On Bots and Fake Followers

This post was published 1 year ago. Some things may have changed since then - use the search function to see if anything has been posted since then, or reach out to us on Twitter if you'd like to see a more updated post!

Everyone’s talking about bots this week, thanks to this post by Victoria from In The Frow.  Victoria wrote about how to see whether followings and engagement are enhanced artificially and why this is a problem we should all be concerned with as bloggers, influencers and the audience.

One agency, Social Chain, has come up with a tool that will help brands see whether a following is legit. The tool, called Like-Wise, will look at past followers and engagement per post to judge whether or not the following is true or if it’s likely that it’s been purchased.

There are other sites that are free to use that also judge whether a following is real, such as IG Audit, HypeAuditor and SocialBlade.

One thing that always comes up in this sort of conversation is “Well, if you know who is buying followers and engagement, why don’t you just name and shame“. There are a few problems with this. 

  • Bot accounts (those used for purchased followings) will follow users who haven’t purchased their service in the hope that they will get followed back. This helps their account look more legitimate. This will obviously mean that those legit users who didn’t purchase the followers will look like they have. 
  • Sites that help you decide whether a following is true usually only take a sampling of the users followers because it would be difficult to calculate the exact percentage (and most likely would not be possible with Instagram’s API access limits). This doesn’t mean that the information they put out is incorrect, but it may not be the complete story.
    For example, we used IGAudit to check an account. We opened three different tabs with the same website in it, put the same username in, and got three different percentages as to how real their following was. The percentages were within 5% of each other which doesn’t seem like much, but if a brand is using this site to decide what influencers to use for a campaign and set a restriction that they were only going to use influencers with a 95% real following, you don’t want the site to be reporting you with a 90% real following! 
  • You do not need to log into your Instagram account most of the time to add purchased followers – you only need the username of the account you want to add followers to. Can you spot the flaw there? It’s totally possible to buy followers for other people to make them look like they are doing it. We’ve heard of this happening to a few people.

Here’s an example of the first point on this list. We’ve pixelated out the profile pictures as some were possibly a bit NSFW!

We are far too lazy and cheap to buy followers but this has been the state of our notifications for at least the past week. It’s been pretty annoying having to go through them one by one and report them or block them, but blocking these spam accounts has helped our statistics on the account (which we’ll be covering in a future post)

  • Reporting an account as being inappropriate for nudity or pornography does not block that account for you. It’s an odd choice, Instagram.
  • Many of these accounts used the same images, and looked like they were posting at the same time – to the extent that as we went through to report and block accounts, we had to check whether they were accounts we had already reported!
  • Instagram’s reporting tools are not amazing. Like we said, some accounts were using the exact same images in the same order, yet Instagram did not always treat them the same – some would be taken down, and some would be left up. 

Instagram users with legitimate followings have nothing to fear from these discussions and tools, but the more that we report and restrict the fake accounts from our own followings, the less we will fear that this is an issue for us.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: