Iteration 32.1's priority backlog

[Gavin Sharp]

Hi firefox-dev,

As mentioned many times before, the core firefox development team is using a new process to track and prioritize work. Part of that process is that we're splitting work in two-week iterations, and each iteration starts with the team picking work from a prioritized backlog (the "c" set that I described in a previous post: https://mail.mozilla.org/pipermail/firefox-dev/2014-April/001551.html).

To get better visibility and broader input into that process going forward, Madhava, Chad and I will be posting updates here prior to the start of each iteration, briefly explaining what we've decided to focus on and why (i.e. summarizing the what and the why of the priority backlog). This is the first of those updates.

The next iteration starts next Tuesday, and is labeled "32.1" (the first iteration of the Firefox 32 cycle). We're going to be focusing primarily on trying to tie up loose ends on goals that we've carried over from Q1 (more on our goals in a followup post). At a high level, that means:

- Getting the translation functionality (https://mail.mozilla.org/pipermail/firefox-dev/2014-April/001556.html) working and ready for the trial
- Enabling default new tab page tiles on Nightly
- Some followup work related to telemetry experiments (which landed on trunk recently! https://mail.mozilla.org/pipermail/firefox-dev/2014-April/001580.html)
- adding some crash data to FHR for better visibility into Firefox stability

More detail about the specific bugs we'll be focused on can be seen in the priority backlog, which is currently maintained as a shared Google Spreadsheet:

https://docs.google.com/spreadsheets/d/10sr6YhDNmO4oimlNtxDZ5fe6IaQKmZ7gqT-ZWqAygrI/edit?usp=sharing

I'm happy to answer any questions about the process or the backlog.

Gavin

[Nicholas Nethercote]

On Mon, Apr 28, 2014 at 8:44 PM, Gavin Sharp <ga...@gavinsharp.com> wrote:
>
> - Enabling default new tab page tiles on Nightly

What will these new tab page tiles look like?

Nick
_______________________________________________
firefox-dev mailing list
firef...@mozilla.org
https://mail.mozilla.org/listinfo/firefox-dev

[Gavin Sharp]

https://bug990713.bugzilla.mozilla.org/attachment.cgi?id=8414784 has a
screenshot. Ed can answer any more specific questions.

Gavin

[Jared Wein]

I am not Ed, but I'm pretty sure that the screenshot in the previous email can be a bit misleading.

Only the tiles with an arrow at the bottom right are the "sponsored" tiles. The arrow is used to provide more information as to why this tile is placed on the new tab page. The other tiles (without the arrow button) would be screenshots of frequently visited sites, but this screenshot shows them as branding logos.

Jared

[Ed Lee]

On Thu, May 1, 2014 at 12:14 PM, Jared Wein <ja...@mozilla.com> wrote:
>> https://bug990713.bugzilla.mozilla.org/attachment.cgi?id=8414784

> Only the tiles with an arrow at the bottom right are the "sponsored" tiles.

All 9 tiles shown in the screenshot are "Directory Tiles" although
only 2 of them are "sponsored." Internally we label some as "organic"
for popular sites and "affiliate" type has been put on Lightbeam,
Mozilla Foundation, Webmaker.

These Directory Tiles appear when the tiles would have shown nothing
or low frecency history pages (e.g., the whatsnew/welcome page). The
current threshold is 1000 frecency (that's half the value of a
typed-in URL page's frecency).

Ed Lee

[Nicholas Nethercote]

So FF32 is planned to be the first one with ads enabled?

On Thu, May 1, 2014 at 12:24 PM, Ed Lee <edi...@mozilla.com> wrote:
>
> All 9 tiles shown in the screenshot are "Directory Tiles" although
> only 2 of them are "sponsored." Internally we label some as "organic"
> for popular sites and "affiliate" type has been put on Lightbeam,
> Mozilla Foundation, Webmaker.

I recall promises that the sponsored tiles would be clearly marked as
such. Is this arrow considered to be this marking?

Nick

[Gavin Sharp]

On Fri, May 2, 2014 at 1:06 AM, Nicholas Nethercote <n.neth...@gmail.com> wrote:

So FF32 is planned to be the first one with ads enabled?

"enabled on Nightly" does not necessarily imply "riding the trains". This distinction will be increasingly important as we experiment more on Nightly (and other channels) in general.

I recall promises that the sponsored tiles would be clearly marked as
such. Is this arrow considered to be this marking?

You might be interested in https://bugzilla.mozilla.org/show_bug.cgi?id=974736

Gavin

[Steffen Wilberg]

The arrow as shown in the screenshot from
https://bug990713.bugzilla.mozilla.org/attachment.cgi?id=8414784 is very
subtle and easy to overlook.

Why don't we mark sponsored tiles more clearly, like e.g. Google.com
does in search results, which display "Ad" in white text on yellow
background?

I don't look forward to the next shitstorm: "Firefox deceives users
about ads and can't be trusted anymore."

Steffen

[Mike Connor]

On May 2, 2014, at 8:52 AM, Steffen Wilberg <steffen...@web.de> wrote:

> On 02.05.2014 07:19, Gavin Sharp wrote:
>> On Fri, May 2, 2014 at 1:06 AM, Nicholas Nethercote
>> <n.neth...@gmail.com <mailto:n.neth...@gmail.com>> wrote:
>> I recall promises that the sponsored tiles would be clearly marked as
>> such. Is this arrow considered to be this marking?
>>
>>
>> You might be interested in
>> https://bugzilla.mozilla.org/show_bug.cgi?id=974736
>>
>> Gavin
>
> The arrow as shown in the screenshot from
> https://bug990713.bugzilla.mozilla.org/attachment.cgi?id=8414784 is very
> subtle and easy to overlook.
>
> Why don't we mark sponsored tiles more clearly, like e.g. Google.com
> does in search results, which display "Ad" in white text on yellow
> background?
>
> I don't look forward to the next shitstorm: "Firefox deceives users
> about ads and can't be trusted anymore."

