Why Monetizing Social Media Through Advertising Is Doomed To Failure (part one)

[synecdochic]

I saw today's lj_2008 post announcing the return of Basic accounts with great interest, because the large print giveth and the small print taketh away. As the comments to that post make clear, they're not talking about bringing back Basic accounts; they're talking about creating two levels of ad-supported account, one with more ads than the other, and calling the one with fewer ads "Basic".

The reason given, of course, is the need to "generate the revenue we need to support the development of LiveJournal while respecting the community", and that's how it should be: any service needs to bring in enough cash to pay the bills, fund development, and make sufficient profit to satisfy their investors and their board.

Right now, the prevailing trend in the industry, which LJ is following, is to use an advertising-centric, sponsor-centric business model that capitalizes on the content being provided by the users of a social media site to provide micro-targeted advertising. In its best implementations, it puts businesses in touch with consumers on a one-to-one, human level, while allowing the hosted service to meet their revenue needs and earn enough of a profit to keep them motivated.

And yet, so many LiveJournal users greet any mention of advertising or sponsorship with an immediate "DO NOT WANT". It's gut instinct, and examining that gut-instinct response is fascinating; if you read the comments to that lj_2008 post, the one thing that becomes immediately clear is that LiveJournal users resent someone else profiting from the content they provide without providing them with what they perceive as "significant enough" return for that use of their content, whether that return is in features, service-stability, or usability. More than that, though, there's a strong undercurrent of LJ users believing that advertising is not the right source of revenue for LJ to be pursuing -- that the service would be better-served by choosing to rely on other sources of income.

Most of the people in the comments have never run any sort of business on the scale and scope of LiveJournal, of course, and so that belief of theirs might be incredibly naive. I've come to believe, though, that the gut instinct is correct: the existing advertising model is inherently unprofitable and unsustainable for a social media service and will never reach a level of profitability and sustainability that will provide a social media service with lasting and ongoing revenue sufficient to cover operating costs and provide a profit, both because the existing advertising model doesn't provide sufficient value to its advertisers and because it's inherently anathema to the people who are the service's most valuable content creators.

It's taken me a while to come to this conclusion, and I'm still not entirely convinced. But I want to talk about the economy of Web 2.0 for a while, and lay out my arguments to explain why I believe that the existing advertising model is not the answer to building a sustainable service. (And, maybe, offer some alternatives.)

Today, I'm going to talk about how we got here, how internet advertising works, and provide some background and history. Tomorrow, I'll lay out the conclusions I've drawn from those premises, and provide my rationale for why I believe that monetizing social media through advertising is doomed to failure in its current implementation.

I wrote the first draft of this a few months ago and left it alone because it needed considerable editing and I got distracted by other things. This is the second iteration, and I want to say thank you to everyone on my friends list who helped me pinpoint the weak spots.



Building our premises

For shorthand purposes while I'm writing, I'm simplifying a lot of terminology; I'm assuming that many of you guys reading this aren't in the advertising or Web 2.0 industries or working in business development. For those of you who are: yes, I know I'm tapdancing over the surface of some of the nuances, as well as retreading some familiar ground. For those of you who aren't in the Web 2.0 industry, I'm going to highlight some occasional Read More links (denoted by ^[?]) that link to further explanation on particular issues.

My cred: I've been working in Web 2.0 before it was called "Web 2.0", including a five-year stint at LJ in which I helped with a few alternate forms of sponsorship (other than the banner/text ad model, I mean). I've never worked in advertising, but I've participated in a lot of projects based around revenue generation and business development. I'm also one of the people behind Dreamwidth, which is an attempt to do Web 2.0 without the advertising -- for the very reasons I'm presenting here.

I apologize to my non-US readers; this analysis does rely on how social media is used in the US and US tech industry, because that's what I'm familiar with, and that's where most of the conversations I've been exposed to arise from.


Defining our terms

If we're going to talk about the economy of Web 2.0, we first need to define what Web 2.0 is. Tim O'Reilly takes a stab at it: "Network effects from user contributions"; "the web as platform". ^[?] Eric Schmidt, the CEO of Google, sums up the Web 2.0 ethos as: "Don't fight the internet."

The social core of Web 2.0, as it's evolved over the past few years, involves the concept of collective intelligence: a Web 2.0 service is one that provides the platform or service and turns it over to its users, who then create and maintain the content of that service. That content is then shown to others, who consume and extend the content. LiveJournal gives you things to read; del.icio.us gives you organized things to remember; Wikipedia gives you things to know. (How many times in the past week have you found yourself saying "Wikipedia knows everything?")

Web 2.0 as I'm referring to it for this discussion involves people: people make the content that other people consume, more-or-less on a one-to-one basis, instead of a centralized service producing the content and the userbase consuming it. In a Web 2.0 world, the things you're reading don't come from people hired to give you things to read; they come from the person next to you, and that person may also be consuming what you produce.

