How to measure brand awareness without a research budget

How to measure brand awareness is a question most small teams answer with the wrong number. They open a dashboard, see views, and call it awareness. Awareness is people knowing your name when you are not on their screen. This guide is a method a small team can run with no research budget: five free signs first, a short survey second, and one sheet you fill in every month.

The order matters. The free signs come from tools you already have. A survey costs time or money and is easy to get wrong, so it comes second. This guide stays on the measuring. Why reach and awareness are different things is its own piece.

Read reach vs brand awareness →

What brand awareness means in numbers you can count

Survey research splits awareness into two kinds. SurveyMonkey's guide puts it this way. Aided awareness is the number of people who show they know a brand when prompted. Unaided awareness is the number who show they know it without prompting.

In plain words, aided is a yes or no question with your name in it: "Have you heard of this app?" The person only has to recognise you. Unaided is an open question with no names in it: "Which budgeting apps can you think of?" The person has to pull your name out of memory with no help.

Unaided is the harder test and the more useful one. A person opening an app store has no list in front of them. They pick from the names they can recall.

Both kinds need a survey. A small team also has a third reading that needs none: what people do when they remember you. They type your name into a search bar. They come straight to your site. They tell a friend, who shows up without having seen an ad. Each of those leaves a count behind.

Measuring brand awareness starts with five free signs

Measuring brand awareness for free means reading five things. None of them is perfect alone. Together they are hard to fool.

1. Branded search: how often people search for your name on Google.
2. Direct traffic: visits from people who typed your address or used a saved link.
3. Customers without an ad: the share of new customers no ad can claim.
4. Mentions and saves: people naming you, tagging you or keeping your posts.
5. "How did you hear about us": one question on your signup or checkout page.

Branded search is the strongest of the five, so it comes first, with the steps.

How to read branded search in Google Search Console, step by step

Nobody types the name of a brand they have never heard of. That is why a count of name searches is the closest thing to a free memory test. Google Search Console is free, and it shows that count for your own site.

1. Open Search Console and go to the Performance report for Search results. Google says the default view shows clicks and impressions for the past three months.
2. Click Add filter and choose Queries. Pick "Queries containing" and type your brand name the way people type it, in lower case, with no extra words.
3. Catch the misspellings. Choose "Custom (regex)" in place of "Queries containing". Google's help page says to separate several values with a vertical bar and put the whole thing in parentheses, like this: (yourname|your name|yourname app).
4. Read impressions before clicks. Google defines impressions as how many times your site appeared in search results. If your site shows up when someone searches your name, every one of those searches adds one impression. Clicks are how many of those people came through to your site.
5. Compare two periods. Add a Date filter and choose Compare. Google suggests the weekly or monthly view when you compare two date ranges.
6. Write down one number per month: branded impressions. Put branded clicks next to it.

There is also a shortcut. Google added a branded queries filter to this report in November 2025, and said in March 2026 that all eligible sites have it. It splits your queries into branded and non-branded for you. Two limits apply. Google says it only works on top level properties with enough queries and impressions, so a young site may not see it. Google also says an AI-assisted system decides what counts as branded, and that some queries may be misidentified. If you do not have the filter, the manual one above does the same job.

One caution on the totals. Google notes that filtering by query leaves out anonymized queries, so a filtered count runs lower than the real one. That is fine. You are watching the direction over months, so use the same filter every time.

A second caution is your own name. If the brand is called something like Bloom or Atlas, most people typing that word are looking for something else. Filter for the name plus what you sell, such as "bloom skincare" or "atlas app", and count only those.

Small numbers count. For a fine-dining restaurant we worked with, branded search went from 5 to 143 weekly impressions in 5 days, a 28x rise. Those are tiny figures, and they were still a clear reading. You do not need a big brand to see the line move. You need a starting number.

Google Trends is the other place to look, and it covers any name, including a rival's. But Google scales its figures from 0 to 100, so it shows relative interest and gives no counts. For your own brand, Search Console is the better tool.

The other four free signs, one at a time

Direct traffic

Google Analytics describes its Direct channel as users who arrive through a saved link or by entering your URL. Someone who types your address knows you exist. So a rising count of direct visits is a good sign.

Read it as a rough number. Google's own help page lists other visits that end up in the same bucket: links with no tracking tags, redirects that strip the tags, link shorteners, links inside PDF files, and visitors with ad blockers. So direct traffic always holds some visits that have nothing to do with memory. Watch the trend and ignore the exact figure.

Customers who arrived without an ad

Once a month, take your new customers and remove every one your ad platforms claim. Divide what is left by the total. That share is the part of your growth that no ad can take credit for.

It is a blunt number, because ad platforms claim sales generously. Keep the counting rule the same each month and that matters less. If the share grows while ad spend stays flat, more people are finding you on their own.

Read how to calculate customer acquisition cost →

Mentions and saves

Search your brand name inside TikTok, Instagram, YouTube and Reddit once a week. Count posts and comments by people who do not work for you. Add tags of your account and messages that mention you by name.

On your own posts, look past views to saves and shares. A view can be an accident. A save means someone wants to find you again, and a share means they put your name in front of a friend. Neither proves memory, but both are steps towards it.

"How did you hear about us" on signup

Add one question to your signup or checkout page. Give five or six fixed answers and one free text box: a friend, TikTok, Instagram, YouTube, a Google search, somewhere else. Make it optional so it does not cost you signups.

