This is not particular urgent; just some thoughts that have been turning over in my mind for a bit.
Our current top-level marketing strategy is based primarily on the three Fedora editions and their target audiences. We have a secondary marketing strategy around more focused solutions: for example, the Python Classroom Lab has the simple target of teachers and instructors. Or the various desktop spins, which target enthusiasts of the particular desktop technologies.
I'd love for each Edition WG and Spin/Lab SIG to come up with search terms that reflect these goals — for example, ranking high for "desktop for developers" might be a goal for Workstation.
But, specific editions aside, what kind of terms would we like to focus on? I'd love for Fedora to be seen as the leader in community Linux distros; should we target simple terms like just *Linux*? Beyond search engines, do we *want* to market Fedora in that way? Or, do we want to do the Android-style thing and market Fedora's OS offerings as their own stand-alone brand?
On 09/18/2017 01:57 PM, Matthew Miller wrote:
Our current top-level marketing strategy is based primarily on the three Fedora editions and their target audiences.
We talked about this, but right now it's actually product-centric, not audience-centric (and should probably be the latter.)
We have a secondary marketing strategy around more focused solutions: for example, the Python Classroom Lab has the simple target of teachers and instructors. Or the various desktop spins, which target enthusiasts of the particular desktop technologies.
Do we really actively market these though?
I'd love for each Edition WG and Spin/Lab SIG to come up with search terms that reflect these goals — for example, ranking high for "desktop for developers" might be a goal for Workstation.
If I search for "developer desktop" the top non-ad hit is https://www.ubuntu.com/desktop/developers
According to the latest Stack Overflow developer survey (https://insights.stackoverflow.com/survey/2016), Ubuntu represents 12.3% of developers; Windows is #1 @ 52.5%, OS X #2 @ 26.2%.
In 10 pages of results I didn't get any direct hits for Windows or OS X. On page 3 I got a single article that referenced Windows for developers. On page 9 I got a single article referencing the stack overflow survey and OS X's position on it.
Interestingly, the results for "developer":
1. + 2. developer.apple.com 3. developers.google.com 4. wikipedia entry for developer 5. developer.android.com 6. developers.facebook.com 7. developer.servicenow.com 8. lynda.com developer training 9. indeed.com developer jobs 10. developer.com (looks stuck in 1998)
Note in both these instances Windows is the #1 platform by a long shot....
Search engine position is an easy number to get and compare over time, but is there convincing evidence that it's meaningful? Is it meaningful in either of these senses?:
1 - Good position in rankings will help make $THING more popular 2 - Good position in rankings reflects popularity of $THING
Is there something else SEO could potentially provide that I'm missing?
But, specific editions aside, what kind of terms would we like to focus on? I'd love for Fedora to be seen as the leader in community Linux distros; should we target simple terms like just *Linux*? Beyond search engines, do we *want* to market Fedora in that way? Or, do we want to do the Android-style thing and market Fedora's OS offerings as their own stand-alone brand?
Optimizing for "Linux" doesn't seem the right way to go in 2017. e.g.
https://trends.google.com/trends/explore?date=all&q=linux,app
Two of our 3 editions are focused on developer workflows, but we do not go to conferences that are primarily developer-centric, we do not talk about or mention topics that are of interest to developers (many referenced in that survey) on/in any of our external-facing materials such as our brochure site or any of our marketing materials, save for Fedora Magazine and the getfedora.org site (the latter could be much better)
As we've discussed, I think maybe merging getfedora.org with developer.fedoraproject.org and having our main user-focused site centered around developer topics could help.
Reaching out and giving Fedora a presence where developers are would also help - here are some ideas -
- development classes on udemy / skillshare / udacity / youtube / etc showcasing the tech in fedora applied to development problems + skill mastery
- starting an ambassadors program for developers who are interested in Fedora to represent us at developer conferences
- seeking out more and deeper developer content for Fedora Magazine
I think at this point in time, without a coherent narrative about what we have to offer, SEO is not actually useful - we won't target the right terms. We need a tighter and richer feedback loop with our target audience to understand what we have to offer and where we need to improve and we need to work on improving in a visible way towards those unmet needs. Build the narrative on that. Without a narrative, if we promote the right thing but we're deficient, it's not going to help it will hurt; if we promote the wrong thing, it won't help either.
