Twitter, Inc. (NYSE:TWTR) Q3 2019 Earnings Conference Call - Final Transcript
Oct 24, 2019 • 08:00 am ET
[Operator Instructions] Your first question comes from the line of Douglas Anmuth of JP Morgan. Please go ahead, your line is open.
Thanks for taking the questions. I have two . First, just Ned on MAP, can you just help us understand a little bit more about kind of when you learned of the issues around MAP around targeting and then sharing data. And then the degree to which they have been fixed at this point and how that kind of informs the four points of impact that you're talking about for 4Q. And then secondly, it looks like you're on track to do nearly the 20% opex outlook that you've discussed for this year. And I know it's early, but do you expect the same big four investment priorities next year and how do you think about opex growth for 2020? Thanks.
Okay. Thanks, Doug. I'll take both of them and Jack can add anything afterwards, if it makes sense. First, on the question on the product-related issues. So these came up over the course of the quarter and it wasn't one particular day, there were more than one of these things. Let me give you a couple of examples, which can help them come to life. The first is we asked people a series of questions when we put them -- before we put into a timeline when you are new to Twitter. Among the questions, we ask are, if we can use your device settings to figure out the best ads to show you. It turns out that that setting wasn't working as expected and we were using device settings even if people had asked us not to do so. So when we discover that, one, we tweeted about it which we often do to try to be transparent with people when things are working as expected. And two, we turned off the setting, so that it would work as expected. That has a negative impact to revenue, because it's one less input you've got, when you are figuring out which adds to show people. So instead of getting a partial quarter impact you get a full quarter impact in Q4.
A second example is specific to MAP where we typically will share data with measurement partners who will then share it with advertisers, so that they can see the effectiveness of their campaigns, not just on Twitter, but across platforms. And another one of the questions that we ask people before we put them into a timeline is if we can share their data with measurement partners. That setting also was not working as expected and we were passing on data, which we had not intended to. So, we stopped doing that and although we are working on remediation, there isn't remediation yet in place and so the effects of that will continue into Q4.
As you can imagine the remediation would be sharing aggregated data as opposed to personalized data when people have asked us not