Testing Ideas with Ads.. & WHEN to call it a success or failure.

[niftyg]

I recently stumbled upon a technique that I can't believe I missed or didn't think about (zen*******'s Testing Video).

I have a few ideas that would be perfect to be tested this way (as this is the way I would go and get sales) so I go ahead and piece together some ads and a landing page when I realize....

How do I know if an idea is a success or a failure so quickly? Facebook, POF and Google ads, in my limited experience, take LOTS of optimizing to get them to be working right, as does a landing page.

To anybody who has used an ad to gain interest, how much time do you spend optimizing your ad and lander, and WHEN do you call the idea a success or failure?

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Tom.V]

My suggestion would be to pick a single traffic source to start. Then spend based upon your returns, for example, if I were testing a new product/service with paid traffic, I'd run something like this.

30 different ads/banners ->3-4 different landing pages -> 1 checkout page(optimized of course, probably will need more testing but for demonstrations, this works)

On the high end, that's 30*4*1=120 different combinations. Now let's figure out how much you are making per CONVERSION. If you anticipate COGS(Cost of Goods Sold) to be $6 (including shipping, manufacturing, etc.) and you are selling the product for $20, you're looking at a gross profit of $14. Generally, for each conversion you want to spend double to triple that number for testing purposes, just to figure out which ad/lander/offer page combos are the best and worst. So let's take that number, 14(conversion profit)*120(ad combos)*3(testing variable)=$5,040?! This is casting a wide net, and generally with this many different variables, you WILL more than likely find something that works and prove the idea a success, or shun it to the depths of hell and move on. You can do a smaller testing net, and run something like 10 ads*2 landers and just gauge how many people actually click through from the lander to "estimate" your conversion rates on the end of the funnel. So that would be 14(CR Profit)*20(ad combos)*3(testing variable)=$840. Costs less, but not quite as accurate and leaves room for unknowns. Now figure out your COGS+CAC(Customer Acquisition Cost) and subtract it from your sales price ($20) and figure out your ROI.

Either way, you can spend a lot, or a little, and get a general idea of whether or not the idea is viable.

BTW, math may be a little fuzzy, correct me if I goofed anywhere.