Branded versus Non-branded Split

Measure how much of one property's search traffic is brand demand and how much is non-branded SEO, how each side is trending, and what share of the period's growth each side explains. The user's request may contain a property (a GSC site URL or a recognizable name), a date range, the brand name or a list of brand terms to save, or a focus such as "non-branded only for the client report".

Workflow

1. Resolve the property. If the user's request or the conversation names a property, use it. Site URLs use GSC formats: sc-domain:example.com or https://www.example.com/. Otherwise call list_sites and pick the obvious match. If several plausible properties exist, ask the user which one before continuing.
2. Anchor and windows. Call get_site_summary with days: 28, or with startDate/endDate when the user gave dates. Use the last 28 days compared with the previous period unless the user specifies dates.
   • Read current, previous and deltas, and keep range.startDate/range.endDate and comparison.startDate/comparison.endDate: the split tool only compares periods when both comparison dates are passed, so these four dates are what keep its comparison aligned with the anchor.
   • GSC data lags about 2 to 3 days: never treat the most recent 2 days as complete, and if a dip appears at the very edge of the range, say it is likely reporting lag, not a real drop.
3. Is a brand filter defined? Call get_filter_options with siteUrl. This call reads configuration only and fetches no Search Console data.
   • Read brand.configured, brand.keywordCount and brand.skippedKeywords (saved keywords the filter could not compile).
   • When brand.configured is false, or the response carries emptyBecause, no branded keywords are saved: go to step 5 to propose a list before running the split, and say the split cannot run until it is saved.
4. The split. Call get_branded_performance with siteUrl, startDate/endDate set to the current window, comparisonStartDate/comparisonEndDate set to the comparison window (both are required for a comparison; omitting them gives no baseline), and limit: 25.
   • If it answers notConfigured: true, read reason: no_branded_keywords means none are saved (step 5), and no_usable_keywords means none of the saved ones compile; savedKeywords lists the stored keywords verbatim and hint says what to fix.
   • Otherwise read branded, nonBranded and total (each clicks, impressions, ctr on 0-100, position), brandedClickShare (0-100), anonymized (the part of nonBranded with no query row), daily[] (date, brandedClicks, nonBrandedClicks, brandedShare, oldest first), topBrandedQueries and topNonBrandedQueries (query, clicks, impressions, ctr, position), previous (branded, nonBranded, total, brandedClickShare for the comparison window), brandedKeywords (total, used, skipped) and savedKeywords.
   • Two caveats the tool states: nonBranded is the property total minus branded, so the two sides always add up to total; and Search Console leaves anonymised queries out of any query-filtered read, so those rare queries land in nonBranded, brandedClickShare is a floor, and coverage.hiddenClicksShare (0-1) is the share of clicks no query row carries. emptyBecause means the window holds no traffic at all, not that nothing is branded.
5. Brand variants missing from the saved list. Scan topNonBrandedQueries for the brand name misspelled, split or concatenated, the domain name, product or model names only this brand uses, and "brand + login/pricing/reviews" phrasings; each is a variant the saved list misses.
   • Then call query_top_queries with brand: "non_branded" and limit: 500 (the brand scope needs saved keywords and a paid plan; without either, call it unscoped with limit: 500) and scan again. Page with offset set to pagination.nextOffset only if the first page is dominated by brand terms.
   • When no keywords are saved yet, propose a starter list from this scan: the brand name, its common misspellings, the domain without its extension, and product names unique to the brand.
   • Keep generic words out of the list. A keyword is a case-insensitive substring, so "pro" or "shop" would swallow non-branded queries; a term wrapped in slashes like /acme ?corp/ is a regex (RE2 syntax, no lookahead or backreferences, \b supported), so /\bpro\b/ matches the word on its own.
6. Save or extend the keywords (write step). update_site with brandedKeywords replaces the stored list wholesale, so the list to send is savedKeywords from step 4 plus the additions, never the additions alone.
   • ALWAYS list the exact keywords and ask the user to confirm before calling. Explain that the call changes the property's saved branded keywords in GSC Wizard (the web app's reports and this split read the same list) and changes nothing in Search Console.
   • Then call update_site with siteUrl and brandedKeywords (strings of at least 1 character, at most 500 entries); the response echoes site.brandedKeywords. A keyword the filter could never use makes the call fail with nothing saved and the message names it; fix it and resend.
   • This step needs write access, and on a property shared with the user it needs full access. If it is refused because the connection is read-only, say that this step needs write access and continue with the read-only parts. After a successful save, re-run step 4 so the split reflects the new list.
7. Trend and growth attribution. From daily[], compare brandedShare over the first 7 days with the last 7 days, and name any day where either side moved sharply (deploys, campaigns and PR spikes show up as branded jumps).
   • Derive, and say you derived, from get_branded_performance: the branded delta (branded.clicks minus previous.branded.clicks), the non-branded delta (nonBranded.clicks minus previous.nonBranded.clicks), the total delta, and each side's share of the total delta when the total moved in one direction. When the two sides moved in opposite directions, report both deltas and say the split cannot be expressed as a share.
   • Growth carried mostly by brand is demand (campaigns, PR, seasonality), not SEO; growth carried by non-brand is SEO; a non-branded delta near zero while branded grows means the organic programme is not yet contributing.
   • Cross-check the sum of the two sides against deltas.clicks from get_site_summary and say so if the two disagree by more than rounding (the summary's window or scope then differs).

