Blogs Web Design

Step 1: Pick a community designator

I've been using the community name followed by a dot. The prefix goes in front of category description to make a set of unique tags for the aggregator. When someone wants to add something for the site they tag it with this "community.category" tag. In my example, when someone wants to list a new Quaker blog they use "quaker.blog", "quaker" being the community name, "blog" being the category name for the "New Blogs" page.

Step 2: Collect the community prefix and category name in Pipes

You begin by going into Pipes and pulling over two text inputs: one for the community prefix, the other for the specific category.

Step 3: Construct these into tags

Now use the "String Concatenation" module to turn this into the "community.category" model. The community input goes into the top slot, a dot is the second slot and the category input goes into the last slot.

Now, when you have a tag in Flickr with a dot in it, Flickr automatically removes it in the resultant RSS feed. So with Flickr you want your tag to be "communitycategory" without a dot. Simple enough: just pull another "String Concatenation" module onto your Pipes work space. It should look the same except that it won't have the middle slot with the dot.

Step 4: Turn these tags into RSS URLs

Pull three "URLBuilder" modules into Pipes, one for each of the services we're going to query. For the Base, use the non-tag specific part of the URL that each service uses for its RSS feeds. Here they are:

Under path elements, put the correct tag: for Del.icio.us and Google it should be the community.category tag, for Flickr the dot-less communitycategory tag.

Step 5: Fetch and Dedupe

Fetch is the Pipes module that pulls in URLs and outputs RSS feeds. It can also combine them. Send each URLBuilder output into the same Fetch routine.

Since it's possible that you'll might have duplicate posts, use the "Unique" module to deduplicate entries by URL. Through a little trial and error I've determined that in cases of duplicates, feeds lower in the Fetch list trump those higher. In the actual Pipe powering my aggregator I pull a second Del.icio.us feed: my own. I have that as the last entry in the Fetch list so that I can personally override every other input.

Step 6: Sort by Date

With experimentation it seems like Pipes orders the output entries by descending date, which is probably what you want. But I want to show how Pipes can work with "dc" data, the "Dublin Core" model that allows you to extend standard RSS feeds (see yesterday's post for more on this).

Google Blog Search and Del.icio.us feeds use the "dc:date" field to record the time when the post was made. Flickr uses "dc:date.Taken" to pass on the photograph's metadata about when it was taken. Pipes' "Rename" module lets you copy both fields into one you create (I've simply used "date"), which you can then run through its "Sort" module. Again, it's a moot point since Pipes seems to do this automatically. But it's good to know how to manipulate and rename "dc" data if only because many PHP parsers have trouble laying it out on a webpage.

Update: it's all moot: according to a ZDNet blog, "Pipes now automatically appends a pubDate tag to any RSS feed that has any of the other allowable date tags." This is nice: no need to hack the date every time you want to make a Pipe!

Step 7: Output

The final step for any Pipe is the "Pipe Output" module.

In action

You can see this published Pipe here, and copy and play with it yourself. The result lets you build an RSS feed based on the two inputs.

Don't assume that your website is easy to navigate. One of the neatest things about the web is that we have instant feedback on use. With just a little tracking we can see what pages people are looking at, how they're finding our site and what they're doing once they're here.

Javascript Trackers:

My most advanced sites are currently using four different tracking methods. Most utilize javascript "bugs," tiny snippets of code that send individual results to an advanced software tracking system. I put the code inside a Moveable Type "Modules Template" which is automatically imported to all pages. Installing a new system is as easy as cutting-and-pasting the javascript into the Template and rebuilding the site.

AXS Visitors Tracking System
This software installs on your server but don't let that scare you: this is one of the easiest installations I've ever seen. AXS gives you great charts of usage: you can narrow it specific pages on your site, or even particular search engines or search phrases.

There's also a option to view the lastest traffic by visitor. I love watching this! You can see how individuals are using the site and where they're navigating. I've been able to identify different types of visitors this way and understand the complexity of the audience.

It doesn't seem like AXS is not being developed anymore. The latest stable version came out over two years go, which is a shame.

HitTail
This service watches search-engine links and makes recommendations for new keywords. I wrote about this service yesterday in Blogging for the Long Tail.

Reeferss.com
This is a simple simple bit of software. Like every other tracking system it keeps track of referrers: search engines and websites that bring traffic to your site. But unlike the others that's all it does. Why care then? It provides a real-time RSS feed of these visitors. I bring the feed into my "Netvibes" page (a customized start page, see below) and scan the results multiple times a day.

