Something strange is happening in the world of Twitter. For well over a year my average Twitter impressions count has hovered at 2.0K impressions per day. (Twitter defines an impression as anytime a Twitter user sees your Tweet.) I’ve never seen that figure deviate by more than 5% (i.e. between 1.9 — 2.1K) for a long time.
And then, starting at the beginning of March, I’ve watched my daily impression count steadily rise to 2.7K per day. Every time I check, it’s gone up. That’s a 35% increase in one month!
I haven’t changed my tweeting activity in any way recently. I continue to post each day:
One tweet about my latest blog post
One tweet for each of an assorted selection of 8 of my older blog posts;
A tweet or two about my upcoming Participate! workshop; and
A few mentions or retweets of other users I find interesting.
All my other Twitter monthly count statistics: engagements, link clicks, retweets, likes, and replies are essentially unchanged.
Yet Twitter insists that suddenly, 35% more people are seeing my tweets!
I would like to think that I’ve suddenly become 35% more interesting, but I doubt that’s what’s going on.
Is anyone else seeing this? Any ideas about what could be happening?
A social media (SM) platform feed that is not a chronological list of all posts, and only those posts.
I am sick of social media platforms deciding for me what I should see. (Read this entire post for a rare exception.)
Every major social media platform started with a simple chronological feed of all the posts from all the people you chose to follow/be “friends” with.
But every platform subsequently FUBARed their social media feed. They dropped some posts entirely, and added content that you had never asked for.
Facebook’s news feed has been FUBARed since 2009.
YouTube has been messing around with search ranking of videos since 2012.
LinkedIn’s homepage feed became FUBARed in 2016, when the company folded in its “Pulse” content.
Twitter’s timeline got FUBARed in 2016. (You can turn off their adding “tweets you are likely to care about most”, but they still insert tweets “we think you might be interested in” into your timeline.)
Instagram stopped chronologically listing posts in 2016.
Google Plus has fussed around with its home feed algorithm for several years, but it doesn’t matter because — and I wish this wasn’t true — Google Plus is dead.
Here are three reasons SM channels do this. Feel free to add more in the comments. They:
Want to add sponsored paid content (aka advertisements) so they can make money.
Want to create incentives (like “boosting” your Facebook posts) to pay them to let more people see your posts.
Know that some of their users — the ones who “friend” or “follow” everyone — would quickly withdraw from their service if the resulting torrent of requested posts were actually provided.
No one likes reason #1 but we understand why SM platforms do it. They need to make money to stay in business. Fair enough.
Reason #2 is really damaging to the concept of a SM channel as a reliable communications tool. Old media doesn’t have this option: when you buy the paper or watch TV you receive the same content as everyone else. But today, two users who follow the exact same people on Facebook may see very different feeds, thanks to Facebook’s secret and ever-changing algorithm. Essentially, Facebook makes the feed unreliable so the company can make additional revenues. I find this unreliability infuriating, and it’s why I use Facebook as little as possible.
Reason #3 is understandable — but what’s annoying is there’s no way to turn this behavior off! It would be easy for SM channels to default to their algorithmic filtering but provide an option for users to say, “Just give me everything I’ve said I wanted to see. Yes, I know I’ll still get all the ads you insert, but I’d really like not to miss anything else.” I guess that they worry too many people would choose to see everything, and the incentive for organizations to pay them to boost content eyeballs would be reduced. In fact, I suspect that only a small percentage of users (like me) would pick this option.
Ultimately, I want a social media channel that doesn’t filter. I suspect I’m not alone. Let me pick what I want to see, and let me see it. All of it.
Is that too much to ask?
A reward for those who’ve read this far — an unFUBARed SM feed! I know one way to get an unFUBARed SM feed. From Twitter, no less — use Twitter Lists! If you create a private Twitter list of people whose tweets you want to see, you can view the resulting stream on Twitter at https://twitter.com/YourTwitterID/lists/NameOfYourList or via other Twitter clients like TweetDeck and the list is chronological and unfiltered! (Please don’t tell ’em; they’d probably FUBAR it immediately.)
Do you know other ways to get unFUBARed SM feeds? Feel free to share in the comments!
The effectiveness of Twitter as a connective social media channel is declining In July I wrote about why 2017 is a tipping point for Twitter, noting that the rate at which users follow established accounts has slowed dramatically. As the year draws to a close I’m seeing further evidence that conversations in the twittersphere are drying up too.
