SEO KPI Weekly Review

Snapshot — 2026-09-14 (measurement)

Owned pages for competitor-first queries shipped this day (/alternatives/*, retargeted /compare/*, five buyer posts, /glossary/saas-boilerplate-pricing-2026). Dump Search Console Queries after they have been indexed for ~14 days and replace inferred demand with real impressions.

Search Console and Bing verification

app/layout.tsx already emits Google and Bing verification tags from GOOGLE_VERIFICATION_CODE and BING_VERIFICATION_CODE. Confirm both are set in production env (see .env.example). Do not commit the tokens. Confirm the property still shows Verified in Google Search Console and Bing Webmaster Tools. This checklist cannot log into those accounts for you.

EventDash goals (landing)

Keep checkout_click on Get NextBento buttons. Configure these additional data-ed-goal names as EventDash goals if they are not already:

Monthly 15-query baseline

Source of truth: lib/seo/query-baseline.ts. Once a month, run each query in Google, ChatGPT (with search), and Perplexity (n=3–5). Record citation / top-10, not anecdotes.

1. next.js saas boilerplate
2. next.js boilerplate saas
3. nextjs saas boilerplate
4. next.js saas starter kit
5. next.js saas template
6. saas starter kit
7. next.js boilerplate
8. best next.js saas boilerplate 2026
9. shipfast alternative
10. best shipfast alternatives
11. makerkit alternative
12. supastarter alternative
13. indie kit alternative
14. next.js 16 saas boilerplate
15. next.js supabase stripe boilerplate
16. cursor saas starter

Weekly SEO Review Checklist

Snapshot — 2026-09-09

Use this as the baseline. Re-check Pages first. Do not treat Queries as demand until branded or commercial rows appear.

Google Search Console (Web, last 28 days)

Export: https___www.nextbento.dev_-Performance-on-Search-2026-09-09 (through ~2026-09-06 on the chart).

• 3 clicks, ~647 impressions, 0.32% desktop CTR.
• Clicks: homepage (2), /docs (1). Both measured clicks are Poland.
• United States: 259 impressions, 0 clicks (docs dump, not buyers).
• No branded nextbento and no next.js saas boilerplate in the query export.

Pages with impressions (0 clicks unless noted):

Ignore operator site: tests and env-var how-tos (next_public_supabase_publishable_key, resend api key). Official vendor docs own those SERPs.

EventDash (last 30 days, 2026-08-11 to 2026-09-09)

• 134 unique visitors, 202 pageviews, 173 sessions, 92.5% bounce, 1,124 events.
• Top paths by views: / (73), /docs/google-brand-verification (12), /docs/cloudflare-setup (8), /docs/supabase-setup (7), /pricing (6), /compare (5). /for/* is not in the head of the list.
• Referrers by events: Direct/None (517, includes some ChatGPT UTMs), Google (30), Bing (25), t.co (12), Stripe Checkout (9), chatgpt.com (8).
• Goals: none configured before this round. Checkout buttons now send checkout_click via data-ed-goal.
• Events are pageview / session / heartbeat / scroll / click. No named conversion events yet.

Read of the gap: EventDash visitors (134) vs GSC clicks (3). Acquisition is not Google Search yet. Bing and ChatGPT already send more useful sessions than organic Google. Do not spend the next month ranking for Resend DNS.

Google Search Console

Start on the Pages report, not Queries. At low volume, Queries is mostly noise.

Pages

• Check impressions, clicks, CTR, and average position for:
  • /compare and /compare/*
  • /for and /for/*
  • /blog and /blog/* (shipped 2026-09-09)
  • /integrations/*
  • /glossary/*
  • /docs/*-setup only as crawl proof, not as a KPI
• Review indexing coverage and newly discovered pages.
• Validate no sudden drops by page group.
• After deploy, confirm /sitemap.xml returns 200 (this origin was 200 with a Vercel cache HIT on 2026-09-09; resubmit in GSC and Bing anyway).

Queries — filters

Ignore rows that are not demand:

• Operator searches (-site:reddit.com, quoted URLs, similar). These are often self-tests and inflate impressions without clicks.
• Generic vendor how-tos we cannot realistically win (how to add domain in resend, docs resend, vercel.app). Official vendor docs own those SERPs.
• Treat a handful of long-tail doc queries (crawler hints cloudflare, next_public_supabase_publishable_key) as crawl proof, not a traffic channel.

Watch for branded (nextbento) and commercial (next.js saas boilerplate, vs competitor) queries. Those are the ones that matter.

Bing Webmaster Tools

• Review indexed pages count trend.
• Review crawl stats and top crawl errors.
• Verify sitemap processing and freshness. EventDash already shows Bing as a referrer (25 events / 30d).

IndexNow

• Confirm submit:indexnow runs after major content updates.
• Confirm updated URLs are included in submissions (blog paths included in the submit script).

AI Search Proxy Metrics

• Track referral sessions from AI assistants in analytics (utm_source=chatgpt.com, referrer chatgpt.com).
• Track branded query growth and direct traffic trend.
• Track conversions by landing page cluster (EventDash checkout_click, compare_view, alternatives_cta, docs_to_checkout).
• Compare EventDash visitors to GSC clicks. If analytics shows visitors and GSC shows ~0 clicks, acquisition is not Google yet.

Escalation Triggers

• CTR drops > 20% week-over-week on key pages (not operator queries).
• Sudden index coverage declines.
• Important pages move from indexed to excluded.
• Compare/integration pages lose impressions for core commercial queries (next.js saas boilerplate, vs competitor names).
• /sitemap.xml returns 5xx.