This has produced a class of sites known as "social media" sites: sites that exist to provide a platform for networked content creators. (We'll talk about the difference between "social network" and "social media" in a little bit.) In a Web 2.0 world, the idea is not that the service provides the content; rather, they make it easy for you to share the content you've created.

(For "content", I don't just mean writing; I mean anything that someone might want to find on the web, whether that be writing, video, information, organizational taxonomy, etc. For "consuming", I mean any form of use: reading, accessing, learning, etc.)

Web 2.0 -- and specifically, social media -- came, in a very interesting way, out of the first dot.com bubble of the mid-to-late-90s, and specifically, out of the bursting of the dot.com bubble of the mid-to-late-90s. It's a direct (and mostly, I believe, unconscious) response to the old model of content creation, and specifically, to content monetization: how a service makes money from its users.

A while back I found a quote on bash.org that provides a fairly glib description of the Web 2.0 business model. I've quoted it before, but I'll quote it again:

<dsully> please describe web 2.0 to me in 2 sentences or less.
<jwb> you make all the content. they keep all the revenue.

It's cynical as hell, but every Web 2.0 person I know looks at it and nods, a bit ruefully: in Web 2.0, that's how it works. The service isn't selling you content; they're selling a place to put your content, and they profit from the content you create.

Let's look at how it happened.


The old model

The offline economy of content is structured around a broadcast model: the service creates content, and the consumers pay to access the content. This can be done through three models: subscription (or purchase, such as when you buy a book), advertising (such as network television), or a mix of advertising and subscription (such as newspapers). The most common is the subscription-and-advertising model, like cable TV: subscribers pay a fee to access the content, and the shortfall is made up by advertising. You pay for basic cable, and then sit through ads during the program you're watching; you pay to subscribe to your newspaper, and then view ads inserted around the articles.

There are challenges involved in adapting that for the Internet, though. There are many different levels of "web service": those that can only exist online, those that can exist both online and offline, and those that exist primarily offline and are enhanced by being online. ^[?]

The problem is, and always will be, that business needs move faster than innovations in responding to them. We're in a phase now where the fundamental realities of the marketplace -- the concatenation of economic and social factors that define a marketplace -- are moving so fast that it's an ever-shifting target. The 20th century "offline" business model (which is, in turn, derived from the 19th century business model) is bad at adapting to that. The business's needs and requirements, and the marketplaces they move through, will always evolve faster than the tactics the business can come up with to respond to those requirements, and by the time a business has solved the last problem, the next one's come along. Businesses are always fighting the last war.

You can sell physical goods on the Internet, pretty easily: the old business model works on that pretty well. Figuring out how to sell people something completely intangible, something that can only exist on the Internet, is a challenge.

Who wants to pay $20/year for every service they use on the Internet? Subscription-based services behind gated walls are falling. (It works for the Wall Street Journal, for instance; it doesn't work as well for the New York Times; it fails spectacularly when you go further. Who still uses Napster?) And even if a site manages to find a way of producing valuable services that people are willing to spend money on, they're going to run into problems down the line. (More on that in a few moments.)


The rise of Web 2.0

Back during the heady days of the first dot.com bubble, during the first round of sites scrambling to figure out how to pay the bills, content-providing services turned to an advertising-only model, dropping the subscription model entirely -- because people don't pay for things they can't touch. These services were, for the most part, still creating the content themselves; rather than asking their users to pay to consume the content, they put ads on the content and called it a day. The cost to run those services, however, included both technical-infrastructure costs -- bandwidth, hardware, power, etc -- and editorial costs: the people they had to pay to make the content they were serving.

Eventually, advertisers realized they weren't seeing a return on their investment and began to drop out, venture capitalists realized that the sites they were funding had no other sources of revenue, and the bubble burst. The services that were lucky enough to have gone public or been purchased for insanely inflated amounts before the puncture won. The services that weren't so lucky lost.

The services that were lucky enough to survive were mostly the ones who'd bottled lightning: services (like LiveJournal) that had managed to create a compelling reason to subscribe, and services (like Slashdot) that had managed to create a niche-market audience to whom advertising could be hyper-targeted (and thus profitable). (I'll note, in passing, that both of these examples followed the not-yet-prominent Web 2.0 model of serving as platform provider rather than content provider, anticipating the next big thing.)

In the ashes of the bubble collapse, people thought about how they could avoid it happening again. Clearly, the idea of monetizing service-provided creative content had already (mostly) failed; the services that survived were mostly dual-platform, such as physical print media + online content, and a service that wanted to be online-only (thus avoiding the additional costs of physical operations) couldn't afford to view their online service as a loss leader.

Web 2.0 happened because content-providing services wanted to shed some of their costs, and the costs they chose (consciously or subconsciously) to jettison was the cost of keeping a paid editorial staff. They wanted to shift the burden of content-creating to the people who were also consuming the content, and take a cut of the profits.