People answer this badly, and it is still worth asking. They forget the first place they saw you and name the last one. Watch the mix over time. If "a friend" and "a Google search" grow month after month, the name is travelling without you. Do not change the answer options once you start.

When a brand awareness survey is worth running

The free signs only see people who reached you. A brand awareness survey is the one tool that sees people who have not: those who know the name and have not searched yet, and those who have never heard of you.

That makes a survey worth the effort in three cases. You are about to spend real money or months on getting known and want a before and after. Your free signs disagree with each other and you cannot tell why. Or you sell in a category where people buy rarely, so searches are too few to read.

It is not worth running on your own list. A survey sent to your own newsletter will report that you are famous. Everyone on that list knows you already. The people to ask are people who use your category and are not your customers.

Reaching them is the hard part with no budget. Paid survey panels exist for this. The free routes are communities where your category gets discussed, and friends of friends who fit the buyer. Both lean towards people like you, so the result will be skewed. Ask the same kind of people, found the same way, every time. A skew that stays the same still lets you see change.

Brand awareness survey questions you can copy

These are the brand awareness survey questions I would start with. There are seven, and the order is part of the design. Swap the brackets for your own category and names.

1. Screening: "Which of these have you bought or used in the past three months?" List your category among four or five others. Keep only the people who pick yours.
2. Unaided recall: "When you think of [category], which brands or apps come to mind? Type as many as you can." Leave an open text box with no list.
3. Aided recognition: "Which of these have you heard of? Select all that apply." List your brand, four to six rivals and one name you made up.
4. Familiarity: "How well do you know [your brand]?" Offer four answers: never heard of it, heard the name only, know what it does, have bought or used it.
5. Source: "Where have you seen or heard about [your brand]?" Use the same answer list as your signup question.
6. Meaning: "In a few words, what does [your brand] do?" Leave an open text box.
7. Intent: "Next time you need [category], how likely are you to try [your brand]?" Offer five steps from very unlikely to very likely.

Ask the unaided question before any question that shows your name. Once a person has seen the list, you can no longer test what they recall alone. The same goes for the invitation. If your brand is in the subject line, question two is spoiled.

The made-up name in question three is a check on guessing. Some people tick names to be polite or to finish faster. If a share of them say they have heard of a brand that does not exist, take about that much off every other name.

Score each wave the same way. Unaided awareness is the share of people who typed your name in question two, misspellings included. Aided awareness is the share who ticked you in question three.

How many answers you need, in plain words

A survey of a few people gives a blurry number. SurveyMonkey's sample size table makes the blur concrete. At a 95% confidence level and for a large population, about 100 completed answers give a margin of error of plus or minus 10 points. About 400 give plus or minus 5.

Here is what that means for you. Say 12 of your 100 people name you in spring and 17 of 100 in autumn. It looks like growth. With a margin of 10 points on each reading, it could be nothing at all.

Comparing two surveys is blurrier still. Pew Research Center works through an example in which a lead moved by 3 points between two polls. The margin of error on that change was plus or minus 8 points. A change has to be large before you can trust it.

Those margins assume people were picked at random. A group you found through friends is less reliable than that. So with a small or informal sample, believe only big moves, and believe them more when the free signs moved the same way.

This also sets how often to run it. Monthly is too often, because a month of real change is smaller than the blur. Every three to six months is enough, plus once before and once after any large push. The free signs cover the months in between.

Brand awareness tracking: set a baseline, then check monthly

Brand awareness tracking is a habit more than a tool. It is one sheet, one row per month, filled in on the same day by the same person. The first row is the baseline, and it has to exist before you start whatever is meant to make you known.

For the baseline, use the three months Search Console shows by default and write down each month on its own. One month can be odd. Then add a row every month with these columns:

• Branded impressions and branded clicks, from the same Search Console filter every time.
• Direct visits, new visitors only if your tool splits them.
• New customers, and the share no ad platform claimed.
• Mentions by name that you counted by hand.
• The mix of "how did you hear about us" answers.
• Notes: what you launched, posted or paid for that month, and anything outside your control, such as press or a seasonal peak.

People skip the notes column, and it is what makes the rest readable. A jump in March means nothing later unless the sheet says what happened in March.

Read the sheet in runs of three months. One good month can be a single video. Three rising months in a row, across more than one column, is people learning the name. Also look at what happens after a push ends. If the numbers settle higher than where they started, some of the attention stayed.

Here is what that looks like on a real account.

Read the proof →

Mistakes that make brand awareness measurement useless

Most bad brand awareness measurement fails in one of five ways. Each one is cheap to avoid.

• Treating reach or views as awareness. Views count screens. Dashboards put them at the top because they always look good.
• Measuring once. A single reading has nothing to be compared with. The baseline and the second reading together are the measurement.
• Changing the question between waves. A new wording, a new answer list or a new group of people makes a new survey. The move you see may come from the edit.
• Counting your own people. Your team and your current customers search for you all day. Leave them out of surveys, and expect a floor of searches that is only you.
• Giving the last campaign all the credit. Name searches can show up weeks after the work that caused them. Read runs of months before you decide what worked.

The first mistake is the most common. Seven public campaigns show the difference.

Read brand awareness campaign examples →

If you want a second pair of eyes on your own numbers, ask for the free teardown. Send a link to your product and its channels. Within three days you get a written read on what brings customers today and what we would build first. No call, and the plan is yours whether or not we work together.