Testing Offers 101: From Red to Green

by Josh Todd on June 29, 2010

One of the toughest ongoing questions for affiliate marketers is, “How do I know when I’ve spent enough money testing an offer?”  Everybody has a different approach. There are a couple of truths that we need to look at.

  1. It is very rare to find a campaign that is profitable right out of the gate. I’m not saying it doesn’t happen, because it does, but it is very rare. In most cases you will need to test, optimize, test, optimize, then test some more before you really get it dialed in and profiting enough to make it worthwhile.
  2. Some campaigns will not work no matter how much testing and optimizing you do. There is nothing more frustrating than beating a dead horse.

Sound like a Catch 22? It all boils down to your tolerance for pain, or your willingness to lose (or I like to say “invest”) money in a campaign before it sees a profit. How do you test properly without blowing too much money, and still avoid dropping potentially profitable campaigns too early? Here’s what I like to do…

First of all, when you setup a campaign, you want to spend about 2x to 3x the payout of the offer for the initial test. If the offer payout is $5, then you want to spend approximately $10 – $15 on the initial test phase. This is the “gathering data” phase.

Stats

Green is good

Once you have done this, you will have some data to work with (assuming you’ve been tracking everything, with Tracking202 or a similar program). Here’s the type of questions you want to ask in order to evaluate your data:

Did it get any conversions at all? If so, what keywords / urls / referrers / ad copies / banners were they from? Did you make a profit? Would it be profitable if you bid on less keywords? Would it be profitable if you bid on more keywords for a cheaper price? The list goes on, but these questions should give you a start.

For me personally, after the initial testing phase, if I haven’t received any conversions at all, I usually dump the campaign. That’s -100% ROI, and that’s going to be hard to overcome without a drastically different strategy, and it’s probably not worth it. If you are anywhere from -50% ROI to 0% ROI, then you might be able to profit with optimization. Try using a landing page or direct linking (whichever you didn’t do the first time), try different ad copies, pictures, headlines, calls to action, etc. Use the data from the test to see what is working and what is not working.

However you got those conversions in the testing phase, focus on that precise set of circumstances and recreate them in greater volume. Once you understand how and why you got some conversions, you are well on your way to having an optimized and profitable campaign.

Now go make some money!

{ 9 comments… read them below or add one }

Profit Addiction June 28, 2010 at 6:42 pm

The one problem I used to have was getting overly excited about conversions during the initial test. Generally speaking, it was 1-2 conversions and you think you might have something to work with. However, many times the conversion(s) were a fluke and that showed after letting the campaign run.

Just because you see SOME action, don't unload your budget.

Nice post Josh 🙂

Reply

Free Classifieds Blo June 29, 2010 at 12:22 am

Great post. Had I read it before blowing $100 on POF in a fat loss clickbank product. I think it is a good rule of the thumb. No more than 3x the payout.

Is this 3x should be the daily spend? and spread over a couple of campaigns targeting to narrow down the demographics (if I am in Facebook or POF).

Reply

Josh Todd June 29, 2010 at 12:35 am

That depends on how quickly you want to test the offer. Also it depends how much it is. If it's an email submit and the payout is $1.50, then the 3x spend is $4.50, which you could do in an hour. But if it's a $20 offer, then $60 you can either do in one day, or spread it out over a week and do $10 a day. It's really up to you and how much money you have to play with. There is something to be said for testing out an offer on more than one day of the week though, especially weekend vs. weekday.

Reply

browie June 29, 2010 at 1:10 am

I always have a hard time letting things go longer. I like to shut it down way to early. I see the 2x or 3x and it's over. But I bet if I just tweaked some things it would work instead of giving up.

Reply

LC June 29, 2010 at 2:04 am

Fantastic post. When I test I won't go past the 2x.

Some media buy companies should read this. When it's way past the Red – there is no coming back. Stop saying it will work. It's NOT going too.

Reply

Justin Dupre June 29, 2010 at 3:34 pm

You're missing a lot of potentially profitable campaigns. If all you are spending is $20.

Sure if you are on a budget, but if you want real testing and data. Throw $100 at it and split test ads and landers. Sure if its an email submit, and you spend $10 with no conversions, cut it out, but if you spend $20 and get one $4 conversion, cut out the poor shit, and spend another $20.

Reply

Josh Todd June 29, 2010 at 11:45 pm

I absolutely agree. I'm not saying that if it isn't profiting in your first $20 to cut it, what I meant what is you have spent 2x to 3x the payout and haven't received a single conversion, then maybe you should cut it. If you have got a conversion or two, then it's time to look at the data and clean up your targets and spend some more on testing.

Reply

Ryan Eagle July 3, 2010 at 2:02 am

Justin I agree it takes money to make money. One must spend more than 20 dollars to be successful.

Reply

Curt July 12, 2010 at 11:21 am

I tend to have the opposite problem than browie. I ususally let them run way to long before I kill a campaign. It seems like I get just enough conversions to keep me hanging in there and cutting bad targets and optimizing good ones, all the while I am slowly bleeding money. Thanks for the tips Josh.

Reply

Leave a Comment

{ 3 trackbacks }

Previous post:

Next post: