The “Grid” Method of Improving Profit
Now we can manage via the grid method
Let’s take an example. Suppose we have a campaign “template” that we run in 100 major cities. And the goal is to figure out which cities are performing, which ones are not, and what to do with ones in the middle. The Y-axis could be the whole campaign together or just one particular ad (template) that we customize for each city. Or a keyword template. The X-axis would be the markets; the 100 cities.
In each “cell” (market/template intersection) let’s calculate the total impressions and the ASI for a period (say last 30 days). This grid method of reporting, with the “duet” of impressions and ASI, tells you everything you need to know about how a particular element or technique works across markets. Or, how a particular market performs across different ad templates. As long as you have a set of cells with the ASI/Impressions duet.
The impressions count shows you the “potential” that the element (keyword or campaign or ad) is creating and the ASI tells you how it is “farming” that potential. Scale and effectiveness. At a micro or macro level. That is the key to this ASI/Impressions teamwork.
Simple rules that drive the right actions
With a little experience, you can develop your ads action plan based on these “grid-type” results.
So, within a particular row or column, the performance of each “cell” of impressions-ASI data can be classified into four quadrants. The impressions are compared (high or low) on a relative basis (such as comparing ads for the same products) and the ASI is evaluated on an absolute basis relative to your goal or overall ASI performance.
This yields a table of results like this:
|High Impressions – High ASI||Expand, exploit|
|High Impressions – Low ASI||Spend less; improve ASI|
|Low Impressions – High ASI||Spend more; expand impressions|
|Low Impressions – Low ASI||Turn off, delete|
The real magic of the grid type reporting is that you can make decisions with perfect perspective. How does that ad template work across markets? How does it work relative to other ad templates in the same markets? You can quickly see what is generating potential and what is happening with that potential.
Use the CampaignBuilder to create and improve the cells in your “grid”. Fix (or turn off) the cells that don’t work. In no time you’ll be bidding and budgeting more successfully and getting rid of clicks that just don’t help make you profit. And watching your overall ASI get much better.
Notes on the “Cellular Method”
If you think of a campaign as a set of coordinates (Settings, Locations, Keywords, Ads/Extensions), then you can understand what we mean by the “Cellular Method”. Just picture taking the cell concept as granular as possible. Single-Themed Ad Groups (STAG’s) times schedule times device type times any other variation you want to control for.
Even relatively simple online business models (such as the one in medical testing we supported) quickly become millions of cells. It doesn’t really matter here if that’s separate adgroups, campaigns, or variations on both.
Far from seeing this drop off as you become more granular – we experienced many examples where only very particular combinations of the variables displayed unexpected and non-linear results. Good results. Like one conversion for every 5 clicks. Consistently.
Of course, for most users/companies this is not necessarily useful information or opportunity because the system required to create and manage these cells is extremely challenging and the opportunities for conflicts and errors grows quickly with the volume of cells. When we say system, we are including the people, the data, and the mechanism to keep it working out on Google smoothly. And to measure and manage performance.
Nonetheless, the cellular method does make it clear that “granularity” is the holy grail of creating high conversion rates from a few low cost clicks. It was the desire to go “granular” (even deeper than on customer need per adgroup) that caused us to think about developing a universal set of performance measurements.
Most larger PPC agencies have versions of such “cell generators” that they can adapt for certain larger clients. We believe we can use the engine from the CampaignBuilder to create and manage large “cellular” accounts in a very systematic manner.