Recently I’ve been evaluating my blog and online presence, to see what works and what could be improved. To be honest, I don’t pay much attention to the analytics and readership numbers most of the time because I’m not trying to write for a mass audience and I don’t want to get in the mindset of doing things to grow the numbers — I want to stick to trying to write what I find interesting and what I hope others find interesting and entertaining, and let that find its natural audience over time.

The numbers, however, are useful for helping you understand what is interesting, but raw numbers need some thought and interpretation and to be honest, there are lots of numbers that all mean different things, and the ones most people focus on seem to be ones that I don’t think mean very much. So I’ve been chewing on the numbers trying to figure out what they mean and what that tells me about what I’m doing.

This implies, by the way, that I’m starting to redesign the blog; it’s been about a year, I’m ready for a new look, but not a radical change. And in the last couple of weeks, I’ve been doing some quiet tinkering to experiment with some ideas to see if they change certain behaviors in any way (or even if they don’t, if I like the look they bring). most of them are things you aren’t likely to notice and don’t deserve talking about in any depth; a lot of it is trying to understand how to improve discoverability of material by people who come and visit the blog, and to improve interconnections between the different pieces of my online universe.

When I redesigned my blog a year ago, the big weakness that I saw was discoverability. people brought in by an RSS link or a search engine search or a link from another site looked at the page they landed on and then left again. There was very little exploration, and there was nothing on the old site to encourage people to explore. The two numbers I paid closest to on Google Analytics were the average length of visit and the number of pages visited per visit. Before I redid the blog, they were terrible — with average stay on the site under a minute, and average number of pages visited under 1.2. The blog design actively discouraged people from finding anything other than what they came to look at and the front page, so they didn’t.

In the redesign, I tried to bring into view a few key articles and other things that might be of interest to people visiting the site; this is typical split out into a semi-permanent list of “best of” articles and a regularly changed list of most popular pages on the site in the recent past. I haven’t found a wordpress plug-in that does that the way I want to do it, and it doesn’t take long to manually update, so I manually do that every 4-6 weeks. each blog post also has an auto-generated list of related articles, and that’s worked pretty well.

I’ve experimented with a few ways to do the same with photos, and haven’t liked any of them; they all clutter the pages up and/or compete with the front page slide show. I’m currently trying something rather less kowabunga, a couple of widget boxes with one photo each for the portfolio and the wallpapers; I like the look, it’s too early to tell if it’s doing anything useful — but I do like the look, and I can tie it back to the saturday or sunday blog posts and create some linkages there.

Overall, the blog redesign did what I was hoping for; average time on site has grown from under a minute to about 1:40. That may not seem like much, but it’s just under double (55 seconds to 95 seconds); pages per visit has grown from about 1.15 to 1.6, up about 25%. Can this be better? I dunno. I certainly don’t want to do things that get in the way of the users primary goal (which is read whatever they can to look at) or be annoying. I can continue exploring how to make sticking around interesting, but I understand most folks are focuses on their task, and I am not going to interrupt that for the wrong reason.

For instance, you’ll never see me do pop-overs or other things that prevent you from reading the article until you do something (like register) or dismiss the dialog. There are a couple of photo sites who aggressively try to get you to subscribe to their newsletter; I know that annoys me, and eventually, I stop visiting links pointing to their sites when I recognize them. If it annoys me, why would I do it to others? So I won’t. I think this comes from thinking about the wrong numbers — total subscribers vs engaged readers. It’s easy to track total subscribers, but I think it’s basically a meaningless number if the subscription isn’t actually read. And I can only wonder how many people those kind of techniques actively drive from the site. I can’t see it as a net positive.

I don’t want to force you to subscribe. I want to convince you to want to. Hijacking your eyeballs won’t make you happy, I can’t see that as a way to build a positive relationship. Because of that philosophy, I’ve played with how to make sure you can figure out how to subscribe (if you want to), but without making it pushy or obnoxious. I’ll take smaller subscription numbers but actively interested readers any day.

So how many subscribers/readers do I have? That turns out to be a suprisingly complicated number. And how does that compare to a year ago?

That turns out to be even more complicated.

You can get content on my site a few ways: you can bookmark it and come visit any time you want. In reality, very few people do this; I know if your site doesn’t have an RSS feed or some way to get notified of updates, I’ll lose track of you — and I know pretty much everyone I talk to is like that now.

I use RSS, Twitter, and Facebook for new content notifications. Whenever I post a blog entry, it goes on the RSS feeds and Twitter. The Twitter updates flow off to my facebook page.  Stuff that goes onto facebook tends to stay on facebook; it’s very sticky about data, so a comment stream there will not end up tying back to the blog; most of the comments I get on postings happen on facebook now, but it doesn’t integrate well with the rest of the universe (in Facebook’s view, this is a feature, not a bug; I don’t care enough now to try to “fix” it and get that data back on the blog. someday, maybe I will).