Google Analytics
The internet's gatekeeper bought the Urchin analytics company in April 2005 and relaunched the product as Google Analytics shortly thereafter. This is becoming an essential tracker. It's free and it's powerful, though I haven't been as impressed by it as others have. See its Wiki page for more.

Internet Trackers:

It's easy to find out what people are saying about you online.

Technorati
This service tracks blogs but you don't need to have a blog to use it, for Technorati will tell you where blogs are linking. Give it your URLs (or those of your competitors!) and you'll know whenever a blogger puts in a link to you. You can also give it keywords and find out when a blog uses them.

Google Blog Search
Google can also let you follow blog references or keyword mentions on the blogs. Google will also track beyond blogs of course. Type "site:www.yourdomain.com" into the main Google search page and you'll see who's linking to your site (or to the competition). There are lots of other services that track blogs and mentions--Sphere, Bloglines, etc. They all have different strengths so try them and see what you think.

Feedburner
The best RSS massager has always focused on ways to track your RSS feed. They've recently introduced page tracking software too. It looks great but I just installed it this week. I still have to see if it's as good as Feedburner's other offerings.

Keeping on top of this flow of data:

It's easy to get overwhelmed by all of this information. Most of the tracking services provide RSS feeds (See The Wonders of RSS Feeds for an intro). I use Netvibes, a customized start page, to pull these all together into a single page that I can scan every morning. Here's a screenshot of part of my Netvibes tracking page--the full page currently shows fourteen tracking feeds on one screen:

So why is tracking important to SEO?

With tracking you find out what people are looking for on the internet. This helps you create pages and services that people will want to find. You might be surprised to see what they're already finding on your site. Some examples:

• Analyzing one site, I noticed that few pages I thought were obscure were bringing in high Google traffic. I looked at these pages again and realized they did a good job of describing the company's mission. I consequently redesigned the site homepage to feature them and I made sure that those pages contained direct links to its most important services.

• When I started work for another client I looked at their site and suspected that they're most important articles were not being seen--visitors had to click through about four times to get to them. Six months of tracking confirmed my hunch and gave me the hard data to convince the executive director that we made some small modifications to the design. Having this strong content linked right off the homepage helped bring in Google traffic.

About eighteen months ago I started a "links blog" of interesting Quaker links, incorporated as a sidebar on my popular "QuakerRanter" personal blog. I eventually gave the links their own URL (QuakerQuaker.org) and invited others to join the linking. I always stumble when trying to tell people what QuakerQuaker is all about. The best definition is that its a "collaboratively edited blog aggregator" but that's a horribly tech description.

The rise of blogs is creating the necessity for these sort of theme-based aggregators. This morning I stumbled on Original Signal, a new site that organzes the best Web 2.0 blogs. A site called PopURLs does the same for "the latest web buzz." A site called SolutionWatch has written about these in Tracking the web with Single Page Aggregators. We're all on to something here. I suspect that sometime this fall some clever person will coin a new term for these sites.

It's startling to realize that my three year toddler is almost the same age as Myspace and older than Facebook. In just a few short years they've come to dominate much of the online world, especially with under-25 users. The kind of independent blogs that dominate a sites like Livejournal and Blogspot don't have the web of cross-connections--what I called the "folksonomic density"--of the new social networking sites. It seems appropriate that Myspace was founded by spammers: who knows more about sucking people in?

The question: will the net have room for independent niche sites? Myspace is changing its architecture to disable key linking features of third-party embedded plug-ins like the from the popular video site Youtube. The big search sites also want a piece of this market--new features on Yahoo local and the geotagged maps on Yahoo's Flickr are impressive). It all reminds me some of the debates about local food co-ops versus enlightened supermarkets: is it a good thing that organic produce and soymilk can be purchased at the local Acme, even if that cuts into the independent co-op's business? Don't we want everyone to have access to everything? In the end, philosophy won't settle this argument.

Here's some of what Johnson has to say:

Distribution, content creation, and the ability to more easily compete with established local players online... blogging is perfect for that. I mean a blog is chronologically arranged, in columns, divided by categories and changes (in many cases) everyday. That's the broad definition of a newspaper, right? A blog is so much more than that, but the basic structure lends itself very well to developing an online competitor for newspapers.

It was three years ago that I followed Brad Choate's instructions for using Moveable Type as a whole-site content management system. What started as an experiment became a way of life for me. The MT interface lends itself so well to content management that I'm now using it for my non-techie clients: Quakersong.org and Quakeryouth.org are both put together by MT and I've been surprised that there's been almost no learning curve for the client's adoption of this software.

