What it means
A customer searches "change billing address" and sees no article, then opens a support ticket. A search gap report can reveal the missing phrase or content before many others repeat the same path.
Define the denominator as actual search events in the help centre, not total page views, and exclude blank or bot-generated queries according to a consistent rule. Zendesk's search-results reporting distinguishes searches with no results from searches with results but no clicks, and those categories reveal different problems: missing content, poor wording or weak ranking.
An illustrative zero-result rate is searches returning no result divided by total valid searches in the period, so if 120 of 2,000 searches show nothing, the rate is 6%. Intercom's help-centre search guidance describes improving article findability, since an article can exist but fail to appear for the customer's language, synonym or typo.
Track unique query themes, because ten spelling variations of one problem can represent the same content gap and should be grouped while retaining the original words. Separate high-frequency from high-impact, since a rare search about a critical safety issue may deserve faster attention than a common but low-stakes query.
Check language and geography too, as a translated article may be missing or a regional policy may differ, and nobody should route a user to guidance that is wrong for their market merely to reduce no-result counts. Review no-click results, because a search may return irrelevant pages but a user may also read the snippet without clicking, so interpret them with later actions, feedback and sample review.
Use actual customer terms, since internal product names can differ from the language buyers type, and add accurate synonyms and plain titles rather than repeating jargon. Watch content access as well, because an article can be hidden behind a sign-in wall or audience rule, and a valid restriction should not be removed just to improve a public search metric.
Check search redirects, since an automatic redirect may send a user to the wrong topic while appearing to yield a result, so test the journey and not only the count. Preserve a fixed period for comparison, because seasonal launches, policy changes and a new product can create temporary spikes, and similar demand should be compared before claiming the knowledge base worsened.
Assign article ownership so that someone decides whether to write, update or relabel content and confirms the answer against current product behaviour, and avoid unsafe answers, because a topic needing account-specific or regulated advice may need a clear escalation path rather than a generic public article. Measure after a fix, since the exact zero-result phrase may disappear after synonym changes while related ticket rate and customer feedback test whether the answer actually helps, and watch search volume shifts, because a lower gap rate can result from fewer searches if the search box becomes harder to find.
Use privacy safeguards, since search queries may contain account numbers or personal details that should be removed before sharing an editorial report, and separate search quality from deflection, because finding a result does not prove a support ticket was prevented. Keep an editorial record of the query cluster, proposed answer and source used to verify it, since a page that merely matches keywords can make the metric look better while giving stale instructions; for an owner, the search gap rate points to questions the help centre fails to surface, so read it as a content and navigation signal alongside search volume, article feedback and the ticket topics that follow a failed search, not as a standalone measure of customer success.
In practice
Real-world examples.
Example
One hundred twenty of 2,000 valid searches return no results, or 6%. The support lead reviews the failed queries each week and sorts them into themes. The largest theme becomes the next article to write.
Example
Several spelling variants of one billing question are grouped for editorial review. The editor keeps the original wording of each variant so that synonyms can be added accurately. One article and a short list of synonyms then cover all the variants.
Example
A no-click query is investigated rather than automatically called a failure. A sample review shows that the search snippet already answered the question for most users. The team leaves the article unchanged and focuses on queries where users searched again straight away.
Formula
Calculation
Zero-result rate = valid searches with no returned result / all valid help-centre searches x 100.
Worked example: in a month the help centre records 2,000 valid searches, and 120 return no result. The rate is 120 / 2,000 x 100 = 6%. Suppose the team then adds an accurate billing-address article with synonyms, and the next month shows 1,800 valid searches, of which 72 return nothing. The new rate is 72 / 1,800 x 100 = 4%.
The fall from 6% to 4% looks encouraging, but search volume also fell by 200, so the team checks whether the search box has become harder to find and compares related ticket volumes before reporting an improvement.Case study
Seen in the real world.
In this entirely fictional example, Maple Cloud sees repeated failed searches for a billing-address change. It adds an accurate article and synonyms, then checks later queries and related support contacts. The support lead also checks search volume, so that a fall in failed searches is not simply the result of fewer people searching. Billing-address tickets decline over the following weeks, but the lead does not claim every later reader was successfully deflected from support. The article owner schedules a review for the next product release.
Watch out
Common mistakes.
- Treating a result shown as proof the customer found an answer.
- Publishing incorrect generic advice for account-specific issues to reduce zero-result counts.
- Claiming a falling rate is improvement without showing search volume.
Questions
People also ask.
Is a no-click search always a failure?
No. A snippet may answer the question or the result may be irrelevant; investigate.
Can synonyms fix a gap?
Sometimes, when accurate content exists but uses different words.
Does a lower rate mean fewer tickets?
Not necessarily. Measure related support outcomes separately.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%