What is brand recall?
Brand recall is retrieving a brand name from memory in response to a category, need or situation cue, without supplying the brand as the answer. Kevin Lane Keller’s conceptual discussion distinguishes this retrieval from recognition, which supplies a brand cue.
Choose which mention the report will count
Decide which question the result should answer. Do you want to know whether a business is named anywhere, or whether it is named before alternatives? Preserve the answer order so those questions can be read separately. A report labelled simply “recall” leaves the reader guessing what was counted.
Pearson’s Marketing Metrics chapter describes top-of-mind awareness as the first brand named in an unprompted category question. An any-mention record instead checks whether the brand appears anywhere in the answer. Keep the measure’s name beside its coding rule.
Write the unprompted question
Write the category or situation in the words the intended audience will understand. Record who is eligible to answer and the setting in which the question will be asked. If the job concerns local cycle repairs, a question about every kind of retailer would be a different brief.
For an unprompted question, SurveyMonkey’s guide advises asking before the survey mentions brand names. Introduce a later recognition question separately if you need it. Keep the category prompt and the supplied brand name distinct when describing what each answer means.
Code any mention and first mention separately
The table uses constructed answers for a fictional cycle-repair shop called Oak Wheel. The target being coded is Oak Wheel: “any mention” records its appearance anywhere in an answer, while “first mention” records its appearance at the start of the ordered list.
| Ordered answer | Oak Wheel mentioned anywhere? | Oak Wheel named first? |
|---|---|---|
| Fern Cycle, Oak Wheel | Yes | No |
| Oak Wheel, Fern Cycle | Yes | Yes |
| No shop comes to mind | No | No |
Keep the raw answer and declare the denominator
Keep the raw answer as well as its coded version. In the illustration, the order “Fern Cycle, Oak Wheel” stays in the record even though Oak Wheel receives an any-mention code. Keeping the original words lets a reviewer check why the coding differs from the first-mention result.
For this illustrative report, the planned denominator includes eligible respondents who received the unprompted question. “No shop comes to mind” stays as a nonmention. A missing or unusable answer is recorded separately with its handling rule; it is not silently recoded as a recognised or recalled brand. No percentages are calculated from these constructed answers.
Write a codebook before turning responses into a report. Specify how spelling variations or shortened names will be matched, how duplicate mentions are handled and what requires review. If “Oak” could refer to another business, keep that answer unresolved until the record provides enough information to code it.
Keep recall and recognition prompts distinct
A recognition question supplies a brand cue; a recall question asks respondents to produce the name. Pearson’s chapter distinguishes prompted recognition from unprompted awareness. Use Brand Recognition for the prompted concept and Brand Awareness for the wider measurement choice, keeping each question’s prompt in view.
Make study comparisons readable
Pearson also cautions that changing how questions are worded or administered can affect tracking results. Keep those details with the recruited audience and collection period when comparing studies. A change in the result needs interpretation alongside any change in the study method.
Before reporting a percentage, name the eligible respondent group and the answer-handling rules used for its denominator. Label any-mention and first-mention results separately. If the report covers only customers you recruited from a mailing list, describe that audience rather than calling the result a finding about every local category buyer.
Separate recall answers from search activity
Google Search Console’s Performance report records your site’s Google Search clicks and impressions and can group data by query. Brand-name queries provide observations about activity involving the site. They are not responses to the unprompted question or a count of everyone who recalls the business.
Keep a communication activity record separate from the recall study: what material changed, where it appeared and the period it covers. Use the study’s question and respondent definition when reviewing the result. A change after an activity does not, on its own, establish that the activity caused it.
End the report with the question the evidence can now inform. You might need to investigate whether the category wording made sense, resolve an ambiguous name or plan a comparable follow-up. Give the next reviewer the questionnaire and codebook so they can understand the result without reconstructing the study.
Related terms
See Brand Awareness for choosing among the broader awareness questions, and Branded Keywords for searches containing a brand name. This page supplies the unprompted-question and response-coding workflow; those related records serve different jobs.
Sources
Kevin Lane Keller’s conceptual discussion supports the distinction between generating a brand from memory and recognising a supplied cue. Pearson supplies the first-mention and comparison cautions; SurveyMonkey supplies vendor question-sequencing guidance; Google documents its own report. The Oak Wheel answers are constructed coding examples, not an observed field study.