Original research - 3 July 2026
SEO/GEO Traffic Forecast 2026
A falsifiable cohort-level planning model for the Visa Atlas 500k/month SEO/GEO target, with ranges, assumptions, measurement commands and kill gates.
This is a planning model, not a live analytics dashboard. It exists so the 500k/month target can be audited against named cohorts, assumptions, commands and falsification gates.
500,000
Planning target
390,000-600,000
Model range
5
Traffic cohorts
24 months
Planning window
Methodology
Cohort ranges are planning ranges, not additive current analytics. The low/high totals are a portfolio target range after coarse overlap control, and authority lift is treated as a dependency rather than a separate traffic row.
Target-in-range check: passed; the 500k/month planning target sits inside the modeled low/high range.
No current traffic, ranking, CTR, AI-citation or market-share claim is published by this forecast.
Actual sessions, Search Console rows, referrer exports and experiment outcomes stay local unless converted into separately reviewed aggregate evidence.
- The model is falsifiable planning evidence, not a guarantee of rankings, traffic, citations or market share.
- The range must be rebased when reviewed GSC cohorts, AI-citation share, bot logs or authority-distribution evidence contradict an assumption.
- New page production must remain subordinate to source freshness, liveness, index-quality and reviewer controls.
Cohort forecast
| Cohort | Target visits/month | Role | Evidence surfaces |
|---|---|---|---|
Calculators and interactive tools calculators-tools | 90,000-140,000 | Profile-specific calculators and planning tools are the click-worthy surface that answer engines cannot fully replace with a static summary. | |
Fees, salary, processing, settlement and cost data pages data-reference | 110,000-160,000 | Reusable figure pages target high-intent how-much, how-long and minimum-salary queries while feeding answer capsules, citation packs and open-data reuse. | |
Freshness and policy-change pages freshness-policy | 50,000-80,000 | Dated rule-change, draw and figure-movement surfaces compete where stale blogs and model-training cutoffs are weakest. | |
Compare, ranking, corridor and profession pages after pruning longtail-navigation | 90,000-130,000 | Long-tail pages contribute only where each indexed URL has unique sourced value, source-basis blocks and a useful next action. | |
Research reports, open API, data reuse and authority distribution research-authority | 50,000-90,000 | Original research, open datasets, repository mirrors, publisher kits and authority distribution drive direct visits plus the off-site citations needed to lift every cohort. |
Assumptions and gates
Calculators and interactive tools
Assumptions
- High-intent calculator pages retain click demand because users need personal inputs, dated thresholds and route-specific caveats.
- Each flagship calculator keeps visible source provenance, deterministic logic and no approval prediction.
- Calculator pages cross-link to source-backed cost, salary, settlement and route pages so the click creates a deeper crawl path.
Falsification gates
- If flagship calculator/tool pages stay below 0.5% CTR at stable rank after the first reviewed GSC experiment window, rebase the cohort target.
- If calculator audits fail or source freshness goes red, freeze new calculator launches until provenance is restored.
Dependencies
- Current calculator source data remains fresh under source SLA.
- Rendered WebApplication schema and visible provenance stay intact.
- GSC stable-rank CTR improves after title, answer-capsule and internal-link tests.
Measurement commands
npm run audit:calculator-provenancenpm run audit:calculator-verdict-contextsnpm run seo:gsc-cohorts -- gsc-pages.csv --json --period-start=YYYY-MM-DD --period-end=YYYY-MM-DD --property-label=gsc-domain-redacted --search-type=web
Fees, salary, processing, settlement and cost data pages
Assumptions
- Each indexed data page exposes a unique source-backed figure or calculation, not a thin route permutation.
- Answer engines may satisfy some clicks, so value is counted through both organic sessions and citation/share evidence.
- The data layer scales only within source freshness and liveness capacity.
Falsification gates
- If stale red-count rises for two consecutive monthly reviews, stop adding data-reference pages.
- If GSC rows show impressions without citation or click lift after source-backed answer blocks, prioritise snippet and authority work over new page creation.
Dependencies
- Fee, salary, processing and cost records preserve primary source URLs and review dates.
