Earlier this month, Twitter (finally) released a native analytics platform to all users.
If you don’t use a Twitter analytics tool already, this information will help pull back the curtains so you can see what your most effective content on Twitter is.
Here’s how to find and use Twitter’s new analytics. First go to analytics.twitter.com and log in using your regular Twitter credentials. Navigate to the “Analytics” section of the menu and select “Timeline activity” from the drop down.
Below is a screen shot of the “Timeline activity” from my personal handle (@inthekisser.) The very top line tracks how many times my handle was mentioned on Twitter during the last 30 days. Going through the dates, I can see that the spikes in “mentions” were due to one of three things that happened during the past month.
1.) I attended a conference and joined the conversation on Twitter via the event hashtag.
2.) I participated in a Twitter chat.
3.) The content I chose to tweet, garnered higher than average responses.
Tip: Mark the days on your calendar that you might be doing some extra networking like conferences and events. That way you can look back at your metrics and determine what is driving mentions. [Tweet this tip!]
The second line shows your “follows” and “unfollows” over a 30-day period. You can see that I had the same spikes in followership on the days that I had the most mentions (those high-impact event and content days.) If you hover your cursor over a particular day, a short summary of mentions and follows will appear. However, a good amount of the follows/unfollows come from bots or those folks only looking for a follow back.
Tip: If someone you don’t know and aren’t sure you’re interested in follows you, don’t feel obligated to follow back immediately. If a few days pass and they unfollow you, you’ll know they were likely only looking to increase their follower base. [Tweet this tip!]
Your “recent tweets” section is where the best insights are. By default, it will display your last 90 days of activity in chronological order. Sorting by “Best” means the program will display the top 15 percent of your tweets that got the most engagement (retweets, replies, and favorites.) Sorting by “Good” will display the top two-thirds of tweets.
My best performing tweets are the ones that had several times more reach than an average tweet from me. This is mostly driven by the number of retweets but, one tweet of mine with only one retweet had 3x the normal reach. Why? It’s because of who retweeted me. @AaronStrouthas more than 21K followers so his one tweet alone reached more people than two or three retweets from people with fewer followers.
Tip: When you have the chance to mention a person or company by name on Twitter, make sure you tag their handle properly especially if they have a big following! That person will receive a notification that you’ve mentioned them on Twitter, and it will be much more likely that they will respond or retweet you. [Tweet this tip!]
The sharp-eyed people among you might have noticed that Twitter is tracking how many times the links you share get clicked on. And it tracks clicks, whether or not you use a “link shortener” like bit.ly.
Twitter doesn’t seem to be counting these clicks as part of their “engagement score.” This means that your most clicked-on content may not show up in the “Best” or even “Good” categories if it wasn’t accompanied by retweets and/or mentions.
So…why are these clicks not part of their engagement score? In my book, clicks are not only a measure of engagement and interest but also a very important metric because they are a website referral. Social analytics programs like Hootsuite and Simply Measured have reported on click metrics for years and it’s one of the main success metrics my team uses to measure @ConstantContact’s success. Tracking referrals to your website is important but, you can also learn what content is most interesting to your audience by tracking the clicks on links that you curate from around the web.
Tip: Content that gets a lot of retweets is often very different from the content that gets clicked on. Content without links that is humorous, poignant, data-driven, and short (120 characters or less) will typically get the most retweets. If you do include a link within a tweet, make sure it refers to a high-quality destination like your website. There’s a time for clicks and a time for retweets. You need to reach people and drive them somewhere of value. Think about what you’re trying to achieve and tailor your content accordingly. [Tweet this tip!]
The “Followers” section (go back up to the “Analytics” menu and select “Followers” in the drop down) gives you a little snapshot of your follower base.
What this shows is your growth in followers over time, the gender and location of your followers, the topics they are interested in as well as the other Twitter accounts they follow. This information can help you figure out not only what to tweet but when to tweet.
For example, I can see that most of my followers are into startups, technology, marketing, and business. This is great because I know that I’m creating a followership based on my industry and their interests.
The “location” section is where it gets interesting. It’s no surprise that the majority of my followers (20 percent) are from Massachusetts, since I am based in Boston. However, another 20 percent of my followers are based in the UK — a full five hours ahead of my Eastern Time Zone. If I want to reach most of those followers, I’ll have to consider the timing (not to mention the content) of my tweets.
Tip: Get the most reach out of localized content by choosing your timing carefully. If you’re trying to reach folks in Sydney, Australia, don’t tweet at 2 a.m. Sydney time. Regardless of their time zone, early morning, lunchtime, and after 6 p.m. are the best times to tweet. Why? Many folks are less likely to be working during those hours and more likely to have a few moments to check Twitter while they pick up their kids at softball practice, stand in line at the grocery store, or watch a TV program. [Tweet this tip!]