Programmatic has historically been used for display advertising and defined as using software to buy clicks from the advertising platform (i.e. Google, Bing, Facebook, etc.). It is about putting the right message in front of the right buyer at the right time; finding that moment increases relevance for the buyer and increases efficiency for the advertiser. That very same approach works at least as well for Search and Shopping campaigns.
In principle the objective is to profile the person the ad is targeted for and use that information to increase efficiency from the advertising spend. In this blog I will focus on how to use Programmatic as a way to drive efficiency for Search campaigns.
Quality Score is the black box of Google, they use it to decide which ads to place where on the Search Engine Results Page (SERP). In general, Google will place the ad that gives them the probability of making the most money on top. There are two main components that drive Ad Rank:
- Max Cost per Click (CPC) (~50%)
- Quality Score (~40%)
Ad Rank is critical because in general the higher the ad position the higher the CTR - which is critical for capturing more traffic - and because the CTR is the most dominant component of how Quality Score is calculated by Google.
So, how do we influence these aggressively and at scale? How you organize your account becomes fundamental in how this can be done. There are a few ingredients that needs to be in place to enable scaling this.
The Exact Match click is the golden one. It has a higher conversion rate (because of higher relevance) and it has a higher CTR. If you capture each Phrase, Broad and Broad+ search string that triggers a conversion and create exact match keywords from them you will over time enter more click auctions with an Exact match keyword vs the other match types. The only objective of the non-Exact match keywords is to farm out Exact match keywords.
Single keyword ad groups are key to scaling this program. The first advantage is that ads are tied to one keyword and A/B testing becomes absolute (vs. having ads serve multiple keywords). This also enables you to align ad copy directly with the keyword to further improve CTR. The second advantage is an ever bigger one, if done right you can force Google to pick the Ad Group vs the keyword.
The benefit of this is forcing the match type separation (i.e. if you tell Google you want a Broad+ keyword, they will still serve up a mix of match types based on their assessment of relevancy etc. Because the conversion rate is so different between the match types it is key to separate to enable accurate bidding.
In addition, all the bid modifiers are available at the ad group level which enables building a very detailed unique click profile for each auction you enter. The campaign building and maintenance, including building a unique click profile for each auction you enter can be done with Programmatic technology. Keep thinking relevance, getting the right ad in front of the right customer at the right time.
This unique click profile enables laser accurate bidding because you have organized the account in a certain way that isolates all the variables that influence conversion rates and revenue generation. With a unique click profile you can predict with a great degree of accuracy what the conversion, revenues or profit from that next click will be.
The last component that needs to be carefully considered is the attribution modeling you chose. People view conversion data as actual revenues, profit or conversions, it is not. No tracking tool will capture everything, so as a general rule there is always more of the outcome than we is captured. Furthermore, what we are looking for are relationships between ad spend and the desired outcome (conversions, revenues, profit) that can be replicated. The more assertive attribution model the more relationships can we replicated. Think of attribution modeling more as a strategic and competitive decision vs. philosophical decision of who “deserves” credit for the outcome.
Now, what do Exact match, single keyword ad groups, a unique click profile and attribution modeling have to do with influencing Quality Score? A lot! Two things drive Ad Rank, Max CPC and Quality Score. The above is a hyper effective way of building granular click profiles runtime for each auction you enter and you will show up armed to the teeth knowing which the clicks that best drives your business objective are… and bid very aggressively on those because you have the data to support it (minimizing risk). The objective is to drive the Max CPC as high as possible, but inside your cost of revenue or CPA target.
Why would I want to maximize my Max CPCs? Because there is a co-dependent mechanism in play. The higher the Max CPC, the higher the Ad rank, the higher the ad rank the higher the CTR. The higher the CTR the higher the Quality Score. The higher the QS the lower the cost for Ad Rank (relatively).
Here is an illustration of how this works. We call it the Wheel of Google Goodness:
It is a cycle that gets started that will propel the performance in an account, but it takes all of these components to be executed in parallel. Finch is executing our client's advertising programs with this methodology executed fully programmatic. Does it work? Here is a screenshot from a Google custom report from a customer we took over in December 2015:
There is a 358% increase in Exact Match conversions while the other match types stay somewhat constant while farming out new exact match keywords. The obvious question is “what happened to the cost per conversion?,” here is a graph showing exactly that:
As you can see, the cost stayed almost constant with a 358% increase in volume. This is what happens when you execute the above strategy with software and put all the leverage behind it.
Summary: Our industry will go through massive changes over the next few years as Programmatic technologies makes inroads. The above is an example of what is being done today to create huge competitive advantages for those who adapt this next wave of technology.