There’s a fine line here. “Ad” isn’t really the right term, nor do I really think we should get too caught up in the idea that sponsored tiles are bad, and users should be “warned" about them. Users who care should be easily able to tell which are which.

On the other hand, Google’s current design has been shown to attract _more_ attention to the ads [1], which makes the balance trickier. We could equally be accused of trying to drive clicks if we have something that draws more attention.

— Mike

[1] http://blog.eyequant.com/2014/03/20/why-the-new-google-search-ads-design-is-a-subtle-work-of-genius/

[David Rajchenbach-Teller]

Why not simply add a word "Sponsored"?

On 02/05/14 15:22, Mike Connor wrote:
> There’s a fine line here. “Ad” isn’t really the right term, nor do I really think we should get too caught up in the idea that sponsored tiles are bad, and users should be “warned" about them. Users who care should be easily able to tell which are which.
>
> On the other hand, Google’s current design has been shown to attract _more_ attention to the ads [1], which makes the balance trickier. We could equally be accused of trying to drive clicks if we have something that draws more attention.
>
> — Mike
>
> [1] http://blog.eyequant.com/2014/03/20/why-the-new-google-search-ads-design-is-a-subtle-work-of-genius/
> _______________________________________________
> firefox-dev mailing list
> firef...@mozilla.org
> https://mail.mozilla.org/listinfo/firefox-dev
>

--
David Rajchenbach-Teller, PhD
Performance Team, Mozilla

[Mike Connor]

I think the question is “what are we trying to accomplish here?” Are we trying to put a value judgement on sponsored content? Are we simply trying to be transparent? If the latter, is having a known annotation that provides more information good enough, or does it need to encourage users to interact with it?

I suspect the answers to those questions are all subjective, and pretty directly correlated with how individuals view the feature.

— Mike

[Axel Hecht]

I'd prefer something visual, so that we don't need to have 90
discussions on what wording walks close enough to the line, and is short
enough to fit.

That said, did legal do an investigation if we're required to show
particular texts? Discussed the word "sponsored" in #firefox.de, and
Archaeopterix mentioned 'german legal: Anzeige'. Thus, do we have our
ducks lined up on what might be required under local laws?

Axel

[Ehsan Akhgari]

On Fri, May 2, 2014 at 9:47 AM, Mike Connor <mco...@mozilla.com> wrote:

I think the question is “what are we trying to accomplish here?”  Are we trying to put a value judgement on sponsored content?  Are we simply trying to be transparent?  If the latter, is having a known annotation that provides more information good enough, or does it need to encourage users to interact with it?

Yes, exactly.

 

I suspect the answers to those questions are all subjective, and pretty directly correlated with how individuals view the feature.

So who is in charge of answering these questions on our side?  I have personally followed all of the public discussions about this feature since it was first announced but have been unable to find what our answers to these questions are.

--
Ehsan
<http://ehsanakhgari.org/>

 

[Gavin Sharp]

Which questions do you want answered exactly?

I can explain some of the goals:
- educating users about the provenance of the tiles is not a design
goal (so we're not trying to optimize for differentiation, or driving
people to explanation pages)
- making it possible for interested users to know the provenance of
sponsored tiles _is_ a design goal (which I think is accomplished by
the button design from bug 974736, though certainly the design could
be tweaked in various ways)

I think most posts in this thread are seeing the problem from a very
different perspective than most of our users will. A term like "ads"
has very negative connotations. Our success criteria is "users do not
perceive tiles as ads (in the negative sense)", not "users perceive
tiles as ads, are annoyed/surprised by them, and are then prompted and
educated about their provenance and selection process to alleviate
their concerns". If we somehow find that we've ended up in the second
place, we need to find a way to get back to the first, not double down
on making explanations more obvious.

Gavin

[Bryan Clark]

We are going try different variations of the sponsored indication to see how they work in practice.  And just to be clear there is no actual sponsorship in these tiles in the screenshots or tiles that might land in Nightly.  Any tiles that indicate sponsorship are being included solely we can iterate on the type of sponsor indication required.

I encourage people to read the link [1] Mike pointed to as it discusses how creating more of a marker can have side effects  one would not initially expect.  Our goal was to create a clear indication of sponsorship and a way for our users to learn more about what that means.  There are many variations that might achieve that goal better but we're only just starting out.  What we have now isn't near ready for broad release so we are going to sit in Nightly and testing channels until we understand more.

I'm the product manager for Tiles so feel free to reach out to me directly as well.

Thanks,

~ Bryan

[1] http://blog.eyequant.com/2014/03/20/why-the-new-google-search-ads-design-is-a-subtle-work-of-genius/

On Fri, May 2, 2014 at 9:47 AM, Mike Connor <mco...@mozilla.com> wrote:

[Ehsan Akhgari]

On Fri, May 2, 2014 at 10:39 AM, Bryan Clark <cla...@gnome.org> wrote:

We are going try different variations of the sponsored indication to see how they work in practice.  And just to be clear there is no actual sponsorship in these tiles in the screenshots or tiles that might land in Nightly.  Any tiles that indicate sponsorship are being included solely we can iterate on the type of sponsor indication required.

Yes, I understand that.

 

I encourage people to read the link [1] Mike pointed to as it discusses how creating more of a marker can have side effects  one would not initially expect.  Our goal was to create a clear indication of sponsorship and a way for our users to learn more about what that means.  There are many variations that might achieve that goal better but we're only just starting out.  What we have now isn't near ready for broad release so we are going to sit in Nightly and testing channels until we understand more.

I did read that link, but I'm not sure what kinds of conclusions I'm supposed to draw from it.  That study seems to indicate that Google managed to attract more clicks on their ads to increase their ad revenue by making the ads look more similar to non-sponsored content, and still included the word "Ad" next to those results in order to not be evil presumably by disguising content that they're paid to include in their search results as organic search results.

Can you please explain how this related to the current design of sponsored tiles?  (I know that this is a heated topic, but I'm genuinely interested to understand our thinking here before I make up my mind one way or another.)

