The Long-tail Campaign Dilemma

The Relationship between Search Term Length and Conversion

The relationship between Search Term Length and Conversion rate has been discussed by many, for example, Long-tail Campaign Performance Study,  Long-Tail Keywords: A Better Way to Connect with Customers, and Are Long-Tail Keyword Phrases Good for My Business?. And they agree that longer-tail keywords have higher conversion rate. Here is a simple chart that illustrates the relationship between long-tail keywords and conversion rate (CVR) showing that, as the keyword phrase length increases, the search volume for each keyword decreases, but the conversion rate increases. Simply put, the Longer the search term is, the better the conversion. For example, when a shopper is searching for a “jimmy choo women’s black leather jacket” and then actually lands on a page showing exactly those items, the conversion rate is better than a shopper that searches for “leather jacket”. Of course, this happens only if the following conditions are met:

1. The paid search campaign has enough keywords to capture long-tail intent (exact and broad match)
2. The site has content that is relevant to the intent expressed by the provided keywords.
3. The website has a deep linking ability to point to the webpage with the relevant content for the specified long-tail keyword.
4. The campaign bidding does not favor head keywords, so that the long-tail keywords will get the impressions and have the same average position as head keywords.

When these conditions met, there are real long-tail campaign performance improvements

Although many articles and blogs hint at performance improvements, we wanted to quantify the impact. Taking data from real search ad campaigns, the table below clearly shows the major impact of long-tail keywords. And the result is that conversion rate increase of +32% for the keywords that contained size or color information compared to those that did not. Since 20% of the campaign clicks were on this specific set of long-tail keywords, performance of the entire campaign really improved.

Keyword	#Keywords	Clicks	Orders	Conversion Rate	Long-Tail Improvement
Keywords (No Size/Color)	34,833	64,997	1,112	1.71%
Longtail Keywords (include size/color)	14,591	15,778	356	2.26%	+32%

The “Head Heavy” Campaign

In reality though, we often see “Head Heavy” campaigns. This is a campaign that looks like this:

Head Heavy Campaign
	Clicks	Average CVR	Conversions
Head Keywords	22,500	1.67%	375
Long-Tail Keywords	8,845	2.98%	264
Overall Performance	31,345	2.04%	639

In this campaign, 70% or more of the clicks go to a head keywords (with 4 or less terms). The problem in this campaign is that the long-tail keywords that would better match the customer search queries either do not exist or are they exist, but head keywords are getting the click volume (usually due to a bidding issue). But this problem is usually hidden, since there is rarely an analysis of head vs long-tail performance performed on the account. However, if that same account is properly balanced with sufficient long-tail keywords, proper bidding and relevant final destination URLs, then the performance will look more like this — a performance increase of over 20%!

20.9% Improvement for Balanced Campaign

Balanced Campaign
	Clicks	Average CVR	Conversions
Head Keywords	12,375	1.67%	206
Long-Tail Keywords	18,970	2.98%	566
Overall Performance	31,345	2.46%	772

Now this campaign properly captures shoppers search intent and directs the click to the right destination on the retailer e-commerce site. With 60% of the clicks going to long-tail keywords, shopper intent is better captured by the keywords and the shoppers are presented with the variety of products most relevant to their search. This is especially important for mobile search, where the shopper sees only the first few items on the small screen. So, when a shopper searches for the desired ““jimmy choo women’s black leather jacket”, the landing page will indeed show the desired women’s black Jimmy Choo leather jackets that you sell — and not a generic assortment of black leather jackets or all of the Jimmy Choo products.

Why aren’t we seeing more “Balanced” campaigns?

There are many reasons for that. Assuming that the account is diligently maintained, constantly monitored and updated by skilled marketing team, the main issue is the size of the long-tail keyword set and the constant need to the update landing page destination URLs. The long-tail campaign size problem is simple — as the number of words in a keyword increase, the result is that the number of longer-tail keywords grows exponentially. Here is illustration with some typical numbers:

Number of Words in a Keyword	Number of Keywords in Campaign	Relevant Landing Page Destinations
1	20	10
2	340	113
3	5,100	1,133
4	61,200	9,067
5	428,400	42,311
6	2,142,000	141,037
7	8,568,000	376,099
8	29,988,000	877,564

As seen above, providing enough long-tail keyword coverage for the account results in very large campaigns. Of course, there is general reluctancy to create such campaigns, since the time to manage them grows with the campaign size — unless you have appropriate automated tools. Defining the most relevant destinations for hundreds of thousands of keywords and constantly maintaining their destination in light of website, inventory and other changes is not something that can be done without appropriate automation. Hence, the “Long-Tail Campaign Dilemma”. The payback for manual management of the large campaigns with sufficient long-tail keywords may not offset the additional management cost. With automated tools, the ROI can be significant.

Automation.. automation and more automation

While there are many tools to automate bidding for large campaigns, there are very few tools in the market that provide solution to the “Long-Tail Campaign Dilemma”. One of the new generation of intelligent tools is Vioby’s LinkDirector. It scans the retailer site, finds the optimal destination for each keyword, and updates the campaign configuration with any required changes, 24×7 fully automated. Using machine learning and natural language processing, it replaces the need to constantly manually analyze and update landing page destinations. For more info drop us a note here – Contact