End-to-End Workflow: Owned, Client, and Competitor Projects

This guide ties everything together with three complete, realistic workflows — an owned-site audit, a client-authorized engagement, and competitor research — so you can see how projects, targets, crawl jobs, governance, and review fit together, and how the three scenarios differ.

The shared arc

Every StretchSearch engagement follows the same arc:

1. Create a project with the right type.
2. Register targets with honest authorization and robots metadata.
3. Draft a crawl job with tight scope and safety limits.
4. Complete the readiness checklist and get approval.
5. Operator enables execution; pages are indexed.
6. Review SEO findings, candidates, ranking previews, and evidence.
7. Feed the suite — audits (StretchBAS), SEO work (StretchSEM/SEO), or AI grounding.

What changes across scenarios is the governance posture: authorization, risk, scope, and how conservative your limits should be.

Scenario 1: Owned-site audit

1. Create a project of type owned_site (e.g., Company Site — Standing SEO).
2. Register your domain as a target: target type owned_property, authorization owned_property, robots reviewed_allowed, risk low.
3. Draft an owned_site_deep_crawl job. Because you own the site, you can use broader scope (same_domain_standard) and a higher page cap — but still set a courteous rate limit.
4. Complete the checklist (authorization is trivial here) and approve.
5. After indexing, review findings and rankings. Owned sites are where you'll act most directly on fixes.

Posture: lowest friction, broadest allowable scope, focus on remediation.

Scenario 2: Client-authorized engagement

1. Create a project of type client_site (e.g., Northwind Coffee — Q3 SEO).
2. Register the client's domain: target type client_authorized, authorization client_authorized. Paste the written-permission reference into notes. Review robots and set it accurately. Fence with allowed URL prefixes.
3. Draft an seo_content_crawl job. Keep scope moderate (same_domain_shallow), set a sensible page cap, and a 1,000–2,000 ms rate limit.
4. Work the readiness checklist carefully — the "authorization on file" item matters here — and get the engagement lead's approval (recorded in approved_by).
5. After indexing, assemble active, citable evidence for a client-facing audit and review findings/rankings.

Posture: medium friction, permission is documented and auditable, scope is fenced.

Scenario 3: Competitor research

1. Create a project of type competitor_research.
2. Register competitor domains: target type competitor_research, authorization public_research_only, robots reviewed_limited (respect it), risk high.
3. Draft competitor_crawl jobs scoped tightly to seed_urls_only, low page caps (e.g., 100–150), and a slow rate limit (2,000 ms+).
4. Confirm the research is contractually and legally permitted before completing readiness; have a reviewer approve.
5. After indexing, use candidate indexing (entities, phrases) and ranking previews to compare topical authority — not to copy content.

Posture: highest friction, narrowest scope, strictest review. When in doubt, stay public_research_only and seed_urls_only.

A side-by-side summary

• Authorization: owned (owned_property) → client (client_authorized) → competitor (public_research_only).
• Risk: low → low/medium → high.
• Scope: standard → shallow → seed-only.
• Page cap & rate: generous → moderate → minimal & slow.
• Review depth: light → documented → strict.

Tips

• Keep one standing owned_site project you never archive, and spin up time-boxed projects for engagements.
• Always fence client and competitor targets with allowed URL prefixes — it's your strongest scope control.
• Escalate caution with risk: high-risk work gets narrow scope, slow rates, and closer sign-off.
• Assemble client deliverables from active + citable evidence so nothing stale or revoked slips into a report.

FAQ

Can one project mix owned and competitor targets? Technically possible, but don't — the authorization and risk postures differ. Separate projects keep governance clean.

What if permission changes mid-engagement? Mark the target blocked (or its authorization blocked) immediately and let related evidence move to revoked. The record of what was permitted, and when, is preserved.

Where does the indexed evidence go next? Into the suite: StretchBAS for audits, StretchSEM/SEO for optimization work, and the AI evidence bridge for grounded answers.