Banner Exchanges Unmasked

High Exchange Ratio

Commission is the lifeblood of any exchange network. If a network operates honestly, the only source of income it has is the sale of commissions. When considering which network to join, you should always remember to compare the rates offered with their affordability for the exchange owner. Only sign up if the figures meet your expectations and don't look contradictive.

But let's start with the evaluation technique. No prominent exchange can ever be run efficiently (or for long) by fewer than 4 people: an administrator, a moderator/support person, a programmer and a promo/development staff member. And each of these people expects to be paid at the very least a competitive rate for their hard work. As a potential advertiser, you need to find out the price for placing commercial ads with that network, and evaluate the network’s profits against its operation costs.

A word of caution here -- the Internet spreads not only across national borders, but also across different economies. The quality of IT expertise does not consistently correlate with the expense to business of employing those skills; therefore, the cost of what are apparently comparable services to customers can vary widely between countries and service providers. Look at it this way: a high-quality IT professional might be able to earn $10,000 per month in one location, yet another person with the same skills and expertise might be content, if not happy, with a monthly wage of $500 in another country.

Thus, it's a very difficult task to evaluate the value of commissions against the business’s expense figures. Yet there is a rule of a thumb -- no exchange, wherever it might be based, can survive without receiving at least 20% commission at a decent reach.

Note that it's not always an easy task to derive the actual commission rate an exchange delivers, as it can vary depending upon different factors (CTR being the most popular one). Only the relative traffic volumes of differently "taxed" groups of sites can actually help you identify the true rate.

If it’s not possible to estimate the actual (as opposed to advertised) commission rate of a particular exchange, your best bet is to consider a different network. If, on the other hand, you can estimate the real commission, but the figure looks too lucrative to be true, there are two possible explanations:

1. The project is on budget and does not need to break even for the time being. This is most likely the case if another profitable business is behind the exchange. Such cases are usually identified as such quite plainly on the site as "limited time offers", or "promotional offers" etc.

2. The exchange breaks even at the expense of its participants by selling more than it has a right to. This means the actual exchange rates differ from those declared on the site.

For example, if the network only shows 800,000 banners daily (24 million per month) and needs to sell 10 million just to survive, it means that the actual exchange rate for its members is less than 2 to 3. Members may receive 3 credits for each 4 banners shown, but some of them are just accumulated on accounts and cannot be spent. The longer things continue like that, the less likely it is that these credits will ever be useable at all.

Targeting

Another feature on the advertiser’s “most wanted” list, and rightfully so, is targeting for banner exchange members. The general attitude is "the more options, the better". But more options come at a price…

The root of the problem lies in the fact that every member contributes untargeted traffic to the network, but every campaign that’s run within the network targets a particular audience.

Let's assume, for example, that we have a group of Canadian sites in a network oriented heavily towards a Canadian audience. In reality, less than 80% of the combined audience for these sites will comprise Canadians. The other 80% will arrive at the site from all over the world. Assuming the GEO targeting option is available to every exchange member (and each has it set to “Canadians only”), if the sites receive aggregate daily traffic of 100,000 users, 20,000 of these users represent traffic that members produce but do not utilize.

Now let’s add the network itself, which has a sales department that’s also based in Canada, and assume it offers a meager commission of 20% (4 to 5 ratio). It should be obvious that this share is most likely of interest to Canadian companies who want to reach Canadian audience. The network needs to sell the entire lot just to survive. This cuts the total members pool down to 80,000, one quarter of which is 'unwanted' traffic that’s used by neither network nor its members.

The situation is similar to one described earlier: the members receive 4 credits for each 5 banners shown, but only 3 of these credits can be spent without violating the targeting settings; the remaining credits become a dead weight on accounts.

There are 3 ways out of such a situation:

1. Increase the network commission (thus lowering the exchange rate), but do not sell the extras. Instead, use them for balancing purpose only. In our example, this would mean that the official rate goes down to 3 for 5, but the network sells only half the commission and the rest are used to show some general ads (usually promoting the exchange itself) to the untargeted audience.

2. Charge an extra commission for the use of targeting. This is a derivative of the above solution that differentiates between members on the basis of their campaign settings. So, if a member site wants Canadian traffic only, that site gets it at 3 for 4 instead of basic 4 for 5, which provides the network with balancing capabilities. Those sites that are ready to accept any traffic get (and more importantly, spend) 4 for 5, while helping the network balance ads and audiences.

3. Force network members to show some untargeted banners, and provide free targeting option for all other ads. Returning to our example, 3 banners out of the earned 4 are shown to Canadians, while the fourth is forcefully shown whenever a banner request comes from a foreign visitor.

Other kinds of targeting work in almost the same way. The only notable exception is frequency capping, which calls for a separate database (and quite a "weighty" one, if the solution is implemented properly). In addition to aforementioned impacts on a network, this option also requires extra investments into memory, disc space, or even a separate server.