Given this, it seems odd that the kids at Moveable Type haven't taken MT in this direction (even more surprising since they hired Brad himself a few years ago!). I see a big market in my niche sites for this sort of functionality and three years later I'm still having to tweak templates to get this to work. Anil, what's up? If Drupal had better documentation and smoother installation it would have been the brawn behind MartinKelley.com.

It would be fun to follow Until Monday's example and create a hyperlocal site (hint hint to VW if she's reading this). Of course, locality is not just geographically-based anymore. Quakerquaker.org is a local portal of a different kind. I'm a big believer that the hyperlocality of niche and geographic sites are the cutting edge in the next-wave of the social web.

There's a lot of pioneering to be done in this regards. The net has a lot of power to take down culture monopolies by confronting old boy networks and business-as-usual thinking with innovative social networks that harness the talents of the outsiders. The smart newspapers, magazines, churches and cultural organizations will come on board and leap-frog themselves to twenty-first century relevance. Too many of the Philadelphia (and/or) Quaker institutions I know respond to change by shuffling job titles and putting blinders up against recognizing the ever-narrower demographic they serve.

An important part of the redesign was an automatic keyword generator. Posts were run through a script that automatically pulled out keywords from the text. My 2003 article, Going all the way with Movable Type generated the following tags, which appear as links after the post:

• choate
• nonviolence
• curly quotes
• personal blog
• php
• org
• moveabletype
• movable type
• hmtl
• dashes
• textile
• garbage
• blogs

Following the links takes you to similarly-tagged articles. At least that's the conceit. When you follow a tag's link you're simply doing a site search for that keyword. A little htaccess rewrite magic is making the result look like it's a static category page.

"Fine and well" you're thinking, "big deal." Well, here's what's cool. There are 225 entries on the QuakerRanter blog. Google's just gone through and indexed the site and is now claiming it contains 1300 pages. Each tag is being indexed as its own page. Every time I mention any interesting term, it becomes a page that Google indexes and delivers to its searchers.

Which brings us to today's cool piece from the access logs. In December of 2004 a rather innocent post on Quaker Ranter became the center of a mini-whirlwind on the political blogs when it mentioned that I had gotten a call from a CBS News publicist interested in Nonviolence.org. All political blogs get publicity calls from news and opinion think tanks trying to suggest (or plant) stories but no one's supposed to talk about it. I only mentioned it because it was so unusual. One of the blogs denouncing the liberal conspiracy my post revealed was the somewhat slimy Little Green Footballs. After a few weeks the denunciations died down.

But this morning, someone looked up littlegreenfootballs in Google and came to my site. Because of my automatic keyword generator, tags, and static-loooking links, I'm now the number two entry, on two three-year old posts, now relocated to a days old quakerranter.org. Cool.

This mixing and matching of content and rich manipulation of data is sometimes lumped together in the cool bu zzphrase folksonomy. Note that none of what I've done is a tricking of Google. Every tag is really going to a page with that content. These are "natural" and "organic" search results in the lingo of SEO. I'm just presenting my information in multiple formats that appeal that the widest array of audiences.

For what it's worth, I don't think I deserve #2 status for "littlegreenfootballs" and I don't think Google will keep it there for long. It's a bit odd that they have elevated that particular term so high and no others tags seem so stratospheric.

Positive Results:

As of February 2007, Google indexes 3,540 pages on QuakerRanter.org, a blog of only 239 posts. In December 2006 30% of my Google visits were to one of the "tags" page. Reconfiguring the blog in this kind of tag-intensive way has more than doubled search engines visits, again in a very natural and organic way. Adding tags has simply made what I've written more accessible to search engines. Very cool.

Negative Ramifications:

Shortly after installing this new system, my servers started periodically crashing (about once/week). The problem would be multiple MT-Search processes overloading the memory.

My guess is that a search engine spider came along and started indexing all of the tags. Each link initiated a search query in Movable Type. The built-in search for Movable Type is just not able to handle this volume of traffic.

I installed Fast Search to solve the problem (tip of the hat to Al-Muhajabah). It took awhile: Fast Search required a MySQL upgrade at my host. After that I needed to install these plugin fixes. Then it was fine-tuning the htaccess files. It was been more work than I initially expected and the tag results now forward to a funny URL that Google doesn't love as much.

Inspired by Doing Your Whole Site with MT on Brad Choate's site, I started experimenting today with putting the whole Nonviolence.org site into Movable Type. At first I thought it was just a trial experiment but I'm hooked. I especially love how much cleaner the entry for the links page now looks and I might actually be inspired to keep it up to date more now. (I've also integrated Choate's MT-Textile which makes a big difference in keeping entries clean of HMTL garbage, and the semi-related SmartyPants which makes the site more typographically elegant with easy M-dashes and curly quotes).