Thanks!

--
Ehsan
<http://ehsanakhgari.org/>

 

[Nicholas Nethercote]

On Fri, May 2, 2014 at 6:22 AM, Mike Connor <mco...@mozilla.com> wrote:
>
>> I don't look forward to the next shitstorm: "Firefox deceives users
>> about ads and can't be trusted anymore."
>
> There’s a fine line here. “Ad” isn’t really the right term

Let's cut the BS. They are ads, pure and simple. They might be
inobtrusive, and some users may find them useful, but don't pretend
they're not ads; you'll (a) deceive yourself and (b) come across as
phony -- see http://daringfireball.net/linked/2014/02/16/mozilla.

(And if you are about to pull out a dictionary to look up the
definition of "advertisement"... don't do that.)

Ok, now that we have that straight... if this ships, there will be a
minority of users who will be outraged. We don't know how big that
minority will be, nor how they will act on their outrage -- switch
browers? Simmer down over time? Something else? Different people have
different levels of concern about this, which is natural. As far as I
know, we have almost zero data on what to expect, though I'd love to
hear if that's not the case.

Nick

[Nicholas Nethercote]

On Thu, May 1, 2014 at 10:19 PM, Gavin Sharp <ga...@gavinsharp.com> wrote:
>
>> I recall promises that the sponsored tiles would be clearly marked as
>> such. Is this arrow considered to be this marking?
>
> You might be interested in
> https://bugzilla.mozilla.org/show_bug.cgi?id=974736

Is there a single wiki page tracking this work? It's understandable
that the actual work is spread across multiple bugs, but it makes it
hard for outsiders to see the whole picture.

One thing that's entirely unclear to me is the mechanism by which the
sponsored tiles will be chosen. Will choices be baked into each
release? Will the browser contact a Mozilla server to request
suggestions? Something else?

Also, is there any written criteria about what content will be
acceptable in a sponsored tile, in terms of both (a) the visuals, and
(b) the organizations doing the sponsoring? Also, is there any
specific documentation about privacy concerns, such as the promise
that no tracking will be done?

[Eric Shepherd]

Yes, Nicholas is right. Trying to get away with not calling these "ads" is just going to increase the number of people that think we're full of it.

Now is not the time to be doing controversial things, don't you think?

Eric Shepherd
Sent from my iPad

[Nicholas Nethercote]

On Fri, May 2, 2014 at 5:10 PM, Eric Shepherd <eshe...@mozilla.com> wrote:
>
> Now is not the time to be doing controversial things, don't you think?

FWIW, I don't think this is the right way to evaluate this. If it's a
good idea, we should do it. If it's a bad idea, we shouldn't. (You can
guess which way I lean.) Recent history shouldn't affect that
evaluation.

[Eric Shepherd]

I normally would agree with that, but given the specific circumstances, and just how unpleasant things have been, I find myself forced to think in those terms.

That said, I also think it's a questionable plan in general, for all the reasons previously stated, so I only added this minor point to the mix instead of reiterating the rest.

Eric Shepherd
Sent from my iPad

[Gijs Kruitbosch]

On 03/05/2014 00:52, Nicholas Nethercote wrote:
> On Fri, May 2, 2014 at 6:22 AM, Mike Connor <mco...@mozilla.com> wrote:
>>> I don't look forward to the next shitstorm: "Firefox deceives users
>>> about ads and can't be trusted anymore."
>> There’s a fine line here. “Ad” isn’t really the right term
> Let's cut the BS. They are ads, pure and simple. They might be
> inobtrusive, and some users may find them useful, but don't pretend
> they're not ads; you'll (a) deceive yourself and (b) come across as
> phony -- see http://daringfireball.net/linked/2014/02/16/mozilla.

You're poisoning the well here, which I don't think is conducive to
having an honest discussion about the pros and cons of what we're doing.
The gut feeling of many of us about ads is based on tracking,
obtrusiveness (popups) and selling a product, not "here's a blank page
with 6/9/however many pointers to things on the web, pick something or
search using your configured search engine". It's a bit as if you went
into a meeting to decide about hiring someone and said "so before we
begin, let's cut the BS and refer to the prospective employee as 'stupid
person' instead of 'prospective employee', and don't you dare argue with
that". That's not a reasonable way to have a discussion.

Having the gut feeling about "ads" guide our decisionmaking isn't
helpful. We should take the gut feeling of others into account when
considering impact, sure, but discussions on this mailing list are
normally rational and technical, instead of framing a plan full of
negative emotional connotations and then going "any questions?". If you
really think that "If it's a good idea, we should do it. If it's a bad
idea, we shouldn't." then let's evaluate the idea on its merits, not on
how you think it ought to be labelled.

~ Gijs

[Panos Astithas]

On Sat, May 3, 2014 at 2:52 AM, Nicholas Nethercote <n.neth...@gmail.com> wrote:

On Fri, May 2, 2014 at 6:22 AM, Mike Connor <mco...@mozilla.com> wrote:
>
>> I don't look forward to the next shitstorm: "Firefox deceives users
>> about ads and can't be trusted anymore."
>
> There’s a fine line here.  “Ad” isn’t really the right term

Let's cut the BS. They are ads, pure and simple. They might be
inobtrusive, and some users may find them useful, but don't pretend
they're not ads; you'll (a) deceive yourself and (b) come across as
phony -- see http://daringfireball.net/linked/2014/02/16/mozilla.

(And if you are about to pull out a dictionary to look up the
definition of "advertisement"... don't do that.)

I'm not going to pull out a dictionary, but honestly, from the screenshots I've seen these tiles don't look anything like ads to me. Ads in my experience come with a significantly richer content: taglines, attractive photos, spiffy videos. They convey a message, in a way that is enticing and by necessity more verbose. You can't sell many shoes by showing the string "Nike".

Displaying tiles that contain company logos (or page screenshots, not sure what the current plan is) seems more like some films where the main character is drinking beer and "coincidentally" the label is facing the camera so we can see the brand. I don't find that practice bad either (if they don't attempt to twist the plot line to explicitly turn my attention to the product) and most of the time I don't even notice it.

