Optimizing Your Facebook Ads Too Often
Most Facebook advertisers are constantly touching their ads (or “turning knobs” as I like to call it) thinking that constantly changing bids, budgets, and targets is equivalent to making faster progress. What actually happens is that each optimization resets your ad rank and puts your ad right smack back to the start of the learning phase.
This is the equivalent of planting a seed, digging it up after 5 minutes to see how much it’s grown, replanting it, and then digging it up again 5 minutes later.
I’ve heard this lament in almost every one of the last dozen meetings we’ve had with Facebook.
In a similar vein, making multiple ads doesn’t necessarily give you more chances to win. You want to get to 50 conversions per ad set per week. If you’re frittering your budget across dozens of ads, there isn’t one single ad with enough power (a strong conversion signal that overcomes statistical noise) to produce a winner.
That means you need to target larger audiences by pooling custom audiences and relying upon lookalike audiences of the custom audiences that convert.
I realize this is 180 degrees from what I preached 5 years ago. But today, the algorithm is smarter and the cost of traffic is much higher. Additionally, there are more ad placements on Facebook than we can manually optimize. You want Facebook to choose the mix for you based on your setting for your actual cost per lead and cost per conversion.
Facebook summarizes these techniques in something they call the Power 5. The way we think about it is that you build three-stage funnels of Awareness, Consideration, and Conversion. Then find your evergreen winners at each stage and keep putting more money into them.