Mike Rhodes Teaches How to Harness the Power of AI With the Google Ad Profit Curve

Google Ad Profit Curve with Mike Rhodes

The Go To Guy for Google Ads is Mike Rhodes of Websavvy, a digital advertising agency. He co-authored the definitive book on the topic with Perry Marshall: Ultimate Guide to Google Adwords.

Ralph Burns, from the fantastic Perpetual Traffic podcast, which delivers the best Facebook advertising information, interviewed Rhodes about his latest thinking on Google Ads for Episode 194.

Rhodes focused on what he calls “the Google Ad Profit Curve,” which boils down to using math to determine what cost per action (CPA) or return on ad spend (ROAS) you should be shooting for.

Google Ads Profit Curve - Mike Rhodes

If you increase your ROAS too high, you significantly reduce the pool of potential buyers that Google’s artificial intelligence (AI) and machine learning (ML) can show your ads to. So, you might hit that ROAS target, but the total customers will be much lower than you could potentially get with a lower ROAS target. 

Thus, total profit drops.

Example Math To Illustrate the Google Ad Profit Curve

Say your break even is $100. If CPA is $40, you make $60 per sale, but the number of sales you get is capped at 100. So total profit is $6,000. Whereas, if you bump up your target CPA to $60, so that you make $40 per sale, but the machine can now find 175 sales, then you’re making $7,000 profit in the same time frame.

More total profit is the name of the game.

There’s a sweet spot, and Rhodes suggests having a target range rather than exact number, to go after.

Try different targets. Test it. Find the range. Then choose the higher number.

Also, you have to remember that increasing the reach of your ads has a lot of side benefits aside from increasing the number of sales.

Increase Speed of Tests and Machine Learning
You have a lot more data to work with, so you can get to statistical significance much quicker on any split tests you’re running. Google’s learning curve for your offer will be much shorter, and it will hone in on the attributes and data points of people that will most likely purchase.

Google Gives Preference to Shops With Higher Cost Per Acquisition (CPA) Targets
Aside from getting your split testing done quicker with more traffic available at a lower ROAS or CPA target, you will also win more Google auctions if you’re willing to pay $75 per lead versus your competitor, who is only willing to spend, say,  $55 per lead.

On the other side of the curve, if you lower your ROAS target too low, you start to lose money on every sale, and you can’t make that up in volume.

Once you wrap your mind around it, it’s a pretty basic concept, but if you’re not thinking in these terms, your energies can all get focused on getting the highest ROAS, while missing the idea of maximizing profit.

“The thing you want to optimize for is profit. It’s not cost per click. It’s not even cost per sale. Because you have to do the math. And the math is very simple really. At a certain CPA level or cost per sale, you’re going to get a certain number of sales.”  
- Mike Rhodes

Advertisers often arbitrarily pick their target cost per sale. If break even is $100, they’ll just say it would be great to get them for $40 without having any math to back them up.

So many advertisers will say “can you get my CPA lower.” But, that’s not the right thing to focus on.

Thoughts on Scaling Your Business With More Ad Spend

Ralph said: When you’re scaling you have to think about your business differently. As budgets rise, traffic will get less and less profitable. So, you can’t scale to infinity, nor can you expect to have the same ROAS at $250K spend as you will with $1 million spend.

Mike added: The game is to know the math. If you know the lifetime value of average customer, you can even lose money on the first sale. This is the fastest way to grow a business, but if you push it too far you’re losing too much money on every sale, and that’s not sustainable.

13x is so much better than 7x on its face, but you have to ask yourself how many sales can you do at 13x? Is there a better target?

Ralph: You need a portfolio of products, so you can make more money  on the backend and afford to spend more money to acquire a customer.

Mike: Dan Kennedy says that one is the most dangerous number to any business. One channel. One product. One key staff member. One big customer. Too much exposure if you lose that one thing.

Make sure you explore ancillary products. 

“If you have one product you don’t really have a business. You have a promotion. You have an offer. You don’t have a real business.”

Bonus Material on Boosting Your Google Shopping

Feed management is the most underrated or ignored. 

The Google Merchant Center (GMC)  is a halfway house for your data. The best way to improve the results you get from shopping campaigns is to better match what you name the products (product title) with how people are searching for products.

The title is at least 80% of the game, and GMC now allows you to tweak and test it.

And for $30/month Datafeed Watch, available in the Shopify App Store, is a great  platform to see and tweak your products. Make them more descriptive. Match them better to what people are searching.

Image Optimization
Remove the background from product images. There is lots of data that says a product image with a white background outsells an image with cluttered background. Make your images nice and clean. 

Pricing
A big thing on Google is price. 5% cheaper will get way more exposure for your feed. But, be careful. It’s dangerous because cutting price can trigger a race to the bottom. 

  • 2019-03-31
>