Ian Fernando is spending his money so we don't have to.

Ian did a really interesting test this weekend on a topic near and dear to me... halloween costumes He tested out pushing costumes via Facebook ads and he learned something. Cost per sale (CPS) is a really tough way to make money on Facebook.

Even after an impressive array of split tests and targeting tweaks, he couldn't make it pay. He spent about $200 to make $16. A failed campaign by any standard.

In my mind, some of this failure has to do with Facebook in general. Some of this has to do with the costume space. Let's look at both.

Halloween Costume Traffic is Tricky

Everybody is looking for costume ideas right now, and while there are plenty of people buying costumes, most people are just looking for a great idea on what to be for halloween. This translates into lots of lookers, which is the kiss of death for paid traffic.

In my business, we've essentially been forced to stop buying costume traffic this year. Between affiliates figuring this out for themselves the hard way, and the incredible influx of costume merchants driving up prices, the margins are gone.

This year, more than ever, if you have to buy traffic, you'll be making Google, the merchants, and even American Express, more money than yourself. But hey there's always Membership Rewards points.

Facebook is for Branding

For the most part, people on Facebook are not shopping. They're interacting.

It is simply too easy to get people to click on Facebook with a very targeted ad. Ian showed a Dexter costume to Dexter fans. What Dexter fan with Halloween on the brain isn't going to go take a look at a Dexter costume? Especially if you show them a small thumbnail with the promise of seeing it in full detail behind a click. Even with a high conversion due to the targeting, the lookers will click away any profit that you might have had.

There are some things that Ian might have done to dissuade the lookers. One commenter suggested that he put a price in the ad. Good idea. I'd also suggest changing the language to, "Buy a Dexter Costume" instead of "Come take a look at the officially licensed costume". But even with these changes, or any changes, it's too much to overcome.

Ian can also do MUCH better than 10% for costumes, but even if he can squeeze twice that out of a merchant, it's not enough. His current EPC is off by more than 10x for profitability.

It takes an incredibly high EPC to make money with CPS on Facebook. The offers exist, but they're not low-ticket items like halloween costumes.

There is a branding benefit at play here, but Ian is paying for the merchant to gain it. This increases the chances that the visitor will eventually purchase from that merchant.

The good news is that Ian is getting a tracking cookie set for each of the visitors, and there will be some longer-term lift to his numbers if he had the luxury of waiting for the return visits (most halloween merchants have 30 or more return days). But even though the branding brought the buyer, he still may not get the sale...

It's the Economy

Add the economic situation onto all this. More than in recent years people are either making their own costumes or are shopping around for coupons or the best deal they can get. So even if Ian delivered a buyer to the merchant he was promoting. The chances that buyer will be lost to another affiliate from deal and coupon hunting is higher than ever.

Thanks for taking one for the team Ian!