Cheers ~m
On Tue, Sep 19, 2017 at 02:06:38PM -0400, Máirín Duffy wrote:
Our current top-level marketing strategy is based primarily on the three Fedora editions and their target audiences.
We talked about this, but right now it's actually product-centric, not audience-centric (and should probably be the latter.)
Well, the marketing strategy is intended to be audience-centric, but structured around the editions as a way of organizing the project. That's somewhat different from whether the website is audience or edition-centric.
We have a secondary marketing strategy around more focused solutions: for example, the Python Classroom Lab has the simple target of teachers and instructors. Or the various desktop spins, which target enthusiasts of the particular desktop technologies.
Do we really actively market these though?
Secondarily. :)
I'd love for each Edition WG and Spin/Lab SIG to come up with search terms that reflect these goals — for example, ranking high for "desktop for developers" might be a goal for Workstation.
If I search for "developer desktop" the top non-ad hit is
That would be an excellent one for us to improve. Right now, Google webmaster console puts getfedora at 100 for that, although it's possible that https://developer.fedoraproject.org/ (which I don't currently have visibility into) scores higher.
Search engine position is an easy number to get and compare over time, but is there convincing evidence that it's meaningful? Is it meaningful in either of these senses?:
1 - Good position in rankings will help make $THING more popular 2 - Good position in rankings reflects popularity of $THING
I think #2 is _probably_ true. And #1 is probably true if advertising works at all, which it seems to.
We can also get numbers on click-through %. Just being the top result and never having any resulting traffic is less useful.
Two of our 3 editions are focused on developer workflows, but we do not go to conferences that are primarily developer-centric, we do not talk about or mention topics that are of interest to developers (many referenced in that survey) on/in any of our external-facing materials such as our brochure site or any of our marketing materials, save for Fedora Magazine and the getfedora.org site (the latter could be much better)
I definitely agree.
I think at this point in time, without a coherent narrative about what we have to offer, SEO is not actually useful - we won't target the right terms. We need a tighter and richer feedback loop with our target audience to understand what we have to offer and where we need to improve and we need to work on improving in a visible way towards those unmet needs. Build the narrative on that. Without a narrative, if we promote the right thing but we're deficient, it's not going to help it will hurt; if we promote the wrong thing, it won't help either.
Hmmmm. I definitely agree on the importance of getting the narrative right — and on backing it up with real tech. But I think there's also low-hanging fruit we can handle to increase visibility.
2017-09-19 22:43 GMT+02:00 Matthew Miller mattdm@fedoraproject.org:
On Tue, Sep 19, 2017 at 02:06:38PM -0400, Máirín Duffy wrote:
Our current top-level marketing strategy is based primarily on the three Fedora editions and their target audiences.
We talked about this, but right now it's actually product-centric, not audience-centric (and should probably be the latter.)
Well, the marketing strategy is intended to be audience-centric, but structured around the editions as a way of organizing the project. That's somewhat different from whether the website is audience or edition-centric.
I think the same as Mairin, the latter you mentioned, Matthew.
We have a secondary marketing strategy around more focused solutions: for example, the Python Classroom Lab has the simple target of teachers and instructors. Or the various desktop spins, which target enthusiasts of the particular desktop technologies.
Do we really actively market these though?
Secondarily. :)
Probably we want to make this happen for Python? I never saw any marketing message for Security, although this is an important market nowadays, and companies are looking for security specialists. Why not telling them we have a dedicated spin for that. Secondarily probably means, we want but didn't until now? :)
I'd love for each Edition WG and Spin/Lab SIG to come up with search terms that reflect these goals — for example, ranking high for
"desktop for
developers" might be a goal for Workstation.
