It’s commonly cited that ‘head’ terms (a.k.a. short requests consisting of < 4 words) make up only 20% of ecommerce site searches, and the remaining 80% are long-tail requests (> 4 words). Constructor analyzed more than 70M user search requests to find out if this is true or a myth, and what it means for ecommerce teams. We share the results below.
The Truth About Head and Long Searches
Surprisingly, our analysis busted the myth about only 20% of search queries being less than 4 words. The reality is just the opposite: about 75-90% of requests are shorter than four words!
This trend holds true across most industries with some key variances. The longest requests are in the Pet category since people need at least one more word to specify the species (i.e., is it a cat or a snake?). And in the Entertainment category, for example, people often search for media like movies or books using the exact title. So naturally, these requests are also longer.
One possible reason for the high frequency of shorter requests is the prevalence of search suggestions, also known as Autosuggest. Smart suggestions try to predict the end of a request based on the user’s first typed symbols or words. Instead of typing the whole request, the user can just click an AI-generated suggestion.
We also found that the proportion of head and long-tail requests are the same on mobile and PC since suggestions work across devices.
How Specificity Plays a Part In the Equation
If head terms make up 87% of requests, what about the other 13% that are long-tail requests? Is there a reason why people prefer to type long requests despite there being automatically generated suggestions? The answer boils down to one thing: specificity.
Sometimes customers just know exactly what they’re looking for (i.e., "iPhone 16 Pro Max 512 gb”). There are no excess words in that query. The search engine needs every word to show the right result. These cases are often associated with brands, models, and titles of books, for example.
Sometimes, though, customers aren’t exactly sure of the specific item they need and therefore provide as many details as they can in their query (i.e., “best iPhone for bloggers 2025”).
If we look at our sample set of long-tail requests, we unpack a curious phenomenon: there are equally as many long narrow (specified) requests as there are wide (unspecified) requests.
In the data we examined, we observed that wide (unspecified) details are usually connected with things like promotions, specific seasons, or sizes. In industries such as apparel, we also saw a connection with more subjective details that may not appear in the product metadata like "casual", "oversized", or “sexy”.
The Million Dollar Question: Which Types of Requests Drive Revenue?
Now that we better understand the patterns of user requests, which ones drive the most tangible business value?
We can measure the quality and conversion potential of requests by using CTR, or the click-through rate. It’s a proportion of clicks over search requests.
This metric shows that long narrow requests work best, since these searchers are in spearfishing mode. They know what they want and just need the search engine to help them find it.
The CTR of long-tail narrow requests is over 12% higher than head terms. Head terms convert at the second-best rate. And the lowest CTR of the three belongs to long-tail wide requests. Their CTR drops by 10.3%, on average, across industries.
The Definitive Answer?
Although the data gives us a better idea of what works in aggregate, we shouldn’t assume that long-tail narrow requests always drive the best results (i.e. most revenue).
A lot of what’s “right” boils down to what makes sense in the context of your category, your ecommerce store, a user’s unique behavioral data, macroeconomic trends, and many other factors.
At the end of the day, the length of a search request is just one piece of the puzzle. What really matters is how well your search experience uses these clues to understand what shoppers want, whether they’re typing one word or ten.
By learning from real shopper behavior and fine-tuning results in real time, Constructor helps ensure that every search connects your customers with the products they want to buy.
Want to learn more ways to optimize for KPIs like CTR, conversions, and revenue through search?
Watch The New Search Standard: What $9.8B in Revenue Data Reveals About Ecommerce Growth.