- The public data catalog, Dataset/DataDownload JSON-LD and citation blocks remain crawlable.
- No stale or invalid high-risk figure record is published.
Measurement commands
npm run audit:data-citationnpm run audit:freshnessnpm run audit:rendered-seonpm run seo:gsc-cohorts -- gsc-pages.csv --json --period-start=YYYY-MM-DD --period-end=YYYY-MM-DD --property-label=gsc-domain-redacted --search-type=web
Freshness and policy-change pages
Assumptions
- Freshness-led pages can rank and be cited when they publish source-backed updates faster than evergreen guides can refresh.
- Policy updates and figure-change rows remain linked to official sources and affected canonical pages.
- Freshness pages avoid unsupported legal-advice or outcome-prediction language.
Falsification gates
- If fresh pages fail to enter citation-share evidence after two quarterly reviews, reduce the freshness traffic range.
- If a high-risk stale claim escapes the answer-key audit, pause publication until the source and correction workflow is repaired.
Dependencies
- Policy update and fact-change feeds retain old/new values, effective dates and source URLs.
- Citation benchmark prompts continue to cover volatile high-risk facts.
- Rendered policy and research pages keep canonical URLs and Article/Dataset schema.
Measurement commands
npm run audit:policy-claimsnpm run audit:monthly-figure-changesnpm run audit:express-entry-categoriesnpm run kpi:citation-share
Compare, ranking, corridor and profession pages after pruning
Assumptions
- Noindex and sitemap gates continue to suppress sparse fallback permutations.
- Surviving long-tail pages carry source-backed at-a-glance data, answer capsules or route-specific evidence.
- Internal linking sends users and crawlers to the strongest data, route and calculator pages.
Falsification gates
- If indexed-thin count rises or long-tail source-basis audits fail, halt long-tail expansion.
- If 50 thickened long-tail pages do not improve CTR at stable rank, deprioritise authoring until authority improves.
Dependencies
- Indexability manifest stays aligned with sitemap and page-level noindex decisions.
- Long-tail templates preserve source-basis and legal-scope blocks.
- Expansion is earned by verified corridor, profession or comparison data.
Measurement commands
npm run audit:thinnpm run audit:noindexnpm run audit:longtail-provenancenpm run audit:indexable-page-manifestnpm run seo:gsc-cohorts -- gsc-pages.csv --json --period-start=YYYY-MM-DD --period-end=YYYY-MM-DD --property-label=gsc-domain-redacted --search-type=web
Research reports, open API, data reuse and authority distribution
Assumptions
- Original computed datasets produce citable claims competitors cannot easily copy from ordinary advisory prose.
- Off-site reuse preserves canonical backlinks, source URLs, license terms and reviewer evidence before any public success claim.
- Traffic lift depends on earned authority; direct research visits are only part of the value.
Falsification gates
- If two consecutive flagship reports earn no approved citation, repository or publisher evidence, reallocate effort to product and data-page quality.
- If any approved authority row lacks source URLs, canonical backlink or reviewer fields, fail the local evidence command and do not publish the claim.
Dependencies
- Publisher, repository and authority evidence rows require reviewed public evidence before claims are published.
- Every report keeps CiteDataset blocks and Dataset/DataDownload schema.
- External distribution avoids bought links, guaranteed-backlink language and unapproved launch claims.
Measurement commands
npm run audit:data-front-doornpm run audit:authority-distributionnpm run audit:publisher-citation-kitsnpm run audit:repository-mirrorsnpm run seo:authority-distribution -- authority-distribution-log.jsonl --jsonnpm run seo:publisher-citation-kits -- outreach-log.jsonl --json
Cite or reuse this dataset
This forecast model is free to reuse under CC BY 4.0. Cite the JSON endpoint for the machine-readable cohort assumptions and the report page for methodology and caveats.
Suggested citation
Visa Atlas, "Visa Atlas SEO/GEO traffic forecast model", https://visaatlas.org/research/seo-geo-traffic-forecast-2026. Last verified 3 July 2026.
- JSON endpoint
- https://visaatlas.org/api/public/traffic-forecast