Or even like wearing a T-shirt with a logo on it. It doesn't feel like I'm advertising that company when I'm wearing it, although I very likely am from their point of view. Is anyone seriously annoyed by the apple logo in the back of a Macbook? Or by the Google logo I see right now in the search box in front of me?

I'm not saying there aren't any ways to implement these tiles that would be repulsive and contrary to our mission. I'm saying that I haven't seen any indication so far that we are even remotely considering any of those. And from my experience with the Mozillians doing this work, the only way I can see them doing something like that is with a brain transplant.

[Nicholas Nethercote]

On Sat, May 3, 2014 at 12:44 AM, Gijs Kruitbosch
<gijskru...@gmail.com> wrote:
>>
>> Let's cut the BS. They are ads, pure and simple.
>

> You're poisoning the well here

I just googled for "Firefox ads". In the top 20 results, I got the
following 13 news articles about the directory tiles program, all from
the period February 11--16, when the directory tiles program was first
mentioned:

http://www.zdnet.com/mozilla-clarifies-defends-firefox-ad-position-7000026335/
http://www.zdnet.com/mozilla-to-deliver-ads-in-its-firefox-browser-7000026216/
http://www.reuters.com/article/2014/02/12/us-mozilla-advertising-idUSBREA1B1OW20140212
http://www.extremetech.com/internet/176521-dont-panic-mozilla-will-be-incorporating-ads-into-firefox
http://techcrunch.com/2014/02/12/mozilla-will-sell-ads-in-firefox-to-create-a-new-revenue-stream/
http://adage.com/article/iab-annual-meeting/mozilla-sell-ads-firefox/291641/
http://www.cnet.com/news/mozilla-to-sell-new-tab-page-ads-in-firefox/
http://www.theverge.com/2014/2/12/5404174/mozilla-will-start-showing-first-time-users-ads-in-blank-firefox-tabs
http://arstechnica.com/business/2014/02/firefox-to-start-putting-new-ads-in-new-places/
http://www.computerworld.com/s/article/9246389/Mozilla_s_top_exec_defends_in_Firefox_ads_revenue_search
http://www.pcworld.com/article/2096923/ads-are-coming-to-firefoxs-new-tab-page.html
http://siliconangle.com/blog/2014/02/12/mozilla-has-just-sold-its-soul-get-ready-for-ads-in-firefox/
http://daringfireball.net/linked/2014/02/16/mozilla

These are major publications -- I have heard of 12 out of the 13.

As for the other results in the top 20: six were about how to block
ads in Firefox (e.g. via AdBlock Plus). And the other one was Darren
Herman's original post, which is interesting, since that post did not
use the word "ads" or "advertising" (though it did appear multiple
times in the comments).

But maybe I'm being unfair with that search term. So let's try
googling for "Firefox directory tiles -ads -advertisement
-advertising" to find sites that don't call the directory tiles
program ads. In the top 20 results I get only five articles that
mention the directory tiles program from that period:

http://www.brunchnews.com/gsm-arena/blog/firefox-directory-tiles-will-bring-sponsored-content-to-your-new-tab-page-672034
http://internetdo.com/2014/02/mozilla-directory-tiles-to-put-sponsored-sites-in-firefox-new-tab-page/
http://technologybreakingnews.blogspot.com.au/2014/02/mozilla-directory-tiles-to-put.html
http://technewsworlds.wordpress.com/2014/02/12/mozilla-bringing-sponsored-tabs-to-firefox-under-guise-of-assisting-newbies/
http://www.cnet.com/news/mozilla-gets-into-content-creation-with-voices/

I have heard of one of these five sites (CNet), and the CNet article
wasn't even primarily about directory tiles.

Among the other 15 results were two copies of Darren's original post,
three Bugzilla bugs relating to the implementation of directory tiles,
two posts on Mozilla users sites asking how to turn the tiles off, two
non-English sites, and the rest were unrelated to directory tiles.

This was a quick evaluation but the results are clear: for the most
part, the only people using the term "directory tiles" are people
within Mozilla, and the rest of the world is using the word "ads".

>The gut feeling of many of us about ads is based on tracking, obtrusiveness (popups)
> and selling a product, not "here's a blank page with 6/9/however many pointers to
> things on the web, pick something or search using your configured search engine".

You're absolutely right. But even if Mozilla avoids the word "ads",
the rest of the world will use it. And directory tiles will be -- has
already been! -- tarred with these negative connotations, whether
that's fair or not. Overcoming that stigma is probably the single
hardest thing about this whole idea -- the user perception aspect of
this is *much* harder than the implementation.

So basically, your message is this: "we're going to start doing
something that lots of other people have done in majorly nasty ways,
but we're going to do it in a nice way." *Oh boy* that is a tough
sell. As a result, my fear is that even if directory tiles are
implemented in a faultless fashion it'll still end up doing more harm
than good.

Nick

[bobbuun]

On 05/03/2014 03:44 AM, Gijs Kruitbosch wrote:
> Having the gut feeling about "ads" guide our decisionmaking isn't
> helpful. We should take the gut feeling of others into account when
> considering impact, sure, but discussions on this mailing list are
> normally rational and technical, instead of framing a plan full of
> negative emotional connotations and then going "any questions?". If you
> really think that "If it's a good idea, we should do it. If it's a bad
> idea, we shouldn't." then let's evaluate the idea on its merits, not on
> how you think it ought to be labelled.
>
> ~ Gijs

Applying the "duck test", there's no way to come up that sponsored tiles
are anything but advertisements. Unless I'm mistaken, an entity pays for
their link to be placed as a tile? It's not just a list of "places the
Firefox developers thinks are cool [and haven't been paid to place them
here]"?

Even if the Mozilla community somehow comes to some other conclusion,
the users will certainly apply the logical "duck test" conclusion.