The shift in conceptualization of "who makes the content", though, wasn't accompanied by a shift in conceptualization of "where the money comes from". Web 2.0 services generally keep the same model as their Web 1.0 predecessors: subscriptions, advertising, or a combination of the two. And for a whole lot of Web 2.0 services, that model works -- or fails -- about as much as it did the first time around.

There's no question, at least to my mind, that we're in "Bubble 2.0" right now, as well as Web 2.0. ^[?] I'm not the only one: I refer to Wired's "MySpace and Friends Need to Make Money. And Fast", a pretty scathing (for Wired) condemnation of the advertising model as applied to social media sites and social networks.

Like the last bubble, the industry is fueling itself on promise, not results: the funding of our little homegrown empires is predicated on some mythological future where we'll figure out how to make money on all this cool stuff everyone's doing. ^[?]

The South Park underpants gnomes cliché ("Step 1: steal underpants; step 2: ????; step 3: Profit!") is a cliché, but clichés become clichés for a reason: right now, nobody has any idea how to monetize or capitalize on social media. Right now, the business model of nearly every social media site out there looks like:

• Step 1: Build a better mousetrap
  
• Step 2: ?????
  
• Step 3: Profit!

A lot of people have tried figuring out what that "Step 2" is. Despite the example offered by the first internet bubble, a lot of sites still think that Step 2 should be: "put ads on the mousetrap". Why?


How Internet advertising works

Understanding the principles behind internet advertising -- at least, the way it's generally implemented right now -- is pretty basic: for the most part, ads on the internet follow the same model as ads on, say, broadcast television, only updated for the modern world. It's the fundamental equation of advertising: advertisers are willing to pay a service a certain amount of money, based on the return that they think they'll get.

There are two basic models of how the cost of internet advertising is set: "cost per thousand impressions", or CPM ^[?], and "cost per click", or CPC ^[?].

CPM functions on a basis of "how many people have seen the ad". Advertisers bid on the price they're willing to pay for every thousand pageviews; an ad with a CPM of, say, $1 means that the advertiser pays the site $1 for every thousand times their ad appears on a page. The price is set by market forces, and any given site is likely to have ads with different CPMs at any time. An ad that's placed on a site page with high traffic, for instance, might have a CPM of $1, while an ad that's buried in the depths of the site where few people regularly go might only have a CPM of $.25.

The problem with CPM is a problem of targeting: of those thousand people who saw that ad, how many of them will find it relevant enough to want to click through to the site it's advertising? (And from there, how many people are going to actually buy the product or service?)

In the absence of hyper-targeting -- showing ads only to the people who are more likely to find them relevant -- it's possible that those thousand impressions might generate zero clickthroughs. Advertisers want absolute, hard, verifiable numbers to prove that they're getting enough bang for their advertising buck. This means that they want to be able to see a decent rate of clickthrough, and an increase in sales, orders, or revenue on their end, in exchange for the money they've spent. Very, very few companies have a large enough advertising budget to be able to afford brand campaigns, rather than sales campaigns.

Enter CPC to combat this problem. CPC functions on a basis of "how many people have clicked the ad". You probably know CPC as Google AdWords: an advertiser on Google AdWords doesn't pay based on how many times their ad is seen; they pay based on how many times their ad is clicked. Advertisers bid for keywords based on how much they're willing to pay per click, and then only pay when someone clicks the ads.

CPC, however, is better for the advertisers, but it's more nerve-wracking for the service hosting those ads, because they aren't guaranteed a revenue stream. With CPM, you collect the revenue at the time of sale, and you know you have that revenue. With CPC, you're trusting that "if you build it, they will come" -- and that's a dangerous proposition. So generally, a service will want to function on a CPM basis -- because they get the money up front -- and an advertiser will want to function on a CPC basis -- because they only have to pay for impressions that are more likely to end in revenue for them.


It's all about the targeting

As a compromise between this fundamental tension, we get the notion of targeted ads. Ad targeting, at its core, is an attempt to take the CPM model and improve it to the point where (ideally) a service is only showing ads to people who are likely to find them useful enough to click.

A site does this by collecting as much data as possible about the person who's viewing the page, and then -- using that information -- selects an ad to display that the ad-targeting engine thinks will be appropriate. The information collected can be everything from very broad and non-personally-identifiable -- your age, your gender, your location -- to very specific and about you personally: what kind of things you like to do, what kind of things you've previously purchased, what kind of ads you've previously clicked, and what other pages you've viewed.

This is why a lot of sites use third-party ad networks: because those ad networks function on multiple sites, and therefore can build up more information about you (based on what ads you've seen on those other sites, what ads you've clicked on, and your personal browsing history). The ads you see on one internet forum might be served by the same people who serve the ads on another, so when you visit HotBustyBabes and then go visit RecipesRUs, you might get an ad that you'd be ashamed to show your mother: the ad network knows you like both naked women and cooking, and so while you're looking for something about cooking, they'll show you an ad about naked women. (Which, okay, who doesn't like naked women and cooking? But not at the same time. Kitchen accidents are painful if you're not wearing any clothes.)

