Because here's the thing: when people see a pop up of any kind, they reflexively think "fuck you". And it doesn't take a marketing genius to realize that priming people in this fashion isn't the best opening move. It's like stroking a cat backwards; the simple rule is "Don't do this."
The correct approach is to start by provide people with what they want, the the form they request, without interruptions, redirects, etc. Make them happy before you try to sell them anything. Not until they're satisfied should you insert a plug for your mobile app. Knowing that you'll be taking up valuable screen space, make sure the app really is much better than the web app. Then tell people that it's much better, and that you think they'll really like it for this reason. And that's it.
Give people what they want, don't be a dick, and you're golden.
I was just calling out dpe82's absurd absolute statement that they suck because all apps "do little to nothing to add to the content or user experience"... Implying that if they the app did offer substantially more, then it would all somehow be acceptable.
Can an app be superior because it allows other types of content that can't be done well via web technologies? Perhaps. Can an app help with content discovery? Perhaps. But when a user clicks on a link to standard text/images/video they're not looking for either of those.
If you want to advertise your app because it provides other features that a user may find valuable, that's fine. Advertise it like any other product that runs against your content. If nobody uses it there's likely a good reason. Don't force it on them.
But that's all beside the point... you are so quick to argue against me, you don't even see that we agree. Re-read my original post. As I said, the execution sucks. You want to inform your audience that there's an app option, but most of the implementations suck and are too intrusive.