Join or Die: It's Time To Embrace Google Automation

*Shortly after this article was published, we were invited to speak about this topic in greater detail at Google's offices in NYC. A summary of those topics can be found here.

--

‍

Automation? It’s just Google’s way of getting you to spend more money!

I still hear this at least once a week and it’s nails on a chalkboard for me. I also used to believe it. But times have changed.

This is not your father’s Google AdWords. In fact, it’s not even your own Google AdWords that you knew and loved just the other day. That Google AdWords is so long gone that they’ve even changed the name of the platform.

My peers in the PPC industry love to complain about every update and change to the Google Ads landscape. Above all, they are most reluctant to embrace Smart Bidding. Smart Bidding, often referred to as Google’s Automated Bidding Strategies, is too often associated with the thought that It’s just a way for Google to get you to spend more money.

So I turn my attention to you, fellow Google Ads account managers:

Too many of you are stuck in your outdated opinions about Google automation. If you do not embrace automation and dedicate the time to understanding how to best leverage machine learning in your campaigns, 2019 is going to be an ugly year for your business.

Join… Or die.

‍

A New Era of PPC Capabilities

The Pre-Smart-Bidding era was a simpler time. You could train a monkey and/or write a script to do most Google AdWords optimizations for you, and most clients could easily understand the actual work that you were doing on a day-to-day basis.

However, this era came with many limitations. It was much more challenging to profitably scale a campaign beyond a core group of bottom-funnel keywords. Put simply, the platform technology was not advanced enough to give advertisers the right advantages.

That era is behind us.

You are no longer a Google AdWords account manager. You are an algorithm engineer. Your primary responsibility is to have a deep understanding of how platform algorithms work and how to manipulate complex data sets to guide the machine to help achieve your goals.

Today, your job is much more challenging, but there is so much more potential to profitably scale your campaigns. This is great news… for some of us.

But most of you are either too lazy to learn a new skill or you are too afraid to rock the boat.

Let’s take a step back

All advertising campaigns can be classified as either Contextual or Behavioral targeting.

Contextual Advertising is an ad placement that is targeted based on the context of the medium. Geico heavily invests in billboards because people who see billboards are driving cars and likely have or need car insurance.

Similarly, Rolex believes that their wealthy target market is likely to also be interested in golf, and so viewers of any professional golf event are bombarded with tons of Rolex Watch commercials.

I am not currently in the market for new car insurance, but I still see a lot of Geico ads every day. I also watch a lot of golf, but I am definitely not Rolex’s target customer. Contextual targeting is what traditional advertising was built upon, but obviously has its flaws.

Behavioral Advertising is an ad placement that has less to do with the medium, and instead, relies on customer behavior to pick and choose the timing, placement, and audience of an ad.

L.L. Bean uses Behavioral Targeting when they send their catalog to all previous customers, regardless of their physical location. Your Facebook and Instagram newsfeeds are filled with behavior-based ads, depending on whether or not you are a fan of romance novels, training for a marathon, shopping for a new apartment, or any mixture of thousands of other categories.

Manual CPC Google Search advertising primarily relies on Contextual Targeting. The context is the search query that the user searched on Google. Despite all your bid adjustments, match type segmentation, and ad copy tests, you are still limited to the actual query typed into Google.

Limitations of Manual CPC

Imagine you are a licensed retailer for the New Orleans Saints football team. I’d bet that your conversion rate is through the roof when a user’s search term contains something very long tail and specific, such as:

Buy Drew Brees Jersey Online

Unfortunately, you find that about 1% of people in the market for a Drew Brees jersey are actually searching for something that specific.

You realize that you need to bid on broader terms if you want your business to grow, and so you build out middle-funnel and top-funnel keywords that would capture more users that might be interested in buying a Drew Brees jersey:

• Drew Brees Jersey
  ‍
• New Orleans Saints Jersey
  ‍
• Saints Merchandise
  ‍
• NFL Jerseys
  ‍
• NFL Merchandise
  ‍
• Football Clothing
  ‍
• Gifts for a Football Fan

‍

As the list grew, it became more and more apparent that conversion rates for these new keywords would be an issue. Even with NFL Jerseys, it was unlikely that you’d ever convert enough users to find profit.

You had an idea to limit these new broad searches to the greater New Orleans area, but even then, you were still limiting your ability to scale and reach Saints fans around the country.

You then realized that your only chance of survival was to become hyper-focused on your Actual CPC, Quality Score, and your website conversion rate.

No matter how hard you tried, you were never able to make this campaign profitably grow. You were forced to operate within the limits of bottom-funnel contextual keyword targeting.

Smart Bidding Leverages Behavioral Targeting

Under the right management, a Smart Bidding strategy would likely allow you to profitably target at least some of those top-funnel keywords listed above.

Here’s why…

If you're exclusively selling New Orleans Saints merchandise, you'd normally never want to bid on the term NFL Jerseys in the New York City area because most New Yorkers are going to be interested in buying Buffalo Bills jerseys.

That was a joke.

Obviously, there are several disadvantages that would limit your ability to bid on the term NFL Jerseys. Most major markets are dominated by fans of the local team, but also, not a single New Orleans Saints jersey cracked the top 10 in national jersey sales this past year.

Therefore, given that there are 32 NFL teams, there is less than a 1 in 32 chance that the user performing the search for NFL Jerseys is even in the market for the type of product you are offering. Traditional Manual CPC / Contextual Keyword Targeting is never going to work.

Enter Google Machine Learning.

Google knows that Shirley was recently searching for flights from NYC to New Orleans, that she checked the distance between her Bourbon Street hotel and the New Orleans Superdome on Google Maps, that she downloaded a New Orleans restaurant app on her Android phone, and that has watched that clip of Drew Brees throwing footballs to his kids 12 times on YouTube.