Because a lot of people find this creepy, there are also a bunch of sites that build their own ad networks in-house: they contract directly with the advertisers, and build their own ad-targeting system. This involves a lot of overhead: sales people to sell the ads, engineers to build the ad-targeting system, etc.

Sometimes a site will be lucky: their userbase is targeted enough that there's a much greater chance certain advertisers will find a relevant audience. These are generally the smaller sites, for a niche audience -- a site devoted entirely to motorcycle enthusiasts, for instance, will have real luck finding advertisers. For everybody else, though, you're back to square one: advertisers still want to show their ads to people who are likely to buy their product.

Remember this. It'll be critical later.


Pageviews, Pageviews, Pageviews

If a service can't target ads, though, the other option is the "drink from the firehose" approach: accept a lower CPM from the advertisers (because each individual impression is less likely to result in a sale), but make up for it in volume. This is why MySpace is covered in horrible ads: they didn't bother trying to implement targeting well, just decided that they'll throw eighty billion ads at you on a single page. (This is why that Wired article I linked up there reveals that CPM on MySpace is so damn low. They don't care, though; they make up the loss through the number of ads they serve.)

Pageviews are also critical to the advertiser. Advertisers want to target their ads, but they also want to make sure that as many people as possible in that target group see their ads. Assume that you have two sites, both of which can provide you with equally-targeted impressions: you'd go for the site that can show your ad to 50,000 well-targeted users in a month over the site that can show your ad to 5,000 well-targed users in a month.

Unique impressions are also very valuable in Internet advertising. If you show an ad to the same user a thousand times, and they didn't click on it the first time, they're not going to click on it the 999th, either. Unique impressions are also a function of pageviews; if you have a lot of people visiting your site, you've got a greater chance of those thousand views being different people.

Under this model, then, it's critical for an ad-supported service using the CPM model to have both a broad userbase and many pageviews -- especially a service that can't provide micro-targeting of ads. Remember this, too. We'll come back to it again.


The Law of Participation Inequality

So, to pull all this together, we return to the definition of Web 2.0: the way Web 2.0 works is that the service builds a platform that people want to use, the people come to the platform and contribute content, and the service needs to find some way of capitalizing on that content to pay the bills.

The reason that Web 2.0 social media services are still turning to advertising, despite all of the problems it carries with it, is the Law of Participation Inequality: broadly formulated, it holds that user participation in any and every service follows (at minimum) a 90/9/1 distribution. 90% of your audience consumes without contributing, 9% of your audience participates in a on-and-off-again form, and 1% of your audience are the major, passionate contributors.^[?]

In many social media sites, assorted and sundry barriers to entry (bad human interface design, bad assumptions about the nature of the service, unclear value-added proposition, etc) skew these numbers even further. Jakob Nielsen, for instance, runs some numbers on Wikipedia in that read-more article above that suggest Wikipedia's curve runs more 99.8/.2/.003.

This is the second problem I was talking about up there when I was discussing subscription services.

Let's assume you run a service that has free accounts only subsidized by subscription-based accounts that provide extra services, which is a model used by many sites: many message boards, Pownce, Hiveminder, most LiveJournal clone sites, etc. (This also applies to sites like Wikipedia -- although Wikipedia is slightly different, being a 501(c)(3) -- that function on donations.) Let's also assume that subscription revenue is your only source of money -- that you've somehow managed to find a way to produce a product that people find compelling enough to pay for, despite ingrained resistance to paying for things on the Internet.

If you run on a subscription-only basis, your third-tier (smallest) subset of users are the ones who are producing the majority of your content -- and the ones who are dedicated enough to your service to be willing to pay for it. It doesn't necessarily break down that neatly -- there will be people in that 1% of passionate contributors who don't subscribe, you'll get some subscribers from the 9% of intermittent contributors, etc -- but it's still a basic function of the model: the people who contribute most passionately are the people who find enough value in your service not only to use it, but to pay you for it. And if you're lucky, they'll be producing great content that makes your site someplace that people want to visit.

But they're still your smallest possible source of revenue -- and, directly or indirectly, your largest source of cost.

One simple, common-sense rule for the bottom line of an online service as it grows over time: Costs grow exponentially.

The "older" a service gets, the more content it accumulates. The more content it accumulates, the more it costs to store and serve that content. A service that's just started up has minimal operating costs:

• Lower daily traffic, therefore less bandwidth needed
  
• Lower site load, therefore less backend hardware needed
  
• Less hardware needed to power the site, therefore lower colocation/power costs
  
• Less data to be stored, therefore less storage capability needed
  
• Company operates more efficiently, therefore fewer people/resources needed

As a service grows and ages, these advantages start dropping off, one by one. You accumulate more content, and therefore more costs, exponentially (assuming you don't ever destroy old content) -- but the subscription userbase will, at best, grow geometrically. (In reality, most services find that no matter how much their total userbase grows, their active userbase stays relatively flat, as people pass in and out.)

