Almost every search engine marketer (SEM) knows the adage, “Use head keywords to generate volume and long tail keywords to generate profit.”
Typically, query or visit volume is used to determine if a keyword is a “head” keyword or a “long tail” keyword. But what is the relationship between keyword length for the head versus the long tail? Questions we set out to answer were:
- How short and long are head terms?
- How many impressions, clicks and conversions do head terms generate?
- How short and long are long tail terms?
- Similarly, how many impressions, clicks and conversions do long tail terms generate?
- Is it true that long tail keywords are more profitable than the head?
Methodology
We analyzed approximately 1.5 million keywords over a six month period across multiple clients, campaigns, and verticals. Each of these keywords generated at least one impression during these six months.
We didn’t differentiate between exact, phrase, and broad match keywords. We also didn’t attempt to normalize impressions by differing bids on keywords; similarly, we didn’t attempt to normalize clicks and conversions for different ads that were shown across keywords.
To analyze keywords length, we sorted keywords into five-character buckets (i.e., 0-5 characters; 6-10 characters, 11-15 characters, etc.).
Analysis
To understand the impact on volume, we studied what percentage of impressions, clicks, and conversions each bucket of keywords generated. In the lingering spirit of the Olympics, we awarded gold, silver, and bronze to the top buckets in each category.
Keywords that were 11-15 characters generated the bulk of impressions, clicks, and conversions. This is consistent with the adage that the head generates most of the volume.
Note, however, that head terms are not necessarily the shortest terms. Keywords that were 0-5 characters long never generated a meaningful number of clicks and conversions. Even though keywords that were 6-10 characters long generated a significant number of impressions, these keywords were not the top performers as far as clicks and conversions were concerned.
To understand the relationship between keyword length and profitability, we did not look at return on ad spend (ROAS) per se – (i.e., revenue generated from the SEM ad spend, divided by SEM ad spend). ROAS is usually a target, and SEMs can achieve target ROAS by adjusting bids up or down on keywords. As a result, almost any keyword can have any ROAS.
Instead of measuring profitability per se, we looked at the efficiency of each keyword, as measured by click-through rate (CTR), conversion rate, and conversions per 1,000 impressions. Again, in the spirit of the Olympics, here’s what we found:
Here we see the power of long tail keywords in action. In general, we see a steady and almost monotonic increase in CTR, conversion rate, and conversions per 1,000 impressions as keywords increase in length, with the most “efficient” keywords in the 31-35 character bucket. Note that keywords above 40 characters not only generated a tiny number of impressions – they also weren’t terribly efficient.
Key Conclusions
While length of keywords will obviously vary across different verticals, we can make a few broad inferences about keyword length:
- 0-5 character keywords are, more often than not, too short for advertisers to derive any meaningful intent. As such, they don’t create meaningful impressions, clicks, or conversions for most advertisers.
- 6-10 character keywords generate a high number of impressions. They don’t, however, generate a proportional number of clicks and conversions. Consumers are entering queries of this length; however, they still aren’t far enough along in the purchase consideration funnel to actually click and convert on paid search ads.
- Nearly 60 percent of all impressions, clicks, and impressions occur with 11-20 character keywords. Including 21-25 character keywords brings us to 80 percent of call conversions. As such, 11-25 character keywords definitely comprise the “head.”
- While 26-40 character keywords generate significantly fewer impressions, they are much more efficient than head keywords. Specifically, 11-20 character keywords represented 62 percent of clicks and 62 percent of conversions (a 1-to-1 relationship); whereas, 26-35 character keywords represented 6 percent of clicks and 10 percent of conversions – a 3-to-5 relationship. In other words, the long tail keywords were – ballpark – about 66 percent more profitable than the head keywords (ignoring bids), supporting the notion that long tail keywords are more profitable than head keywords.
- The impression, click and conversion volume with +40 character keywords is minimal. Efficiency on these keywords is very low as well. As such, SEMs should probably not spend time generating/managing keywords with greater than 40 characters.