So here's what I'm doing: there are three Movable Type blogs interacting with one another (not including this personal blog):

• One is the more-or-less standard one that is powering the main homepage blog of Nonviolence.org.
• The second I call "NV:Static" which holds my static pages, much as Brad outlines. I put my desired URL path into the Title field (i.e., "info/index") and then put the page's real title into the Keywords field (i.e., "About Nonviolence.org") and have that give the data for the title field and the first headline of the page. It might seem backwards to use Title for URL and then use Keywords for Title, but this means that when I'm in MT looking to edit a particular file, it will be the URL paths that are listed.
• The third blog is my "NV:Design Elements." This contains the block of graphics on the top and left of every page. I know I'll have to redesign this all soon and I can do it from wherever. This blog outputs to HTML. All the other pages on the site are PHP and its a simple include to pull the top and left bars into each PHP page.

Oh yes, I'm also thinking of incorporating guest blogs in the near future and all of these elements should make that much easier.

Here's another site to check out, about how someone integrated Movable Type into their church website using some interesting techniques.

State of the Websites

Report for FGC Central Committee, October 2004
By Martin Kelley, webmaster

It's important to start off with a little editorial about why we need reports like this. We put up a website and we know people use it. Why bother spending time collecting data?

The internet is simultaneously vague and precise. We can say definitively that the FGC website received 114,097 "unique visitors" in the past fiscal year. But how many people does that represent? Is that a high number or low number? How did these users react when they came to the site. Did they think to themselves "whoops, not what I want" and leave, or did they go "wow, what's this FGC?, hey this is great." LESSON: We need data to know if the site is being used well.

Everyone who reads this report is by definition an insider. None of us are able to step into the shoes of an unknowledgeable seeker. In my study of usage patterns, I have found that the differences in website use between Quaker insiders and seekers is so great that they might as well be looking at different websites, if not different media altogether (see How Insiders and Seekers Use the Quaker Net. Because of this gap we cannot design the site based on whims or personal preferences. It is incredibly difficult to imagine how newcomers might navigate the site. We can only consider the design of the site after we've examined in usage, both in detail (actual users moving through the site) and in aggregate (pages and links visited over periods of time). See also: How to measure the peace movement. LESSON: We can only effectively design the site if we incorporate sophisticated and detailed data about how the site is being used.

Part 2, Googlization

By far the most significant change in our websites over the past year has been the "googlization" of Quakerbooks and Quakerfinder, both of which now have over four times the visitors they were getting last year.

The Google Problem: Both Quakerbooks and Quakerfinder have had great content from their start. The former lists the entire inventory of FGC's bookstore, along with book descriptions and reader commentary. The latter has our list of meetings--addresses, worship times, and contact information. But on both sites the bulk of the content was locked up in databases. Before users could benefit from the sites, they had to find them. This limited much of the use to people who already know about FGC and our resources. Because internet search engines can't search website databases (a problem known as the hidden or deep web), they could index only a limited number of pages on these sites and they made referrals on only the most generic search phrases (e.g., "quaker bookstore" "quaker meeting directory").

We made various changes to both sites (technical details below) that have made them searchable by Google and the other search engines, which now return our sites for very specific search queries, e.g., "Quakers in conflict Ingle" and "Quakers Poughkeepsie".

A Wider, More Inclusive Audience: What's great is that this has given us not just a bigger audience, but our target audience. Most of these visitors don't know enough about how Friends are organized to even know where to look for information. With Quakerfinder and Quakerbooks, we're now be visible on their terms.

We're giving them the basic information they're seeking and we're doing it when they are actively seeking it. This last point is important. I spend a lot of time watching how people use websites. If you email someone out of the blue with a link to a website, they might follow it but only half-heartedly. They might be doing five other things at the same time and they rarely stay to full use the website's resources. When someone comes to a site via a search engine they're much more likely to look around: this is the visit that they are initiating because they have something specific they're trying to find.

Having a "googlified" Quakerfinder means we're actually reaching people who are ready to try out a Quaker meeting and we're giving them that most basic information that's often hard to find. With a searchable Quakerbooks we're selling books to people who might not even have thought about Quakers as a possible spiritual path. I suspect that both sites are doing more outreach about Quakerism than any of us expect.