When you combine this with the law of participation inequality, you quickly run into a situation where -- for a purely subscription-based service -- the majority of your operating costs are resting on the shoulders of a small (sometimes fractional) percentage of users. At this point, any company with any sort of fiduciary responsibility is going to look around and say: hey, we need to capture some revenue from that 99%.

This is why sites -- especially sites with a rich pool of existing content -- still use the advertising model, even if they also offer a subscription model. The content that people have created becomes a mechanism for ad delivery. (Remember those pageviews?)

Under this model, the subscribers become an active revenue stream, paying the site directly via their subscriptions (or donations), while the viewers -- the people who are viewing the ads -- become a passive revenue stream, generating profit for the service through their viewing of the content produced by those others.


The Economy of Web 2.0

This is how we've gotten to the familiar economy of Web 2.0; this is how we've arrived back around (again) at sites being so dependent on banner/text advertising, despite the limited success of banner/text advertising during the first round of the dot.com bubble. It's actually an attempt to shift the burden of supporting the service away from resting entirely on the subscribers; the ideal theory is that passive revenue should come from the passive viewers, to enhance and support the active subscription revenue.

Advertising -- banner/text advertising -- is the method of passive-revenue collection that most sites choose, because for all the flaws in the advertising model, it's still a model that's well-understood. Advertisers understand how Internet advertising works; management understands how Internet advertising brings money in; sales and marketing people understand how to make Internet advertising as successful as possible, given its inherent limitations.

Some sites are innovating, inventing, or using new passive-revenue-collection models that aren't banner-and-text-advertising-based, but those innovations are few and far between, and nobody's really hit it out of the ballpark yet. Advertisers distrust these new ad products, because they don't have enough experience to be able to measure what a successful campaign would look like. Management distrusts these new ad products, because there's no guarantee of bringing in enough revenue to cover their costs. Sales and marketing people distrust these new ad products, because they're not yet comfortable enough to be able to guarantee a good return and a successful experience to their advertising clients.

Most Web 2.0 services will be relying on the banner/text advertising model for their passive revenue stream for a very long time, until someone manages to come up with some form of advertising that really is a quantum shift forward. For the forseeable future, Web 2.0 is tied to the advertising model as we know it.

Next, I'll explain why I think this is a time bomb waiting to explode.


[ Part one | Part two | Part three ]

Comments

[tango.livejournal.com]

Nice. I will need to read it a few times more to make sure the big things stick. But excellent writing.

[rho]

On the occasions that ads actually get past my adblocker, the thing that always strikes me is that quite a few of them are ads for other sites on the net that aren't actually selling tangible stuff. Sure, some of them are trying to flog t-shirts or holidays to Ibiza or whatever, but there always seem to be a decent number of web 2.0 sites who are advertising on other web 2.0 sites.

If that isn't a sign that something fishy is going on and that there's a bubble just waiting to burst then I don't know what is. From my perspective (very much looking in from the outside), it just looks like one great big house of cards that's ready to tumble at any time, just as soon as the investors realise that the money isn't really coming in from anywhere other than them. Again.

[ataniell93.livejournal.com]

I agree with this. I really seriously do.

[sinboy.livejournal.com]

Two terms that are essential to understanding media buying/selling online: Remnant Inventory - (inventory not sold at rate card prices)

and CPA - Cost Per Action or Cost Per Aquisition - the calculated spend on each user who clicks through actually taking the action you want them to (signing up for a service, ordering something, going to a specific page, etc...)

Also, your example of a $1.00 CPM is incredibly low. Premium sites are closer to $10 or $20 on rate cards.

Are you in the ad sales biz? This is pretty interesting commentary. Not sure if I agree 100%, but it's pretty well informed.

[truwest.livejournal.com]

Yeah, ad rates vary incredibly (I'm talking across the entire web space, not just social media alone).

There are a large number of established consumer-content-specific sites (basically functioning like online magazines/resources for readers) that get a respectable $4-8 cpm. None of these $4-8s are specifically social network/content creation sites (that I know of)...they all have a large amount of professional content, even if they have some end-user-generated content allowed somewhere on the site, to encourage engagement and stickiness (the so-called mullet strategy).

For a few really, truly primo sites/pages, with very specific, high-value audiences -- can be $100+ cpm. Of course these are not general-consumer sites...they're usually very narrow business/industry-specific sites.

On the low end, is social media. Social media sites may have the user volume but their rates are very low. Avg cpm on social networking sites is 27 cents...which means 2.7 billion page views to generate $1mm in revenue. Ouch.

Article w/ rates:

http://www.microsoftstartupzone.com/blogs/the_next_big_thing/archive/2008/07/15/advertising-rates-dropping-social-networks-lowest-cpm.aspx

[nopseud]

For a long while, I kept waiting expecting ads to disappear from the net, because unlike cable TV and newspapers it's so straightforward for consumers to get rid of them. But if the blaring ugliness of Myspace hasn't driven everyone who uses it to adblocking, then nothing will. I guess that at the end of the day, no matter how annoying and intrusive the ads are, most people just don't care enough about them to install adblocking software.

