MetaMuse

The Blogosphere's Loooong Tail

In stark contrast to the Business 2.0 article on the popularity of blogs (and the related ad dollars flowing into them) is a post called The Great Unread, which gives in my opinion a MUCH more realistic view of the blogosphere.

In it, the author Nicholas Carr speaks the truth ... that every blogger struggles to be heard, to have been read, and most fail to reach others relevantly and meaningfully. It simply reinforces my theory that the 50% abandonment rate for blogs is simply due to "nobody reads it". It's really really hard to get lots of people reading your blog. This is not a problem for just bloggers, but rather anyone with something to say on the Web. There is way too much content on the Web, and it continues to grow exponentially. Not so much, people.

Notable comments include:

• "What we tell ourselves about the blogosphere - that it's open and democratic and egalitarian, that it stands in contrast and in opposition to the controlled and controlling mass media - is an innocent fraud (a lie that makes most everyone happy)." (Carr)
• "The powerful have a greater stake in the perpetuation of an innocent fraud than do the powerless. Long after the powerless have suspended their suspension of disbelief, the powerful will continue to hold tightly to the fraud, repeating it endlessly amongst themselves in an echo chamber that provides a false ring of truth. "(Carr)
• "I had delusions of influence. The random-payoff of attention makes it seem far more effective than it actually is. It's painful to admit you've wasted so much time and effort and pretty much nobody is listening." (Seth Finkelstein)
• "As the blogophere has become more rigidly hierarchical, not by design but as a natural consequence of hyperlinking patterns, filtering algorithms, aggregation engines, and subscription and syndication technologies, not to mention human nature, it has turned into a grand system of patronage operated - with the best of intentions, mind you - by a tiny, self-perpetuating elite."  (Carr)

I love that "echo chamber" comment. It's so true, and I think today's consumer-generated news sites are also complete echo chambers.

Blogging for Dollars

Just read an interesting article in Business 2.0, which I heard about via GigaOM, about the seemingly endless flow of ad dollars into popular blogging sites. It just seems so "bubble-icious" to me.

Big brands will always have sizable ad budgets to focus on “brand advertising” — which means they’re looking for (and measuring) reach and not so much conversion. But so much of the growth in online advertising has been based on conversion, where tirelessly monitored ad metrics demonstrate that the $ benefit exceeds the $ cost. I get that. This is the art of contextually targeted advertising and it works. I go to Google or Yahoo, I see ads that are relevant and it makes sense.

What I don’t get, and what reminds me of the late 90’s, is why many struggling companies are blowing ad budget on non-targeted CPM advertising. It’s like deja vu all over again (as Shaggy would say). I go to the sites mentioned in this article and I see their non-targeted (”Hail Mary?”) advertising, which is likely generating the usual 5-6 clicks for each 1000 impressions. (Then of course, it’s the matter of whether they can convert those clicks into customers/sales.) This won’t last — it can’t. Sooner or later they’ll realize that they spent $5K to generate $1K of profit.

The old adage is “I know half my marketing efforts are working; I just don’t know which half”. I believe the future of online (and offline) advertising is one-to-one and knowing exactly what is working, which is what Others Online offers. It’s my honest-to-God hope that these blog advertisers are seeing a positive ROI on their marketing dollars. If so, THAT is incredible. If not, then here we go again.

My First Podcast

I was recently featured in a podcast, being interviewed by Eric Mattson of the blog MarketingMonger.  Those of you who are already sick of hearing me talk should just skip it, but for those of you learning about Others Online, it's a good informational interview.

Jeez though, I talk way too much. I need to work on 15-30 second answers!

"On Page" Chat/IM Has no Chance

There are a number of companies, such as Gabbly, Userplane, Chatsum and Meebo that have created Web page specific chat applications, where you can join in a group chat based on the page you're on.

In my opinion, it's not a model that will ever succeed, for the following reasons:

• There are way more pages on the Web than people, so you need a HUGE number of users for there to be any meaningful connections, even if you show folks that were there. Plus Web content is multiplying at a much faster rate than humans are.
• The page-level "connection" is asking too much of users. People skim the Web, they don't spend time on ONE page. And their interests persist over their online experience; their pageviews don't -- with the exception of bookmarks. But it's my opinion that people view 100x (or more) pageviews than they bookmark.
• It's hard for me to think of many situations where I'd care to get (or send) an IM message or email from (or to) a stranger in the context of a single Web page. I believe to reach the mass market, you need a lot more context to the connection -- the more the better.
• What does "same page" mean anyway? Is it same content (potentially syndicated) or same URL? With the latter, you have to take into account dynamic pages, multiple web servers, sessionIDs in the query strings, etc.