As advertisements, it's only proper, and fair for the users, that they
clearly be indicated as such. Not doing so will be seen as an even
greater affront than including them at all. Goodwill is a finite resource.

-Daryl

[Ed Lee]

On Fri, May 2, 2014 at 4:57 PM, Nicholas Nethercote
<n.neth...@gmail.com> wrote:
> Is there a single wiki page tracking this work? It's understandable
> that the actual work is spread across multiple bugs, but it makes it
> hard for outsiders to see the whole picture.

+clarkbw? Or a public doc? Various parts of the original doc were
broken down into multiple story bugs that have development attached to
each requirement.

> One thing that's entirely unclear to me is the mechanism by which the
> sponsored tiles will be chosen. Will choices be baked into each
> release? Will the browser contact a Mozilla server to request
> suggestions? Something else?

For now they're baked into each release:
https://bugzilla.mozilla.org/show_bug.cgi?id=990713

But there's already progress on moving things to a remotely hosted file:
https://bugzilla.mozilla.org/show_bug.cgi?id=986521

I don't have the details of how which tiles are chosen as which types,
but there's some updates in the former bug (990713#c14) noting sites
like reddit, NPR, EFF at that point were being considered but not yet
ready.

> Also, is there any written criteria about what content will be
> acceptable in a sponsored tile, in terms of both (a) the visuals, and
> (b) the organizations doing the sponsoring?

There's this bug describing the image sizing requirements:
https://bugzilla.mozilla.org/show_bug.cgi?id=975199

Although you're probably looking more for the content guidelines,
e.g., animations, text, promotions. As far as I can tell from the
images provided so far, they've been logos and/or wordmarks.

> Also, is there any specific documentation about privacy concerns,
> such as the promise that no tracking will be done?

In regards to privacy, I believe both the Privacy/Public Policy and
Security/Privacy teams are looking to evaluate how the community
understands tracking to best message what we're doing. E.g., would
users understand if we said these Tiles behaviors observe the user's
DNT intent even if there was no tracking.

Ed Lee

[Eric Shepherd]

That's odd. I see all-text ads all the time. Google shows them everywhere. For that matter, they're in the newspaper every day. Have you not seen classified ads? Oh hey, there's that word. :)

I'm not trying to be flippant; this is an important point. You can try to argue the fine point of semantics all you want, but the very fact that this debate is happening means that the problem of perception is a real one. It doesn't matter what you think; it matters what others think.

This is a lesson we have to repeat to my daughter all the time: it doesn't matter what you think you're doing; it matters what the people around you think you're doing.

Eric Shepherd

Sent from my iPad

[Tom Schuster]

Just came across this quote.

>>I am critical of this. The most obvious problem is that they don't indicate in any way which of the tiles are organic and which are sponsored. That is problematic!

>I think the image in the article is just an early mockup. I'm sure there will be some differentiation in the actual release. This is Mozilla we're talking about here, they've kinda got a track record for being transparent about this sort of >thing.

http://www.reddit.com/r/linux/comments/24lv63/firefoxs_sponsored_tiles_will_look_like_this/ch8imvc

Tom

[Chris Hofmann]

On 5/3/14 8:30 AM, Tom Schuster wrote:

Just came across this quote.

>>I am critical of this. The most obvious problem is that they don't indicate in any way which of the tiles are organic and which are sponsored. That is problematic!

>I think the image in the article is just an early mockup. I'm sure there will be some differentiation in the actual release. This is Mozilla we're talking about here, they've kinda got a track record for being transparent about this sort of >thing.

http://www.reddit.com/r/linux/comments/24lv63/firefoxs_sponsored_tiles_will_look_like_this/ch8imvc

Gavin's quote from earlier is a good way to think about this.

- making it possible for interested users to know the provenance of
sponsored tiles _is_ a design goal (which I think is accomplished by
the button design from bug 974736, though certainly the design could
be tweaked in various ways)

From the reddit comments I'd say that  that our new icon "->" and button design could be failing the test of users being easily able to identify the provenance of the tile.  The extra step of having to click on that symbol is not making into their heads so they aren't seeing that the content is sponsored and its helping mozilla...   Its not clear to me that users would ever just make a connect of this icon "->" to Sponsorship or Ads or not.

I'm confused that we just don't say "Sponsor ->" on the title bar  to identify that the content provider is helping to sponsor Mozilla's activities.   That seems it would be a benefit to the content provider (creating a closer connection to Mozilla's public benefit), and the user who likes to know the provenance of the content.

-chofmann

[Gijs Kruitbosch]

On 03/05/2014 12:24, Nicholas Nethercote wrote:
> And the other one was Darren
> Herman's original post, which is interesting, since that post did not
> use the word "ads" or "advertising" (though it did appear multiple
> times in the comments).
>
> But maybe I'm being unfair with that search term. So let's try
> googling for "Firefox directory tiles -ads -advertisement
> -advertising" to find sites that don't call the directory tiles

I suspect this will also exclude pages that get linked to from pages
that use the phrase "ads" or "advertisement" and so on for the link,
because of how PageRank works. That means your comparison isn't fair,
but nevermind that. I also think that the press loves a story, and based
on Darren's original post, "ads" might have been a fair critique. I'm
hopeful the eventual implementation won't be.

> But even if Mozilla avoids the word "ads",
> the rest of the world will use it. And directory tiles will be -- has
> already been! -- tarred with these negative connotations, whether
> that's fair or not. Overcoming that stigma is probably the single
> hardest thing about this whole idea -- the user perception aspect of
> this is *much* harder than the implementation.

I'm not sure this is true. It's true for me, and probably for you -
we're engineers. If we had marketing do the implementation, I'm sure
they'd feel the opposite way. :-)

Having seen the proposals back in January, and noting Ed's post here:

> All 9 tiles shown in the screenshot are "Directory Tiles" although
> only 2 of them are "sponsored." Internally we label some as "organic"
> for popular sites and "affiliate" type has been put on Lightbeam,
> Mozilla Foundation, Webmaker.

