usdigitalmarketingindex.com SEO
Which SEO metrics are appropriate for a campaign primarily intended to increase brand awareness?
Updated November 15, 2026
Measure exposure to people who did not search for you, and recognition among people who then do. In practice that means four metrics: non-branded impressions, the number of distinct non-branded queries producing impressions, branded impressions and clicks over time, and new users arriving from organic search. Leads and revenue are the wrong primary metrics for this goal, because an awareness campaign succeeds when someone learns the name, which can happen without a click. Every one of these metrics also moves for reasons unrelated to SEO, so each needs its caveat written next to it.
Brand discovery signals
Discovery happens when your page appears for a search that did not contain your brand. Two numbers describe it.
- Non-branded impressions: how many times your pages were shown for queries without the brand. This is reach.
- Count of non-branded queries with impressions: how many different needs led to an appearance. Growth here means the brand is being seen in new contexts, not shown more for the same query.
Clicks matter less at this stage than they do in a lead campaign. A person who sees the name in results three times this month has been reached whether or not they clicked.
Search Console and analytics indicators
The dashboard below names where each of the four numbers comes from and what it does not show. Its last row is an optional fifth line for sites whose links appear in generative AI features.
| Metric | Source | Definition | Caveat |
|---|---|---|---|
| Non-branded impressions | Search Console performance report, non-branded view | Appearances for queries not classified as branded | Classification is automated and can misidentify queries |
| Non-branded queries with impressions | Search Console query table | Count of distinct queries | The table is limited in rows and leaves out rare queries |
| Branded impressions and clicks | Search Console performance report, branded view | Appearances and clicks for brand-related queries | Rises with any marketing, not SEO alone |
| New users from organic search | GA4 acquisition reports | First-time visitors in the Organic Search channel | Counts visits, so it misses exposure without a click |
| Impressions in generative AI features | Search Console generative AI performance report | Times links to the site were shown in a generative AI feature | Reported as impressions |
For the channel row, Google’s definition of Organic Search in GA4 includes visits from AI Overviews and AI Mode (GA4’s default channel definitions), so those visits are already inside the number.
Here is how a quarter might read. The figures are hypothetical and exist to show the interpretation.
| Metric | Start of quarter (hypothetical) | End of quarter (hypothetical) |
|---|---|---|
| Non-branded impressions | 40,000 | 62,000 |
| Non-branded queries with impressions | 310 | 520 |
| Branded impressions | 2,100 | 2,300 |
| New users from organic search | 1,150 | 1,480 |
In this hypothetical quarter reach grew by 55 percent and the number of distinct queries by about two thirds, while branded searches moved little. The honest reading is that more people were exposed and that recognition has not yet shown up as searches for the name.
Separating branded from generic exposure
Search Console provides the split. Google’s announcement of the branded queries filter describes two views, branded and non-branded, and says the classification is “NOT based on a regular expression method” and comes from an “internal, AI-assisted system” in which “some queries may occasionally be misidentified” (Search Console branded queries filter announcement). The same announcement limits the filter to top-level properties with enough query volume.
Two consequences follow for an awareness dashboard. Report both views side by side, because a total that mixes them hides which one moved. And treat the branded line as an association. A rise in branded searches during a television campaign, a product launch, or a news story is real and is not evidence about SEO. Note concurrent marketing activity on the same timeline as the chart.
Setting up and maintaining that split as a regular report is a task of its own, described in the guide to reporting branded and unbranded search clicks separately.
Reviewing qualitative outcomes
Numbers from search tools cannot show whether people remember the brand. Add evidence from outside them.
- Ask new customers how they first heard of you, and record the answers in fixed categories.
- Ask sales and support staff to log when a prospect mentions having read something of yours.
- Track mentions of the brand on other sites that appear after content is published.
- Review which pages new visitors read first, to see what the brand is becoming known for.
None of this is proof of cause. Taken with the dashboard, it shows whether the direction is consistent.
One caution governs the whole stack: agree on these metrics before the campaign starts. A campaign judged afterward on leads it was never designed to produce will look like a failure, and a campaign with no agreed measure cannot be judged at all.
Add your comment to this answer