Psychology of the Searcher

Psychology of the Searcher: Patterns in How Searchers Formulate Queries

Executive Summary:

Marketers see visitors from a wide variety of search queries coming to their site.  This data is valuable in guiding a search strategy, but it has existed in a vacuum, with little known about how searchers make decisions about how to phrase their search that lead up to the visit.  New research from Blue Nile Research surveys searchers about how they choose to form their searches in a variety of different scenarios, and helps Marketers see the patterns in how searchers formulate their queries.

Blue Nile’s research shows an exact 50-50 split between users who search in fragments (e.g. ‘swollen ankle’) and those who search in more fully formed terms (e.g. ‘causes of swollen ankle during sleep’).  When it came to questions vs. statements, 27% of respondents phrased their query in the form of a question, with ‘How’ being the most commonly used prefix.

With the research showing no clear clustering in how users phrase their searches, Marketers who wish to be well prepared to reach their target audience must be thorough in first understanding how their audience chooses to search before developing a strategy and by crafting content that closely maps to their pain points.

Humans are Distinct Individuals  

The possible human DNA combinations are as varied as the snowflakes that fall in winter.  Scientists estimate that there are trillions of distinct DNA combinations.  These wide and varied combinations of DNA strands that make up the human means how people react when faced with uncertainty is unpredictable.  Nowhere is this more relevant than in search.

Research shows the average person performs 129 searches per month.  Little research exists that shows how searchers choices differ when it comes to the way they form and phrase their queries.

When faced with the same information retrieval scenarios, how if at all, do searchers’ queries differ?  What patterns can be gleaned from analysis of the responses?  What can be learned and applied to the practice of search marketing from this analysis?

Like this study? To commission your own attention-getting, lead-generating research study  for your company from a former Forrester Research Analyst, contact Nathan Safran at  It is likely to be more affordable than you think and a conversation is free.

Real Implications for Marketers

While searchers’ phrasing choices are an interesting question on their own, this research has very real implications for search marketers.  Normally search marketers use keyword research tools such as Google’s Keyword Planner to discover keywords relevant to their industry they want to optimize for.  While this approach can be fruitful in the discovery of new keywords, the flaw with this sole approach is that keyword tools generally suggest phrases similar to the one(s) started with.  This means initial assumptions about how the user might search for your product/service are perpetuated throughout the keyword research process.

Further perpetuating the problem is that keyword tools generally do a poor job of offering suggestions for more specific (long tail) queries.  In an example taken from our research, entering ‘coffee maker won’t turn on’ into the Google Keyword Tool returns generic ‘coffee maker’ terms.


None of the variations other respondents used in our survey were suggested:


This is why a market research based approach that first seeks to discover how a user might search—the phrasing, language, etc. that distinct audience segments use–will ensure that whatever keyword research path is subsequently followed is based on empirically grounded insight rather than guesswork.


Our approach to identify distinct ways in which searchers approach the same searcher scenarios was to ask respondents to imagine several distinct scenarios that create an information gap.  Respondents were then asked to input their online search they felt would best resolve the scenario.

The scenarios we tested were:

  • Technology Problem:  Coffee maker does not turn on
  • Health Problem: Swollen ankle
  • E-commerce: In the market for a new laptop

This approach gives us multiple distinct scenarios for analysis and sets us up with a rich data set to identify how a base of respondents phrase their query when faced with the same situations.  We collected 183 responses, randomly selected from a group of online users.

People Really are Different

The starting point for this paper, that online search is an expression of the human it emanates from and human makeup varies widely, comes into sharp focus when we analyze responses by length of query.  There is an exact even split between queries we are calling ‘fragment queries’ (2-3 words ) vs. ‘full queries’ (4+ words).


This split, and the follow up analysis of the individual responses across all searcher scenarios, reveals two distinct approaches searchers take in resolving an information gap:

  1. Throw it Against the Wall and See What Sticks (fragment query):

    This searcher is focused on ‘speed of search’ and inputs the minimum amount of information into the search box.  They are willing to peruse the search results, click into multiple links to discover the information they are looking for, and follow up with a more specific search if necessary.
  2. Be Specific out of the Gate (full query):  This searcher is focused on ‘depth of search’ and takes an extra moment to best phrase their query, with the hopes they will find what they are looking for in one click, towards the top of the search results.

A respondent said it best in a post-survey interview:

Different people search in different ways.  Some people are very short with their search and like to click around.  Me, I like to figure out the best way to phrase my search and find my answer right away rather than click around.


Like this study? To commission your own attention-getting, lead-generating research study  from a former Forrester Research Analyst  for your company, contact Nathan Safran at  It is likely to be more affordable than you think and a conversation is free.

Searchers Phrasing Varies Widely

Looking at how respondents search by number of words from a slightly different angle than above, we can see that the highest percentage of participants use two words in their query.  Interestingly, the spread of responses was broad, no single query length had as much as a third of respondents.  This finding again reinforces our contention that individuals search differently, phrasing their query in distinct ways when faced with an uncertainty gap.

This means Marketers must be prepared with a strategy to be visible in the search results for the varied and distinct ways in which their audience might choose to search for their product or service.  Marketers must avoid ‘searcher bias’ by not tunnel-visioning on how they might choose to phrase a query, recognizing it varies, often widely, from one searcher to the next.

number-of-words png

People Search in ‘Fragments’

When we dug deeper into the makeup of the search queries, we found that 27% of searchers phrased their query in the form of a question (‘how, ‘why’, ‘where’, ‘what’, ‘which’) as opposed to a ‘statement query’.  This is another way in which the distinct ways in which searchers choose to approach their information gap is highlighted; some choose to ask a question of the search engine while others choose to make a statement, each believing that their own approach will best serve to resolve their information gap. Q-vs-non-Q question-queries-table

With ‘How’ Searchers Seek a Cure, not a ‘Why’

Of the responses that were questions, searchers most often used ‘How’ (44%) while ‘What’ was used least (12%).  This revealed a desire by searchers for immediate resolution (‘How  do I fix my coffee maker?’) rather than an investigation as to why the issue might be occurring (‘Why is my coffee maker not working?’).

This finding should encourage Marketers to think like their audience and consider where the individual might be in the buyer’s journey.  In the research phase, buyers want information—and they want it to be specific.  They want resolution rather than a detailed explanation or to be sold on a product.  This creates an opening for opportunistic marketers to create content that educates and informs, and accurately caters to the place in the buyers journey that the buyer is in.


Conclusion: Develop Distinct Approaches for the Individual Ways People Search

Although some may have begun this journey believing the data would show large groupings of respondents searching in similar ways, in reality human identity is alive and well.  Marketers that wish to be well prepared to appeal to a wide and varied audience will be thorough with their research process and craft content that maps to the answers their audience is looking for.

Like this study? To commission your own attention-getting, lead-generating research study  for your company from a former Forrester Research Analyst, contact Nathan Safran at  It is likely to be more affordable than you think and a conversation is free.
Posted in:


  1. Hi Brian,

    The searches were made in a simulated search box on the survey form. So searchers were asked to imagine a scenario (e.g. wake up with a swollen ankle, coffee maker doesn’t work etc) and then asked to enter in the box the exact search they would make on Google to resolve the scenario.


  2. Great article – good to see some stats behind searcher query preference in particular. For clarification, do 27% of the 50% performing full queries ask full questions or do 27% of total search queries?

  3. Thanks for sharing results of the study Nathan. Pretty interesting question type distribution, I guess that shows why so many people love how-to’s. Would be interesting to see this type of data broken down by demographics, at least men vs women, to see how we think differently and alike.

    • I agree–common sense is a good guide and there are findings that line up with common sense, but I think there are findings that are also counterintuitive–e.g I wouldn’t have predicted an exact split of those that search in fragments and those that search in fully formed queries.

