Branded and nonbranded search can answer different growth questions. A query naming your company may reflect existing awareness, while a category query may reflect a broader discovery task. Report them separately before interpreting an overall traffic or conversion trend.
Build a documented classification rule rather than guessing from a campaign name. Keep the raw query, assigned category, rule version, and ambiguous cases. A campaign labeled nonbrand can still require a query-level review of what it actually captured.
Google's Search Console performance overview, checked September 9, 2026, describes query and traffic analysis. The taxonomy below is an original reporting method that can be applied to appropriate exported data; it is not a claim that every report exposes identical rows or dimensions.
Create an explicit brand vocabulary
Record the verified company name, product names, domain forms, meaningful misspellings, and former names that belong in the analysis. Separate competitor names and ambiguous common words.
For a fictional approval product called North Ledger, the analyst might include the exact brand, its domain, and a verified common misspelling. The word ledger alone would remain a category or ambiguous term depending on context, not automatically a branded query.
| Invented query | Classification | Reason |
|---|---|---|
| North Ledger pricing | Branded | Names the fictional product |
| purchase approval software | Nonbranded category | Describes the job without naming the brand |
| ledger approval | Ambiguous pending review | Could be a generic phrase |
| Competitor B alternatives | Competitor-related | A separate discovery context |
Keep exclusions visible. An internal employee query, support request, or unrelated homonym may need its own treatment rather than being forced into the acquisition comparison.
Report the groups with matching units
Use the same period and metric definitions for each group. Show impressions, clicks, relevant conversions, and qualification evidence only where the source supports those measures. Keep paid and organic observations separate unless the report explicitly reconciles them.
Do not divide a classified subset by an unrelated total and call the result complete coverage. State which rows were available, which were classified, and which remain unclassified or absent from the export.
If a brand rule changes, preserve its version and recalculate comparable periods when appropriate. An apparent increase in branded demand may otherwise be caused by adding a new product alias to the rule.
See how a changing mix improves the overall rate
Consider two hypothetical periods. In the first, branded search produces 200 clicks and 20 conversions, while nonbranded search produces 1,000 clicks and 30 conversions. The combined rate is 50/1,200, or about 4.17%.
In the second period, branded clicks double to 400 with 40 conversions. Nonbranded results remain 1,000 clicks and 30 conversions. The combined rate rises to 70/1,400, or 5%, even though each group's own rate remains unchanged: 10% branded and 3% nonbranded.
The overall improvement is a mix effect in this invented example. It does not establish that the landing page or nonbranded acquisition became more effective. The breakout makes that distinction visible and points the review toward the reason branded demand increased.
Interpret the mix before celebrating the total
Suppose a hypothetical month has rising branded clicks while nonbranded clicks are flat. That can be useful evidence of stronger brand-related demand, but it does not show that category discovery improved. Other marketing or business activity may contribute to the brand interest.
Likewise, a higher nonbranded click count is not automatically better if the queries are less relevant. Review the buyer task and downstream quality alongside the group totals.
Use the decision-focused reporting guide to connect the breakout to action. A branded view should help the team distinguish awareness, category discovery, and audience quality without attributing every change to the same cause.