[rho]

I've wondered about this, and wondered if adblockers are going to end up being included as standard on browsers in the next round of browser wars.

[truwest.livejournal.com]

Ooh, I get to ask one of my favorite questions: what're the alternatives?

It'd be great to discuss a comparison list of pros-and-cons across the various monetization alternatives for social media sites. I'm not in love with advertising. But what *are* the viable alternatives re: revenues/biz models?

(Now I don't disagree that the current advertising biz model is flawed -- in general, not just for LJ specifically. And part of the problem is that investments are very out-of-whack per actual returns...just look at Facebook's purchase valuation vs its current lousy cpm rates. That said, advertising isn't going away; it does work very profitably for some sites as-is; the ad model needs to evolve but it's not going to disappear.)

(btw I'm in web 2.0 biz myself, altho I'm more involved careerwise in saas / b2b)

[synecdochic]

I don't have a perfect alternative for passive-revenue collection. (If I did, man, I'd be raking in the bucks.) I think that the ultimate solution is something that has to be designed for every site individually, and what works for one site won't work for another, and that everybody -- advertisers and sites alike -- is going to have to be willing to accept some short-term wobbles in order to create a better ad model in the long term.

The problem, of course, is that most sites are beholden to outside investors in some way: venture capitalists expect a completely unreasonable short-term profit. The way most sites generate that short-term profit, of course, is to sell for ridiculously inflated prices to other ownership and throw the hot potato into someone else's lap; the purchasing entity then has to recoup that investment cost somehow. That's one of the major reasons why I think we're in Bubble 2.0 and headed for a meltdown; there will be some very unhappy people when the music ends and everyone discovers the severe shortage of chairs.

[azurelunatic]

I still think a journaling service could achieve something closer to micro-targeted ads by allowing its users to self-select either individual advertisers or fine-tuned categories of ads. I may not know what would fly on the rest of the site, but I do know what I, myself, buy, and I have a good idea of what quite a few of my friends are interested in. I would be thrilled to have a tasteful "Powered by Mountain Dew" badge on my journal, because it's very true now that I'm nightshift. Tech swag, books, and chocolate wouldn't be a bad thing to see sponsored recommendations advertisements for.

But that's hardly the traditional "plaster banner ads all over it" model, and it would take rather a lot of doing to be able to do it all over the site for every user. Not everybody has the time or inclination to figure out what they want to see in ads, or what their readers want to see in ads. (While I'd be very interested in seeing, say, sex toy ads, I know that it would put off a bunch of readers, in no small part because people are cruising through at work.)

[synecdochic]

Yeah, the biggest problem with a model like that is twofold: one, advertisers don't care what ads you see on your content, because (presumably) in that model you'd be selecting ads for companies whose products you were already going to be purchasing; the purpose of advertising is to gain new customers, and there's no guarantee that your readership would be interested in their product. (Micro-targeting to the viewer is much better for advertisers, because it goes by the viewer's preferences, not the content creator's.)

Two, it goes back to the "unique pageviews" thing: they'd very quickly exhaust your readership and have to serve the same ad to the same reader multiple times.

[miggy.livejournal.com]

Fantastic post, and I'm looking forward to tomorrow's.

[synecdochic]

It seems to have turned into three parts. Alas, for my teal deer run rampant.

[admiralmemo.livejournal.com]

Great post. Looking forward to tomorrow.

All this will be moot when a replicator is invented and the only units of currency will be energy. :-D

[synecdochic]

Man, wouldn't that be nice?

[bluemeridian]

Dear God, you have me reading and commenting on an advertising essay. *sigh* If I wasn't going to Hell before... :P

Anyway, as I halfway mentioned a bit further up, I've come at advertising via selling ideas (social consciousness - namely recycling, anti-littering, etc.) and I had to argue strongly at the time for approaching it as a campaign: making it recognizable, keeping it fresh and updated, and spreading it across multiple venues (website, displays, print and radio advertising, etc.).

The psychology is relatively simple in that there are fairly strong theories on how a person processes a page of information (where to put the most important information, etc.), or who is likely to be listening to any give station, or who's going to see the ad in the university paper versus the Guide to Local Officials.

I know there is some of this out there regarding the net, but wow. There are so many variables, it's positively mind-boggling. Adding in the complications of an international consumer base just makes me whimper! The psychology of color alone...

On top of all this, "page view" counts and "visitor" estimates have always seemed one step short of imaginary numbers for me - there's an absolute number built on a framework of half a dozen guesses. As I know it, the science of gauging network traffic is so elementary that the whole thing is vaguely insane. Advertisers are taking these numbers and putting far more stock in them than they deserve, mainly because they don't understand the complications involved and because it's familiar ground from print advertising.

I'll just be over here staring at the internet in horror. Pay no attention.

[synecdochic]

Do you study behavioral economics at all? I find it fascinating; a lot of my conclusions about online advertising come out of watching the sociology of how we decide what to buy.

Dan Ariely's Predictably Irrational is a good one, and I've got Rob Walker's Buying In next on the list -- as soon as Best Beloved finishes it. *g*

One of the interesting things about online advertising -- for me, at least -- is how much of it really is call-to-action advertising and not branding. Brand campaigns are very, very rare online, because advertisers want to see hard numbers and return on investment.

[dragojustine.livejournal.com]

This is a fantastic post!

I can't wait to read the next installment. I wonder to what extent you're going to talk about the spread of adblockers? No one I know has seen too many ads on their personal machine in two years, and while I know my circle may be ahead of the population average, I've recently set my mother and grandmother up the same so the average can't be too far behind. Or is the timebomb something else?

[synecdochic]

Adblocking technology is an important part of it, yeah, but it's only a minor part -- the greatest adblocking technology ever invented is the human brain, after all.

The big reason, though, that I think advertising is underperforming on social media sites is that people just flat-out aren't there to buy things; they're there to hang out with their friends...

[solar-cat.livejournal.com]

Thank you SO MUCH for posting this. I'll confess to being woefully ignorant about most of this stuff, so all the arguments about it tend to go right over my head and I can never figure out which side I'm on. :\

So thank you for writing up such a nice, clear description of what's going on. ^^ *waits for the next part*

[synecdochic]

It took me a while to come to a lot of these conclusions, and my ultimate conclusion is sort of a collection of economics, sociology, and marketing. There's a lot of fascinating little things coming into play ...

[ljmouse.livejournal.com]

I basically agree with you, however, one thing to keep in mind here is that some advertises don't give a damn if anyone ever clicks on their ads. They're just advertising for the branding.

One of the most common advertisers on Firefox News right now is Buick, via Tribal Fusion. Tribal Fusion pays CPM. After Buick, there's a company that is promoting their solar and wind powered ideas -- they're not selling anything, the ads just lead to an informational page about their company. They very clearly just going for the name recognition with a pretty ad. I cannot imagine either ad gets many clicks ... they're purely going for name recognition and that's it.

It's no different than buying an ad in a magazine. You can't click on a magazine ad either. But the person reading the magazine ad may (if the ad is effective) remember the company later when it comes time to make a purchase.

I also suspect at least part of the reason Myspace has such a low CPM is that so many of the customers are teens. Online advertisers who are actually selling things online don't like sites that target teens; the conversion rate for sales is low.

On the flip side, I've had major advertisers come courting me for Firefox News because the reader base tends to be adult fen. Advertisers love fandom; we buy stuff.

And if online advertising didn't work, there wouldn't be people making millions from affiliate sales.

In the end, a social networking site needs multiple sources of income. Advertising, IMO, is one of them. Charging for services that can't be profitably supported by ads but which are desirable for users is another.

[ljmouse.livejournal.com]

A bit of topic, but you might find this interesting -- if you want to give yourself a headache, google Phorm.

Short explanation? ISP contracts with Phorm. Phorm inspects the packets. Phorm spots ads from your web site. Phorm swaps your ads out with the ISP's own personal ads. End user sees their ISP's ads appearing on your web site, and never knows anything is worng, and all the revenue belongs to your ISP.

Illegal? Yeah.

Hard to detect? Very.

[wyomingnot.livejournal.com]

...it's just not the same in print as sitting outside on a warm Maryland night, listening to you go on, complete with hand gestures and such. ;)

[synecdochic]

And vile imprecations. Don't forget the vile imprecations. *g*

[buddleia.livejournal.com]

Fascinating. I picked up Wikinomics recently and dropped it out of annoyed boredom and your explanation of Web 2.0 is a breath of fresh air.

[synecdochic]

Most writing on Web 2.0 makes me want to throw things at the author. *g*

[laitaine]

Excellent writing - thank you so much for laying all this out and making it easy to swallow.

[synecdochic]

Thank you! I'm trying to make this easy to understand for people who don't have the background in the industry, because by the time I get to part three, I hope to be able to explain to people just why they get so pissed off about advertising on social media. *g*

[executrix.livejournal.com]

Thank you, great essay!

But on the Web or off, the underlying problem is an economy made up in large part of businesses that think it doesn't matter if you earn a profit or how many resources you devour as long as you keep introducing new products that nobody wants and can't afford anyway--just make the advertising louder and more obtrusive!

[synecdochic]

Amen to that, man!

Have you read Gotcha Capitalism? Some really neat things in there.

[wovenindelibly.livejournal.com]

Fascinating!

[synecdochic]

Thank you!

[nearly-there.livejournal.com]

Thanks for this. I've heard many of these terms and concepts before, but I feel like I have a much better understanding of how they work together (or not, as the case may be) now.