Twitter has links that point back to the blog; not all twitter clicks have referer data that point back to twitter, so the exact percentage of clicks twitter generates isn’t accurate. In general I’m finding about 15% of my blog visits start with Twitter these days. That’s pretty good (and the real number is somewhat higher; 35% of my visits have no incoming referer, and a chunk of those are from Twitter; it’s not unreasonable to think that 1/3 of my blog visits come from Twitter)

RSS? It all goes through Feedburner, and since I use full feeds, you can read my postings without actually visiting the blogs. This is convenient for readers, inconvenient for trying to maximize pageviews and complicates generating analytics and statistics no end. Since I don’t care about pageviews for advertising and I can live with complicated statistics, I’ll stick with with making it convenient for the reader. I have considered putting some limit to the size of an article in the RSS feeds (say, 750 words) or tweaknig RSS and how it presents the photos, but deep down inside, my view is — not broke, don’t fix.

So really, people who read my blog read it in two places; one is on the blog itself, and one is inside their RSS reader via the RSS feeds. So to answer the question of how many readers I have, I have to figure both of those out. And the data on the RSS feed doesn’t include how long they took to read it, so unless they actually click on a link, it’s hard to tell if they read it, skimmed it, or merely marked it as read while answering the phone…

So how many subscribers are on the RSS feed? Somewhere around 550. the number is variable (if you don’t fire up your reader, you don’t get counted). And that number is down from a high of 700 in January, but in clearly fell off a cliff and so that change was a change in how Google manages or interprets numbers, not a massive die off. It looks like the data slowly inflated and then was corrected (if Google told anyone they did this, I missed it). If I look at that number from a year ago, it’s down from about 625.

But as I’ve said, I find that number rather bogus, and there are clearly problems with that number given the obvious artifacts I see in my two years of data. Feedburner gives you another number, called “reach”, which is an indication of how many subscribers actually interacted with an article (active vs. passive) — and a year ago, my reach number was typically in the 50-75 range. It sometimes spiked into the 200s, and if i wasn’t actively posting, dropped down around 5, so reach is the number I use to gauge how many users are actually looking at the content or doing something with it, not just subscribing to it.

Think about these numbers from a real world perspective. If you hire the post office to deliver a flyer, they’ll stick one in every mailbox on the route for a modest fee. You can claim a huge subscriber base — but a 1-2% response rate is considered great in the bulk mail world, and 90 or 95 of the 100 flyers gets tossed in the trash unread. That’s what your feedburner subscriber number is, it’s a bulk-mail number, and says nothing about whether someone actually looks at your article, just that they have a mailbox it gets stuffed into.

Reach is a number that gives some indication that the reader did something — looked at it, clicked at a link, something. It’s the indication there’s actually a reader there, not just that someone subscribed sometime in the past. So to me, that’s what matters. I spent too much time at Apple fighting the “wives and cattle” mentality of amassing huge subscriber lists of people who mostly wanted off those lists to ever see that as a good thing (I was known to say “Anyone who comes and tells you how large their list is, instead of how many users did something based on that email, should be fired” — which did not sit well with people who liked to brag about how large their lists were).

So a year ago, my average reach was about 50-75. Today, that number is around 200; tripled or more. Even though my “total subscribers” is down (in theory), the number of users actually reading or clicking on a given article is significantly larger. Not remotely “engadget” numbers, but still, that’s a huge grown in active readership.

On any typical posting, about 200 RSS users view the article within the first 18 hours, and that number will grow to between 260 and 400 over the next week. It’s interesting that the time between posting and first read is that short and the curve this sharp; RSS is still very much about finding content fast and then moving on; because there’s always more content arriving. After two weeks, I’ve seen 99% of the RSS reading I’m going to see for any given article.

Based on all of this, I’m now consciously trying to write fewer posts, but longer and with more original writing and less “drive by blogging” crap. My target now is 5-7 pieces a week. two of those are photoblog entries. I’m trying to do 1-2 more extended (over 1000 words) pieces if I have time, and the rest of the articles I’m trying to do 500-700 words. What really matters is that whatever I write about gets the attention it deserves and that it’s me writing, not me just filling space by cutting and pasting from what others are blogging.

If I get too busy to do that properly, I don’t blog. No filler crap any more.

So that’s my goal: two photoblog entries a week; one long-form review piece around 1500 words. one long-form piece on something worth writing about. and 1-3 shorter pieces that don’t require as much writing time.  I’ve said for a long time that successful blogging isn’t about posting every day (one of those things “experts” say I’ve always griped about) — it’s about writing consistently with good content. I made a commitment this last year to try to do that, and I think the readers are showing me I was right. So moving forward, I’ll be trying to do it even better, and more consistently. But I like this format and I like this structure.

And added benefit of sticking to the “one piece a day and a link summary” format is I have to think through what to write about. There are always things I put in the “to do” pile that get thrown out. That’s a good thing — it forces me to think about what the more interesting/important issues or topics are, and that also leads to better content and writing. The idea of throwing everything against the wall and hoping some of it sticks just doesn’t work very well. IMHO.