Update, 11/29/04: I recently met someone who came to Friends after reading the Quaker entry in Wikipedia. He had gone through the list of religious denominations in the U.S. till he found one that spoke to his condition. In the past month FGC has gotten 57 visitors from Wikipedia.

The Fixes

In the official committee report I tried to steer clear of too many technical details since I wanted people to read it. So I'll expand on them here on the website version.

Unique Domains: I don't think it really helped to give Quakerfinder.org and Quakerbooks.org their own domains, at least initially. In last year's report I noted that most of the traffic to those sites came from the main FGCQuaker.org site and that the separate domains weren't particularly useful. Now the sites do have their own sort of identity, thanks to the "googlization," which was a different process for the two sites.

Quakerbooks.org: Visitors to the Quakerbooks.org site are given session IDs to allow us to follow along with them as they make their selections. Since some users don't allow cookies, this ID sometimes appears in the URL (it appears as something like "?sessionid=1514" appended to the end of the address). Google really hates session IDs because its automated software doesn't know if the different URLs are different pages (to be indexed separately) or merely different sessions looking at the same page. So Googles just ignores anything that looks like this. The easiest fix is to have the software look to see if the visitor is Google and take of the session IDs (Google is okay with this workaround; I also used this method to allow them to index my Nonviolence.org discussion board.)

Quakerfinder: On Quakerfinder.org, the problem was that visitors had to type in a zip code to get to any of the content. Google's not that interactive and only follows links. Until recently, it thought there was only three pages to the site. To fix this we set up an alternative way to navigate the site: from the homepage you can now follow a link to lists of Quaker Meetings by state. The zip code lookup is so much more convenient that we don't suspect many live people will look up by state, but Google will and because of this it now lists 808 pages on the site. Now Google acts as a alternate lookup service, one that doesn't depend on people finding our site beforehand.

Part 3, Comparing the Sites

Visitors

The basic measure used to measure website traffic is that of the "unique visitor," which counts user sessions. Here are this year's comparisons to last year's. Numbers represent the monthly average "unique visitors" to each of our three websites.

     Site	      FY 03/04 total  FY 02/03 total  Increase
     FGCQuaker.org    114,097         82,747           38%
     Quakerfinder.org  48,084         23,964          100%
     Quakerbooks.org   69,924         19,332          262%

The last two sites have truly remarkable jumps. The numbers are a little misleading, however, as the increase in traffic hasn't been gradual but sudden and climbing. Compare the last full month (September 2004) with the same month the previous year and all three sites have higher jumps.

     Site             Sept 04         Sept 03         Increase
     FGCQuaker.org    9459            8254             15%
     Quakerfinder.org 8782            1997            340%
     Quakerbooks.org  7498            1611            366%

While the internet grows in use every year, the increases on Quakerfinder and Quakerbooks represent a quantum leap over that incremental increase. They represent "search engine optimization" of those sites, or what we all refer to the "googlization" of the sites.

Links:

One way of measuring the visibility of a website is to count how many other webpages link to it. Here are

     Site              October 2004    October 2003    Increase
     FGCQuaker.org     496             396              25%
     Quakerfinder.org  196              46             326%
     Quakerbooks.org   151              96              57%

For comparison: Quaker.org is up to 11,900 links, Phila. Yearly Meeting is 248, PendleHill.org is 420, FCNL.org is 10,200, Nonviolence.org is 20,900 and AFSC.org is 21,800. See Miscellaneous & Notes at end to see how numbers were obtained. See How Can We Measure the State of the Peace Movement? for more on this method of measurement.

Part 4, The FGCQuaker.org Site

Visitors

Use of FGCQuaker.org continues to grow at a good clip. We have a 38% increase this fiscal year compared with last's. The site received over 114,000 unique visitors from October 1, 2003 to September 30, 2004.

To the right is the chart showing unique visitors by month for the past three years:

Referrers: Where did visitors come from?

In September 2004, there were 9459 "unique visits" to the FGCQuaker.org site, still our most-visited site. Here's where they came from.

1021 from Quakerfinder.org. One surprise this year is the jump in Quakerfinder-referred visits. This is due of course to the phenomenal visibility of that site. In a recent one-month period, FGCQuaker received 983 visits from Quakerfinder links, two-thirds of which came from the "googlized" Quakerfinder pages. About one in ten visitors are now coming to FGCQuaker through Quakerfinder. Up 288% from last year.

842 from Google. We get a lot of Google traffic because we have a lot of content on our site: dozens of pamphlets, years worth of FGConnections, large parts of the old Fostering Vital Friends Meetings resource binder. Visitors via search engines often don't know FGC exists but they want to know about our programs and work. Because FGC does such great work (and because we publicize it online!), many of our resources answer questions people have. I think this is great outreach.

