Closed Bug 1170392 Opened 9 years ago Closed 5 years ago

After I clicked "X" on a suggested tile, a different suggested tile for the *same site* appeared, with a different "Suggested for..." target-group

Categories

(Firefox :: New Tab Page, defect)

defect
Not set
normal

Tracking

()

RESOLVED WONTFIX
Tracking Status
firefox41 --- affected

People

(Reporter: dholbert, Unassigned)

References

Details

In my normal browsing profile, I just saw a "Firefox for Android" tile, which said "Suggested for [Mozilla] visitors".

I clicked its "x" icon to remove it.

It was replaced with a *different* "Firefox for Android" tile, which said "Suggested for [Technology] visitors".

This seems like something we should prevent... If users have X'd out a tile for a particular URL, we shouldn't show them different tiles for the same URL, simply because they also match a different target-group.
Flags: needinfo?(edilee)
Summary: After I clicked "X" on a suggested tile, a different variant of the same tile reappeared, targeting me for a different reason → After I clicked "X" on a suggested tile, a different suggested tile for the *same site* appeared, with a different "Suggested for..." target-group
In this case, there are 3 different "Firefox for Android" tiles that have 3 different urls (the parameters are different), but they take the user to the same page. The 3 different tiles also have 3 different creatives, so I think it's somewhat fair to show something different if the user blocked one creative.

If the urls were actually exactly the same, blocking one would have blocked the other ones. I believe we do have some policy around when to reuse urls vs having slightly different urls.
Depends on: 1169658
Flags: needinfo?(edilee) → needinfo?(jterry)
(In reply to Ed Lee :Mardak from comment #1)
> The 3 different tiles also have 3 different creatives, so I think
> it's somewhat fair to show something different if the user blocked one
> creative.

So... I suppose I can *imagine* cases where a user may block Tile #1 for example.com, and wouldn't be bothered by subsequently seeing Tile #2 for the same site. (maybe if Tile #1 has a really annoying graphic on it, and the user blocked it purely for that reason, whereas Tile #2 is more tame.)

But, I feel like such a scenario is a bit of a stretch. In general, I think it's safe to assume that a user who has just blocked a suggested tile for $WHOEVER would be upset/unnerved by being subsequently served a different tile for the same $WHOEVER.

Maybe we need some sort of cool-down period after a tile has been X'd out, and during that cool-down period, we don't show any other suggested tiles for the same thing?  (Not sure how to define "the same thing", but it seems like the same URL with different parameters would fit the bill.)
(Note that when a user blocks a tile, they're *taking action* to shape their new tab page. If we follow that up by showing a different tile for the same thing they just blocked, it's likely to make them feel like they're having to fight with their browser, & reduces their sense of agency.)
(So, I think we should err on the side of being sensitive to user frustration here, even if there are thought experiments where a partner provides multiple tiles and users might maybe only want to block one of them.)
We have discussed smarter blocking on a per-advertiser or as a proxy per-site basis. Or even blocking an interest category or per campaign. Not quite sure if the complexity of having the user need multiple clicks to choose the appropriate type of blocking is better though.

The per-site blocking is tricky because two unrelated content could happen to be on the same site, e.g., perhaps a suggested tile for a general news site and another suggested tile for a specific article.

I definitely agree that we don't want the user to feel like the browser is broken or having to fight with it. I think showing the same creative after just blocking it definitely would give that feeling because it looks almost exactly the same.

If we do site level blocking, should blocking any one of

> https://www.mozilla.org/firefox/android/?utm_source=suggested-tiles&utm_medium=tiles&utm_content=androidenthusiasts&utm_campaign=firefoxforandroid
> https://www.mozilla.org/firefox/android/?utm_source=suggested-tiles&utm_medium=tiles&utm_content=mobileproviders&utm_campaign=firefoxforandroid
> https://www.mozilla.org/firefox/android/?utm_source=suggested-tiles&utm_medium=tiles&utm_content=mozillafans&utm_campaign=firefoxforandroid

prevent any future mozilla.org suggested tile? Or as suggested, perhaps there could be a timeout/cool-down period.. 24hrs? before showing another suggestion from mozilla.org.
Two thoughts: 
1) Blocking at the campaign entity level could help. Oyiptong?
2) I know this is an internal Mozilla campaign, but we should not allow advertisers to use unique click through URl trackers per category. This is unlikely, but could potentially fingerprint the user based on the combination of interests and click throughs. As a policy, we should not allow category names within the click through string. Cc: merwin
Flags: needinfo?(merwin)
Depends on: 1171107
(In reply to Kevin Ghim from comment #6)
> Two thoughts: 
> 1) Blocking at the campaign entity level could help. Oyiptong?
> 2) I know this is an internal Mozilla campaign, but we should not allow
> advertisers to use unique click through URl trackers per category. This is
> unlikely, but could potentially fingerprint the user based on the
> combination of interests and click throughs. As a policy, we should not
> allow category names within the click through string. Cc: merwin

I agree with Kevin about this. It seems to me that the URl should be unique to the campaign but not the category.
Flags: needinfo?(merwin)
What would the user expect from blocking an Ad?
It feels like the blocking the campaign is a sensible solution.

Once we implement campaigns on the server-side, we could pass on a campaign id.
That way, the URL wouldn't have to be unique to the campaign.
Flags: needinfo?(jterry)

Hello!

This bug has been closed due to inactivity and/or the potential for this bug to no longer be an issue with the new Discovery Stream-powered New Tab experience.

Please help us triage by reopening if this issue still persists and should be addressed.

Thanks!

Status: NEW → RESOLVED
Closed: 5 years ago
Resolution: --- → WONTFIX
You need to log in before you can comment on or make changes to this bug.