After years of predictable behavior, Twitter analytics reveal that something strange is going on with how Twitter is used.
Something is happening to Twitter, but you don’t know what it is. Do you, Mr. Jones?
I started tweeting 12 years ago. Though I didn’t know it at the time, Twitter would turn out to be the most important way for people to discover my work and for me to connect with thousands of kindred souls all over the world who share my specialized interests. Over time, Conferences That Work grew into a website with ten million page views per year.
But as 2016 drew to a close I noticed that something was changing in the Twitter world. Here’s a graph of my follower count over time:
What I’ve noticed about my Twitter analytics over the last nine months Since I began posting in June 2009, the graph shows that I consistently added between two to three followers per day — until around September 2017. At that point, highlighted by the red circle, there was an unusual increase to ~six followers/day for the remainder of 2016, followed by a sudden flattening that has persisted through the first half of 2017 to less than one follower/day.
In 2017 I’ve also noticed a dramatic reduction in the number of retweets I’ve been receiving. Though I haven’t had time to develop quantitative statistics, it looks to me as though in 2017 retweets have been replaced to a large extent with likes, (though the frequency of mentions seems more or less unchanged).
Why are these changes happening?
I’ll begin with a caution that everything that follows is ultimately speculative. I can’t say definitively what is going on, and can think of multiple plausible reasons for these significant changes. For example:
My experience may not be representative of other Twitter users. A sudden surge of engagement with my posts during Q4 2016, was followed by a rapid loss of interest.
The US election results caused more people to visit Twitter for a few months, but attention eventually shifted to the continuous torrent of breaking news at the expense of general engagement.
Twitter user growth has been flattening for some time as per the graph below; my 2016 EOY bump is reflected in the graph’s Q1 2017 bump, but future official statistics will show little continued active growth.
In retrospect, 2016-2107 will be seen as a period when late adopters continued to join Twitter, but a critical mass of active users concluded that engagement on the platform was not for them and moved to other social media platforms (I’m thinking Instagram for one). Although Twitter seems to be doing well in percentage market share of social networking site visits, as per the statistics below, it’s becoming more a site that users visit for breaking news — engagement is moving to other platforms.
What do I think is actually going on?
I’d put my money mainly on #4 above. Perhaps this Twitter analytics trend has been accelerated by #2’s associated flood of U.S. breaking news (61% of my followers are in the United States). It will be interesting to see if the trend continues, which may help to shine more light on what, to me, are changes that are interesting and important for anyone who uses Twitter for connection, content marketing, and engagement.
What do you think is going on? Add your ideas in the comments below!
Are you attempting to build social media followers by drive-by following—i.e. following a batch of new accounts every day, waiting a day or two, and then unfollowing the accounts that don’t follow you back? STOP THAT! You are trashing your brand.
Many people with social media bios designed to project a professional image destroy their credibility by using this “strategy”.
I suspect these are people that would never stoop to buying followers or likes. And yet ~30% of my daily new Twitter followers are drive-by followers.
Drive-by following backfires because it ensures that I’m extremely unlikely to want to have any kind of social media connection with you.
Here’s how it works on Twitter, my most important social media platform. I do my best to read the profile of every new follower. Rarely will I follow back right away unless you’re someone I know. Birdbrain, the excellent app I use to track Twitter followers, also shows anyone who’s unfollowed me. That’s where I get to notice that you’ve drive-by unfollowed me, typically within 48 hours of your initial follow. That’s when I make a mental note that you’re not a serious user of social media, just someone chasing a high follower count.
Instead, follow for a bit and post interesting stuff (I admit that mentions and RTs of me are nice too!) I may well follow you back.
What’s worse than drive-by following? Repeated drive-by following! I routinely see accounts commit multiple drive-bys, usually a week or so apart. My conclusion: either you are using a second-rate automated drive-by service, or you have a memory even worse than mine (which is saying something). Either way, your attempt to get me to follow you back is even less likely to succeed.
If you want to use social media as an effective marketing platform, don’t broadcast stuff about yourself all the time. Don’t implement elaborate plans solely designed to maximize your followers. Instead, post interesting stuff (both yours and others) and interact with people. Keep doing this. Over time, if you’re doing a good job, your followers will grow and be genuinely interested in your social media presence, and your brand recognition and value will increase.
