Buffer · Follow
16 min read · May 4, 2017
--
We social media marketers are a lucky bunch; not only are there a gigantic number of free social media tools at our disposal, there are free tools from the social networks themselves: Facebook Insights, Pinterest analytics, and now Twitter analytics, too.
We now have tons of information on hand.
The ideal next step: Knowing that we’re looking at the right numbers and drawing the right conclusions from the data, simply and easily.
When it comes to Twitter analytics and Twitter metrics in general, I’ve done a bit of research into what exactly we can learn — as straightforward as possible — from all the awesome numbers Twitter displays. Read on to see what I’ve come up with and if you can gain some insights from these stats, too.
I’ve found a lot of inspiration in the way that others have used Twitter analytics to find the stats and insights that help them tweet better. Here is my collection of 17 favorites, as well as how you can find these stats for your Twitter profile.
- Monthly performance overview
- Trend insights
- Average tweet performance for benchmarking
- The top 10 interests of your followers
- Type of engagement
- Your most significant sharers
- Engagement rate
- Twitter Like rate
- Tweet length vs. engagement
- Tweet reach percentage
- Tweet reach, including your retweets
- Hashtag comparison
- Impressions by time of day
- Engagement rate by time of day
- Best days for total engagement and engagement rate
- Video completion rate
(Bonus: I’ll be happy to share all the Buffer insights, too.)
What are your top tweets?
On the homepage of your Twitter analytics, you’ll get a 28-day summary of your Twitter account and a summary of each of the previous months.
The monthly summary is a great way to get a quick overview of your Twitter performance for each month. On the left, you’ll see your tweet highlights, which includes:
- Top tweet: The tweet that received the highest number of impressions
- Top mention: The tweet that mentioned your @handle and received the highest number of impressions (This can include other people’s tweets)
- Top follower: The account with the highest follower count that followed you in the month
- Top media tweet: The tweet with photo or video that received the highest number of impressions
- Top card tweet (if you use card tweets): The tweet with a Twitter Card that received the highest number of impressions (This can include other people’s tweets)
Here’re a few ways you can use these insights:
- Top tweet/media tweet: Reshare it in the following month — with a new copy
- Top mention: Retweet the tweet
- Top follower: Reach out, thank the person for following, and start a conversation
On the right, you’ll see a few key stats of your Twitter account for the month. These numbers are great for reporting purposes — to show your monthly performance and determine month-on-month growth.
If the month is not over, Twitter will show you the data for up to the current date of the month.
How to find this for your profile: Go to analytics.twitter.com, and it’ll be right there!
How are your social media strategies playing out?
When we focus on the day-to-day activities of social media, we can sometimes miss out on the bigger picture. Taking a step back and looking at the trends of our social media performance can inform us if our strategies are working out in the long run.
Is there a general upwards trend? Or is it flat?
Once you spot any trends of your Twitter performance, it can be helpful to dig into and understand the trends. Here’re a few helpful questions you can try to answer:
- What caused any spikes or valleys?
- When will we reach the next milestone?
- Where will we be in the next (x) days or months?
- What factors are we not accounting for?
How to find this for your profile: In the Analysis Report of your Buffer Analytics tab, you can get graphs of several metrics such as clicks, Likes, and followers.
(The spike in clicks was due to the launch of our new online social media strategy class.)
If you prefer using the native Twitter analytics dashboard, you can find the trend data for five engagement metrics on the right of the “Tweets” page.
How are your tweets performing against your goals?
As seen in the graph above, Twitter stats tend to fluctuate on a daily basis. This makes it challenging to compare your performance for a period (e.g. a month) with the period before.
A good solution for this is to use averages.
Averages smooth out the fluctuations to make comparisons easier. You can easily compare your average tweet performance for this month with the previous month and quickly determine if your Twitter performance has improved.
How to find this for your profile: You can find your averages on the right of the “Tweets” page in your Twitter analytics dashboard. They are right below each of the graphs.
Once you have found your averages, you could use them to set your social media benchmarks and analyze your performance.
What topics resonate your followers?
Alex Bossenger of Social Media Examiner wrote a great overview of what all you can get from Twitter analytics. I often think of Twitter analytics as the data behind the tweets. But there’s complete data behind followers, too!