I genuinely think that the reality of the implementation will help solve
a lot of the concerns, and equally, that if it doesn't, it won't be too
late to ditch the sponsoring part of the project if your fears turn out
to be realized.

In the meantime, however, I don't think calling these things "ads"
contributes to having a healthy discussion about it. Neither you nor
most other people, for instance, have discussed the motivation/necessity
(and corresponding benefits) of having sponsored tiles. It's not exactly
fair to discuss the proposal only based on its negative press due to
imperfect communication from our side.

> So basically, your message is this: "we're going to start doing
> something that lots of other people have done in majorly nasty ways,
> but we're going to do it in a nice way."

I'm not saying that - you said I couldn't discuss why these aren't ads,
so I didn't! I don't think we're doing "something that lots of other
people have done in majorly nasty ways".

I'm saying that if we're going to have another discussion on the same
subject, it ought to be a fair one. However, I'm generally skeptical
that that will be happen without having a concrete implementation (and
policy guidelines about the content of the (sponsored) tiles) about
which to discuss.

~ Gijs

[Gijs Kruitbosch]

On 03/05/2014 13:42, bobbuun wrote:
> On 05/03/2014 03:44 AM, Gijs Kruitbosch wrote:
>> Having the gut feeling about "ads" guide our decisionmaking isn't
>> helpful. We should take the gut feeling of others into account when
>> considering impact, sure, but discussions on this mailing list are
>> normally rational and technical, instead of framing a plan full of
>> negative emotional connotations and then going "any questions?". If you
>> really think that "If it's a good idea, we should do it. If it's a bad
>> idea, we shouldn't." then let's evaluate the idea on its merits, not on
>> how you think it ought to be labelled.
>>
>> ~ Gijs
>
> Applying the "duck test", there's no way to come up that sponsored
> tiles are anything but advertisements. Unless I'm mistaken, an entity
> pays for their link to be placed as a tile? It's not just a list of
> "places the Firefox developers thinks are cool [and haven't been paid
> to place them here]"?

As Ed said earlier in the thread, the plan is for there to be organic as
well as sponsored tiles, and to have the user's own history/bookmarks
come to determine the tiles in the end. So no, I don't think the feature
as a whole passes a "duck test".

I'm also confused, because certainly (a) certain search provider(s)
pay(s) Mozilla for their placement in and/or searches/"clicks" from the
search provider list, and others (e.g. Wikipedia) don't. Are you arguing
those also fail the "duck test" for advertisement?

(I'll note that all of these points were raised the previous time all of
this was discussed, and as I said in my other message 5 minutes ago, I
don't think having another in-depth discussion about this feature is
productive without an implementation)

> As advertisements, it's only proper, and fair for the users, that they
> clearly be indicated as such. Not doing so will be seen as an even
> greater affront than including them at all. Goodwill is a finite
> resource.

Nobody's arguing they won't/shouldn't be distinguishable. People are
arguing over *how*, what word/icon, etc. Not the same thing.

~ Gijs

[Nicholas Nethercote]

On Fri, May 2, 2014 at 5:38 PM, Ed Lee <edi...@mozilla.com> wrote:
>
>> One thing that's entirely unclear to me is the mechanism by which the
>> sponsored tiles will be chosen. Will choices be baked into each
>> release? Will the browser contact a Mozilla server to request
>> suggestions? Something else?
> For now they're baked into each release:
> https://bugzilla.mozilla.org/show_bug.cgi?id=990713

Thanks for this link and the others. I see in
toolkit/content/directoryLinks.json that Wired and Trulia are the two
sponsored links for the en-US locale. Are these placeholders, or have
business deals of some kind actually been completed for these sites?

Nick

[Nicholas Nethercote]

On Sat, May 3, 2014 at 1:19 PM, Gijs Kruitbosch
<gijskru...@gmail.com> wrote:
>
> I genuinely think that the reality of the implementation will help solve a
> lot of the concerns, and equally, that if it doesn't, it won't be too late
> to ditch the sponsoring part of the project if your fears turn out to be
> realized.

How will that decision be made, and who makes it? What happens if
contracts with sponsors have been signed?

Nick

[Panos Astithas]

I agree that all-text ads are very much ads, but I've never seen a text-only ad that didn't include a tagline or a copy of some sort to make it work. The difference I'm trying to highlight is between a classified ad for a Volvo where the copy probably describes how safe they are, and a classified ad that consists of the single word "Volvo". I'm arguing that there are no ads of the second kind.

I agree that perception is important and that this is a useful debate to be had, which is why I'm contributing my point of view. I don't agree with the view that some have expressed that this is such a controversial topic that we should avoid it at all costs.

[Gavin Sharp]

On Sun, May 4, 2014 at 12:29 AM, Nicholas Nethercote
<n.neth...@gmail.com> wrote:
> On Sat, May 3, 2014 at 1:19 PM, Gijs Kruitbosch
> <gijskru...@gmail.com> wrote:
>>
>> I genuinely think that the reality of the implementation will help solve a
>> lot of the concerns, and equally, that if it doesn't, it won't be too late
>> to ditch the sponsoring part of the project if your fears turn out to be
>> realized.
>
> How will that decision be made, and who makes it? What happens if
> contracts with sponsors have been signed?

The desktop product owners (Gavin, Madhava, Chad) will make this
decision, with input from a lot of other stakeholders.

We're not going to sign any contracts that make it impossible for us
to do the right thing.

Gavin

[Gavin Sharp]

We certainly have a "perception problem". I think we still have an
opportunity to correct that problem with action. I still think that
it's possible for us to eventually ship a feature that will not only
offer benefits to, but also be perceived as beneficial by, the
majority of our users. We're trying to prove that with some
experiments.

You seem to disagree - you think we're doomed to fail no matter what.
It's fine for you to hold that opinion, and maybe you'll ultimately be
right. But I don't think we know enough to make that judgement yet,
and so we're going to move forward with the experimentation.

