http://img181.imageshack.us/img181/4738/typewriter11ie6.jpg

Starting from scratch
I started my professional blog in late 2006 as I was packing my bags and moving from Seattle to San Francisco. In the first month, I was pleasantly surprised to see a couple ex-coworkers subscribe via e-mail, and didn't think it would ever lead to more readership than that. Two years later, I have a nice community of a couple thousand subscribers, and I occasionally get the question of "how would you start a blog, if you were to do it over again?" I thought I'd share my thoughts on that.

Here's the quick summary, for those who want a quick skim:

1. Carpet bomb a key area and stake out mindshare
2. Take time to find your voice
3. Stay consistent on your blog format and topic
4. Just show up
5. Go deep on your topic of expertise
6. Meatspace and the blogosphere are tightly connected
7. Embrace the universal reader acquisition strategies for blogs
8. Come up with new topics with brainstorms, news headlines, and notes-to-self
9. Look at your analytics every day
10. Don't overdo it

Extended discussion below...

Carpet bomb a key area and stake out mindshare
Like all products and markets, the blogosphere has its own set of existing products and channels. For example, if someone asks me about a "VC blog" I might refer them to Fred Wilson, Mike Speiser, Venture Hacks, and others. If someone asks me for a "games blog," I typically recommend folks like Raph Koster, Nabeel Hyatt, Daniel Cook at Lost Garden, and others. The point is, just like companies, blogs tend to achieve 30-second elevator pitch status, and it makes sense to figure out a theme for whatever you're going to be writing. IMHO, as long as the space you're writing about is growing, you can never be "too vertical" since you'll easily attract a couple thousand supersmart people who care passionately about your particular sub-vertical.

In general, I find people describing this particular blog as "the viral marketing blog" more than anything else. I write about a bunch of other stuff other than that, but people seem very interested in the topic so I'll take what I can get ;-)

Take time to find your voice
As I said before, it's good to find a key area. That said, it takes some time to get there, and I spent the first couple months switching between a couple topics - be it personal stories, product design, and advertising. It wasn't until almost a year in that I started writing about viral marketing, which this blog is probably most widely read for.

I found that as I wrote more consistently, and learned from other bloggers, I began to change the tone and voice of my articles. While some of the key elements were always there - essays rather than links, certain topical themes, etc. - I added much better formatting within the blog posts, photos, linking to other blogs, etc.

Stay consistent on your blog format and topic
Related to finding your voice, it turns out that blog format really matters. To completely oversimplify, there seem to be two very different kinds of blogs that are successful. Either you're a "curator" of news, or you're a primary content source.

The curator is someone who blogs often and throughout the day with links and snippets, and I would consider someone like Robert Scoble (or iJustine!) to be the Michael Jordan of this approach. The style is often more conversational and casual, and includes lots of little updates on what they are doing or reading or trying out. These guys can really "cover news" and are widely read because they can provide the first opinion on new stuff coming out.

The bad news is that the curator model requires you to stay on top of things ;-) For a guy like me, with a full-time job and blogging as a dirty habit, being a pure content creator is much more appealing. I will never get the traffic of the news curators, but I can go deep on a specific topic and get a laser-focused audience that just cares about the topics I write about.

It's obviously good to experiment and leave the door open for any and all topics that interest you, but obviously once you begin to settle and find your voice, it's good to focus since then your readers will know what to expect from you. There's nothing worse than that guy that writes one really good essay about the industry and then spends the rest of the time writing about his dog!

Just show up
Hands down, the hardest thing about writing a blog is doing it regularly. I often just don't have the energy to write, and have to consider it as a core part of my job in order to get it done. It's especially true once you get past the first couple months and you've hit the top 90% of topics you wanted to get off your chest. Then coming up with new ideas is much harder.

In general, it seems like you have to maintain a tempo of at least 1 essay a week to be relevant. Any less than that, and people stop reading (or at least you'll have all subscriber traffic and no one will just check your site). This blog is averaging about 1.5 posts a week, which I should probably work on, but it seems enough for at least some group of people to follow it. If I weren't so lazy, I'd try to get at least 3-4 posts up per week, and possibly make them a little shorter. (Or one long one, and 2-3 news-related items)

Go deep on your topic of expertise
In general, I've found that you can never go "too deep" when covering a vertical. Some of the things I would have thought were the most esoteric - like viral loops and sharkfin graphs - have become the most widely read and widely linked posts. I originally hesitated to even post those since I thought they would be too obscure, but instead I found that people either appreciated it more. My guess is that it has to do with the fact that either they're learning something completely brand new that they think is interesting, or at the very least you're build "street cred" by not being the typical super-high-level analyst.

