# Some statistics about topics NOT created?
**URL:** https://meta.discourse.org/t/some-statistics-about-topics-not-created/6186
**Category:** Site feedback
**Created:** [April 26, 2013, 10:40pm UTC](https://meta.discourse.org/t/some-statistics-about-topics-not-created/6186 "2013-04-26T22:40:48Z")
**Posts on this page:** 5
**Page:** 1
### Author:  [@PabloC](https://meta.discourse.org/u/PabloC)
#### Post date: [April 26, 2013, 10:40pm UTC](https://meta.discourse.org/t/some-statistics-about-topics-not-created/6186/1 "2013-04-26T22:40:48Z")
I’m interested to see a any metric about how many people does not create a post because finds a relevant answer in the Topic Popup, specially in this community, or in HowToGeek.
I mean, this popup is fantastic, it helps a lot to answer new questions and minimize duplicates.
@codinghorror can you share some numbers? Is there an easy way to measue this in Analytics?
Tks!
---
### Author:  [@codinghorror](https://meta.discourse.org/u/codinghorror)
#### Post date: [April 26, 2013, 10:50pm UTC](https://meta.discourse.org/t/some-statistics-about-topics-not-created/6186/2 "2013-04-26T22:50:08Z")
Unfortunately I don’t have any numbers, nor can I think of a way to get them off the top of my head.
I’m not a big numbers guy, really.
I am a big fan of the feature, though! If anyone has ideas on improving it – beyond “add Godlike search prediction of duplicates” – would love to hear about them.
---
### Author:  [@cregox](https://meta.discourse.org/u/cregox)
#### Post date: [May 11, 2015, 3:51pm UTC](https://meta.discourse.org/t/some-statistics-about-topics-not-created/6186/3 "2015-05-11T15:51:10Z")
Awesome idea, @PabloC! 😍
Ironically enough, I can tell you I just did not create 2 topics thanks to it. One was a duplicate from this, the other one [about `/email_preferences`](https://meta.discourse.org/t/reference-generic-preferences-link-solved/12223/7). 😃
> [@codinghorror](#):
>
> nor can I think of a way to get them off the top of my head.
What do you mean? How about just checking if user clicked on any “similar links” and then canceled the new topic, at any point in time? It shouldn’t be too hard to at least think of ways! 😮
I’d be interested in making it happen even into the interface if you tell me this is something you might accept a PR on. 😊
But I do have an awful history of wanting to do PR’s (to discourse) and never following through.
> [@codinghorror](#):
>
> If anyone has ideas on improving it
Just like with stackoverflow, I often use this instead of searching, as it works so much better.
Why not having a link to “more” and then a page full of similar results?
As for background work, if users and mods could mark topics as duplicated (or at least _related_) it would make awesome data for this feature to find similar topics.
---
### Author:  [@PabloC](https://meta.discourse.org/u/PabloC)
#### Post date: [May 11, 2015, 4:18pm UTC](https://meta.discourse.org/t/some-statistics-about-topics-not-created/6186/4 "2015-05-11T16:18:01Z")
I think this is one of the best features that Discourse brings on the table. I’ve been doing some consultancy to customer services companies, and this particular case, answering customer questions without agent intervention, is gold.
---
### Author:  [@cregox](https://meta.discourse.org/u/cregox)
#### Post date: [May 11, 2015, 5:25pm UTC](https://meta.discourse.org/t/some-statistics-about-topics-not-created/6186/5 "2015-05-11T17:25:22Z")
Couldn’t agree more!
That’s also yet one more reason why [I’m trying](http://talk.cregox.com/c/support) so hard to use it as a [community ticket system](https://meta.discourse.org/t/using-discourse-as-a-community-ticket-system/20607/17). 😉
Having stats on things like this can help other people realizing it (or eventually we focusing on what really matters the most, if this is not it).