Gavin

[Bryan Clark]

What has landed are kind of placeholders, we're experimenting with what how sponsored tiles should be indicated. There are no sponsorship deals in place for anything yet, we're looking to iterate with the design of our sponsored tile indicator and measure if people find these valuable before considering a business deal.

Bryan

[David Rajchenbach-Teller]

On 05/05/14 19:20, Gavin Sharp wrote:
> The desktop product owners (Gavin, Madhava, Chad) will make this
> decision, with input from a lot of other stakeholders.
>
> We're not going to sign any contracts that make it impossible for us
> to do the right thing.

This is a good idea.

Fwiw, I would be more comfortable with a presentation that explicitly
writes "Ad" or "Sponsored" or at least makes the small arrow more
understandable. However, I'm willing to give the current a try.

Cheers,
David

[Gavin Sharp]

On Fri, May 2, 2014 at 7:52 PM, Nicholas Nethercote
<n.neth...@gmail.com> wrote:
>> There’s a fine line here. “Ad” isn’t really the right term
>
> Let's cut the BS. They are ads, pure and simple. They might be
> inobtrusive, and some users may find them useful, but don't pretend
> they're not ads; you'll (a) deceive yourself and (b) come across as
> phony -- see http://daringfireball.net/linked/2014/02/16/mozilla.

"ads are ads and they're all bad" is too reductionist to be useful.
There are different types of ads, and that word carries different
meanings/connotations to different people. That's the core of the
debate over terminology here, and I think it's mostly a distraction.

What matters is what users experience and what they perceive - we're
on the same page there. If users overwhelmingly come away from the
feature with negative sentiment (for whatever reason), then clearly
we've failed. As mentioned in my previous post, I disagree with your
opinion that we're doomed to failure here, and so am willing to invest
in some further experimentation. I am asking that you have faith in
the Firefox leadership team to not fuck it up.

Gavin

[Chris Peterson]

On 5/5/14, 10:29 AM, Gavin Sharp wrote:
> You seem to disagree - you think we're doomed to fail no matter what.
> It's fine for you to hold that opinion, and maybe you'll ultimately be
> right. But I don't think we know enough to make that judgement yet,
> and so we're going to move forward with the experimentation.

Reading Nicholas' posts, I don't think he believes we're doomed to fail
no matter what. I think he is saying we should not ignore that people's
wide-spread perception is that directory tiles are ads.


chris

[Chris Peterson]

On 5/3/14, 12:59 AM, Panos Astithas wrote:
> Or even like wearing a T-shirt with a logo on it. It doesn't feel like
> I'm advertising that company when I'm wearing it, although I very
> likely am from their point of view. Is anyone seriously annoyed by the
> apple logo in the back of a Macbook? Or by the Google logo I see right
> now in the search box in front of me?

Apple and Google don't have to pay to place their own logos on their own
products. Those are not ads, but a paid-placement of an Apple or Google
logo on a non-Apple or non-Google product is an ad. Mozilla about:home
showing a Firefox logo is not an ad, but a paid-placement of (say) a
TechCrunch logo and link is an ad.

chris

[Gavin Sharp]

The "wide-spread perception" you're referring to is:

a) held by a subset of the people who are even aware that we're
considering this (i.e. who read blogs/hacker news/tech press, etc.),
which is obviously a subset of "people who use Firefox"
b) is largely based on having read misleading/confusing articles and
posts, not from having used or experienced the feature

What matters most is what people perceive while using Firefox, and
while what they read can certainly influence that, it's not the most
significant factor. It's premature to conclude that we can't succeed
based only on feedback we've seen so far. That's where I think
Nicholas is wrong.

Gavin

[Axel Hecht]

Doing an almost-top-reply.

There's a host of back and forth about what directory tiles maps to in
the the real world.

IMHO, the best match is "product placement" for the tiles we do with
financial benefits, and "co-branding" for those that we include for values.

If our financial deals will be about "product placement" in a mozilla
way, we'll come out fine. The risk we're facing is that we're giving out
data as if we'd do ads, just because that's what people are used to pay
for. I've seen too many people of ours argue that way. We need to boldly
own this.

Another part of mozilla boldly owning this is that we listen and engage
with local mozillians about the values and brands we expose across the
globe.

Axel

On 5/2/14 2:52 PM, Steffen Wilberg wrote:
> On 02.05.2014 07:19, Gavin Sharp wrote:

>> On Fri, May 2, 2014 at 1:06 AM, Nicholas Nethercote
>> <n.neth...@gmail.com <mailto:n.neth...@gmail.com>> wrote:
>> I recall promises that the sponsored tiles would be clearly marked as
>> such. Is this arrow considered to be this marking?
>>
>>

>> You might be interested in
>> https://bugzilla.mozilla.org/show_bug.cgi?id=974736
>>
>> Gavin
>
> The arrow as shown in the screenshot from
> https://bug990713.bugzilla.mozilla.org/attachment.cgi?id=8414784 is very
> subtle and easy to overlook.
>
> Why don't we mark sponsored tiles more clearly, like e.g. Google.com
> does in search results, which display "Ad" in white text on yellow
> background?

>
> I don't look forward to the next shitstorm: "Firefox deceives users
> about ads and can't be trusted anymore."
>

> Steffen

[Nicholas Nethercote]

On Mon, May 5, 2014 at 10:52 AM, Bryan Clark <cla...@gnome.org> wrote:
> What has landed are kind of placeholders, we're experimenting with what how
> sponsored tiles should be indicated. There are no sponsorship deals in place
> for anything yet, we're looking to iterate with the design of our sponsored
> tile indicator and measure if people find these valuable before considering
> a business deal.

Thanks for the info. I guess Wired and Trulia essentially random
choices that somebody made, and they don't have any particular
meaning?

It looks like the sponsored tiles will be tied to the locale. It's
worth noting that this can lead to some unhelpful suggestions. For
example, I'm in Australia so I get the en-US locale. Most of the
suggestions are fine, but Trulia is entirely uninteresting, because it
doesn't work outside the US. Canadians, New Zealanders, etc, will be
in the same boat, and I imagine there could be similar cases with
other locales.

