Great ideas from all.

I think Alex has definitely gotten a good start, but we should  
definitely consider Owen's points about analytics.

On May 26, 2009, at 9:26 AM, Owen Winkler wrote:

>
> I'm glad that people are interested and passionate about the hp.o  
> site.
>
> My current major concern:  If we do not state what we are trying to
> accomplish with the site, then we have no means to evaluate the  
> success
> of any changes.

That's definitely true.

>
>
> Alex gets off to a decent start here:
>
>> I think the hompage needs to do two main things:
>>
>>    * Show that our software is amazing
>>    * Show that we have an active and welcoming community of users and
>>      developers
>
> But I think that this only really applies to one type of user who  
> would
> visit the site.  There are actually FOUR types of users who visit our
> site.  (I documented this long ago at
> http://wiki.habariproject.org/en/Designs/Habari_Home)

Right, I think we should make all our decisions being informed by  
those personas.

> I certainly don't need to know that our software and community is
> amazing.  I'd nevertheless appreciate direct links to the API docs,  
> the
> wiki, or support channels.  We might say, "yes, of course," to these
> issues, but we do need to remember that our biggest contributor to the
> goal of bringing in new users right now is word of mouth.  Satisfying
> the needs of users who are willing to try and stick with Habari is  
> just
> as important as bringing in new people.

Agreed, but I'm not sure the front page is where that needs to happen.  
I of course value those direct links, but I think we should break down  
the users into 3 categories:
1) First-time visitors/evaluators: for these visitors, we should focus  
on what Habari offers. Features and screenshots should be displayed  
prominently, as well as links to comparison features.
2) Current users: if you are logged in, then the homepage should  
feature information on the latest release, recent activity (across all  
of the Habariverse), support information, and a link to file a bug in  
trac.
3) Developers: if you are logged in and have marked your account as  
"developer" then the homepage includes recent Habari activity, links  
to the wiki & trac, and a "quicksearch" field to look up information  
in the API.

> The above are also hard goals to measure.  Is there a click-through  
> that
> serves as verification that a user agrees our software is amazing, or
> that the community is welcoming?  Hard to say.  "We're awesome!  Click
> here if you agree!"  Some more down-to-earth goals could be
> quantifiable, like, "Turn X% of visitors into registered users."  Or,
> "Have X% of visitors continue from the home page to at least one more
> page."  In analytics terms, these are called, "conversions."  We  
> should
> assemble a nice list of these and discover a way to measure them.

Agreed. I think 3 of the best metrics are:
1) Downloads/home page visits
2) Clickthrough to anywhere else on site
3) Registration

> We need also to keep in mind that our home page (as distasteful as it
> may seem) must include some SEO.  If a goal of the site is to attract
> new users, and they never find our site when searching, we've already
> failed.  One of the primary reasons to have the blog on the home  
> page is
> to keep fresh content available for search engines.  We're currently
> doing a poor job of keeping that content fresh and engaging -- oh,  
> well.

I like Alex's idea of featuring a "stream" of content, including:
1) Trac updates
2) The blog
3) Forum posts
4) Mailing list threads

If we use good, semantic markup (which I'm sure we will) this should  
definitely help to drive traffic.

Also, I think *all* homepage versions should be accessible to search  
engines. Perhaps at /user and /dev respectively, with logged-in users  
being automatically redirected.

> There are some claims being made by folks about what users want to  
> see.
>  I made this Crazyegg heatmap a while back, showing where people
> actually click on our site:
>
> http://farm4.static.flickr.com/3178/2817018229_663dc042c1_o.png

Thanks for putting that together! It's definitely useful in our  
evaluation.

Based on those links, these are some things I think would do well on a  
new page:
1) High quality documentation, including the ability to search it.
2) Enhance the demo with a variety of themes, making it easy to switch  
between different themes. I'd also like to (once again) offer that we  
use the "tour" I developed.
3) Screenshots *must* be on the homepage. The UI is one of the best  
things Habari has going for it, and people are clearly interested in  
finding out more.
4) We should compile a 2 different "why Habari" pages, and link them  
from around the site.
— Users
— Developers

> I'd suggest looking at actual analytics to see how people are using  
> the
> site, but someone insists on loading Woopra onto the site instead of
> something useful, and I'm not sure how to access the Woopra stats.  To
> be clear, I don't want to access the Woopra stats -- I want real,  
> public
> analytics.  We should choose a single, web-based analytics provider,  
> and
> open up access to review the statistics there.  Sadly, none of our
> historical stats are going to be in there.

Agreed. Does GA seem acceptable? We could have a shared profile  
accessible to anyone who wants it. In fact, we could even have  
periodic reports sent to the list automatically (if desired).

> In any case, I'm sure that we can do the site better, but it would be
> folly to take the knowledge that is available and research that has  
> been
> done and discard it in the name of a cool idea.  Whether it's true, it
> seems to me like people are coming at the "hp.o sucks" problem with no
> review of the past work, perhaps simply because they don't like the
> site's current design.

Agreed, but I do think Alex at least is doing a good job of looking at  
past information. The wiki pages and linked threads are definitely  
helpful.

I think the most important part is nailing down the what and the why,  
then focusing on the how (design).

> Nonetheless, I like the idea of portals for the separate user types,
> because it caters to the idea that there is more than one type of
> visitor to the site.  I'm not sure that those should be automatically
> presented based on user preference, but a "developer center" that
> appears after you log in would be useful.

As I briefly referenced earlier, I think we should have 3 homepage  
versions. Anyone can access any of them, but based on your login it  
will direct you to one. They should *all* be linked to from each of  
the homepages.
1) /welcome — basic information, "why habari," etc. default for non- 
users
2) /user – support and community information, default for registered  
users
3) /dev — community and reference, default for registered developers

> On the other hand, I oppose the idea of draggable, personalized,
> configurable interface elements.  Habari is not Netvibes, and I don't
> see that we have that much information to drag around anyway.  Please
> also keep in mind that this site will require maintenance.

Agreed. Seems like bloat and web 2.0 just for the sake of web 2.0.

The only setting you should have as a registered user is default page:  
user or developer. In fact, that wouldn't even be a setting. It would  
simply use the last version you clicked.

Thanks for all the thoughts,
Morgante


--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/habari-users
-~----------~----~----~----~------~----~------~--~---

Reply via email to