Here's an example. This Spring I noticed that we were getting visits on fairly generic searches for racism. Here's a list of search inquiries that brought people to the CMR pages on FGC:

"ending racism"
"racially diverse communities"
"quaker racial diversity"
"diversity in friends"
"ethnic diversity"
"responsibilities to racism"
"pastoral care racism"
"activities for ending racism"
"testimonies racial unity"

This is a fascinating list precisely because these are generic searches. People aren't looking for "Quakers ending racism," they're looking for anyone "ending racism" and Google is bringing them to us (we're number 6 on that search term). This is surprising: I would think the much bigger denominations would all have committees ending racism that would come up higher just because of their larger institutional clout. That we are so high suggests that this work is not as common as I we might hope and that Friends might have the opportunity to play a role in larger faith dialogues.

When people use search engines, they get results from all over the FGC website. Searches might pull up some four-year article on FGConnections, or one of the "Friends And..." pamphlets that we've put online. Google up 12% from last year. There were about 83 more visits from regional Google sites.

434 from Quaker.org. Most of these people are coming directly from the Quaker.org homepage to the FGCQuaker.org homepage. I estimate that about 60% of these visitors leave the FGC site without clicking on any links. They're probably just superficially curious about us, but not enough to look around the site. Up 39% from last year.

253 from other search engines: 118 from Yahoo (118), MSN (74), AOL (42), Ask (19).

81 from Beliefnet. Beliefnet has a popular "Belief-o-Matic" quiz that will magically tell you what religious faith you should join. It's rigged in such a way that a lot of people unexpectedly come up as Quaker. The qui zthen directs people to an information page on Friends, which includes some links to FGC. Most of the Beliefnet visitors are coming from that information page directly to the FGC homepage. Up 200% from last year.

69 from UVa's Religious Movements site. This is a pretty good description of Quakerism

60 from Quakerbooks. Our own bookstore website attracts a lot of new people who aren't part of the established Quaker networks and many of them first learn of FGC this way.

53 from Religious Tolerance. A popular website from a Canadian Unitarian that profiles religions..

52 from QuakerInfo.org. This is the Philadelphia Quaker Information Center, a joint project of a number of Quaker organizations, including FGC.

Where did people go?

Top Destinations in September 04:
* To the homepage: 2396;
* Library's "Welcome to Quakerism" pages: 463;
* A&O "Resources for Meetings": 320 (prominently linked from Quakerfinder);
* Gathering pages: 309;
* "Silent Worship Quaker Values" tract on the Library section;
* Gathering's pictures from last year: 149;
* Religious Ed: 149;
* FGConnections articles: 129;
* Ideas for First Day School": 127;
* Advancement & Outreach homepage: 124;
* Young Quakes: 118;
* Publications: 100;
* Development 97.

These are pretty typical numbers. The only significant variation over the year comes in Spring, when traffic to the Gathering pages goes up. In May 2004, 961 people visited the Gathering homepage, and 355 visited the workshop listings.

Forget the Aggregates: How Do People Use the Site?

So far I've looked at tallied-up numbers: how many people visited, how many pages were looked at. The problem with this sort of statistic is that it doesn't give us a feel for how individuals are actually using the site. Looking at usage explodes the preconceptions that many of us "Insider Quakers" might bring to the web.

The first lesson: most people don't come into our site via the FGC homepage. Even more shocking: close to half never even see the homepage! This blew me away when I first realized it. We spend so much time designing the homepage and wondering how we're going to direct seekers from it but a lot of this work is in vain.

Of that 45% or so that enter the site via the FGC homepage, most of them leave the site immediately without following any link whatsoever.

Let's splice this another way: 70% of the people who hit our site (wherever they enter) don't look at any page other than that first one. They don't click on anything but the back button.

What are some of the lessons on this: one is that content is all important. Those majority of visitors who bypass the homepage to parachute directly inside the site are coming for specific information. Many of them don't know anything about FGC and most of them don't care to learn about FGC the organization. They're looking for some specific piece of information on Quakers ("painting of Pennsylvania Abolitionist Society Quakers" and "Quakers prison reform"), or on religious education in general ("religious meeting"), or on how churches are dealing with racism ("racial diversity" and "do blacks worship with only blacks"). These are all search phrases that have brought visitors to FGCQuaker.org. So it's great that we have our pamphlets online and FGConnections and RE materials and A&O brochures. There are hundreds of pages on our site, most of which we probably forget are there, but Google knows them and will display them up when the query is right.

