There are a large number of Web startups relying on an advertising revenue model, many of whom may not understand the landscape. For them, I think I have a fairly sound and business-oriented way to describe the online ad space, which consists of Web sites and advertisers, and of course the myriad ad networks and companies connecting the two. This isn’t a “how the online ad industry works” primer, but rather a simple analysis of how your Web site fits into the online advertising ecosystem. Let’s think in terms of the following two axes and follow with an explanation.

All Advertisers (horizontal line) are  interested in reaching the right people. Simply speaking, there are big $ advertisers – the folks that have huge online ad budgets — and many many more long-tail advertisers who might only spend $50/month. For all advertisers, the more targeted their campaigns the better their ROI (regardless to how you define “targeting”). Advertisers often use online agencies to buy on their behalf, but for this post let’s assume “advertisers” and “agencies” are interchangeable.

Web sites (vertical line) are the proxy by which advertisers reach people. Advertisers evaluate Web sites (among other ways) on the basis of their content, since it’s the easiest way for advertisers to understand the audience/people.  Vertically-oriented sites are those with narrowly focused content, which offers advertisers more straightforward targeting. Advertisers in the auto market, for example, buy campaigns on sites about cars.  Edmunds.com is an example. Horizontally-oriented sites are those with widely diverse content/focus. It’s much more difficult (and often inefficient) for these sites to sell advertising on a targeted basis, because they have content about anything/everything.  MySpace.com is an example.

For big $ advertisers, volume/scale is the 2nd most important consideration. If you don’t have a sufficient audience size, it doesn’t matter how good your targeting is as a Web site — it’s more efficient for them to allocate each $100K they spend across 1-5 sites instead of 50-100. And of course, Web users visit all kinds of different sites so it’s hard for any one site to offer a ton of volume.  Ad networks therefore serve the necessary purpose of aggregating many sites together, organizing them by content type, and offering advertisers scale at some level of targeting.

While some advertisers certainly buy direct from Web sites, the majority of advertisers buy through ad networks of some sort (roughly defining an ad network as any company connecting the buyers and sellers of ad inventory). Let’s mark them as red below, and explain further from the perspective of the Web site publisher.

Vertical Content Sites Make More Money

If you have a vertically-focused Web site then you’ll have a much easier time selling your online ad inventory AND earn the best eCPM rates (defined: the value per thousand impressions, regardless of rate type — CPC, CPM, CPA, etc.).  Medium/large Web sites (more than 1M unique users per month) fit in the lower left quadrant. Since they offer both targeting and scale, they’re able to sell direct to advertisers (making the most money), and they have their pick of ad networks to monetize whatever they can’t sell directly (which is called remnant inventory). This quadrant is where most traditional ad networks hang out — since they primarily consist of a direct ad sales force, they make the most money as a business by focusing on big $ advertisers and vertically-focused sites that can deliver volume.

If you have a vertically-focused Web site with a small audience then you most likely won’t gain any traction with big $ advertisers or traditional ad networks; you are in the lower right quadrant.  Google AdSense is your best option, and you may even earn better eCPM rates than medium/large publishers. Google and a variety of others created this market by inventing a self-service advertising model which provides you as a Web site with a ready supply of advertisers and revenue. By eliminating the need for a direct ad sales force and automating campaign targeting, sales and delivery, they’re aggregating and matching long-tail advertisers with long-tail publishers.

Horizontal Content Sites Make Less Money

If you have a horizontally-focused site (content about “anything/everything”), as is often the case with social media and user-generated content, then you’ll have a tough time monetizing your ad inventory. Average eCPM rates on social media sites are low and decreasing.

If you offer a massive audience — as in many millions — then you’re in the upper left quadrant and will have some luck selling directly to advertisers, but only if you can segment your audience in some way. Advertisers require some understanding of WHO their campaign reaches. Even traditional ad networks with experienced sales teams and solid advertiser relationships shy away from horizontally-focused content sites, especially social media and user-generated content.

There are several startup ad networks attacking this upper left quadrant however, given the massive eyeballs/volume and extremely low eCPM rates. These networks, such as Media6° and Lotame, aggregate and package social media behaviors for the benefit of large advertisers. Unless you have a massive audience, they won’t work with you. There’s nothing long-tail about them. Like all other  ad networks,  they can only support their direct sales efforts by focusing on big $ advertisers and large-volume sites.

