Blog Tags: 

Website monetization primer: how online advertising works

Money. A tricky subject to bring up. It's been proven that merely activating the concept will induce selfishness. But bear with me here. This is a public service announcement, not a get rich quick scheme.

If you have a website that gets a reasonable amount of traffic, you've probably considered adding advertising at one point or another. Maybe you're dreaming of quiting your day job to work full time on the site or maybe you just want to cover some basic costs.

Though TurnKey has been primarily a labour of love since the project was launched in 2008, we realized a couple of years in that achieving some kind of financial sustainability would be key in ensuring the project survived and even thrived in the long run. We also wanted to be able to bring more people on board as full-time developers, but obviously the project would only be able to do that after we figured out how to solve the funding problem.

That got us looking seriously into advertising, but unfortunately when we started looking we didn't really understand how advertising on the Internet actually worked. It took a bit of research to tease out the essentials. For what it's worth, we decided in the end that advertising wasn't the way to go. But still perhaps someone else will find these notes useful.

Best types of advertising

By order of expected conversion rate and ultimate value per thousand visitors:

  1. Inhouse tie-in products and services

  • targeting: you know your audience. You can engage them naturally. You don't have to distract them from what they are looking for.
  • higher conversion: your audience knows you. They're much more likely to trust you than some random third party vendor they've never heard about.
  • accumulated relatioships: this way you own the relationship with the customer and can continually up-sell new premium services throughout the lifetime of these relationships. When successful this ultimately captures much more of the lifetime value of the relationship.

  • up-front costs: before you make the first bit of revenue you have to pay for the expense (and risk) of product development. Sometimes that doesn't work out and you won't get a good return on investment. Sometimes you can't even try because creating a good, competitive product for your target audience is out reach.

  • no overpaying: advertisers sometimes waste money overpaying. This isn't economically sustainable in the long run but sometimes it will temporarily push up advertising prices

  1. Targeted affiliate deals

    Pros: high payout if it works and the ability to switch offerings quickly if it doesn't until you find an offering that your audience is interested in.

    Cons: you bear all the risk of it not working out and you still don't own the relationship with the customer.

  2. Direct advertising sales

    Pros: roughly 10X pay-outs compared with ad networks.

    Cons: opportunity risk, you have to be able to set a realistic price point for ads, potentially labor intensive to negotiate. Third party services that help manage and streamline direct sales cost money ($50-$100 for our level of traffic).

    Rates (by CPM):

    • $3-$5 for most sites on buysellads
    • $15 for TechCrunch

    Implementation alternatives:

    • hosted service (e.g., buysellads, isocket, trafficspaces)
    • self-hosted - open X ad server. Also gives you access to a marketplace. It's a lot of work to setup though.
  3. Ad networks (e.g., Carbon Ads, AdBard)

    Pros: attractive and unobtrusive ads, little to no fuss, low risk (e.g., guaranteed some level of payment)

    Cons: low rates ($1-$3 CPM)

Alternatives with unknown payouts

The following are options where the payouts are unknown:

  • Adsense

    Pros: can be combined with direct sales, so it's only used to sell the "remnant" impressions you haven't sold directly, easy to to implement. For a few rare sites AdSense works very well ($20-$60 CPM), but $1-$5 effective CPM seems to be the norm.

    Cons: ugly, payout in proportion to obtrusiveness, sometimes works very poorly (e.g., $0.1 CPM)

    Somewhat ironically AdSense seems to work best on pages where the user is already on the way out because what he reached isn't what he's looking for. If the user actually is interested in your site, they're less likely to click on an ad and leave.

Ads come in standard sizes

One of the first things you should realize before you begin adding third-party advertising into your site's web design is that advertisements come in standard sizes.

The IAB (Internet Advertising Bureau) publishes a list of recommended sizes. The idea is to limit the number of creatives that advertisers need to create and publishers need to support.

The advertiser's perspective

Advertising is dominated by inefficiency and waste. This effects prices. As one famous advertiser said: "I know half my advertising budget is a complete waste. The problem is I don't know which half".

Sales oriented advertising (short-term)

Sales oriented advertising is that it is tightly focused on getting measurable results.

You have something to sell so you try to buy a sliver of the attention of the target audience hoping that you will be able to persuade some of them to become your customers.

The objective is to acquire targeted traffic and directly convert that into sales. Every step in the funnel is tracked, measured and optimized:

CPM => CPC => CPA => CAC

CPM: cost per thousand

CPC: cost per click

CPA: cost per action

CAC: customer acquisition cost

The cheaper the advertiser can acquire new customers the more money they make. If they spend too much (e.g., by misjudging the lifetime value of a customer) they may actually acquire customers at a loss. I've noticed you can often tell when a field is saturated by VC funded startups when the bid prices for clicks in AdSense are so high it's very unlikely they make any economic sense.

A lot of effort is spent trying to home in with increasing effectiveness on the target audience. This involves some educated guesses but also a lot of trial and error.

Branding oriented advertising (long-term)

Successful branding creates positive mental associations with the brand that help everything from sales to hiring. Advertising for branding purposes introduces additional considerations besides immediate sales.

Results are much harder to measure in the short term so you mainly have to go on theory and gut feeling. But there is no mistaking the power and value of a strong brand (e.g., Coca Cola, Nike, Apple).

If an advertiser believes is a branding benefit to "sponsoring" a web site, they may be willing to pay more for the perceived branding benefits.

The publisher's perspective

Publishers have inventory, usually measured in CPM (thousands of impressions). They're trying to maximize how much they get for it while at the same time they have to avoid putting off their audience, which come for the content, not the commercials.

I think users don't hate advertising so much as they hate ugly, poorly implemented irrelevant advertising that distracts them from their goals. The ideal ad would be relevant content (e.g., a review) from a trusted source that just happens to recommend a product or service a user is in the market for.

Poorly implemented advertising damages the user experience and that will erode the publisher's brand. An eroding brand won't get as many link backs and tweets backs which may impact the long term growth of the publisher's audience.

The best mutually beneficial economical relationship between publishers and advertisers is to match advertising for high value products and services with the target audience as closely as possible in a way that minimizes the impact on the user experience.

Matching up advertising to content will never be perfect until computers can understand context and read minds (shudder), but the closer you get, the better it works out for both sides.


Add new comment