If I search for "developer desktop" the top non-ad hit is
That would be an excellent one for us to improve. Right now, Google webmaster console puts getfedora at 100 for that, although it's possible that https://developer.fedoraproject.org/ (which I don't currently have visibility into) scores higher.
Search engine position is an easy number to get and compare over time, but is there convincing evidence that it's meaningful? Is it meaningful in either of these senses?:
1 - Good position in rankings will help make $THING more popular 2 - Good position in rankings reflects popularity of $THING
I think #2 is _probably_ true. And #1 is probably true if advertising works at all, which it seems to.
We can also get numbers on click-through %. Just being the top result and never having any resulting traffic is less useful.
Two of our 3 editions are focused on developer workflows, but we do not go to conferences that are primarily developer-centric, we do not talk about or mention topics that are of interest to developers (many referenced in that survey) on/in any of our external-facing materials such as our brochure site or any of our marketing materials, save for Fedora Magazine and the getfedora.org site (the latter could be much
better)
I definitely agree.
Me too.
I think at this point in time, without a coherent narrative about what we have to offer, SEO is not actually useful - we won't target the right terms. We need a tighter and richer feedback loop with our target audience to understand what we have to offer and where we need to improve and we need to work on improving in a visible way towards those unmet needs. Build the narrative on that. Without a narrative, if we promote the right thing but we're deficient, it's not going to help it will hurt; if we promote the wrong thing, it won't help either.
Hmmmm. I definitely agree on the importance of getting the narrative right — and on backing it up with real tech. But I think there's also low-hanging fruit we can handle to increase visibility.
Yes, but this returns on your first question. If you like to get the users we are targeting we should not go for Linux terms or researches. Fedora has a different target, and if we really go for that, then also ranking will be better.
-- Matthew Miller mattdm@fedoraproject.org Fedora Project Leader _______________________________________________ council-discuss mailing list -- council-discuss@lists.fedoraproject.org To unsubscribe send an email to council-discuss-leave@lists. fedoraproject.org
Well, the marketing strategy is intended to be audience-centric, but structured around the editions as a way of organizing the project. That's somewhat different from whether > the website is audience or edition-centric.
"Marketing strategy" sounds more formal and organized than what I am familiar with and aware of. Are there any pointers to specifics on this, outside of the website efforts? I am interested because a formalized strategy would help us make design decisions about collateral / materials we're asked to create on the design team.
We have a secondary marketing strategy around more focused solutions: for example, the Python Classroom Lab has the simple target of teachers and instructors. Or the various desktop spins, which target enthusiasts of the particular desktop technologies.
Do we really actively market these though?
Secondarily. :)
This thread is the first time I heard of the Python Classroom Lab. It sounds like an awesome project.
Search engine position is an easy number to get and compare over time, but is there convincing evidence that it's meaningful? Is it meaningful in either of these senses?:
1 - Good position in rankings will help make $THING more popular 2 - Good position in rankings reflects popularity of $THING
I think #2 is _probably_ true. And #1 is probably true if advertising works at all, which it seems to.
A bunch of data I referenced was snipped wrt specific terms and results I saw; I am wondering if you can corroborate them w the tool you have?
I dont understand how #2 can be true if the #1 ranked dev desktop by far and #2 ranked desktop by quite a lot arent reflected in the results at all. How can it reflect whats going on when 75% of the dev desktop base (win 50 + os x 25) isnt reflected?
Of course advertising works, but SEO isnt advertising. None of this works without data. You need data to drive SEO. You need information about the audience youre targeting to even know what you're optimizing for so it's relevant and not unnaturally forced into your content. Do we have these marketing data resources?
I think cleaning up the metadata on our site and keeping content fresh etc - these are good things. There's a good oppty here to clean things up, make our message clearer, update things and maybe even institute better processes / initiatives for keeping them fresh. However, lacing our site with keywords and stuffing our page titles with buzzwords to vy for the attention of a bot whose algorithm changes daily is not ok. The core function of the website is to communicate a readable message to humans, right? Let's not compromise the consistent and coherent voice we've established....
~m
My two cents on marketing of DE spins, I think that this question needs to be clarified first, who do we actually target, before we're talking about marketing.
Or the various desktop spins, which target enthusiasts of the
particular desktop technologies.
Do we target enthusiasts for desktop technologies or just people "who want to take a look?" In recent devel thread "A less bloated KDE Spin" it was specifically said several times, that we target only people who want to take a look. Not an enthusiast for the desktop environment in question. If our marketing is indeed aimed towards said enthusiast, they're gonna be upset after installing a few gigabytes of unwanted software.
If we do want to target said enthusiasts to provide an easier way to install their favorite DE, we need a version for every spin without all the show-off-software that people don't actually use.
Regards, Radka
------------------------------ *Radka Janeková* .NET & OpenShift Engineer, Red Hat *radka.janek@redhat.com radka.janek@redhat.com* IRC: radka | Freenode: Rhea
On Wed, Sep 20, 2017 at 5:52 AM, Máirín Duffy duffy@fedoraproject.org wrote:
Well, the marketing strategy is intended to be audience-centric, but structured around the editions as a way of organizing the project. That's somewhat different from whether > the website is audience or edition-centric.
"Marketing strategy" sounds more formal and organized than what I am familiar with and aware of. Are there any pointers to specifics on this, outside of the website efforts? I am interested because a formalized strategy would help us make design decisions about collateral / materials we're asked to create on the design team.
We have a secondary marketing strategy around more focused solutions: for example, the Python Classroom Lab has the simple target of teachers and instructors. Or the various desktop spins, which target enthusiasts of the particular desktop technologies.
Do we really actively market these though?
Secondarily. :)
This thread is the first time I heard of the Python Classroom Lab. It sounds like an awesome project.
Search engine position is an easy number to get and compare over time, but is there convincing evidence that it's meaningful? Is it meaningful in either of these senses?:
1 - Good position in rankings will help make $THING more popular 2 - Good position in rankings reflects popularity of $THING
I think #2 is _probably_ true. And #1 is probably true if advertising works at all, which it seems to.
A bunch of data I referenced was snipped wrt specific terms and results I saw; I am wondering if you can corroborate them w the tool you have?
I dont understand how #2 can be true if the #1 ranked dev desktop by far and #2 ranked desktop by quite a lot arent reflected in the results at all. How can it reflect whats going on when 75% of the dev desktop base (win 50
- os x 25) isnt reflected?
Of course advertising works, but SEO isnt advertising. None of this works without data. You need data to drive SEO. You need information about the audience youre targeting to even know what you're optimizing for so it's relevant and not unnaturally forced into your content. Do we have these marketing data resources?
I think cleaning up the metadata on our site and keeping content fresh etc
- these are good things. There's a good oppty here to clean things up, make
our message clearer, update things and maybe even institute better processes / initiatives for keeping them fresh. However, lacing our site with keywords and stuffing our page titles with buzzwords to vy for the attention of a bot whose algorithm changes daily is not ok. The core function of the website is to communicate a readable message to humans, right? Let's not compromise the consistent and coherent voice we've established....
~m
council-discuss mailing list -- council-discuss@lists.fedoraproject.org To unsubscribe send an email to council-discuss-leave@lists. fedoraproject.org
Hi team,
I see several things going on here.
1. Targeted oriented campaigns: We already have 2 tickets to talk/discuss this [1][2]. We had feedback for some weeks but now both are kind of dead. People looks like just discuss things when Matt write an email, not in the tickets or in the meetings.
2. Interest: I don't want to sound like if I'm complaining but this is an example:
This thread is the first time I heard of the Python Classroom Lab. It
sounds like an awesome project. Of how people inside the project is not reading neither the Talking Point nor the annoucements, so: What can we wait outside the project? (please, don't take this like an attack, it's just quotting an interesting thing that we supose to highlight, even when there is an article in the magazine for it [3]) If we really want to push things better in the marketing strategies and in the team in general a good idea is to concrete ideas, attend meetings and read/create/use the ticketing system.
3. Involvement: We don't have an specific topic in the marketing strategies, we are focused in finishing the Talking Points that supose to help people to promote the use of the distro. Please read this carefully "promote the use of the distro". I never felt an urgent desire to promote developing environments, or specific topics; we just cover the changes and the bleeding/leading tech included in each release. One of the more important things about why this is happening, IMHO, is because there isn't a lot of people doing Marketing. The Marketing team have a lot of things, like Social Networks presence, with Twitter, Facebook, the Magazine and Telegram, also we supose to have contact with media (news blogs or anything here), that I have never seen. Normally you can see the news sites copying from the ML or the release announcements. I'm probably wrong on this, but really if these contacts exists are not reflected in ML or anywhere (wiki page about media contacts is outdated). I wasn't here when the strategy about "Fedora <3 Python" came out or when it was planned, but that was the last strategy really planned.
4. Topics: We need to define what we want to do and call people to help in each specific topic. For example, to do this:
I'd love for each Edition WG and Spin/Lab SIG to come up with search
terms that reflect these goals But honestly, even for collecting the Talking Points, reach each team is really a pain. Nobody answer anything, we normally intersect teams in their meetings and some of them are responsives, others aren't. So, ideally we need to recruit people to help in marketing that works in each WG/SIG. Also, if we are going to extend our topics to promote Fedora in the "target audiences" we need people that can write about each topic, with technical knowledge. I mean, it isn't like just say: "we want to promote Fedora as a great distro for developing in python", we need people that can help with writting about how to setup a python environment, what tools are included in Fedora for that, how is the Python Classroom Lab ready to work with python, that is not in the Workstation edition; and these kind of questions need to be answered for each topic that we can come out.
Conclusions: If we really want to help to promote Fedora for a specific topic, we need to involve people that can really talk about that topic. It won't be to just come up with the idea of make Fedora promoted like the best for do X o Y. I'm really agree with the idea of promote Fedora in different environments, like developers, media producers, different desktops, and all the targets we can come up with, but we need the help of the poeple that is using Fedora in those environments.
Br,
[1] https://pagure.io/fedora-marketing/issue/245 [2] https://pagure.io/fedora-marketing/issue/248 [3] https://fedoramagazine.org/introducing-python-classroom-lab/
On Wed, Sep 20, 2017 at 01:30:32PM -0300, Eduard Lucena wrote:
- Targeted oriented campaigns:
We already have 2 tickets to talk/discuss this [1][2]. We had feedback for some weeks but now both are kind of dead. People looks like just discuss things when Matt write an email, not in the tickets or in the meetings.
Lesson learned: write more email. :)
On Tue, Sep 19, 2017 at 11:52:39PM -0400, Máirín Duffy wrote:
Well, the marketing strategy is intended to be audience-centric, but structured around the editions as a way of organizing the project. That's somewhat different from whether > the website is audience or edition-centric.
"Marketing strategy" sounds more formal and organized than what I am familiar with and aware of. Are there any pointers to specifics on this, outside of the website efforts? I am interested because a formalized strategy would help us make design decisions about collateral / materials we're asked to create on the design team.
Agreed, an overarching, authoritative strategy, written down and maintained, would help a lot for these types of discussions and supporting work.
Search engine position is an easy number to get and compare over time, but is there convincing evidence that it's meaningful? Is it meaningful in either of these senses?:
1 - Good position in rankings will help make $THING more popular 2 - Good position in rankings reflects popularity of $THING
I think #2 is _probably_ true. And #1 is probably true if advertising works at all, which it seems to.
A bunch of data I referenced was snipped wrt specific terms and results I saw; I am wondering if you can corroborate them w the tool you have?
I dont understand how #2 can be true if the #1 ranked dev desktop by far and #2 ranked desktop by quite a lot arent reflected in the results at all. How can it reflect whats going on when 75% of the dev desktop base (win 50 + os x 25) isnt reflected?
Of course advertising works, but SEO isnt advertising. None of this works without data. You need data to drive SEO. You need information about the audience youre targeting to even know what you're optimizing for so it's relevant and not unnaturally forced into your content. Do we have these marketing data resources?
I think cleaning up the metadata on our site and keeping content fresh etc - these are good things. There's a good oppty here to clean things up, make our message clearer, update things and maybe even institute better processes / initiatives for keeping them fresh. However, lacing our site with keywords and stuffing our page titles with buzzwords to vy for the attention of a bot whose algorithm changes daily is not ok. The core function of the website is to communicate a readable message to humans, right? Let's not compromise the consistent and coherent voice we've established....
There are already quite a few searchable bits in the source behind the page. We could of course work with those without affecting the content. But doing so without clearly supporting a coherent strategy probably isn't going to have a huge effect on Fedora's popularity.
I also don't think anyone's saying the website is deficient -- IMHO it's the best it's been in a *long* time. I'm all for improving or reworking the content to be fresh, timely, and high value. It needn't be an artificial function of trying to "game" SEO. Maybe we should start with the authoritative written strategy mentioned above.
On Tue, Sep 19, 2017 at 11:52:39PM -0400, Máirín Duffy wrote:
I think cleaning up the metadata on our site and keeping content fresh etc - these are good things. There's a good oppty here to clean things up, make our message clearer, update things and maybe even institute better processes / initiatives for keeping them fresh. However, lacing our site with keywords and stuffing our page titles with buzzwords to vy for the attention of a bot whose algorithm changes daily is not ok. The core function of the website is to communicate a readable message to humans, right? Let's not compromise the consistent and coherent voice we've established....
The core function of a website doesn't matter if people looking for it can't find it. As a case in point, people looking for "Fedora Media Writer" get https://developers.redhat.com/blog/2016/04/26/fedora-media-writer-the-fastes... as the first hit. That's at least associated with our sponsor and not entirely foreign, but I'd think we'd want https://fedoramagazine.org/make-fedora-usb-stick/ or even https://getfedora.org/workstation/download/ to outscore it.
On 09/26/2017 09:42 AM, Matthew Miller wrote:
On Tue, Sep 19, 2017 at 11:52:39PM -0400, Máirín Duffy wrote:
I think cleaning up the metadata on our site and keeping content fresh etc - these are good things. There's a good oppty here to clean things up, make our message clearer, update things and maybe even institute better processes / initiatives for keeping them fresh. However, lacing our site with keywords and stuffing our page titles with buzzwords to vy for the attention of a bot whose algorithm changes daily is not ok. The core function of the website is to communicate a readable message to humans, right? Let's not compromise the consistent and coherent voice we've established....
The core function of a website doesn't matter if people looking for it can't find it. As a case in point, people looking for "Fedora Media Writer" get https://developers.redhat.com/blog/2016/04/26/fedora-media-writer-the-fastes... as the first hit. That's at least associated with our sponsor and not entirely foreign, but I'd think we'd want https://fedoramagazine.org/make-fedora-usb-stick/ or even https://getfedora.org/workstation/download/ to outscore it.
Sure, like I said, using this as an opportunity to clean up / better the metadata to improve those results is a great idea. I just want to make sure we don't start working keywords unnaturally into the text or gaming for territory that is too generic.
I would think the first step would be to come up with an inventory of those things we want to ensure come up in the results - what are the terms, and what pages do we want to appear prominently for them. Maybe a good approach would be to come up with a task list of things we want people to be able to do with the website(s), then test them and see what users search for / how they approach it, and use that to develop the inventory of keywords to optimize?
~m
On Tue, Sep 26, 2017 at 10:03:56AM -0400, Máirín Duffy wrote:
I would think the first step would be to come up with an inventory of those things we want to ensure come up in the results - what are the terms, and what pages do we want to appear prominently for them.
Yes — see back to my first message in this thread. :)
Maybe a good approach would be to come up with a task list of things we want people to be able to do with the website(s), then test them and see what users search for / how they approach it, and use that to develop the inventory of keywords to optimize?
That sounds awesome.
council-discuss@lists.fedoraproject.org
-
Eduard Lucena -
Matthew Miller -
Máirín Duffy -
Paul W. Frields -
Radka Janekova -
Robert Mayr