By Adam Ward

There are hundreds of affiliate networks (which are like ad agencies or exchanges for online ads) on which you can run your ads. Because the barriers to entry are so low for online publishers (often referred to as “affiliates”), it makes sense for publishers to sign up on many networks. After all, they usually don’t have to pay to sign up, the more networks they sign up for, the more advertisers (at least in theory) they’ll be exposed to, and financially, they aren’t out anything for being on multiple networks.

For advertisers, though, it is a different story. Running the same ads on multiple networks can actually be counterproductive, and costly, because if you’re not careful you could end up paying double, triple or more for each sale or lead.

The problem isn’t from overexposure. Rather, it’s just a downside inherent to what makes online advertising so desirable in the first place: the ability to track leads and sales. With traditional advertising you can benefit from overexposure. By placing your ad on TV, the radio, billboards, etc. you get your message out to a wide target. You hope to drive traffic from those ads, but you are essentially branding your product or company. With online ads, your approach is more one-on-one targeting. A single customer sees your ad, clicks through it, buys the product, and only then do you pay for that ad. The plus side of this is you don’t pay for the exposure you might be getting if people don’t buy. The downside is you either penalize publishers who can’t claim sole responsibility for sending you the lead, or you end up paying multiple publishers as though they were the only ones sending you the lead.

Here’s how it works. Let’s say you run a single ad (the same one) on three different networks. Each network places the ad on one of their publishers. The networks also have tracking software to keep track of who clicks through the ads on their publisher sites, and whether it resulted in a sale or lead. If while your customer is researching products he or she clicks through your ad on all three publisher sites, then ultimately buys your product (either right after clicking through the ads, or comes back directly to your site a month or two later), all three of those networks log the sale. Their publishers expect to get paid the full commission for referring that customer. And they have the tracking numbers to back it up. But for you, that means you have to pay full commission to three separate publishers, while you get the revenue from just one sale. You can see the dilemma.

This is why consultants (called outsourced program managers, or OPMs) try to convince their advertising clients to choose a single network and stay on it. Although that reduces your payout risk, it could also limit the publishers you might be exposed to. And to a certain extent, limiting your ads to publishers on a single network might be like running ads just on TV and ignoring print and outdoor.

Besides, from a branding standpoint, it might be a good thing to get your message on as many sites as possible. And if multiple publishers helped to get you a sale, shouldn’t they be compensated for that? The kicker is, how much should they get compensated, and how do you track that? Right now there isn’t good technology for sorting this issue out. Advertisers who join multiple networks usually have to manage this manually to a certain extent. They have to clearly spell out in their terms and conditions, what they will pay, and how. They also have to negotiate that the tracking statistics they’ll go off will be their own (which means they have to create some scripts on their own servers that can track where the leads came from, and when). Networks are hesitant to do that because they have their own tracking software and they like to be the one holding the cards when advertisers and publishers dispute tracking numbers.

One advertiser, CSN Stores, has come up with a tiered way of paying for sales generated by two different networks. They have a way of distinguishing the publisher who referred the lead first (and rewarding them at a higher payout) and the publisher who referred the lead last (and rewarding them at a lower payout). You can see that they have to actively manage this approach to work, and do a lot of manual work. But they recognized that there were publishers on one network that would not sign up with their preferred network.

Maybe someday there will be tracking software sophisticated enough to account for multiple referrers, make sure the referrals were legitimate, and spread the payout from a single sale to those multiple referrers. That seems to be the equitable way to do it.

Editor’s Note: Since I originally wrote this post, I’ve discovered TagMan, which purports to do exactly what I mentioned in the last paragraph. I haven’t seen it in action or talked to anyone using it, but I thought I should mention it here.