There are a lot of users on Twitter; reports say there are 271 million monthly active users (MAU) on Twitter as of the end of June alone. There's just no getting around that one. But even Twitter knew last month, reportedly, that not all of Twitter's users were represented by actual breathing human beings, and now, the true extent of Twitter's bot problem may have been revealed: as many as 23 million of Twitter's users may actually be bots.

The latest word out of the company says that “up to approximately 8.5 percent” of Twitter's accounts have one thing in common: the accounts are updated, automatically, “without any discernible additional user-initiated action.” There are several kinds of account that might be included here, ranging from mobile apps not owned by Twitter to accounts that request data from Twitter to be displayed elsewhere. When these and others are factored together, that represents that 8.5 percent figure.

While earlier figures suggested that as many as 14 percent of MAU were coming from services outside the official website and accompanying mobile apps, this got a note of clarity in its own right as Twitter revealed that that 14 percent included things like Twitter for Mac, as well as TweetDeck. But around 11 percent of MAU were coming from places like Tweetbot and Flipboard, places that Twitter doesn't specifically own.

That certainly helps matters, but part of the matter is a particular problem, known as “spam accounts,” accounts that fire off meaningless advertising messages or exist to inflate a follower count. Twitter estimates that less than five percent of its MAU count qualifies in this bracket, but that's really a secondary concern, particularly for advertisers. Whether the automated user in question is there as a useful and helpful tool or as an odious advertising mechanism, the key point to advertisers is that there is no human behind that user to receive advertising at all.

Now, this is a problem. But how much of a problem exactly, is something different. Consider for a moment the numbers involved; estimates from the end of June, as noted previously, are calling around 271 million MAU. Around 23 million of these are automated, which means that around 248 million MAU are red-blooded human beings. How concerned can advertisers really be with this development? While in an absolute sense, this is a serious issue—paying for access to 271 million and really only getting access to 248 million isn't exactly a bargain—but there are still 248 million MAU here. It should be a comparatively simple problem for Twitter to fix here; offer access numbers based on the human count, and charge accordingly, while either not sending the advertising to the bots or including the bots at no extra charge. But it's quite clear that some kind of adjustment needs to take place here; Twitter really can't continue—at least, from an ethics standpoint—to sell access based on 271 million MAU when it's pretty much clear that's not the case. If it can, then advertisers can't continue to patronize Twitter and keep getting around 8.5 percent less than is being paid for.

Only time, of course, will tell how this all comes out, whether Twitter modifies its numbers, its prices, or simply carries on business as usual and just how the  advertisers and others who depend on social media will ultimately react. Still, this disconnect doesn't bode well immediately, and needs to be addressed, somehow, lest Twitter lose quite a bit of otherwise interested business.