Does Manual Bidding Actually Work?

Question

Hey Jon, it's Nate. I have a campaign or a couple campaigns that are working well and kind of my theory in my head, and I'd like your feedback on it, could you just set a cost cap? You know, like, Hey, I'll pay 13 bucks for a customer. And just set your campaign budget to like a ton, right? So if it has a good day, it can scale big. And if it has a bad day, then it just keeps the costs low. Thanks.

Answer

Okay, Nate, so this is a question that’s been asked by many creative advertisers over the years.

Advertisers have three main bid control options if they don’t want to let Meta optimize for volume or value of conversions: Cost Per Results Goal, Bid Cap, and Return On Ad Spend Goal. The main thing is advertisers are trying to control costs or value.

Here’s the best way to look at how a Cost Per Result Goal or Bid Cap, in particular could change things.

By default, Meta is trying to get you the most results within your budget. So if your performance goal is to maximize the number of conversions where the conversion event is a purchase, Meta will try to get you the most purchases.

But the key is that the secondary goal is to spend your budget. So while Meta is trying to get you the cheapest purchases possible, they’re also trying to spend your entire budget while doing so.

As a result, for a given day of ad spend, there’s a mixture of expected costs to get a conversion from a given person. One of those factors is simply the competition for that person.

Of course, Meta can’t guarantee a purchase out of anyone, and that’s an important key here. Meta can show your ads to the people most likely to convert, but ultimately Meta can’t force anyone to buy.

So even if you set a Cost Per Result Goal of $10, Meta can’t guarantee you’ll be able to accomplish that.

A good way to illustrate what might happen is if you are optimizing for link clicks, rather than a conversion.

Meta’s already trying to get you the lowest cost per link click by default, but they’re also trying to spend your entire budget. You can get the cheapest link clicks from the Audience Network placement, and we all know that these tend to be low quality.

If you’re spending $1,000 per day, Meta’s going to spend a decent chunk of your budget in Audience Network, but there’s some limitation in inventory. So you will still have your ads shown elsewhere.

Let’s say that you’re getting an average cost per click of 10 cents in Audience Network, so you set a Cost Per Result Goal of 10 cents.

The likely impact of that is your ads may only be shown in Audience Network to try and maintain that 10 cents cost. And if Meta can’t get you into enough auctions from that placement, you may struggle to reach your budget.

But that’s because you’ve eliminated the requirement of Meta needing to reach your budget.

Now while Meta can predictably get cheap clicks from Audience Network, cheap purchases aren’t as simple.

Meta will try to keep your Cost Per Purchase around your goal, and part of that may be avoiding highly competitive auctions. But ultimately, Meta is bidding in the auction for your ad to be shown, and you’re going to pay for those impressions regardless.

So if you’re unable to maintain that goal, you’ll likely get delivery throttled.

Admittedly, I haven’t used manual bidding in years, but trying to outsmart the algorithm with manual bidding isn’t a new idea. It usually fails.

I’d worry about the quality of results if not optimizing for a purchase. It seems most viable when optimizing for a purchase.

But even then, it would likely lead to very inconsistent delivery and results.

Is it worth an experiment? Absolutely.

But I wouldn’t expect miracles.

Thanks for your question, Nate!