Twitter shows you the top 10 interests of your followers. There’s a little bit of mystery behind how Twitter finds this number. The official word from Twitter’s documentation is:
The top interests that distinguish your followers from the Twitter average
Use these interests as a good guideline for content and ideas, even if the scientific explanation isn’t obvious.
How to find this for your profile: Go to analytics.twitter.com and click on ‘Audiences’ in the menu at the very top of the page. Interests will be on the main followers dashboard.
If you want to learn more about your followers, you can click over to the “Demographics” or “Consumer Behavior” tab to see their age group breakdown, country, buying styles, and more.
Are you being retweeted or liked?
If you happen to see that your engagement numbers are really high for a particular handful of posts, does that mean that each of these posts performed the same? There’s likely more to the story.
You can get deeper into engagements by noticing what specific types of engagements took place. For instance, were they retweets or likes?
Retweets can be a sign of value. Someone found your tweet valuable enough to share with their audience.
Likes can be a sign of appreciation. Your tweet resonated with someone else, and they wanted to give a virtual high-five.
Both metrics count as engagement. Both tell a unique story.
How to find this for your tweets: You can access this breakdown quickly and easily for an individual post straight from your Twitter analytics dashboard; click on ‘Tweets’ in the top nav bar and then click on ‘view tweet activity’ for any tweet from that list.
Outside of your Twitter analytics dashboard, you can also get this breakdown by clicking on the bar chart icon on the tweet you are interested in.
Which influencers are sharing your content?
The example I used in the above stat — the tweet with a high number of retweets — the total potential audience was over 1,000,000 users.
We certainly do not have 1,000,000 followers on Twitter, so there must have been some influential folks who retweeted the content. Fortunately, there’s an easy place to check this — and then work on follow-up and thanks to those who shared.
Here’s how to find this for your tweets: From the Posts Report of your Buffer’s Analytics tab, find the tweet you want to analyze. Hover your mouse over the Retweets stat and click.
Bingo! The stat expands to show everyone who has retweeted you plus their follower count.
For instance, in the example below, we owe a huge thanks to all who retweeted, and I can see that a few people made a big impression.
Which posts resonate the most?
There are a couple murky stats that could trip you up in Twitter’s tweets dashboard. Engagement rate isn’t one of them.
It’s super helpful because it shows the number of engagements divided by impressions.
In other words, out of everyone who saw the tweet, what percentage of people did something with it.
Twitter counts engagement anytime someone clicks anywhere on the tweet, including:
- retweets
- replies
- follows
- favorites
- links
- cards
- hashtags
- embedded media
- username
- profile photo
- tweet expansion
How to find this for your tweets: Visit the tweets dashboard, and look in the far right column in the table of tweets. Engagement rate will be listed after Impressions and Engagements.
You can also get this information — and much more — in the exported data offered by Twitter analytics.
The rest of the stats here will reference the exported Twitter data, so get your spreadsheets ready!
From the dashboard, go to “Tweets” and click the export button in the upper-right corner.
Take the downloaded file and import into Excel or Google Sheets. It will look something like this spreadsheet.
Twitter will show you data for up to 3,200 tweets, including a breakdown of all impressions and other engagement numbers.
Pro tip: You can remove all of the @-replies from the spreadsheet by sorting the “Tweet text” column alphabetically and deleting the rows that begin with an @ symbol.
What are your best tweets?
Dan Shure of Evolving SEO came up with a unique stat that can help you predict the success of a piece of content. He calls it favorite rate (or perhaps now, Like rate).
Twitter users hand out favorites for any number of different reasons, and one of those could be that they quite like the idea you express in a tweet. Like rate helps surface the best ideas. It measures favorites by way of impressions, highlighting the posts that received the most love relative to how many people saw it.
Here’s the way that Dan ties it all in to content creation:
For Twitter, it’s probably safe to assume that tweets that people have favorite would make for good article, blog posts, tools, ideas — because your audience has told you they like the idea with their engagement. At the least, the contents of those tweets for some reason got their attention.
How to find this for your tweets: Divide likes by impressions. Format the resulting data as a percentage. Voila!
What is the ideal length for your tweets?