Meatspace and the blogosphere are tightly connected
Surprisingly (or unsurprisingly?) the world of San Francisco conferences, events, hanging-out, etc. are very much correlated with the blogosphere. A lot of readers of my blog either know me personally or know of me through a mutual friend - I get this sense since many of the inbound emails I get start with, "hi I read your blog and am friends with X" or I have lots of friends introducing me to FOAFs who want to talk about a topic from my blog. I think the point of this is just to say that you can grow your readership by being part of the conference circuit - either organizing or speaking or otherwise - and also your blog readership will lead to opportunities to get more visibility in meatspace. It's all useful, so embrace it. And move to SF if you haven't already ;-)

Embrace the universal reader acquisition strategies for blogs
When it comes to blogs, the user acquisition is pretty boring. You basically have the following sources of traffic, by importance:

1. SEO, specifically Google
2. Blog aggregators (like delicious, digg, etc.)
3. Social platforms (like twitter, friendfeed, etc.)
4. Individual blog links

Given that a lot of your blog traffic will come from SEO, it's a good idea to try to own some keywords for a topic if you can. I get a ton of searches on viral loops and other viral marketing terms. It's a good idea to add whatever your main keywords are to your blog title, blog topics, etc. I sometimes use blog titles like: "Facebook marketing: X" to get people to link back to me using those terms.

Similarly, whatever your expertise is, you might find vertical aggregators that drive a lot of traffic. For me, it's Hacker News, Techmeme, and others. Identifying those key aggregators and submitting your articles is key.

Come up with new topics with brainstorms, news headlines, and notes-to-self
As mentioned above, it's very very easy to run out of new topics to write about. Writer's block seems to afflict me almost every week ;-) This is especially true once you hit the 2-3 month mark, since many of the topics that you might want to write about have already been covered in one angle or another.

My usual remedy to this is to employ a set of tactics that generate a healthy list of blog topics in my inbox, to be written one day or another. The first tactic is that when I'm in a good creative mood, I'll often do a quick brainstorm of many potential topics and ideas. Some can be simple and explanatory, like "how to do X" or specific companies, or recollections of specific conversations that are worth blogging about. Similarly, I'll also peruse sites like Techmeme and look for headlines that catch my attention. In particular, I often look for things that I think are either wrong, need clarification, or otherwise would compel me to rambling if someone told me about it in person. All of these ideas I will write up in very short outlines and e-mail to myself. Having a short outlin or starting a paragraph or two of the post helps me sketch out the idea in enough form to easily execute it at a later date. Otherwise, if you just have a fun headline but no body, going from 0 to 60 can be quite rough.

Look at your analytics every day
I look at the small amount of analytics on my blog on a frequent basis to understand what's going on. It's really nothing fancy, and certainly pales to the kind of instrumentation I'd do on an actual web project, but it's good enough. More importantly, it helps you get some interaction with your passive users that aren't leaving comments, and helps you figure out how to serve them better.

In general, I start by looking at my referrers every day, along with daily visitors/pageviews. I have sitemeter bookmarked on my phone so that I can glance at this all the time. I'll look at what searches people are making via Lijit, and what searches are drawing people to this site. Another thing is to look at my Feedburner numbers and subscribers to see who is subscribing and how things are trending. I also get my top referrers and top content e-mailed to me on a weekly basis by Google Analytics, so I have an understanding of what people are looking at.

In general, in looking at this information I'm trying to assess a couple things:

• Are there specific topic areas people are coming to the site for, that I should write more about?
• Are there certain traffic sources that I should try to "develop" more? (Twitter is one good example of this)
• What are the e-mail domains and companies that are visiting this blog, and how would I better serve those readers?
• What are the searches people are making on the site, and are there any that aren't returning any results?

The point of all of this is looking at your blog not as a "diary" as many people do it - instead of being fuly focused on yourself and what you want to write, you can think of your readers as your customer-base, and you're trying to collect whatever knowledge you can to cater to their needs. Obviously, for part-time bloggers like myself, it's important to balance your interests with the interests of your audience, but in general I think the philosophy holds.

Don't overdo it
Finally, have fun ;-) After all, you can always quit! I often find myself not blogging for a week or two just because I don't feel like it. I think that's OK, since this isn't my full-time job and I'm just doing it for fun. I think if I felt a lot more pressure to do this consistently, regardless of my enjoyment, I'd probably stop since it wouldn't be fun anymore.