Welcome to the second of four excerpts from “Open the Black Box”, OptiMine’s white paper about finding transparency and control in paid-search automation. If you’d like to catch up, you can read the first excerpt here. Today we look at context which, in this case, means it’s not enough to know the results your paid search system is producing if you can’t do know the why behind them.
Seeing the Context
As noted, bid automation is essential for companies with complex paid-search programs, but with black-box bid automation systems (that is, most of automation systems), it’s hard to determine the “why” behind their paid-search performance. The performance might be acceptable, but there’s no way of knowing if it could be better.
The problem is especially acute with cluster- and rules-based automation systems, where individual keyword performance is lost because groups of keywords are treated the same. Does your system allow you to filter and sort individual keywords based on highest or lowest cost, sales, revenue, profitability (and any other metrics you can think of), and drill down into the factors – such as day-of-week effect, or days until the next holiday – to determine the variables that are driving those individual keyword results? Some systems do provide bid values and results for individual keywords but limit access to the variables that were used to drive those results.
This absence of context might be acceptable to some advertisers, because set-and-forget automation is superior to manual bid optimization, but it will be inadequate for advertisers looking for the significant performance improvements attainable with systems
that provide detailed and filterable data on individual keywords.
How do these systems achieve such results? Consider the case of two keywords: “silver amethyst rings” and “sterling silver amethyst rings.” Most automation systems would cluster such similar-sounding keywords and calculate an average bid amount based on the value per click for the cluster as a whole. Other systems apply the same rule to each, potentially masking unique performance attributes. A system that provides transparency into individual keyword performance might reveal these strategies to be sub-optimal and identify very different driving forces that make these keywords convert. One might be more influenced by a particular day-of-week effect, while the other is weighted more heavily on a one-week average cost per click. The reasons for these differences might be obscure or even unknowable, but the differences are nevertheless very real.
Software automation systems that employ advanced multivariate regression analysis modeling at the individual keyword level will pick up such differences and bid each keyword appropriately, based on a historical analysis of the factors that drive individual keyword performance. This feature is especially important in managing mid-tail and extreme-tail keywords, which often account for as much as 90 percent of a campaign or ad group keyword portfolio, to maximize whatever revenue is available in the tail – revenue that is typically lost with clustering or general rules.
With better transparency giving providing context, control is the next step and, coincidentally, the next post in the series. If you don’t want to wait for excerpt 3, you can download the white paper at OptiMine.com now.