We’re always keen to learn data on the ideal length of tweets, so this one could prove really insightful.
What you’ll learn from this is exactly how the length of your update impacts the way your audience interacts with it. Are they blind to long, 140-character updates? Do they prefer short or medium? Once you pick up on the general trends, you can even drill deeper to see what kind of engagement (for instance, retweets vs. favorites) happen at each particular length.
How to find this for your tweets: To find this stat, you’ll need to add a new column to your spreadsheet.
Create a new column, and copy this formula into the cells:
=LEN(A1)
(Change “A1” to be the cell that contains the tweet text.)
The above formula will give you the character count of the tweet, including any article URL and image URL. (Image URL are not counted in your tweets so this character count might exceed 140, which is okay since we are mostly interested in any general trends than a precise character count).
To get the word count, try this formula:
=IF(LEN(TRIM(C2))=0,0,LEN(TRIM(C2))-LEN(SUBSTITUTE(C2,” “,””))+1)
Armed with this data, you can then compare your engagement with your tweet length. One neat way to do this is to run a scatter plot with the data in the Engagement column and the Length column.
Here’s the data for our last 150 Buffer tweets (excluding viral tweets).
How many of your followers do you reach?
Facebook reach is a rather ballyhooed statistic. Now that we have impression data from Twitter, we can essentially find the same stat for tweets.
One of the best takeaways from this stat is that you’ll discover whether or not you should be resharing your tweets multiple times in order to hit more of your audience (chances are, you should be resharing).
How to find this for your tweets: Since you already have your impression stats, you simply need to divide impressions by total followers.
Make a new column in the spreadsheet, and add this formula for each tweet:
=Impressions/Total Followers
(802,920 is the number of followers we have during the calculation. Your follower count will likely have increased throughout the period of your analysis. Use your latest follower count to give you more conservative calculations.)
Once you have this column, you can calculate the average reach for all your tweets. On our @Buffer account, we reach about 2 percent of our followers on average with each tweet.
(One area this statistic won’t take into account is the effect of viral sharing. Impressions may include people who do not follow you, so the actual percentage reach of your followers can be a bit skewed, especially if you’re promoting tweets. We’re also making a general assumption that a person is only seeing our tweet once.)
How many people can your tweet potentially reach?
I mentioned above that Twitter analytics might not account for a complete view of potential reach, especially if you’ve been retweeted. Here’s where the Buffer dashboard and Twitter dovetail nicely.
How to find this for your tweets: In your Buffer’s Analytics tab > Posts Report, you can view the potential reach of your post — your followers plus the followers of those who retweeted you.
Buffer for Business provides an export of this data that includes the specific number of potential users who saw the tweet (e.g., 1,513,289 as opposed to 1.5m).
Armed with this data, you can do one of two things:
- Figure out the specific reach of an individual post.
- Figure out the specific reach of all your posts.
To accomplish this, divide the “Impressions” data from Twitter analytics by the “Potential” data from Buffer. For individual posts, you can compare 1:1 — the individual post’s impressions divided by its potential reach (i.e. out of the actual potential reach, how many people saw the tweet).
For the total of all posts, you can collect the information in a spreadsheet, total each of the columns, and perform the impressions per potential calculation on the totals.
How do Impressions, Engagement, and Engagement Rate compare?
Let’s say you often tweet with a go-to group of hashtags, for instance #socialmedia, #seo, and #inbound. With Twitter analytics, you can see which of these tags performs best.
How to find this for your tweets: Filter your spreadsheet using a quick search of the tweet text for the desired hashtag.
For instance, this is what the filter would look like for the #socialmedia.
With the resulting cells, grab the average impressions, engagement, and engagement rate (and whatever other metrics you may want to track). Mark these averages in a separate worksheet.
Then do the same filter and averages for your other hashtags. Add these stats to your worksheet and compare the hashtags to one another.
What is the best posting time to maximize impressions?
Tomasz Tunguz of Redpoint came up with some really useful ways to break down his Twitter stats. The following handful of stats come courtesy of Tomasz. (You can see his personal Twitter insights in his full blogpost.)
First off is impressions by the time of the day.
Basically, what time of day is best to post in order that the most people will see a tweet?