[Nicholas Nethercote]

On Mon, May 5, 2014 at 10:29 AM, Gavin Sharp <ga...@gavinsharp.com> wrote:
> We certainly have a "perception problem".

Yes!

> I think we still have an
> opportunity to correct that problem with action.

That's a big part of our disagreement. You're optimistic about this.
I'm deeply pessimistic about this.

I'll finish with one concrete suggestion. When this is enabled for
evaluation in Nightly, *please* ensure that there is a prominent FAQ
somewhere (a small link from the new tab page is a possibility) which
explains how it works, how to turn it off, that it doesn't track you,
etc. In fact, I'm happy to file a bug for this, though I'm not sure
what component it would be in.

[Bryan Clark]

As a person living in Canada running the en-US build I understand.  Using the locale as the only key for choosing Tiles won't ever be very useful.  Actually it is a problem in a number of other ways as well.  You don't have the correct spell check dictionary, search engines, and general localization of the menu items. 

We might have the opportunity to offer a local experience but right now there isn't a good system in Firefox for that.  Are there solutions out there that others are aware of?  We've been looking into ways we can correct for the incorrect locale but I'd appreciate input from people here.

As for choosing Tiles we're working on an open process, I'll give you my vision of it right now.  (This is as I see it, not necessarily indicative of what it will be)  The goal I have sketched out is a set of Tiles localized by the community but using whatever data metrics we can get back (via Telemetry or other methods) to verify our choices.  Each region would likely have a list of more than 9 Tiles that rotate into a given build.  After the community votes to add a new Tile we do a "test run" where all builds include the new Tile for some period of time (lets say a couple months) after which we look back at the usage data.  This doesn't mean we have to live and die by data but ideally reinforce good choices with real information (if possible).  Tiles which get well used could stay and poorly used Tiles which have run for a certain period could be removed from the list. 

Assuming it's possible to create a dynamic system like this I would hope it gives each community much more control and awareness over the choices for their region.  What we have now is just a baseline guess that needs validation.  I've ignored a number of details here including getting clearance from websites to link to them and all the behind the scenes work of getting the correct images etc. but that boring stuff I can detail out later.

Thanks,

~ Bryan

[Chris Hofmann]

On 5/7/14 4:34 PM, Bryan Clark wrote:

As a person living in Canada running the en-US build I understand.  Using the locale as the only key for choosing Tiles won't ever be very useful.  Actually it is a problem in a number of other ways as well.  You don't have the correct spell check dictionary, search engines, and general localization of the menu items. 

We might have the opportunity to offer a local experience but right now there isn't a good system in Firefox for that.  Are there solutions out there that others are aware of?  We've been looking into ways we can correct for the incorrect locale but I'd appreciate input from people here.

As for choosing Tiles we're working on an open process, I'll give you my vision of it right now.  (This is as I see it, not necessarily indicative of what it will be)  The goal I have sketched out is a set of Tiles localized by the community but using whatever data metrics we can get back (via Telemetry or other methods) to verify our choices.  Each region would likely have a list of more than 9 Tiles that rotate into a given build.  After the community votes to add a new Tile we do a "test run" where all builds include the new Tile for some period of time (lets say a couple months) after which we look back at the usage data.  This doesn't mean we have to live and die by data but ideally reinforce good choices with real information (if possible).  Tiles which get well used could stay and poorly used Tiles which have run for a certain period could be removed from the list. 

This is where it gets hard, and leads to the skepticism that we can't do a good job of picking the "right" content when people use the internet for so many different purposes,  and the things that aren't "on-purpose" are often viewed as getting in the way.  Separation on locale is just one of the dimensions that need to be matched to get the right content to the right set of people.

This research shows how differences in gender and age might also be desirable content and activity on the web.
http://www.infoplease.com/ipa/A0931238.html

That's the great thing about the web.  Its not like commercial TV where a few people get to pick what everyone should see, or even like cable TV where there might be hundreds or thousands of choices.   The web means unlimited choice and everyone gets to tune their own experience.

How exactly will the data tell us if we've made good choices, or bad?  And how do users get to "tune" their own choices in this feature. Those are the hard question to be thinking about.

-chofmann

[Bryan Clark]

Agreed, we'll be only able to offer broad appeal content but regional content would be a good step forward.  This comes across with other our other content like search engines.  Easily searching the AARP site isn't something useful for myself quite yet but I'm sure a lot of (likely under represented) US Firefox users might find that handy.  The UP (User Personalization) team has spent much more time thinking of how we can offer content that is "tuned" using privacy respecting methods and I think they might be able to offer some great options in the future.

[Chris Hofmann]

On 5/8/14 9:28 AM, Bryan Clark wrote:

Agreed, we'll be only able to offer broad appeal content

How do we define that?  I've talked a bit to Darren about needing to do this, but we need to get that discussion going.

"Broad Appeal" is hard when you look at the diversity in the ways that people use the web.

If you look at even the most popular kinds of content and activities you still mostly end up serving or attracting less than half of users in the general population.   Take a broad appeal segment like "get news" for which 66%-77%  have interest in doing.   Then try to figure out if we are talking about NPR, FoxNews, CNN, or some local news channel, etc....   offering any of those suddenly looses broad appeal.

Interested in how people define "broad appeal" and how that translates into specific content.

-chofmann

[Darren Herman]

chofmann, you are completely right when it comes to “broad appeal” and how people use the web.  With 9 tiles and no personalization, there is absolutely no way we’re going to satisfy everyone.  So what we’ve done is looked at the Alex Top 500 sites list and reached out to potential partners who would be relevant.  And for the partners who responded back affirmatively, they are participating with us.  In addition, we wanted a varied group of partners (not just all news, sports, tech, etc) to understand how they perform.

Teams like UP (led by PM Kevin Ghim) are working on ways to think thru personalization.  And maybe personalization can lead into Tiles.   But this is putting carts ahead of the horse.  Way ahead.

Darren