Google also knows that searchers that are traveling across state lines for sporting events are most likely to buy at least $100 worth of merchandise in the weeks leading up to their trip. I actually don't know if this is true, but Google would.

Google knows better than you that this is an incredibly valuable search to bid on. If I were you, I would welcome the opportunity to pay just about anything to get Shirley to click on my ad when she searches for Football Merchandise.

Similarly, if someone was searching for Gifts for a Football Fan, and was listed on the same IP Address that Shirley typically uses, I would also be willing to bid aggressively for that click.

Putting Smart Bidding Into Practice – Some Things to Know

Your CPCs Will Go Up

And it’s not that big of a deal. It’s just supply and demand. The highest quality traffic (supply) is the most valuable to advertisers, and advertisers (you and your competitors) are therefore going to express more demand in that traffic.

You need to stop thinking of your CPC as the be-all-end-all and see the forest through the trees. When executed correctly, Smart Bidding will increase your conversion rate.

Conversion Rate Is More Important

If your CPC goes up by X%, and your conversion rate goes up by anything more than X%, your campaigns will be more profitable.

Here’s an example:

‍

Client: We’re happy with our current CPA of $75, which is a huge improvement from where we were last year. However, our CPCs are up 15% year-over-year. I can’t help but think that we are overpaying for clicks and that our CPA could still be much lower if we brought down the CPC.

AMG: I agree that there are likely still opportunities to lower our CPA, but it’s crucial to consider the fact that Conversion Rate is up 25% year-over-year. This is the primary reason why we are more profitable, despite an increase in CPC.

Client: But obviously our CPA would be lower with a lower CPC…

AMG: Not necessarily. Our conversion rate is up in part because we are spending more of the budget, and more per click, on higher quality traffic that is more likely to convert. We pay a premium for the best traffic, and so long as conversion rate increases at a greater rate than CPC, overall profitability will continue to increase.

It’s also likely that if our cost per click was much lower, we would only be acquiring the low-quality scraps that were left after our competitors scooped up all the premium traffic. If that was the case, our conversion rate may have plummeted, and our CPA would be higher than it was last year.

Think of it this way – If our conversion rate was 100%, we could pay up to $74.99 for a single ad click and be more profitable than we are now.

Client: That makes sense, but are you trying to say that CPC doesn’t matter and that there’s nothing we can do about it?

AMG: Of course not. There’s always room for improvement, but I wouldn’t recommend we optimize just for a lower CPC when a higher CPC might result in better performance overall.

Client: OK that’s fair. Can we talk about literally anything else now?

Automation Does Not Mean Hands Off

Your campaigns are not in coast mode just because you are no longer manually setting bids.

The individual tasks that you conduct will change, but the amount of overall effort required will not.

Here’s what will not change:

• Understanding your client’s goals.
  ‍
• Competitive analysis and industry research.
  ‍
• Attribution modeling and working with your clients to understand the true profitability of their advertising efforts and lifetime value of their customers.
  ‍
• Emphasis on Search Term analysis. In fact, this is going to be more important than ever as you begin to target broader top-funnel keywords. A negative keyword should be viewed as another data-point for Google's algorithms to use as part of their optimization.
  ‍
• Testing new ad copy and landing pages. Running usability tests and heatmap analysis.
  
  

Here’s what will change:

• You’ll spend more time managing remarketing lists. Not just for quality users, but also of low-quality users that you want to exclude from future targeting. Google’s algorithms will not understand that your blog traffic converts at a lower rate than average unless you set up your campaigns to collect that data.
  ‍
• You will place more emphasis on Similar Audiences and In-Market audiences in search.
  ‍
• You’ll realize that YouTube ads are the most valuable (i.e. cost effective) ad placement you can currently buy. Most advertisers are slow to embrace YouTube, and so with an unsaturated market, the cost-per-YouTube placement is a fraction of what it should be.
  ‍
• You’ll spend more time developing Custom-Intent YouTube Audiences – a targeting method which marries Google Search with YouTube placements.
  ‍
• You’ll spend more time developing video content, obviously.
  ‍
• You will spend a ton of time analyzing data sets with the goal of predicting dramatic swings in demand and conversion rate. Google is introducing more features that allow account managers to guide the machine and add more valuable data points into the algorithm, or proactively warn the system that a change in the market is about to take place. We're not allowed to go into detail about any of these features just yet, but they're coming soon and they will be awesome.

• You’ll spend more time thinking about which information you should be feeding into Google, and how your data strategy should be changing over time. You’ll run tests to see which conversion actions you should be tracking in Google Ads and therefore optimizing for.
  ‍
  ‍

For example, let’s say your New Orleans Saints Retail Shop averages $100/order, and so your primary conversion is Sales. However, you also want to keep track of how many ad clicks result in Email-Newsletter Signups.

If you don’t manage your settings properly, Google will optimize for both Sales and Email Signups with equal weight. A Target CPA bid strategy could potentially be a disaster.

Putting It All Together

Through its use of Behavioral Targeting, the Google Ads Smart Bidding strategies clearly have an advantage over Manual CPC Contextual Keyword Targeting. Smart Bidding will allow you to scale your campaigns in ways that were previously not possible.

If you are not interested in scaling your campaigns, your performance will still suffer over the next year. The quality of your Manual CPC traffic will continue to decline as your competitors leverage Smart Bidding and can afford to pay a premium for the highest quality traffic.

It’s time to stop making excuses. Change your old habits and to embrace this new reality.

‍

For more information on this topic, check out our follow up post, Our Q&A at Google: Automation and the Future of Ad Agencies.