Interpretation rules

• ctr values are already percentages on a 0-100 scale, and so are brandedClickShare and the daily brandedShare. Never multiply them by 100. coverage.hiddenClicksShare is a 0-1 fraction, so multiply that one by 100 to present it.
• position: lower is better. A change toward a lower number is an improvement; never color it as a decline.
• Judge by clicks and impressions at stake, not by percentage alone: a brand variant with 3 clicks is not worth a keyword entry unless it is a misspelling the list should cover anyway.
• Present every multi-row result as a compact markdown table (top rows only, typically 5 to 10); keep prose for interpretation, not for restating numbers.
• Do not fabricate rows the tools did not return. If a side has no top queries in the window, say so.

Output

Present, in this order:

1. Anchor: one sentence with the window, the comparison window and the property's clicks, impressions, CTR and position with deltas (get_site_summary).

2. Split table (get_branded_performance), followed by the branded click share now versus the prior window and the keyword count (used of total, with skipped named):

   Side	Clicks	Prior clicks	Delta	Impressions	CTR	Position
   Branded / Non-branded / of which anonymised / Total

3. Trend: branded share in the first and last week of the window, and any sharp day, from daily[].
4. Top queries per side: two tables (query, clicks, impressions, CTR, position) from topBrandedQueries and topNonBrandedQueries, with the suspected brand variants in the non-branded table flagged.
5. Growth attribution: the derived deltas and shares from step 7, with the sentence that says whether the period's growth is brand demand or SEO.
6. End with a short prioritized list of 3 to 7 concrete recommendations: the keywords to add (or the starter list to save), the non-branded queries and pages to grow when SEO is not yet contributing, branded queries with a weak CTR or position to protect (a competitor or a marketplace listing above the brand's own page), and whether the client report should carry the non-branded series alone.

Error handling

If any tool call fails with an authentication or subscription error, tell the user to run /mcp to (re)authenticate, and that a GSC Wizard account with an active plan or trial plus a connected Google Search Console account is required. If the property is not found, re-run list_sites and offer the closest matches instead of guessing a URL format. If a single tool fails for another reason, report the gap in the relevant section and complete the rest of the workflow with the remaining data.

Grounding

Every number you report must name the tool it came from and the date range it covers, for example "clicks 12,340 (get_site_summary, 2026-09-21 to 2026-09-27)". For a table, state the tool and date range once on the line above it. If you derive a figure yourself (a delta, a sum, a share), say so and name the tool outputs it was computed from. Never report a number no tool returned.