In a filing with the US Securities and Exchange Commission, it said it had calculated a new metric which showed that, in the three months to the end of June, "approximately 8.5% of all active users used third party applications that may have automatically contacted our servers for regular updates without any discernable additional user-initiated action".
Many of these will include automatic updates from legitimate sources as varied as news service, retailers or earthquake alerts. Twitter estimated that false or spam accounts made up less than 5% of its monthly active users.
Marketers will obviously be interested in the new metric as these accounts carry no possibility for interacting with brand advertising or content.
At the same time, however, the filing reported that "11% of all active users solely used third-party applications to access Twitter". This compared to an earlier figure of 14% meaning that more people than it had thought were able to view ads served on the platform, 89% of the total active user base against 86%.
Previously Twitter has spoken of monetising visitors to the site who are not logged in – those who arrive, for example, after seeing tweets on news programmes or in newspapers.
"You get a great signal for the kind of content they like to consume and we think those signals will provide the data that we need to deliver the right kinds of monetisation experiences to that audience," said CEO Dick Costolo
Data sourced from SEC, Marketing Week; additional content by Warc staff