“What does matter, however, is how many people notice you, either through retweets, favorites or the holy grail, a retweet by someone extremely well known, like a celebrity.” —Jenna Wortham
She then laments: “Twitter is starting to feel calcified, slowed down by the weight of its own users, cumbersome, less exciting than exhausting“.
Most of the comments on her post go even further than Jenna, smugly dismissing Twitter as a waste of time—unless you’re a narcissist.
“I can handle Twitter because it is irrelevant.” “…this writer sums up exactly how I feel about social media in general, not just Twitter. This whole idea of likes and followers — it’s like setting up one’s business based on some vacuous high school popularity contest. Are we grown ups or not?” “Brevity may be the soul of wit, but I find little soul in twit. (er)” —The three most popular comments on Jenna’s article
When you see Twitter solely as a broadcast tool, you are overlooking its most important use: as a tool for discovery, conversation and connection.
On this site I write about a niche topic: participant-driven and participation-rich events. For me, Twitter has turned out to be the most important way for people to discover my work and for me to discover and connect with thousands of kindred souls from all over the world who share my specialized interests. When I began this website 12 years ago, I discovered that traditional search engine optimization was useless because no one was searching for the new ideas I was writing about. Today, with ten million annual page views, I’ve found that the core value of Twitter comes from its ability to discover and connect with geographically dispersed individuals with whom I have something important in common.
If you’re not a celebrity, Twitter becomes powerful when you use it for appropriate two-way communication and connection, not broadcast.
You can’t have a conversation with a million people on Bieber’s antics, but you can have a valuable conversation with smaller numbers of people who are interested in a more specialized topic, and who find each other through appropriate use of hashtags.
For example, there is a community of event professionals on Twitter who tag their tweets with #eventprofs—as well as a host of other hashtags related to their interests, professional affiliations, upcoming events, etc. This soup of appropriately tagged tweets provides a great way for those interested to check up on what is happening and talk about it. One beauty of Twitter is that all these tweets are public and searchable, so it’s easy for newcomers to the profession to discover interesting information and peers on their own schedule.
Yes, over the years the #eventprofs hashtag has been used increasingly by people who view Twitter as a broadcast medium, pumping out “listen-to-me” tweets while rarely or never responding to anyone else or retweeting interesting material. So Jenna is right that the amount of noise on Twitter has increased: the inevitable tragedy of a social media commons where posting costs nothing but the poster’s time. I don’t dismiss this noise lightly—it makes finding interesting tweets harder, and there can come a point when you decide that the effort to filter is just not worth it any more.
What has happened in the event community as a result of increasing noise is the creation of more specialized hashtags for smaller niche groups. Because anyone can create and use a new hashtag at any time on Twitter, it’s possible for a community to coalesce around a useful hashtag. Hashtags are flexible Twitter tools that can be used freely by anyone or any group that finds them useful.
I like that I get to decide how Twitter works for me. Unlike Facebook there are no secret, ever-changing algorithms deciding what I should see. Yes, it’s work to filter the fire hose of information that Twitter serves up; the daunting collective output of currently over 200 million monthly active Twitter users sending 500 million tweets per day. But by discarding the biggest myth of Twitter, you can reap the benefits of meeting and connecting successfully with people who are of value—value that you get to choose.
Last month Twitter quietly rolled out some very useful free analytics tools. I say quietly, because I’ve seen very little discussion of them online. Perhaps that’s because they have not been made available to everyone yet. Whatever the reason, they have been eye-opening for me, so I’m sharing my initial impressions here.
Here’s how to access Twitter’s new tools. You’ll find them on the Analytics menu on the Twitter Ads page, as shown in the above screenshot. This page requires you to turn off any ad-blocker you have running in order to load—fair enough!
If you see an Analytics menu (at the time of writing not everyone does) you’re in luck. There are two options: Timeline Activity and Followers. Pick the former and you’ll see something like the graphic above.
Timeline activity The Timeline shows statistics about your last thirty days of tweets. What is most interesting to me is the number of clicks on any link included in a tweet. Traditionally, social media mavens tend to focus on how many times tweets are retweeted and mentioned, and obviously that’s important. But I had no idea how popular some of my tweets were in terms of people clicking on embedded links. Much of my content is narrowly niche-focused, so I generally don’t see a lot of retweets. But conferencesthatwork.com receives over two million page views a year, and those hits come from somewhere. What these Twitter analytics show me is that many people are clicking on my links, even if most of the time the associated tweets are not being subsequently retweeted. And, most important, I can now see which tweets were popular. This is valuable information!