Another lesson is that we shouldn't rely on our homepage to help visitors navigate. We shouldn't even worry much about using how its design will work for both insiders and seekers: most of the seekers never even go there. Most of the people coming to the FGC homepage are looking for FGC the organization.

Committee Page Case Study: One committee, Advancement & Outreach, is considering redesigning their committee page. In preparation I've looked at the usage and I think it makes a good case study. The A&O committee gets the most visible link on the FGC Homepage (top left, it gets this position because the committee list is alphabetical). Despite this prominence, almost no visitors actually follow this link. Only 1.5% of visitors to the FGCQuaker.org site ever get to the A&O homepage and even at that it's the most visited committee page on our site!

Most of the visitors that did get to the A&O page left without clicking on anything. It is safe to say that most of those visitors didn't thoroughly read through the page. The most-followed link is the first one, for the "Inreach/Outreach" review. In the one-month period I examined only 9 people followed this link! This doesn't mean A&O material isn't used: Quakerfinder is very successful and the pamphlet "Resources for Local meetings" is popular. And over 300 people in this month came to some part of the A&O site. Committee pages are useful for the relative trickle of Quaker insiders who visit the page, but we should focus more on the content committees are producing.

The lesson is clear: visitors are primarily looking for 1) good useful content from the "Quaker Library" resources and 2) practical information about the Gathering. Pages about committees and internal FGC workings are not well used. We need to continue the focus on practical resources. We also have to accept that people will not be looking at what we think they should be looking at. Through these visits we will slowly build up FGC's reputation but many people only dimly know what they're looking at.

What I didn't say in the report

In my official FGC report, I only hinted at the differences between institutional websites and focused online new media sites.

One surprising find that didn't make it into the report is that the three most-viewed pages on my own Quaker Ranter site were seen by more people than all but the two most-viewed FGC pages. The most viewed pages on FGCQuaker are the homepage and the Welcome to Quakerism page. Three of the pages on "Quaker Ranter" are seen by more people than any other page on the FGC website. FGC's Religious Education and Advancement and Outreach and Publications pages all are more obscure than my homepage or my "resources on plain dress" directory.

Institutional websites by their very nature have too many conflicting audiences and too timid a voice to act as much more than a reference resource. The Friends General Conference website is probably more friendly to seekers than most other institutional websites out there but even it gets a lot of people hitting the "back" button as soon as they hit the homepage.

Religious seekers are looking for individual voices with something to say and I suspect new media seeker websites will only become more important as time goes on. I suspect this will come as a surprise to institutional insiders as it happens. Sort of relatedly, see my Peace and Twenty-Somethings for some of the generational aspects of this shift. My Books and Media section collects similar sorts of essays.

One more piece in this: the FGC websites didn't get a lot of blog traffic. If all I were was the webmaster of Friends General Conference, I'd assume that all this blog talk in the media was hype. But as the "Quaker Ranter" I know that a popular blog and/or personal site can get a lot of readers. The lesson here is that there's little cross-over. Blogs seem to send little traffic to institutional websites and vice versa (actually institutional websites can't really send people to bloggers for a variety of reasons). I've had a number of people read my blog and declare they'll be coming to the next FGC Gathering so I know personal blogs can help raise organization profiles but that interest doesn't manifest itself as an immediately-followed link. I suspect the community being formed by the blogs is far more important than the raw number of referral links.

Part 5, Quakerbooks.org and Quakerfinder.org

Quakerbooks.org

The first of our two sites to be "googlified" was Quakerbooks.org. I had long hoped to have our book listings show up on the search engines, especially since we carry a lot of hard-to-find ones. I had opened up the discussion board of my peace site to Google and been happy with the results.

Back in early 2003 we installed new software by Steve Beuret to power the bookstore website, one that would allow easy transfer of information between the website and our inventory program. The website could now list whether a book was in stock, and orders would go directly into the system (no more retyping them!). Once the new system was running smoothly, I emailed Steve about optimizing it for Google. There were two parts to this: having the books show up (Steve) and linking them in such a way that Google would index them properly (me). It took awhile to get ito all working but on December 17, 2003 Google came through and indexed the site.

The most visited pages are the introductory ones:

• Welcome to Quakerism
• Becoming a Member
• Basics for Everyone

The search phrases that are bringing in visitors used to be generic ("quaker bookstore") they now are very specific. September's list is typical:

• crash by jerry spinnelli
• Andrew Goldsworthy
• celebration of discipline
• the misfits by james howe
• rufus jones

I knew we'd show up high in the Google rankings for obscure books but I've been pleased that we're right up there with Amazon and Barnes and Noble even with mainstream books.

