nowhere to be found. You’ll have to use your calculator or Excel spreadsheet and take the time to do the math yourself.  Aspirin again!

At the end of the day, where Google Analytics allows you to create pre-defined reports that are automatically sent to your inbox at the end of each defined period of time, Facebook only leaves you with the time consuming copy and past option.

Here again, only a third party analytics tool will save you unnecessary hours of very low value added work.

11 - Agorapulse reports

Third party Facebook statistics tools will provide reporting features that will save an awful lot of time to the Social Media Managers who have to produce reporting for their clients or bosses.

11-1 - Fan content moderation stats

Third party analytics tools will help you monitor your community management activity. Pretty helpful if you have to report that to someone (a boss or a client).

 

 

    o
  • No more viral data

Virality is what made Facebook so attractive for brands in the first place, along with the promise that whatever you did on Facebook, your fans were likely to tell their friends about it.

The one and only native metric measuring the effectiveness of the viral power of Facebook was viral reach- the number of people who saw a story from your brand because their friends (your fans) engaged with it.

For some pages, that viral reach was so effective it even surpassed their organic reach, giving them double the number of content views than their fan base!

12 - viral reach

In this example, the page Being Liberal reaches 100% of its fans during a 30-day period, but it reaches 50 times more friends of fans through viral reach! This is the kind of viral amplification you would like to know if you were the manager of that page. As of today, Facebook will not provide that information, only a third party analytics tool will.

That metric is now gone from the new insights. The only metrics left for measuring reach are organic and paid. Is organic now encompassing viral? I don’t know. What I do know is that you’ll have to download the Excel spreadsheet to access your viral reach, but I’m not even sure this will last!

Was this decision linked to the fact that viral reach has been mistakenly reported as being very low from Sept 2012 until February 2013 due to a bug ? Maybe.  Maybe not. No communication from Facebook on this issue. Even though they promised they would communicate more on what they do now.

Bad news Facebook, really bad news.

 

    o
  • No percentage values

The only way to compare each metric for one given post or a given period is to create a percentage out of it. That way, you get a number that will be consistent post after post (or day after day) and you can benchmark them against each other.

For example, If you rely on the raw number of engaged users, you’ll never know if a high engagement is due to the quality of the specific content or if it was just shown to more people in the first place. What you need to assess the level of engagement accurately is to divide it by the number of people that piece of content reached and multiply it by 100. That way you get a percentage of users who engaged with a post that was displayed to them and you can easily compare that percentage with the same for other posts, or a different period.

7-formula

Evaluating the engagement on a specific post should always be made by comparing the number of engaged users with the number of users actually reached to obtain a percentage of engaged users. That way, you can compare each post engagement to a benchmark and determine accurately which ones are winning the battle.

14 - Post metrics on Facebook

When looking at the post view on the new Facebook insights, you only get the global reach and the total number of clicks on the one hand, and likes + comments + shares on the other hand. You’ll have to do a lot of math to benchmark each post against one another.

 

Facebook will only give you raw numbers. Taking the engagement metric as an example, you’ll get that post N°1 got 1,678 people engaged while post N°2 got 1,898 people engaged. So post N°2 did a better job right?

Well, if post N°2 has reached twice as many users as post N°1, it did pretty poorly compared to post N°1. And that’s the kind of information you need. Facebook will not help you with that, only a third party analytics tool will help.

15 - post performance

The post performance analysis we have built within Agorapulse shows each post’s metrics as percentages and compares them with the average performance for the concerned metric. That allows us to generate an easy-to-understand heatmap that shows you the top performers in a matter of seconds. No need to do the math to find out what posts got the best results. Also, you can get that view for more than 20 post metrics where Facebook only gives you 3: total reach, clicks, likes+comments+shares. Too limited if you want to dig a little.

 

    o
  • Lots of data, but not enough actionable data

OK, ok, I know, I’m complaining a lot lately, and that’s right after I praised Facebook for providing much more data than they used to with better visuals.

It’s true, there’s more data, and it looks better. But, we don’t need more data or more beautiful graphics, what we really need is more data we can understand and act on.

And at the end of the day, even if I think they did a better job this time, I still think there are far too  few metrics in these new insights that can help us make decisions on what we should keep doing, what we should stop doing and what we should do differently. Actionable data.

 

    o
  • Only 3 months of history

Going back in time is one of the best ways to understand the evolution of the performance of an online asset such as a website or a Facebook page. This is one of the things I keep doing when looking at my Google analytics account. And sometimes, I appreciate the fact that I can go back several months in the past to look at the evolution of my traffic or conversion and identify the impact of my marketing efforts over time.

With the new Facebook Insights, you are limited to 3 months of past data. This is probably enough in most cases, but you’ll find yourself frustrated if you want to verify if things have grown since January 1st of this year, 9 months ago.

Here again, with a third party analytics solution, your data can be kept for a much longer period of time, and this will allow you to report on your progress over periods longer than 3 months. That can be very useful.

 

    o
  • Only 7 days worth of overview data

This isn’t my biggest complaint, that’s why I kept it for the end.  When you look at the overview page, the one you land on when you get to your insights, the only period available is the last 7 days.

17 - last 7 days

Like most of us, I’m used to conducting and providing monthly reports.  A month’s worth of this type of overview sure would beat the hell out of capturing screen shots, saving them, then mixing them together somehow.

Sigh.  One last aspirin?