We say it all the time, “your website is important for Google ads.” We run landing page tests and suggest copy tweaks for keyword density. And as PPC specialists, we know Google doesn’t place ads on bid alone. But over the next few paragraphs, we’re going to show exactly why the Google auction considers Quality Score and Ad Rank. And we’re going to show why most people aren’t paying nearly enough attention to their paid media clients’ websites. 

Bear with us. This is gonna stray beyond marketing for a while. We’re going to explore some of the game theory behind auctions, and it can get a little dense, but we tried to make it simple and understandable. 

 

Table Of Contents

  • The TL:DR
  • Game Theory In 360 Characters
  • Modeling An Ads Auction
  • Auction Equilibrium
    • Breaking Down Second-Price, Sealed-Bid Auction Equilibrium
  • Modeling The Google Ads Auction
  • Thinking Like The Auction Algorithm
  • A Good Website Is More Important Than Your Bid
  • Summary

 

The TL;DR

Quality Score and Ad Rank level the playing field by accounting for the qualitative attributes of a good website like content relevance and user experience. They protect the everyday paid search marketer from facing Google Ads auctions in which big-budget companies dominate. 

Google has continued to call for relevancy and content, and our findings reaffirm that content is king. Even in the world of paid search, with these metrics baked into its auction algorithm, Google prioritizes advertisers with content-rich, relevant websites over big-budget competitors. 

Now let’s get nerdy.

Game Theory In 360 Characters

Game theory models strategic interactions among decision makers. In game theory, interactions are modeled as a “game” in which players make a decision — at the same time or in turns — based on an expected payoff.

We can think of the Google auction as a game. In that case, we’re the players, and the payoff is an impression, which can lead to revenue or leads.

Modeling An Ads Auction

If the Ads auction is a game, then we can model it. To understand an auction, we want to answer the questions: 

  • what is the best bid I can make?”
  • And, “If I make that bid, will I win or lose the auction?

The Google auction is a second-price, sealed-bid auction. That means two things:

  • A second-price auction means that you do not pay your bid. You pay the bid of the second-highest bidder.
    • In google’s case, you pay the second highest price plus one cent.
  • The sealed-bid component simply means that you cannot see what other bidders are bidding. 

As we answer our questions, we need to consider all of our strategies as well as all of our opponents’ strategies. Which, in an auction, boils down to four options:

  • We can bid our exact value for an item
  • We can refuse to bid
  • We can bid above our value
  • We can bid below our value

Auction Equilibrium

The best outcome for an auction is when nobody wants to change their bid. When that happens, the auction is in “equilibrium.” Let’s dive into that a little more. 

Equilibrium is a state in which nobody gets a higher payoff from changing their move, assuming your opponent plays the same move. For an auction like Google’s, equilibrium happens when everybody bids their exact value for a product. 

Breaking Down Second-Price, Sealed-Bid Auction Equilibrium

As an example, say you are bidding on an item you feel is worth $500, and that’s your absolute maximum price. You should bid $500.

But say you bid $550, which is above your maximum value. With that, a couple things could happen. 

  • If there is a bid higher than $550, you lose regardless. 
  • If the 2nd highest bid is lower than $500 ($300, for example), you pay $300 and win anyway. You didn’t need to bid above your max price. 
  • Finally, if the second price is higher than your max bid, but lower than your $550 bid ($525, for example), you lose $25, the difference between your $500 value and the second highest bid.

Similarly, bidding your value is also a better strategy than biding under your max price.

  • In this case, if you bid $300, someone can still bid above your $500 value. You would have lost either way. 
  • If you bid $300, and someone bids under your valuation, say $250, you would have still won had you bid your true value. 
  • And if somebody bids between your $300 bid and your true value of $500, say $400, you risk losing the auction when you could have won.

That’s a lot, but in the context of the Google auction, it means that the best strategy for advertisers is to bid their clients’ true value for a keyword. 

Modeling The Google Ads Auction

What does “bidding your true value” mean in PPC? Knowing our clients’ true value for a keyword is easy to think about, but hard to do. That’s one thing automated bidding strategies solve, right? It’s complicated, time-consuming, and honestly really, really hard to get the bid right every time. But let’s take a step back and revisit the auction from the perspective of a single impression. 

This is a thought experiment. So for simplicity’s sake, let’s ignore automated bidding. Let’s also assume all of our advertisers are running manual cpc, and that, in general, ad rank, conversion rate, and quality score are comparable across all advertisers. The only thing that varies are business-side factors like margin and product cost.

In this pruned-down scenario, we can see how Google Ads’ skeleton is the same as the second-price, sealed-bid auction. Advertisers are sealing their bids and paying the second-highest price (plus one cent).  

If our best strategy is to bid our exact value, we need to calculate it. It’s easy to bid on a physical good; we know how much we’re willing to pay. When it comes to a click, it’s trickier.

That said, with the old, profit-maximizing formula marginal cost = marginal revenue,  we can find our target cost per click. In our case, if we sub-in our client’s expected profit for marginal revenue. And we can use cost per click as marginal cost. At this value, we can expect a quantity of impressions with a proportional conversion volume that will maximize our profit.

In that case, your target cost per click should equal your expected profit per click, which is a function of click through rate, conversion rate, and your client’s product margins. When CPC = Profit Per Click, expected payoff is maximized. 

We recognize that this is a huge simplification, and it doesn’t directly consider the feedback loop of changes to CTR and CVR that may come with a change in bid. But it tees-up an important discussion.

 

Thinking Like The Auction Algorithm

If we all bid our true value, we are all better off in a simple auction. Now let’s apply that thinking further; let’s introduce some actual Google auction complexities. 