The exception to this theory is online support, where I believe there IS a strong model around having customer support professionals "lurk" on their Web site and answer any questions that may arise.

But if you're going to bring strangers together while they're browsing, it's so much better to match people by topic or keyword, because then you open up the population to everyone instead of just those on the page. Also, tag-based matching studies have shown that at only 229 people, there are no more unique tags -- all users overlap in one or more ways.  This means you need very few users comparatively in order to start to see meaningful connections. We're validating this theory today with Others Online.

Revolution in Ad Targeting

We will see a revolution in ad targeting over the next few years, where PEOPLE are targeted, not the content they're reading, and it's more of an opt-in basis than guesswork.

Behavioral is striving to be like this now -- targeting ads based on an interest profile of  content that the user has gone to in the past. But by and large, contextual targeting like AdSense is currently based on content. The problem is that more and more content is consumer-generated and while interesting, is not consistently narrow in terms of its topic area.

There will always be a place for contextual and behavioral targeting. Search targeting will always work well (and the cost/competition for keyword-based inventory will only increase). Online advertising will not go away. But we're almost to the point where people will determine how they want to be targeted, and Others Online will play a role in that.

Core Obstacle for Online Advertising?

I recently listened to a podcast by Om Malik and Niall  regarding online advertsing, where Om points out the that "the shortage of advertising sales talent is the Achilles heel of the new Web 2.0 ecosystem".

I disagree that the availability of sales talent is the major obstacle facing Web 2.0. That was a Web 1.0 problem.

The key is targeting and ad relevance. With relevance, you don’t have a user backlash, and everyone wins — the user, the publisher and the network.

The long tail of Web site publishers is getting longer, and they’re using ad networks and ad marketplaces to generate ad revs (the difference being that in ad marketplaces, commissions are lower and advertisers have visibility to the publisher network. With ad networks, the publisher network is a black box.).

At the same time, we’re seeing a longer and longer tail of advertisers (even with budgets of $100/month or less). You cannot match the supply and demand with talented salespeople. The online ad industry is successfully(?) addressing this with deeper and deeper targeting (keyword and behaviorial) and automated ad buying. I believe these two factors alone are driving the increase in online advertising.

While the likes of Google and others are basking in the profits from online sales automation and keyword-based inventory, the large traditional networks from Web 1.0 (like ValueClick, 24/7 RealMedia, etc.) are still primarily offering their publishers’ inventory across a dozen categories. Take a look at their revenue growth. Take a look at the eCPM they offer publishers.

The long tail of advertisers want to buy keywords, not categories, and they want to pick their publisher (marketplace model) for CPM buys. And they are all empowered with the tools to exactly measure their results.

Publishers are currently less empowered. They have many choices, can swap networks easily, and can negotiate the commissions down as their traffic grows. But they can’t profile their user base in order to contextually/behaviorally optimize their ad inventory.

Therefore, the holy grail is a mechanism (which users will accept) by which Web site publishers gain access to the desktop.

Now THAT is cool ...

Others Online launched today, and TechCrunch wrote about it. Very very cool. Much more press to come, I'm sure.

More Keyword Seeding ...

I reviewed another set of queued keywords today for inclusion into the Others Online "context" database, as well as our "active" keywords which are already connecting users to each other. There were over 22,000 keywords and phrases "suggested" by our users in the past 2 weeks, but since they were sorted by frequency of occurence I was able to start with the ones that recurred most frequently and work my way down the list.

I got through almost 7,000 keywords, right up until I was reviewing keywords that had only recurred 3 times. More than 60% of the 22,000 keywords had occurred only once. Of these 7,000, I added 242 to the active list, so that we now have a grand total of 804 active keywords to launch public beta with! I'm very pleased with this -- they are solid keywords, organically grown by our private beta users. Some of the interesting ones added today include (by rough grouping):

• topple bush, bush jokes, george bush, christine gregoir
• facebook, tagged, wikipedia, motley fool and alexa
• portland craigslist, tacoma craigslist, portland citysearch
• cabernet sauvignon, vancouver wine, wine talk, chardonnay
• home theater, canon powershot, nikon, xbox 360, tag heuer and powershot s2
• labradoodle, goldendoodle, puppies, golden retriever
• justin timberlake, britney spears, kevin federline, jessica simpson, homer simpson
• aston martin, ferrari, tt roadster, hummer h3