Yes, these analytics has been available for some time via other mechanisms. All the URL shortening services provide similar statistics for individual shortened links. But in practice, you’d need to use a) a unique short URL for every tweet and b) only one shortening service. a) is cumbersome, and b) is impractical because some services that auto-generate tweets from posts, like LinkedIn and Google Plus, insist on using their own link shorteners, requiring manual amalgamation of clicks over multiple services.
The timeline of mentions, follows and unfollows at the top of the page provides a nice overview that can be helpful for noticing interest peaks, but I prefer to monitor this information using the excellent Birdbrain IOS app.
Followers The Followers option displays a graph of follower growth plus some demographics on interests, location, and gender.
This is interesting but less useful to me, though you may find it valuable. It would be great to be able to drill down further into the location demographics so I could see my followers in, say, the Netherlands, and then be able to reach out to them when I was visiting.
Conclusion For me the gold here is the clicks per tweet statistics. Although I don’t write blog posts based on what I think will be popular, this information gives me a much better picture than I’ve had before of how interesting specific tweets are to others. Over time it should help me understand better how my tweet content and timing affect what people read, allowing me to reach more people with better-marketed content. For free, what’s not to like about that?
Are these new tools of interest to you? How would you use them?
Here’s a one hour video of a Hangout On Air that Jenise Fryatt & I held July 24, 2012 on Moderating Twitter Chats.
To decide whether this is a valuable use of your time I’ve listed below the topics and tips we covered from our summary notes.
Introduction – 5 minutes Jenise & me intros – poll of participants; brief answers; tweet if watching – Q1) who wants to start a new chat? moderate an existing chat; chat name? – Q2) what’s the most important thing you’d like to get out of this hangout?
Set up – 12 minutes Presence on the web – 2 minutes helpful to have permanent place for chat on web: wiki, WordPress site – include schedule, format, rules, chat archives (use Storify)
Chat formats – 3 minutes fixed or rotating moderators 1+ moderators on busy chats topic based guest(s) pre-announced questions moderator asks pre-determined questions to all to guests first, and then opens up discussion
Choosing a topic – 2 minutes something that can be usefully covered in an hour appealing “how to do something” “tips for doing something” controversial current topic
Tools – 3 minutes Tweetchat http://tweetchat.com/ TweetDeck/HootSuite columns chat hashtag; mentions; DMs columns use when you don’t want hashtag at end of tweet keep as a backup in case Tweetchat goes down/is slow (rare)
Preparation – 2 minutes gather up topic links in advance crowdsourced topics http://www.allourideas.org/epchat write out Qs in advance so you can paste them into your Twitter client
Running the chat – 23 minutes Protocol – 2 minutes welcome as many participants as you can encourage first-timers, lurkers to tweet
Welcome everyone – 4 minutes moderator intro (write out in advance include welcome, your name, who your with, topic for today and welcome guest if any) participant intros, including ice-breakers possibilities: names, company, location ice-breaker question: favorite candy, unusual experience etc.
Heart of the chat -12 minutes asking questions concentrate on making them clear (in advance?) make tweets stand-alone participants often RT questions number them Q1), Q2) and ask participants to answer w/ A1) A2) keep track of time; have a plan for time available to get through Qs you’ve prepared but be flexible if circumstances dictate don’t be rushed by anything; don’t feel bad if you miss a tweet or two, we are human; can always go back after the chat & respond then consider ignoring trollish/annoying behaviour
end of chat – 5 minutes ask for takeaways thank moderators, guests mention next topic/guest(s)/time describe where/when archive will be posted
Post-chat – 8 minutes use Storify for archives (login first, click on save regularly, laggy!) Jenise: can add rich media (videos) to Storify; create threads (subheads, move Tweets around)
Adrian is am actively looking for moderators for the #eventprofs chat crowdsourced topics http://www.allourideas.org/epchat
With the recent demise of the wthashtag service, it has become increasingly difficult to create a text archive of Twitter chats. As the organizer of the popular twice-weekly #eventprofs chats, I have been looking for a replacement. Tweetreports offers a free pdf report, but other output formats cost $9+/month.