So over on the RSS side, the change has been massively to the good; tripled active readership, lots more engagement and interaction. The link summaries turn out to be really popular, and they create some visibility and help spread the word on other sites and writing of interest without a lot of clutter; I’ve experimented with ways to do that, and I think I finally have a setup that works well. I do have some wish for a short form, where I could maybe write 25-100 words on a link and give some context; I haven’t found a way to do that I like yet, and it’s not a huge priority. But some day…

What about on the blog itself?

Looking at the last two weeks of the blog, here’s what I see: compared to the same two weeks a year ago (and both two week periods are “typical” of blog performance at the time — you need to be careful about grabbing data that isn’t representative of overall usage); visits are up 30%; pageviews are up 67%. 25% of traffic is coming in with no referral, and I think a large part of that is twitter. 25% from search engines. 30% from twitter with a referer, and about 12% from RSS, and 7% from Facebook. The rest is from random sites.

quick digression: a meme out on the net is “RSS is dead”. Mostly, I think, from people addicted to always going off and playing with the new and bleeding edge: my response, based on my numbers: “no, it’s not. but it’s been commoditized”. And the twitter traffic I see driven to the blog seems to be close to the amount of activity I see on the RSS feed (since I use full feeds and they don’t necessarily click over to the blog to read, there’s some guessing and handwaving here), but I think close to 35% of my total reader engagement on a given post is now via twitter, 35% RSS, 25% search and everything else combined is “other”. Most of what is happening on twitter probably would have been RSS subscriptions 2-3 years ago, so to that degree twitter has replaced RSS as a primary interaction channel. Twitter, however, is having its own fight to prever commoditization of twitter as it tries to capture its own feed and give it an economic value; if I had to guess, I’d say they’ve already lost that fight and twitter will end up like RSS, something in the background that moves data around where other things turn it int information of value. Whether twitter can also be the thing that creates that value (and monetizes it) I’m not so sure.

The last thing I’ve looked at is what kind of content generates the most interest. One first problem with that concept is defining “most interest”. I decided that meant two things: how many folks read it (on the blog; sorry, RSS readers, but it gets too crazy otherwise; if you like something, CLICK THROUGH TO THE BLOG; easy, and it leaves an indelible mark on that article for later analysis) and how long they spent reading it. Since I’m trying to focus more on long-form writing, I think it makes sense that the stuff people spend the most time reading is the kind of stuff both I and the readers want more of. Make sense?

This is all really subjective. Without going down a rathole with lots of lists of individual articles without any real context, here’s what I’m finding.

The photo essays I’ve done about my trips seem to be popular (such as this one); people seem to spend some time enjoying those.

I’m a little surprised to say this, but the pieces I write about thinking through the planning process of a shoot or trip seem to be a lot more popular than I’d expect (this one is a good example). This to me indicates there’s a lot of pent up interest not so much in the geeky aspects of photography, but in learning the logistics of being a photographer (shoot planning, visualization, trip planning, etc). There’s definitely interest in the geeky stuff as well, but it seems a lot of us are trying to figure out the mental aspects as much as the bits involving a camera. I know that’s why I post it — it helps me focus my musings and structure the ideas.

It is still too early to make any real judgements, but the “wednesdays in review” series is seeing some nice responses. People seem to be reading them in good numbers, and spending some time reading the pieces. And I’ve made FOUR whole dollars in Amazon affiliate referrals. John Scalzi owes me for the six books of his I’ve sold so far… I thought this would be an interesting thing to do on a regular basis, and the early response is justifying that. Another couple of months and I’ll be able to afford to go to Starbucks. Once.  (as I have more experience with this under my belt, I’ll write in some detail about the whats and hows).

And yes, photo geeking is fairly popular.

If I wanted to maximize readership, I’d write about two things: hockey and Apple. Writing about Apple is off the docket as long as I work for HP/Palm due to conflicts (and honestly, my massive enthusiasm for that is a bit past as well), and while I’ve picked up again on the hockey side again, I’ve done it only to the degree I want to write about it. It’s about writing what’s interesting to me and not just maximizing pageviews. And there’s no reason I should be seen as an expert in the photo field — but it’s really where my head’s at much of the time, and so it’s where i spend a lot of my time and energy. And so that’s why it shows up on the blog so much. Maybe, over time, it’ll help build a reputation, but mostly, it’s a chance for me to learn so I can share, and occasionally to teach. And sometimes such as my lightroom keywords piece, they catch on and become something that benefit large numbers of people. That piece now accounts for about 6% of the pageviews on my blog (and I’m damn proud of it…). Now, I need about 20 more pieces that good and popular.

Which means it’s time to shut up about this and get writing… But hopefully, if you’re trying to think through how to evaluate your own site and you are looking at feedburner or Google Analytics and going “none of this makes any sense at all”, this will help give you some ideas on how to view your own data. And if not, at least give you some comfort that you’re not alone at looking at this stuff and thinking to yourself “this is all gibberish!”