It's interesting to think about the concept of ads in parallel with alternate models of generating revenue, like the In Rainbows experiment or the Wikipedia donations model. Is the model there still 90/9/1, where 90% of people consume the content for free, 9% give a small donation, and 1% are passionate and well-off enough to send you $50 for a downloaded album (or the equivalent)? I'm guessing that the answer is that there's really not enough data to tell yet.

[synecdochic]

I'm going to guess -- gut instinct here -- that the model's the same there. In general, it's hard to get people to open their wallets and give you money on the intertubes, no matter what you're providing them; in the end, it's going to be the people who care most passionately about the thing you're providing that give you money.

I think there's probably an additional factor playing into those models, too: people who want to demonstrate their support for the alternative model by rewarding it, financially. I'm guessing there are people who look at something like Shadow Unit and give more than they would have under the traditional publishing model, as a way to encourage the adoption of the alternate model.

[rez-lo.livejournal.com]

I agree absolutely (as another veteran of the first one) that another shakeout is coming up. Thanks for this cogent summary.

[synecdochic]

Bubbles are so damn easy to spot, aren't they? And everyone wants to get theirs before the bubble bursts.

[mecurtin.livejournal.com]

Subscription-based services behind gated walls are falling. (It works for the Wall Street Journal, for instance

I wonder if the fact that it *does* work for venture capitalists' favorite newspaper gives them a completely skewed idea about what does & doesn't work online.

[msilverstar]

I'm pretty sure Rupert Murdoch took down the subscription-only access when he bought out the WSJ. Yep, as of January 10, 2008, they changed to free access for the last 90 days. That's not the Dow Jones high-value content, and archives are 4.95 per article, but it does mean that the most high-profile successful subscription service decided it wasn't worth it.

[princessofg.livejournal.com]

thank you so much for expanding this and explaining all the jargon... i'm very much looking forward to part 2; I come to the net not from the technology side but from the mass media side, the "journalism in a capitalistic society" side, and thus I've been very interested in this whole fact that internet sites have indeed so far been "Loss Leaders" for Old Media companies, and wondering how that will play out.

Only a very few businesses have been able to make money off the content side by selling services or products that people will actually pay extra for.

I'm sure you read the Wired feature on Apple called "Evil Genius." Vertical integration is doomed... and people will pay for certain kinds of downloads, but not all....
Thanks for the thinky. Looking forward to more...

[synecdochic]

Behavioral economics is a fascinating, fascinating science. In short: people are weird, people's buying patterns are weird, and unless you grok sociology and psychology, you're never going to make it in the advertising industry ...

[dunv-i.livejournal.com]

... that was very educating.

I have to say, while I don't like them, I've always been of the 'whatever' camp when it comes to ads because it strikes me as something that a company has to do to support themselves (no different from my mom as much as doubling the photocopying for the programs for her concert so that she can promote the one next month, right?), but this is making me think differently, perhaps. While, in my mom's case, it probably does actually help, I'm very intrigued by your essay and am awaiting tomorrow.

Perhaps not quite as related, but I'd love to hear your opinion on this:

There's a company called QTrax (and it's better-established sister SpiralFrog). They both offer free music. (QTrax is getting a lot of flak for not offering stuff it says it is, but it's really because it's not set up fully.) It's supported entirely by ads - whenever you update a page, you see a new ad (incidentally, it's a little unwieldly sometimes trying to find music, because it needs to show you those ads on each page. The ads are well targeted, actually, since they're promoting tickets to concerts and bands' new CDs, etc. Does this strike you as the sort of thing to work in the long run?

[synecdochic]

Something like that, because the service itself is very targeted, is probably more likely to work out okay in the long run: advertisers are going to see high return on investment on those sites, because the people viewing that site have common interests and are there for a specific purpose. If you've got a targeted niche market to deliver advertising to, you're going to be able to charge higher rates for your ads, and advertisers are going to find their ad campaigns more successful.

[imkalena.livejournal.com]

tl;dr. I wish I wasn't in the middle of a family crisis; I always appreciate your concise and smart analysis. I miss fandom and the incredible people in it -- like you.

[synecdochic]

I hope the crisis gets better!

[msilverstar]

This is really clear and interesting, as a person who spent quite a lot of time on The Well in the 90s, it never did make real money. And VC expectations of reward are non-trivial, they seem like they'd rather have the companies risk it all in trying for the big wins than grow slowly and steadily.

Have you thought about affiliate marketing as an alternative income source? I don't know enough about the business to know if it's possible.

[synecdochic]

The problem with affiliate marketing is that there's a limited pool of dollars to chase, and you still have to worry about pageviews et al. Check out Web 2.0 Economics 101 for a quick overview of some of the problems.

[crucially.livejournal.com]

There is going to be a larger need for brand advertisement.

[synecdochic]

As time goes on, yeah, absolutely. It's going to take advertisers being willing to redefine the 'success' of an ad campaign, though, and I think that's a pretty hard sell. Advertisers want to see return on investment, and if a service can't provide them hard numbers leading to success, they're going to back away...