Using TwitterCounter, I looked at the follower growth rate curves of the Twitter accounts of a number of prominent marketers who are active Twitter users. I chose these accounts because they are people known to me personally, and watching their behavior on Twitter over the years, I believe they are all people who have grown their followings organically, without tricks, purchasing followers, or auto-followback schemes.
What I saw was fairly consistent.
Starting in late April 2018, nearly all of these accounts showed a distinct trend of slower follower growth. Sometimes the decline was minimal (but still discernible); in other cases it was dramatic.
After showing some examples, I’ll discuss what I think may have caused this. All examples cover the past 90 days.
I checked back further, and there was no substantial difference in the growth curves prior to March, so this recent change across many accounts appears to be a trend.
Examples of Twitter Follower Decline
Below is a sampling of accounts I looked at that showed a discernible decline in follower growth since late April.
All of the accounts in this section showed every sign of normal, “white hat,” organic Twitter usage over at least the last several months. They all:
- Tweeted and engaged regularly and at a consistent rate.
- Added new followers at a slow rate (in most cases, they added few-to-no new followers in recent months).
- Regularly show evidence that they are personally engaged in their accounts.
- Have high reputations in our industry.
To help you see the change, I’ve placed an orange line on the slope of the graphs prior to late April and a red arrow at the exact point where the decline began for the account.
You will also notice a sharp dip in many of the graphs the first week of March. That comes about two weeks after Twitter announced a massive bot purge, but I’m guessing it took that long to affect some accounts.
I’ll start with my own account (only fair!):
As you’ll see with some of the other examples, the effect on my account was not dramatic (but still tangible).
Here’s the graph for Rand Fishkin, founder of the new startup Sparktoro:
Similar to mine, but Fishkin’s decline is a little more dramatic.
Lee Odden’s follows a similar trajectory to Fishkin’s.
Like my own, Eric Enge’s decline is less steep than others, but still very evident and beginning around the same time.
Another less steep yet clear decline in growth rate in the account of marketing strategist Dorie Clark.
Now for a change up, how about a brand account?
Here the drop off was even more significant than in the personal accounts above.
Moz got hit hard too.
Were All Twitter Accounts Hit By the Decline?
No, not all accounts were impacted by this follower growth rate decline, but enough were to see it as a significant pattern.
I found many more examples of accounts that began to drop off in late April than I showed above, but it also wasn’t hard to find as many that appeared to have suffered no effect.
Here are a couple of examples:
As you can see, Cyrus Shepard’s rate of growth has remained steady for the past three months, including right through the period of growth decline experienced by the accounts shown above. Shepard also seems to have taken no hit from the early March bot purge. Cyrus, tell us what you’re doing right!
My second example of an account showing no April growth rate decline is that of Bill Slawski (@bill_slawski).
For Slawski’s graph I included the purple line that shows the number of people he is following. I wanted to point out that in early May he began adding people he followed on a daily basis. It appears he did so organically (i.e., by “hand” without automation). That will be important when you compare with the auto-follow account in the next section.
It may be that adding followers organically (many of whom probably followed him back) at that time mitigated the effect of the decline. But then again, Shepard’s follow line remained flat throughout, and he also showed no effect. So it may be that Slawski was just an exception like Shepard. for reasons hard to discern.
A Growth Hacker Example
I’ll keep this account anonymous, but here’s what happened to someone I know who did use automation to follow accounts with the hope that most of those accounts would follow back. This has been a common growth hack for Twitter accounts going back to the early days of Twitter.
Again, the purple line is the growth of people the account was following. You can see that it roughly follows the slope of the added followers line, up until late April.
The occasional dips in the purple line are exactly what you expect to see when someone is using a service that automatically follows accounts. Such services are typically designed to purge those who don’t reciprocate the follow within a set period.
Notice both lines flatten in late April. I saw this in a number of auto-follow/auto followback accounts.
It looks like Twitter has finally killed this hack, perhaps for good.
Why Is the Twitter Growth Rate Decline Happening?
We can’t see inside Twitter’s algorithms, so it is impossible to know all the reasons this happened, or why it happened to some accounts and not others.
However, the pattern of decline starting in late April shows up in enough accounts to give me confidence enough to call this a real action by Twitter, and not an anomaly or coincidence.
The interesting thing about the late April timing of this is that it comes right after some significant announcements by Twitter.
Twitter Cuts Down on Automation
On February 21, Twitter announced it would be severely limiting the kinds of automated tweeting accounts could do using third-party apps and services.
Specifically, accounts are no longer allowed to repost a tweet that is substantially the same as a previous tweet, tweet the same tweet from multiple accounts, and auto-engage with other’s tweets.
Apps were to comply with this by March 23.
Obviously those new rules wouldn’t necessarily cause a drop in the rate of new followers by themselves, but they did indicate the seriousness of Twitter’s intentions to eliminate gaming and abuse of its system.
Twitter Shuts API to Thousands of Apps
On April 25, Twitter revealed that it had closed access to its TweetDeck API to over 140,000 apps that had been tweeting through it.
Notice that date in the second announcement: April 25. That’s almost exactly when we see the follower growth rate decline begin for many accounts.
It is my speculation that part of the April 25 API purge included access to the ability to autofollow accounts using these apps. That would account for the growth rate decline seen in so many accounts since the final days of April.
How Should Twitter Accounts Respond?
First, while no one likes slower growth, this is not a disaster by any reckoning. What we now see in the accounts which experienced a growth rate decline is their real follower growth.
That is, it is now more likely that most of the followers an account gains are from real accounts where someone has made a conscious and intentional decision to follow the account.
If you are on Twitter not just for social proof numbers, but to help build your business and bring you personal value, then this is good news.
Social media users, and marketers and businesses on social media in particular, need to shift to this more mature mindset, in which the value of social media is not so much how many follow you as who follows you.
Notice that in every case shown above, the accounts continue to grow their following, even if it’s at a slightly slower pace. That doesn’t surprise me. As I said in my introduction, these are all high quality, legitimate accounts that provide great value to their followers.
Those kinds of accounts will always grow, and they will grow with followers who matter.
Want to know more about how to grow a Twitter following that matters for your business? See my Search Engine Journal post 9 Proven Tips to Get More Social Media Followers.
More Twitter Marketing Resources:
- Twitter Cracks Down on Automation, Use of Multiple Accounts
- 22 Terrific Tips to Optimize Your Twitter Business Profile
- 6 Ways You Can Use Twitter in Your Customer Feedback Loop
Featured Image & In-post Photos: Created by author, June 2018