Brand awareness: what it is and how to measure it

Brand awareness describes whether an audience recalls or recognizes a brand. To measure it, you can ask a defined audience directly; reach and mentions show exposure or conversation, not recognition by themselves. For some video campaigns, Google Brand Lift compares responses from people who saw an ad with responses from eligible people who did not.

Separate recognition from exposure

Reach estimates how many accounts or people a platform delivered content to under its own definition. Impressions count exposures. Mentions show how often a brand appears in a defined conversation set. None of these measures proves on its own that an audience remembers or recognizes the brand.

AMEC’s evaluation framework places reach and post counts among communication outputs. Audience recall appears among outtakes, while outcomes include changes such as understanding, trust, or preference. The framework helps distinguish activity from audience response; it does not prescribe one universal social metric. See AMEC’s Integrated Evaluation Framework.

Share of voice also has a clear boundary: SOV in a defined set is not the percentage of people who know a brand. Read what social media share of voice measures before using it as a supporting signal.

Choose a direct measure that answers your question

Start by deciding what you need to learn about a defined audience. Do people recognize the brand name? Do they recall it when thinking about a category? Did responses change after a campaign? The question determines the evidence you need.

For a paid video campaign, Google Brand Lift can survey people who saw the ads and people who were eligible but did not see them. It compares their responses about ad recall, brand association, awareness, and consideration. The feature is for video and Demand Gen campaigns and is not available to every account; Google says an account representative may be required. Google explains how Brand Lift works.

For organic social activity or when Brand Lift is unavailable, your own research can ask a sample of the intended audience directly. To make two readings easier to compare, keep the question, population, recruitment method, survey mode, and reference period consistent. Record who answered and who was left out. This is a research option, not an automatic metric supplied by social platforms.

Ask about unaided recall first

Define the audience and category. If your intended audience extends beyond current followers, a survey of followers alone does not represent that wider audience. Describe the sample size and recruitment, survey mode, and full questions; no single sample size guarantees representativeness.

Start with an open question, without showing the brand name, logo, or a brand list beforehand:

When you think of [category], which brands come to mind?

Then ask about aided recognition. Show the brand with relevant alternatives and ask, for example:

Which of these brands had you heard of before this survey?

You can also ask, “Before today, had you ever heard of [brand]?” If you use a list, rotate its order when practical and keep the same order or method across comparable waves. Unaided recall and aided recognition answer different questions; a yes to the latter does not show that someone would have recalled the brand without a prompt.

Question order matters: earlier questions can affect later answers, and a list can prompt responses. General guidance from the Pew Research Center on question order and AAPOR on survey design and reporting supports clear questions, a logical sequence, and describing the method. A methods reference on aided recognition also recommends measuring unaided recall before showing brand names. These sources guide general method; they do not validate these exact questions or guarantee a representative sample.

A fictional example of how to read the result

Suppose a brand asks two comparable samples of 100 people from its target audience whether they recognize its name. In the first survey, 32 say yes; after an activity, 38 do. The observed change is 6 percentage points. The example is invented to show the calculation; it is neither a real result nor proof that the activity caused the change.

Without a comparison group, you cannot tell whether the change came from that activity or another factor. Record the question, each sample’s size and profile, the date, recruitment method, and any differences between waves. Do not report a percentage without explaining whom it represents.

Use social signals as context

Platform metrics can add context when you label them as indirect signals:

• Reach and impressions: describe delivery or exposure under the network’s definitions.
• Profile visits, branded searches, or mentions: help track interest and conversation, but do not confirm what the audience remembers.
• Share of voice: describes a share of mentions within the set you defined, not awareness across a population.

Connect each number to a different question instead of combining them into a universal awareness score. For help choosing a measure based on your team’s goal, see social media KPIs by goal.

The cover illustration was generated with AI.