In the auction, we can think of each advertiser as a player with a set of specific moves (remember: bid our value, bid up, bid down, or don’t bid).

In the following example, advertisers can bid up or down by 30%. They can bid their true value, or they can simply not bid. Because this is a second-price auction, payoff will vary depending on if a player wins or loses. If they win, they earn the difference between their true value and their actual bid. This can be negative if they bid higher than their value.

Our expected profit per click equals our target cost per click, so players will always lose a value equal to their cost per click when they lose the auction. Note that this does not mean they spend money; they just forego possible profit.

We model auctions with a table in which every player’s move has an associated payoff. In our example, the auction will be a 4×4 table. 

Here is how that table looks for two players, Agency 1 and Agency 2 with respective target bids of $5 and $6. The payoffs are represented like this: (payoff for agency 1, payoff for agency 2).

 

Agency 2
Agency 1 Bid Up Bid Down Do Not Bid Bid Target
Bid Up (-5,-.5) (.8,-6) (-1.5,-6) (-1,-6)
Bid Down (-5,2.5) (-5,2.5) (1.5,-6) (-5,2.5)
Do Not Bid (-5,-1.8) (-5,.8) (-5,-6) (-5,0)
Bid Target (-5,1) (.8,-6) (0,-6) (-5,1)

 

It’s not as complicated as it looks.We want to find equilibrium outcomes. To do that, we can go through each move and consider the expected payoff of playing that move for all of your opponent’s possible opposing moves.. 

We call this concept dominance. A strictly dominant strategy always has a higher payoff than any other strategy, no matter what your opponent does. A strictly dominated strategy always has a lower payoff. Looking at our auction, we can eliminate the moves that are strictly dominated, because neither of the players will ever make them if they can help it. 

For each player, there is one move that is strictly dominated.

 

Agency 2
Agency 1 Bid Up Bid Down Do Not Bid Bid Target
Bid Up (-5,-.5) (.8,-6) (-1.5,6) (-1,-6)
Bid Down (-5,2.5) (-5,2.5) (1.5,-6) (-5,2.5)
Do Not Bid (-5,-1.8) (-5,.8) (-5,-6) (-5,0)
Bid Target (-5,1) (.8,-6) (0,-6) (-5,1)

 

“Do Not Bid” always provides the lowest payoff, which makes sense. If you expect to make money from a placement, you want to rank for it. You want to bid. If we cut that option, each player really has three moves.

 

Agency 2
Agency 1 Bid Up Bid Down Bid Target
Bid Up (-5,-.5) (.8,-6) (-1,-6)
Bid Down (-5,2.5) (-5,2.5) (-5,2.5)
Bid Target (-5,1) (.8,-6) (-5,1)

 

Remember that equilibrium is where every player has made their best move, given their opponent’s move. When possible, a player will choose the move with the highest payoff, but if two or more moves have the same payoff, a player is considered indifferent between them.

To find equilibrium, we can go through each row and each column and answer the question, “What would you do if you knew your opponent’s move?”

For example, if Agency 2 knew that Agency 1 were going to bid up, they could look at their moves and choose the one with the highest payoff. Agency 2 would therefore bid up, because it earns a payoff of -5, as opposed to -6 for all other options. 

If we repeat that process, we get the different equilibria for this auction! In this case, there are four equilibrium outcomes, in which our agencies will have made the best possible moves. 

Agency 2
Agency 1 Bid Up Bid Down Bid Target
Bid Up (-5,-.5) (.8,-6) (-1,-6)
Bid Down (-5,2.5) (-5,2.5) (-5,2.5)
Bid Target (-5,1) (.8,-6) (-5,1)

 

Take a look at those for a second. Notice that Agency 2’s best strategy is to bid up, and that Agency 2 always has a higher payoff than Agency 1. Also note that Agency 1 will almost always lose. In this auction, the Agency with the higher bid dominates.

There’s a clear answer to why that is, and it comes back to our target bids.

In our scenario, target bids were set where CPC = Profit Per Click. That means more profitable companies can afford to spend more per click. Agency 2 was more profitable. They had a CPC of $6, whereas Agency 1 had a CPC of $5. 

That suggests that high-budget competitors should sweep the floor with other advertisers by outbidding them every time. 

 

A Good Website Is More Important Than Your Bid

Bids don’t dominate because, much like in it’s search algorithm, Google has designed the Ads auction to support website quality & content over everything else. 

That is precisely why Google doesn’t consider bids alone. If it did, CPCs would be astronomical. Advertisers with bigger margins encroach on the SERPs with big-budget campaigns that crowd-out the rest of us. 

 Instead, as we know, the auction is designed around a mixture of ad rank and quality score. And that is very intentional. It opens the door for small-to-midsize businesses to succeed on the platform. 

By considering metrics like expected CTR, landing page experience, and ad relevance, the auction shifts the payoffs in our table. It balances the outcomes and favors those whose website is more relevant and offers users a better experience with good content.

 

Summary

Google has said it time and time again to SEOs, but it’s time we PPCs started listening. Good content and UX are important. The fact that the Google Ads auction is adjusted to consider these factors only reaffirms this. 

If we don’t look at a website’s holistic content as part of our paid search efforts, we might just be abandoning one of our best strategies for cutting costs, increasing conversion rates, and driving consistent new customer acquisition.

The truth of the matter is that no marketing happens in a vacuum, especially not on search. Creating all-around solid, authoritative content should be a focus for all search marketers. Its spillover benefits will set the stage for a long-term, profit maximizing strategy that transcends paid and organic. 

With that in mind, it’s your move.