We only make profit from advertisement if the overall money we are paying for ads is less than the profit we make from the sales of our products. 

Cost per action otherwise known as CPA is the total Amount spent on ad divided by the total number of result obtained.

For instance in ad set 1 above I spent N2857.99 to get 645 link clicks. Therefore cost per link click =2857.99/645 = N4.43.

But how do we reduce cost per result consciously? 

These are the methods we use to reduce cost per result.

  • Prepare to run ads for a long time without editing any of the parameters. Prepare a reasonable amount of money for your ads and run ads for a little longer period before editing any of the parameter. Because the moment you edit the ad parameters like audience, biding ad creative etc the learning process starts clocking all over and during learning period the cost per result is always very high.

Ensure there are no grammatical or spelling error before publishing your ads so your error would not be detected by your audience. Because if they start commenting on your error you will be tempted to edit the ad and publish it again. This will lead to the learning process clocking all over again and eventually increase your cost per result at the initial stage.

When you publish an ad the cost per result is usually high because Facebook algorithm is still learning to determine the right people to serve your ads to. Similarly if you edit any parameter of the ad and republish  it the learning process starts all over again which otherwise increases your cost per result. 

The more frequent you edit and republish your ad the more the the learning period and the more the cost per result. Do your best to limit the number of times you edit and republish your ads.

  • Let the size of your audience be considerably large enough so as to avoid changing target audience frequently which will invariably increase the learning period. 
  • Publish a high quality ad creatives and landing page.
  • Use hidden audience targeting if possible.
  • Use a lookalike audience closer to purchase if you don't have enough data from purchase. eg lookalike audience of lead, View content, add to cart, initiate checkout etc.
  • Ensure you have adequate data in your custom audience like 50 or 100 before you create the lookalike audience this way the algorithm will be able to match it correctly and get it right faster.