by Oren Netzer, Op-Ed Contributor, June 27, 2016
Part I of this series showed how bidding directly on mid-tail and long-tail Exact match keywords can significantly improve click-through rates, average position and cost per click. The study revealed that this is likely due to a better match between user intent (as expressed through the search query) and between the ad copy and landing page. In Part II, I will discuss how similar results can be achieved by bidding on mid-tail and long-tail Phrase match and Broad Match Modified (BMMs) keywords.
Long-Tail BMMs/Phrases can have the same benefits as Long-Tail Exact Match
What I refer to as mid-tail and long-tail Phrases and BMMs are simply phrases and BMMs that are more specific and that match a smaller number of more specific search queries. For example, while a BMM such as +open +bank +account can match a large and diverse number of search queries looking for different types of accounts, a more specific BMM such as +open +bank +account +children will only match search queries by users looking to open a bank account for their kids. Using relevant ad copy and landing page for the situation will significantly increase performance of this keyword.
It is not necessarily the use of the Exact match that is driving improved performance — it is the more accurate match to user intent driving performance. The same approach would work if Phrase matches and BMMs were used that are narrower in scope and target more mid-tail and long-tail search queries.
The following analysis shows the differences in performance of BMMs based on how many unique search queries they match.
Long-Tail BMMs had a Better Average Position
Figure 1 shows the average position of the keyword as a function of the number of unique search queries that it matched. The BMMs were divided into 3 groups for simplicity: BMMs that matched 50 or more search queries, BMMs that matched between 10 to 50, and BMMs that matched less than 10 search queries.
The average position of the long-tail BMMs matching less than 10 search queries was 2.44, compared to 3.4 for BMMs matching more than 50 search queries — a 30% improvement. This kind of difference in average position is very significant — it can make the difference between being at the top of the page or the bottom of the page and can translate into huge differences in click-through rates and overall clicks.
Long-Tail BMMs had a Better Click-Through Rate
Not surprisingly, very similar behavior can be seen when observing click-through rates. Figure 2 shows a scatter graph plotting the distribution of click-through rates of each BMM keyword (on the Y axis) as a function of the number of unique search queries the BMM was matching (X axis).
Keywords matching a small number of search queries (on the left side of the graph) showed a large diversity in their click-through rates, with many of the keywords at above 5% click-through rates. However, as you move to the right on the graph, once keywords passed 20 or so matching search queries, their click-through rates seemed to stay consistently below 3% (with the possible exception of one or two keywords).
These results demonstrate how executing a solid long-tail keyword strategy can significantly improve front-end metrics. In our next installment takes a look at how bidding on long-tail keywords directly can affect your overall share of voice and how it can impact your back end metrics such as conversions and cost per conversion.