Followerwonk and Tweriod are a couple of Twitter tools that can help you find this answer, based on when your followers are likely to be online. Digging into your Twitter analytics will tell you the best time based on how your tweets have performed in the past.
How to find this for your tweets: In your exported spreadsheet, change the formatting of the Time column to view the numbers as minutes/hours instead of the full timestamp.
Here’s a formula that will pull out just the date and time from the full timestamp.
=ArrayFormula(mid(D2:D,1,16))
You can paste this into the top row of a new column. The ArrayFormula will fill down the data into the entire column. Once this data is there, copy it to your clipboard and “paste as values” into a new column (preferably on a new sheet to keep things clean). This will allow you to format the numbers as you see fit.
Once you have the date and time formatted correctly, you can begin analyzing the data.
Another trick you might want to try is rounding the time to the nearest hour and then comparing the impressions at the hour level. Here’s the formula:
=ROUND(A2/(1/24),0)*(1/24)
(Change “A2” to be the cell that contains the tweet time.)
Once you have the time in hours, you can put together a pivot table (with the row set to Time and the value set to average impressions) to compare the time to average impressions of tweets.
What is the best posting time to maximize clicks, retweets, or replies?
Is there a particular time of day when your followers enjoy clicking to read what you’re sharing? How about retweeting?
In Tomasz Tunguz’s case, he found that people clicked at a higher rate over breakfast or during a commute to work.
Readers are most likely to click on an embedded link in the morning. 6am just as successful as 8–10am.
How to find this for your tweets: Same as above, except sort and filter from the clicks, retweets, or replies column.
What is the best posting time to maximize engagement rate?
How well-liked are your posts at different times of day? Impressions and views are always a nice metric to keep an eye on, and you’ll probably also be interested in how engaged your audience is with your sharing. This stat can be helpful in figuring out when your followers are most engaged during the day.
How to find this for your tweets: Same as above, except sort and filter from the engagement rate column.
Which day of the week is best for engagement?
A.J. Kohn of Blind Five-Year-Old has a number of neat stats to seek out, many based on the best day to tweet. The next few ideas below are inspired by A.J.’s testing and tweaking.
Starting with the best days for average engagement per tweet, this can tell you when to maybe bump up your sharing schedule on a certain day to take advantage of a more fully engaged audience.
You can also check engagement rate to see which day of the week an individual post stands to perform best in terms of engagement rate.
How to find this for your tweets: Like we did above for the impression by time of day, we’ll need to make a new column in the spreadsheet and add a formula:
=CHOOSE(weekday(A2), “Sun”, “Mon”, “Tue”, “Wed”, “Thu”, “Fri”, “Sat”)
A2 should be changed to reflect the cell where the tweet’s date is stored. The formula will extract the day of the week from the date.
Now you can sort and filter the engagement column and the day column. Try placing this data into a pivot table, with the column set to Day and the value set to Engagements.
Here’s how average tweet engagement per day looks for Buffer.
For engagement rate, you can also perform the same set of steps as above, except sort and filter from the engagement rate column.
Here’s how engagement rate per day looks for Buffer.
What percentage of viewers watched the whole video?
Videos are becoming a popular content format on many social media platforms, including Twitter. To help you create better Twitter videos, Twitter created a new analytics dashboard for your videos. As this feature is still under beta testing, it might change (and get even better) in the future.
Apart from the video view count, Twitter provides another powerful stat: Completion rate — the total number of complete views divided by the total number of video starts.
In other words, out of everyone who watched the video, what percentage of people watched the entire video.
How to find this for your tweets: In your Twitter analytics dashboard, select “More” from the top nav bar and then “Videos (beta)”. Look at the far right column in the table of tweets. Completion rate is listed after Video views.
First off, it’s amazing that Twitter has handed over such detailed analytics to users on the engagement and statistics behind tweets.
If you haven’t yet, access your Twitter analytics dashboard now (it’s free and easy), and have a look around.
When you’re ready for the next step, try out one of the 17 vital Twitter stats above and see if what you discover might help you tweet better. Need a starting point? I found the statistics per day and per hour were most fascinating to me.
What Twitter stats have you come across for your profile? Which of the above have you tried or are you interested in digging up? I’d love to hear from you!