So here are step-by-step instructions for using the two-year old TwapperKeeper service, together with a copy of Excel, to create a text archive of your Twitter chat.
Note: Please don’t use TwapperKeeper excessively. Twitter’s Terms Of Service and rate limiting can affect their ability to offer their service for free. Such issues caused wthashtag to shut down. Let’s not inflict the same fate on TwapperKeeper.
To create a #hashtag archive before your first chat (one-time only)
Click on the “search for an archive” button to see if there’s already an archive for your chat. (Enter the hashtag for your chat without the hashmark.) If there isn’t, click on the “create #hashtag archive” button to create one.
Go to the archive link bookmark you created above (it will have the form “http://twapperkeeper.com/hashtag/xxxxxxxx” where “xxxxxxxx” is the hashtag for your archive).
You’ll need to enter the start and end time for your chat in Greenwich Mean Time (GMT). You can use TimeZone Converter to convert the local start and finish times for your chat into GMT. Enter start & end date and times, change the “View Limit” to a number larger than the number of tweets in your chat, and click on “query”.
Select all of the report that is relevant to the chat, then copy (Ctrl + C).
Open a blank spreadsheet in Excel, with the A1 cell selected.
Choose Paste Special… from the Edit menu and select “Text”. Click OK.
Select all the cells in Column A that contain data. You should be looking at something like this:
Note the first few characters of the rows containing the dates (in the above example, they would be “Thu “).
Now we’re going to delete any rows that contain “tweet details”, the date of the tweet, or blanks by using Excel’s filter command. First, select Filter from the Data menu and then AutoFilter. A checkmark will appear next to AutoFilter in the menu, and you’ll see a small double-arrow scrollbar appear in the A1 cell, like this:
Click on the double arrows and choose (Custom filter…). From the first drop down, choose “begins with” and type “tweet details” into the text box, like this:
Click OK. Now select all the rows shown that start with “tweet details”. Make sure row 1 is not selected. Choose Delete Row from the Edit menu.
Click on the double arrows again and choose (Custom filter…). From the first drop down, choose “begins with” and type the first few characters of the date you noted in step 8 into the text box. Click OK.
Select all the rows that start with the date. Make sure row 1 is not selected. Choose Delete Row from the Edit menu.
Click on the double arrows and choose (Custom filter…). From the first drop down, choose “does not contain” and type a “?” into the text box. Click OK.
Select all the highlighted empty rows. Make sure row 1 is not selected. Choose Delete Row from the Edit menu.
Finally, click on the double arrows for the last time and choose (Show All). Success! Each row contains one tweet from the chat.
If you wish, scan the rows and delete any that contain non-chat tweets.
Select the remaining rows and copy (Ctrl + C).
Congratulations! A text archive of your Twitter chat is now stored on your Clipboard, ready to be pasted into the web page or document of your choice. (Final tip: You may need to use Paste Special to transfer the information so it formats correctly.)
Is there a better way of archiving Twitter chats? Please let us know when you find one—but test it first to make sure that it 1) reliably includes all the tweets and 2) can produce text output.
Yes, the #eventprofs chats are back! These popular, one hour, Twitter chats on a wide range of topics of interest to event professionals will be once again held twice-weekly: on Tuesdays 9-10pmEST/6-7pmPST and on Thursdays 12-1pmEST/9-10amPST/7-8amGMT starting on May 3, 2011.
Got questions? Here are some answers.
What is #eventprofs? #eventprofs was founded in February 2009 on Twitter by Lara McCulloch-Carter. The #eventprofs chats were one of the earliest Twitter chats—find out more by reading Lara’s history of #eventprofs.
Who will be moderating the chats? Twenty(!) members of the #eventprofs community have each committed to moderating a chat every 6-7 weeks. Our current volunteers are:
Please thank these sterling volunteers at every opportunity! I have volunteered to act as a moderator manager, working to keep the chats scheduled as regularly as possible.
How are chat topics chosen? Anyone can suggest and vote on possible topics for #eventprofs chats at our new AllOurIdeas page. We urge you to do so! The more suggestions, and the more votes, the better our chat topics will be. Moderators will occasionally use their discretion to choose chat subjects, particularly when there are topical events or issues to discuss.
How do I know what chat topics are scheduled? There are two ways to stay informed about upcoming #eventprofs chats: