Outreach techniques -- what works

[Rob Weir <ro...@apache.org>]

Some quick observations based on recent experience and metrics.  I
think this is important when we consider ways of reaching out for
volunteers.

Facts:

1) Adding something an the www.openoffice.org homepage, as a news
story, gets around 80K hits/day

2) Adding something to the header on all of ooo-site gets 800K hits/day

3) Putting out a blog post on blogs.apache.org gets around 1K hits/day

4) Sending something out by Twitter gets maybe 1K, but it is a one-time thing

5) Sending something out via announcement list reaches 8K users, but
this is also one-time

6) Other mailing lists, like ooo-dev and ooo-users reach a few hundred users


What is effective?  What isn't?  What gets the eyeballs?


Case 1:  Google Moderator

Over we had 1,116 users submit 910 ideas and cast 13,354 votes.  This
was promoted via mailing list, social networking, blog post,
announcement list, but it did not really take off until I linked to it
from the website header (method #2 above).  This massively increased
the number of people participating.


Case 2:  Danish and Polish translators.

I put a brief note on the Danish and Polish homepages, in English,
saying that we would welcome volunteers:
http://www.openoffice.org/da/

This was something so simple, so low tech that I never bothered to do
it before because I was not sure it would be effective.   But then I
noticed that these NL home pages were getting nearly 5K hits/day.
Although this is a much smaller audience, it is a very targeted
audience.

Within 48 hours of putting these notes up we now have multiple
volunteers starting to work on completing the Polish and Danish
translations.  In fact now we need to worry about how we coordinate
multiple volunteers on the same language, a good problem to have.


Case 3: QA volunteers (a negative example)

We've had a lot of good information on helping test AOO, on the wiki,
automation code checked into SVN, test procedures, test reports, etc.
All of this is happening in the open on ooo-dev and ooo-qa.  But I
don't think we have a really attracted any more test volunteers.
Why?

Maybe this is because we have only asked on our lists, which have
relatively few subscribers -- a few hundred -- compared to the how
many we can reach out to via other means.


So based on what I've seen, in this example and others, I'd recommend
thinking like this:


1) Are we looking for a broad or targeted outreach?

2) If narrow, look for targeting specific pages on the website that
will be seen by those users

3) If broad, consider something on the home page or the header on
every page, like we're doing now with ApacheCon.

4) Even though a blog post gets less traffic, it still might make
sense to start there.  It gives you something that you can then think
to from other places, as well as a way to engage with the reader via
comments.

Re: Outreach techniques -- what works

[Rob Weir <ra...@gmail.com>]

On Oct 11, 2012, at 9:48 PM, Kevin Grignon <ke...@gmail.com> wrote:

> On Tue, Oct 9, 2012 at 4:22 AM, Juergen Schmidt <jo...@gmail.com>wrote:
>
>>
>> Am Montag, 8. Oktober 2012 um 19:03 schrieb Rob Weir:
>>> Some quick observations based on recent experience and metrics. I
>>> think this is important when we consider ways of reaching out for
>>> volunteers.
>>>
>>> Facts:
>>>
>>> 1) Adding something an the www.openoffice.org homepage, as a news
>>> story, gets around 80K hits/day
>>>
>>> 2) Adding something to the header on all of ooo-site gets 800K hits/day
>>>
>>> 3) Putting out a blog post on blogs.apache.org gets around 1K hits/day
>>>
>>> 4) Sending something out by Twitter gets maybe 1K, but it is a one-time
>> thing
>>>
>>> 5) Sending something out via announcement list reaches 8K users, but
>>> this is also one-time
>>>
>>> 6) Other mailing lists, like ooo-dev and ooo-users reach a few hundred
>> users
>>>
>>>
>>> What is effective? What isn't? What gets the eyeballs?
>>>
>>>
>>> Case 1: Google Moderator
>>>
>>> Over we had 1,116 users submit 910 ideas and cast 13,354 votes. This
>>> was promoted via mailing list, social networking, blog post,
>>> announcement list, but it did not really take off until I linked to it
>>> from the website header (method #2 above). This massively increased
>>> the number of people participating.
>>>
>>>
>>> Case 2: Danish and Polish translators.
>>>
>>> I put a brief note on the Danish and Polish homepages, in English,
>>> saying that we would welcome volunteers:
>>> http://www.openoffice.org/da/
>>>
>>> This was something so simple, so low tech that I never bothered to do
>>> it before because I was not sure it would be effective. But then I
>>> noticed that these NL home pages were getting nearly 5K hits/day.
>>> Although this is a much smaller audience, it is a very targeted
>>> audience.
>>>
>>> Within 48 hours of putting these notes up we now have multiple
>>> volunteers starting to work on completing the Polish and Danish
>>> translations. In fact now we need to worry about how we coordinate
>>> multiple volunteers on the same language, a good problem to have.
>>>
>>>
>>> Case 3: QA volunteers (a negative example)
>>>
>>> We've had a lot of good information on helping test AOO, on the wiki,
>>> automation code checked into SVN, test procedures, test reports, etc.
>>> All of this is happening in the open on ooo-dev and ooo-qa. But I
>>> don't think we have a really attracted any more test volunteers.
>>> Why?
>>>
>>> Maybe this is because we have only asked on our lists, which have
>>> relatively few subscribers -- a few hundred -- compared to the how
>>> many we can reach out to via other means.
>>>
>>>
>>> So based on what I've seen, in this example and others, I'd recommend
>>> thinking like this:
>>>
>>>
>>> 1) Are we looking for a broad or targeted outreach?
>>>
>>> 2) If narrow, look for targeting specific pages on the website that
>>> will be seen by those users
>>>
>>> 3) If broad, consider something on the home page or the header on
>>> every page, like we're doing now with ApacheCon.
>>>
>>> 4) Even though a blog post gets less traffic, it still might make
>>> sense to start there. It gives you something that you can then think
>>> to from other places, as well as a way to engage with the reader via
>>> comments.
>>
>> thanks for this very interesting and useful observations and hints how to
>> reach a broader audience.
>>
>> Juergen
>
>
>
> Can we set up the header content to cycle through important notices. With
> so many daily hits we should optimize this opportunity.
>

I like this idea. It would require some changes but one way would be
to have the header markup be static rather than a SSI, but have that
static markup reference a JSON object in an external JS file that
contains the various messages, URL's and weights. Could even be
locale-specific.


> Perhaps we could assign priorities to the messages so core messages have
> high visibility, while still leaving some impressions for other messages,
> such as recruiting, surveys, etc.
>
> Thoughts?
>
> Regards,
> Kevin

Re: Outreach techniques -- what works

[Kevin Grignon <ke...@gmail.com>]

On Tue, Oct 9, 2012 at 4:22 AM, Juergen Schmidt <jo...@gmail.com>wrote:

>
> Am Montag, 8. Oktober 2012 um 19:03 schrieb Rob Weir:
> > Some quick observations based on recent experience and metrics. I
> > think this is important when we consider ways of reaching out for
> > volunteers.
> >
> > Facts:
> >
> > 1) Adding something an the www.openoffice.org homepage, as a news
> > story, gets around 80K hits/day
> >
> > 2) Adding something to the header on all of ooo-site gets 800K hits/day
> >
> > 3) Putting out a blog post on blogs.apache.org gets around 1K hits/day
> >
> > 4) Sending something out by Twitter gets maybe 1K, but it is a one-time
> thing
> >
> > 5) Sending something out via announcement list reaches 8K users, but
> > this is also one-time
> >
> > 6) Other mailing lists, like ooo-dev and ooo-users reach a few hundred
> users
> >
> >
> > What is effective? What isn't? What gets the eyeballs?
> >
> >
> > Case 1: Google Moderator
> >
> > Over we had 1,116 users submit 910 ideas and cast 13,354 votes. This
> > was promoted via mailing list, social networking, blog post,
> > announcement list, but it did not really take off until I linked to it
> > from the website header (method #2 above). This massively increased
> > the number of people participating.
> >
> >
> > Case 2: Danish and Polish translators.
> >
> > I put a brief note on the Danish and Polish homepages, in English,
> > saying that we would welcome volunteers:
> > http://www.openoffice.org/da/
> >
> > This was something so simple, so low tech that I never bothered to do
> > it before because I was not sure it would be effective. But then I
> > noticed that these NL home pages were getting nearly 5K hits/day.
> > Although this is a much smaller audience, it is a very targeted
> > audience.
> >
> > Within 48 hours of putting these notes up we now have multiple
> > volunteers starting to work on completing the Polish and Danish
> > translations. In fact now we need to worry about how we coordinate
> > multiple volunteers on the same language, a good problem to have.
> >
> >
> > Case 3: QA volunteers (a negative example)
> >
> > We've had a lot of good information on helping test AOO, on the wiki,
> > automation code checked into SVN, test procedures, test reports, etc.
> > All of this is happening in the open on ooo-dev and ooo-qa. But I
> > don't think we have a really attracted any more test volunteers.
> > Why?
> >
> > Maybe this is because we have only asked on our lists, which have
> > relatively few subscribers -- a few hundred -- compared to the how
> > many we can reach out to via other means.
> >
> >
> > So based on what I've seen, in this example and others, I'd recommend
> > thinking like this:
> >
> >
> > 1) Are we looking for a broad or targeted outreach?
> >
> > 2) If narrow, look for targeting specific pages on the website that
> > will be seen by those users
> >
> > 3) If broad, consider something on the home page or the header on
> > every page, like we're doing now with ApacheCon.
> >
> > 4) Even though a blog post gets less traffic, it still might make
> > sense to start there. It gives you something that you can then think
> > to from other places, as well as a way to engage with the reader via
> > comments.
> >
> >
>
> thanks for this very interesting and useful observations and hints how to
> reach a broader audience.
>
> Juergen



Can we set up the header content to cycle through important notices. With
so many daily hits we should optimize this opportunity.

Perhaps we could assign priorities to the messages so core messages have
high visibility, while still leaving some impressions for other messages,
such as recruiting, surveys, etc.

Thoughts?

Regards,
Kevin

Re: Outreach techniques -- what works

[Juergen Schmidt <jo...@gmail.com>]

Am Montag, 8. Oktober 2012 um 19:03 schrieb Rob Weir: 
> Some quick observations based on recent experience and metrics. I
> think this is important when we consider ways of reaching out for
> volunteers.
> 
> Facts:
> 
> 1) Adding something an the www.openoffice.org homepage, as a news
> story, gets around 80K hits/day
> 
> 2) Adding something to the header on all of ooo-site gets 800K hits/day
> 
> 3) Putting out a blog post on blogs.apache.org gets around 1K hits/day
> 
> 4) Sending something out by Twitter gets maybe 1K, but it is a one-time thing
> 
> 5) Sending something out via announcement list reaches 8K users, but
> this is also one-time
> 
> 6) Other mailing lists, like ooo-dev and ooo-users reach a few hundred users
> 
> 
> What is effective? What isn't? What gets the eyeballs?
> 
> 
> Case 1: Google Moderator
> 
> Over we had 1,116 users submit 910 ideas and cast 13,354 votes. This
> was promoted via mailing list, social networking, blog post,
> announcement list, but it did not really take off until I linked to it
> from the website header (method #2 above). This massively increased
> the number of people participating.
> 
> 
> Case 2: Danish and Polish translators.
> 
> I put a brief note on the Danish and Polish homepages, in English,
> saying that we would welcome volunteers:
> http://www.openoffice.org/da/
> 
> This was something so simple, so low tech that I never bothered to do
> it before because I was not sure it would be effective. But then I
> noticed that these NL home pages were getting nearly 5K hits/day.
> Although this is a much smaller audience, it is a very targeted
> audience.
> 
> Within 48 hours of putting these notes up we now have multiple
> volunteers starting to work on completing the Polish and Danish
> translations. In fact now we need to worry about how we coordinate
> multiple volunteers on the same language, a good problem to have.
> 
> 
> Case 3: QA volunteers (a negative example)
> 
> We've had a lot of good information on helping test AOO, on the wiki,
> automation code checked into SVN, test procedures, test reports, etc.
> All of this is happening in the open on ooo-dev and ooo-qa. But I
> don't think we have a really attracted any more test volunteers.
> Why?
> 
> Maybe this is because we have only asked on our lists, which have
> relatively few subscribers -- a few hundred -- compared to the how
> many we can reach out to via other means.
> 
> 
> So based on what I've seen, in this example and others, I'd recommend
> thinking like this:
> 
> 
> 1) Are we looking for a broad or targeted outreach?
> 
> 2) If narrow, look for targeting specific pages on the website that
> will be seen by those users
> 
> 3) If broad, consider something on the home page or the header on
> every page, like we're doing now with ApacheCon.
> 
> 4) Even though a blog post gets less traffic, it still might make
> sense to start there. It gives you something that you can then think
> to from other places, as well as a way to engage with the reader via
> comments.
> 
> 

thanks for this very interesting and useful observations and hints how to reach a broader audience.  

Juergen