So what do you do if your horizontal Web site attracts an audience of <5-10M unique users per month. You have two options: Google AdSense and ad exchanges. The problem with Google Adsense is that it often doesn’t work on these pages — the ads are targeted only to page content/context, and many social media pages (pictures, videos, social networks, etc.) are contextually poor.  You’re better off targeting each visitor based on THEIR MySpace profile, instead of the MySpace profile they’re looking at!

As a result, I believe the vast majority of this ad inventory is being monetized by arbitrageurs via ad exchanges. Ad exchanges such as Right Media differ from ad networks primarily in their pricing model, which works like a stock exchange. This facilitates an arbitrage model where those who know something about your audience will buy your ad inventory at very low rates and resell at higher rates. Of course, this doesn’t always work well for you the publisher — they don’t share what they know about your audience with anyone, because the lower their price the more money they make.

In my opinion, this upper-right quadrant represents the biggest opportunity in online advertising today, given the massive volume and extremely low eCPM rates being earned by publishers today. Just as Google AdSense grew the entire online ad market by matching long-tail advertisers with long-tail vertical content, there exists the opportunity to match long-tail advertisers with people instead of pages.

Jordan Mitchell, the author of this post, is the CEO and Founder of Others Online, his 4th Internet startup. Others Online helps Web site publishers and networks better understand, target and monetize their audience through the use of proprietary affinity profiling software. Jordan blogs regularly and tweets even more!

DeSilva+Phillips wrote a great whitepaper providing an overview of online ad networks. A must-read for anyone trying to familiarize themselves to the world of online advertising, covering:

• Why ad networks exist, and the value to publishers and advertisers.
• The differing objectivs of brand and direct response advertisers.
• Ad network business models.
• Recent mergers, acquisitions and venture investments in ad networks.
• Rank, reach and ownership of the top ad networks.

Jordan Mitchell, the author of this post, is the CEO and Founder of Others Online, his 4th Internet startup. Others Online helps Web site publishers and networks better understand, target and monetize their audience through the use of proprietary affinity profiling software. Jordan blogs regularly and tweets even more!

I was one of the panelists at the TiE event Opportunities in the Web 2.0 Ecosystem, and moderator John Cook asked whether a new startup ought to build their company around the new Facebook platform or choose to build their own platform and network effect.

For a while there (after May 2007, when the platform first launched), I kept getting questions from investor-type people asking me if I was going to build a facebook app — there seemed to be a lot of “enthusiasm”. We chose not to, for a variety of reasons. It’ll be tough to ever know whether that was a bad decision or not (it’s not like the opportunity is forever gone), and clearly there are many companies building their business around a facebook app.

But in my opnion, the value of a facebook app is as mere advertising and/or a feature, not as distribution, a standalone product or a business. Here’s why:

• When you’re offering a free service within another free service, it’s damn hard to arrive at a sustainable business model. The advertising revenue model with Facebook apps is still unproven.
• You don’t own your facebook app users, facebook does. Users/customers form the core value of a business, and you’re not going to have any unless you get a facebook user to register separately for your service.
• The biggest value to web 2.0 products, besides users, is the network effect — which is more valuable than the sum of all users. With facebook apps, you don’t own the network or any of the connections between the users, facebook does.
• Facebook is free to compete with your app, shut you down, etc. They are already doing so.

As a startup, a core value in your company centers around your distribution — your users. Jeremy Liew wrote a good post on how to value a social media company, and # of users is his first variable for good reason — assuming they’re YOUR users. I guess maybe I’m just old-fashioned and want to build my company earning my own users instead of someone else’s (who can be taken away from me within 24 hours). I just wouldn’t do a distribution deal unless it added tangible value to my core business.

Now there are times when a facebook app makes sense to me, such as:

• As a feature of your core product/service, which is offered across multiple channels. Any social media company is going to have a percentage of their users on facebook, so why not make it easy for them to use it within their fb community?
• Promoting your core product/service. Just as widgets ought to be viewed as advertisements promoting your company, a facebook app ought to be viewed the same.
• Optimizing your viral coefficient. Startups need to continually hone their messaging and features in order to improve their viral capacity. Doing this on a closed network can be very effective for ongoing iteration/testing.

The companies that focus on a core product/service, earning their own user base and network value, utilizing Facebook (or any other closed platform) as just one channel of promotion and testing will be in the best position to create real shareholder value. The companies that are defined by their Facebook app incur a greater risk that they end up creating no shareholder value at all.

Jordan Mitchell, the author of this post, is the CEO and Founder of Others Online, his 4th Internet startup. Others Online helps Web site publishers and networks better understand, target and monetize their audience through the use of proprietary affinity profiling software. Jordan blogs regularly and tweets even more!