      Thanks for the comment Andy.

  4. Hello Nathan,
    Great article. I’m new to the search marketing world, and I’ve been having trouble finding the right long-tailed keywords for customers. I found the fragmented vs. non-fragmented searchers info very helpful. And that most searchers seek immediate solutions or value adding information based on where they fall in the sales funnel. I found this article helpful and I will be sure to keep an eye out for future articles. Thanks Again. – Will

  5. Hey Nathan,
    Great post in uncategorized category, though I was expecting to see top search patterns by people. During keyword research I usually use product category option in keyword planner. For swollen ankle research I would have looked for adgroups and exact keywords ideas under Health to see ankle causes, swollen painful, swollen foot adgroups and long tail keywords such as [swollen ankle treatment], [swollen ankle pain], [how to treat a swollen ankle], [what to do for a swollen ankle] etc. I would like to thank you for the research which is very helpful for us all, now I can think about user in terms of fragment and full queries to develop better content to target both behaviors. Thanks again Nathan!

  6. Great article, and the fact that Google rolled out the “Knowledge Graph” only supports it. But really, great insight on how to form our wording / content as marketers and also SEO


  7. Hi Nathan – great article! Do you have any opinions on tools like long tail pro to help search marketers with this process?

    • Thanks! I think long time practitioners would be better versed in the specific of the tools, but what I did learn from the study is that starting a search strategy with specific research about how searchers search for their product or service can be very valuable in guiding a strategy.

  8. Great article and information. Since Google’s keyword planner offers generic keyword results, do have any suggestions on getting long tail keyword search results that are in the form of a question? Or just use common sense and add the who, what where and when with your keyword in a logical sentence?

  9. Yeah, that is one valid way to do that. You can set up a spreadsheet that auto fills in question modifiers to your queries.

    Another approach is via market research–similar to the process used in the study, to survey respondents about how they search for your product/service. This will give you a much more thorough view of how your audience searches for your product/brand including question modifiers.

  10. In my SEO workshops I teach that we need to understand User Intent in search; so it’s not just about ‘what they want to know now’, but also “What else might they want to know?” So for example, aside from just ‘Why won’t my coffee machine turn on?’, the user may also want to know, ‘Can I fix it myself, or, How much will it cost to fix’ etc. That way, when we package these things up into content, we move away from Keyword Centered Content into Customer Centered Content – exactly what your study shows. I would have liked to see it with a higher sampling, but still, the initial results are positive. Nice article.

56 Pings & Trackbacks

  1. Pingback:

  2. Pingback:

  3. Pingback:

  4. Pingback:

  5. Pingback:

  6. Pingback:

  7. Pingback:

  8. Pingback:

  9. Pingback:

  10. Pingback:

  11. Pingback:

  12. Pingback:

  13. Pingback:

  14. Pingback:

  15. Pingback:

  16. Pingback:

  17. Pingback:

  18. Pingback:

  19. Pingback:

  20. Pingback:

  21. Pingback:

  22. Pingback:

  23. Pingback:

  24. Pingback:

  25. Pingback:

  26. Pingback:

  27. Pingback:

  28. Pingback:

  29. Pingback:

  30. Pingback:

  31. Pingback:

  32. Pingback:

  33. Pingback:

  34. Pingback:

  35. Pingback:

  36. Pingback:

  37. Pingback:

  38. Pingback:

  39. Pingback:

  40. Pingback:

  41. Pingback:

  42. Pingback:

  43. Pingback:

  44. Pingback:

  45. Pingback:

  46. Pingback:

  47. Pingback:

  48. Pingback:

  49. Pingback:

  50. Pingback:

  51. Pingback:

  52. Pingback:

  53. Pingback:

  54. Pingback:

  55. Pingback:

  56. Pingback:

Leave a Reply

Your email address will not be published. Required fields are marked *