Our online best sellers are pretty

• Grounded in God: Care And Nurture In Friends Meetings
• Friends for 350 Years
• The Quaker Way
• Philadelphia Faith and Practice
• Listening Spirituality Volume 1
• Silence and Witness
• The Journal of George Fox

The bookstore inventory software is not very good at pulling marketing statistics. While it's very good at telling us what books have sold and what books need to be reordered, it won't tally up things by type of sale (phone vs. web vs. mail-order). The bookstore report should include more information on actual web sales.

Anecdotally it appears as if about half our web orders are new customers. Many of them are from geographic areas which are not traditionally Quaker. A&O has produced a flyer which goes into orders for new customers.

Quakerfinder.org

After we saw how successful the "googlization" of Quakerbooks was, I thought we should try it for Quakerfinder. It took a little seasoning to get everyone on A&O to sign off on the project but I am delighted to say they saw their way clear. The result has been nothing sort of amazing. Use of the site has grown by 340%. But the actual numbers are even more important: by my best estimate, over 6000 a month are using Quakerfinder who would not have even found the resource if we hadn't made it search engine friendly. That's 72,000 people a year--twice FGC's membership, and these are the EXTRA people coming. Altogether at our current rate, this site is being used by over 100,000 unique visitors. Even if only one in ten of them make it to a Meeting, that's a lot of people.

In last year's report I pointed out that most of Quakerfinder's traffic was coming from the FGC site. At that point, it didn't looking like giving the location look-up utility it's own domain name was paying off in any tangible way. Now it's clearly worth it. Just the extra 600 or so visitors Quakerfinder is throwing to FGCQUaker.org site makes it worth it! Horray!

Twenty Times the Google-Linked Visits: I compared two typical months, one before and the other after the "search engine optimization." In May 2004 Quakerfinder received 241 visitors from Google searches (footnote 1). In September, it received 3813 visitors--that's over twenty times the visits. Overall visits almost tripled, from 2292 to 6037, with 60% of those extra visitors directly attributed to the Google bounce. The chart to the left shows daily Google-referred visits since the middle of March.

More Than Just Google: Other search engines were affected too: all together search engine visits went from from 311 in May to 4134 in September. For those interested, the top five search engines for Quakerfinder traffic are:

• Google.com 83%
• AOL: 5%
• Google Canada: 3%
• Yahoo: 1%
• Comcast: 0.8%

As you can see, Google far overwhelms everyone else, which is why we often just call this "the googlization" of Quakerfinder!

Part 6, Miscellaneous and Notes

Miscellaneous

Mailing Lists

Late in the fiscal year, we purchased bulk email software. No, we're not going to try to sell Viagra or a new home mortgage. This program will help us get information out to our bookstore customers and committee lists. Our occasional bookstore emails ("Book Musings from Lucy") have been very well received, with only a tiny fraction of recipients asking to be taken off the list.

Web Host Changes

A big project, though not very exciting, is that we're changing our web hosting company. FGCQuaker.org is with the new company (OLM)and Quakerfinder.org and Quakerbooks.org will be moving shortly. The new company organizes our accounts better and we hope that their service is better. (We'd recommend avoiding Data Realm also known as Serve.com.)

Notes

Programs I Use to Collect Stats:

• For overall numbers, I used a extremely-common program called Webalizer, which gives useful monthly summaries.
• For details I used a program called AXS Visitor Tracking Program, which lets me watch individual users as they navigate the site. With AXS I can also get details on where visitors to specific pages come from.
• I have a list of key words which I watch on Google; every few weeks I record where our sites stand on those phrases and watch how navigational changes I make affect our Google rankings.
• I also use Google to see what other websites are linking to us. I look at what they link to (often not our homepage) and how many sites there are linking.
• I also follow links using more specific search engines such as Technorati, which indexes blogs ("web blogs" or personal diary-like sites).

Measuring Links:

I use Altavista's search engine to measure how many links a site has. For good reasons, Google doesn't list obscure websites and also counts how a site's links back to itself. Here's a sample Altavista query:

See How Can We Measure the State of the Peace Movement? for more on this method of measurement.

Unique Visitors:

The most standard measure of website usage, here is a definition: "A real visitor to a web site. Web servers record the IP addresses of each visitor, and this is used to determine the number of real people who have visited a web site. If for example, someone visits twenty pages within a web site, the server will count only one unique visitor (because the page accesses are all associated with the same IP address) but twenty page accesses."