As a trusted search engine, a premier advertising medium, and a lucrative investment, Google become one of the most influential companies to ever grace our planet. Google’s reach has grown so substantially over the past decade that they’ve even inspired societal transformations, not to mention their name has literally become the modern eponym for “search.” Yet in spite of this ubiquitous like influence, most individuals that affiliate regularly with Google fall into one of three groups: searchers, advertisers, and/or stockholders.
Although these three groups are broadly defined, each is important because of the unique roles they play in Google’s “digital ecosystem,” commonly referred to as: Google.com. I find the word “ecosystem” to be a good description of Google.com because of the symbiotic relationships formed as a result of its presence. Allow me to elaborate so that this very important concept is clearly understood.
Each of the three groups mentioned above have different expectations resulting from their affiliation with Google. Searchers expect Google to deliver search results when they perform a search. Advertisers expect Google to show their ads to searchers based on predefined criteria. And lastly, stockholders expect their investment in Google to provide some kind of future return. Stated another way, advertisers need searchers to search, searchers need advertisers to show up in the results, and investors need both to engage with one another.
These interdependencies create several unique challenges for Google, the biggest being the development and maintenance of a system that successfully balances the interests of all three parties mentioned above. A slightly more descriptive summary of Google’s paid search model exemplifies this point very well:
- Searchers expect Google to show advertisements that are both high quality and relevant to their individual search queries.
- Advertisers expect Google to show their ads to highly targeted searchers that have a statistically higher probability of becoming their customers.
- Stockholders expect Google to maximize the value of each transaction that occurs between advertisers and searchers so that their investment continues to grow.
From the descriptions above, it’s easy to see that one party’s unique expectation has the potential to easily conflict with another party’s unique expectation because of the differences inherent in each parties’ goals. This is where Google’s role is really made manifest in the ecosystem. Google has successfully created a system that protects each parties’ interests by actively promoting equitable transactions for each unique search performed on their search engine. They’ve done this by creating a new “quasi-equity type equalizer” variable they call “quality score.” This proprietary score, and its underlying algorithms, evaluates how appropriate it is for an advertiser to participate in an ad auction, and at what level, for any given search. Google represents quality score to advertisers on a 1 to 10 scale, with 1 being the absolute lowest possible score and 10 being the highest.
On its website, Google defines quality score as the usefulness and relevance of an ad, a keyword, and a landing page to a user’s search query. More is known, but in the end, Google keeps its descriptions of quality score pretty vague in an effort to protect the interests of its owners’. For this reason, it often helps marketers to think of what quality score represents rather than focusing solely on its definition. Consider an analogy with the stock market.
A stock listed on a stock exchange creates value in two ways: (1) cash now in the form of dividends; (2) cash later in the form of stock price increases due to the underlying company’s growth over time. These two attributes help investors to decide what price they’re willing to pay, and therefore bid, for any given stock available. This is similar to the consideration Google gives each bidder every time there’s an Adwords auction.
Google looks at advertiser bids like investors look at dividends. Like dividends, these keyword bids result in payments to Google now rather than in the future. On the other hand, Google looks at an advertiser’s quality score like investors look at the growth potential in a company’s stock price over time. Quality score is therefore a representation of the value Google estimates a bidder, campaign, ad group, or keyword to contribute to its own growth over time.
As you can imagine, Google assigns high quality scores to advertisers that provide “high quality” experiences to the individuals using their search engine. Google’s intent is to promote the long-term use of their search engine through continued usage from their users, which also supplements future revenues expected from paid clicks, and a marketplace that continues to be highly competitive. Together, an advertiser’s quality score and keyword bid combine to form Ad Rank, or going back to our stock market analogy, an advertiser’s “stock price.”
As a search advertiser, Ad Rank is one of the most important concepts to fully wrap your head around. Google uses Ad Rank in each of its ad auctions to do two things:
- Assign ad positions in its sponsored results to participating advertisers
- Determine the cost-per-click of each participating advertiser in the event an ad is clicked
It’s important to not that an advertiser’s ad is positioned in decreasing order of Ad Rank, while the cost-per-click is determined by this formula:
Bidder One’s CPC = Bidder Two’s Ad Rank / Bidder One’s Quality Score
There are two real takeaways here. The first is that an advertiser’s Ad Rank ultimately determines an advertiser’s position for any given ad auction. And roughly speaking, the better an advertiser’s Ad Rank, the better their position in the sponsored search results. The second key takeaway is that an advertiser’s shortcoming in quality score has to be made up by the cost-per-click they’re charged for the position their ad placed in. Because quality score is believed to be a base number and not an exponent, the following table can be constructed and used to illustrate the additional charges an advertiser incurs when a keyword scores below a 10 in an ad auction relative to what the charges would have been had they scored a 10 instead.
Quality Score CPC Increase
The stated cost differences in table above are astounding, but the real costs are likely even greater than what the advertiser initially realizes. Consider the two primary options available to an advertiser that has a keyword with a quality score that’s actively increasing:
- The advertiser can maintain their ad position for the keyword by decreasing their bid, resulting in a lower cost-per-click and a higher profit margin.
- The advertiser can increase their ad position by maintaining their bid for the keyword, resulting in additional clicks from searchers.
The first point is self-evident for the majority of people reading this article, so I’ll turn my attention to the second point. Research has shown that search engine click-through-rates have a tendency to follow a Zipf-like distribution, with ad position on the x-axis and click-through-rate on the y-axis. This means that an increase in an advertiser’s ad position generally results in an even greater increase in click-through-rate. This type of relationship results in either “exponential growth” or “exponential decay” for the advertiser, depending on the direction he’s moving. This is enough of an understanding to now define both costs incurred by an advertiser plagued with a poor keyword quality score:
- A linear decrease in a keyword’s profitability.
- An exponential decrease in an ad’s click-through-rate.
Both of these effects combined together spell disaster for an Adwords campaign. Why? Receiving fewer clicks at a higher rate generally leads advertisers into a self-perpetuating problem they cannot work their way out of. Once caught in this devastating cycle, advertisers sometimes have no choice but to decrease their keyword CPCs in an effort to maintain a profit, but this also lowers their ad positions even further as a result. At the same time, these same advertisers also have a tendency to start vigorously searching for additional keywords they can add to their campaign in an attempt to either substitute or supplement existing keywords.
In my experience, an advertiser’s initial keyword selection almost always outperforms those keywords added to a campaign later. What’s worse is that these new keywords then have the potential to create additional quality score issues for everything else at the campaign or ad group level.
So how does an advertiser resolve this problem once they’re in it, or even better, how do they avoid it altogether? The statistics team at Netmark recently spent some time studying this issue and found some interesting correlations. I’ve provided 6 of the variables they examined in their study, and their associated correlations to keyword quality score, separated into two groups below:
1. Click-Through-Rate: 0.4002
2. Bounce Rate: -0.4576
3. Pages-Per-Visit: 0.5928
4. Average Visitor Duration: 0.4393
1. Maximum CPC (Keyword Bid): -0.5458
2. Average CPC: -0.5535
Although Netmark’s statistics team could not determine causation in their study, it’s believed that group one is made up of variables more likely to influence a keyword’s quality score, while group two is made up of variables more likely to be influenced by a keyword’s quality score. These results provide keen insight into the constituents of a keyword’s quality score and their effects. From the data above, it appears that an advertiser’s best strategy is to be relevant to a searcher’s query and to provide a good user experience. We would be wise as advertisers to learn from these results. What do you think?