Require six workstreams: baseline, technical controls, query-to-page architecture, content and evidence, authority/distribution, and measurement/governance. For AI discovery, require sound SEO foundations and useful original content—not a collection of “GEO hacks.”

1. Baseline and decision map

Ask the agency to define:

• priority audiences, markets and buying questions;
• current Search Console queries, pages, clicks, impressions, CTR and position;
• analytics and conversion events that can be trusted;
• branded versus non-branded demand;
• existing service, product, documentation and editorial URLs;
• known migrations, redirects and indexing risks.

The output should include a query-to-page map and a list of unknowns. Aggregate keyword volume alone cannot reveal which Crynet page already owns an intent.

2. Technical controls

The RFP should cover crawl access, indexability, canonicals, redirects, status codes, rendering, sitemaps, robots directives, internal linking, structured data and performance where relevant.

Require reproducible evidence: affected URLs, method, severity, expected decision impact, recommended change, owner and validation test. A list of automated-tool warnings is not a technical roadmap.

3. Content portfolio decisions

Every important page should receive a job: keep, improve, consolidate, redirect, create or monitor.

The agency must check proposed content against service pages, legacy posts and planned articles. If two pages answer the same query with the same promise, choose one canonical destination.

For new content, require a brief with reader, situation, primary question, one-sentence answer, original contribution, evidence, internal links, CTA, claim risks and refresh trigger.

4. Evidence and entity consistency

Blockchain and crypto topics frequently involve volatile products, policies, regulation and market data. Require a claim ledger with source, date, geography, owner and limitation.

The agency should also reconcile company, product, protocol, people and proof across the website and credible external profiles. Structured data must match visible content; it cannot repair contradictory claims.

5. AI-search deliverables without gimmicks

Google’s current guidance says that established SEO best practices remain relevant to generative search features. It prioritizes unique, useful, non-commodity content and explicitly warns against superficial tactics such as unnecessary AI files or inauthentic mentions.

A defensible AI-search scope can include:

• crawl and index eligibility;
• clear definitions and entity relationships;
• original evidence and labelled frameworks;
• citations placed next to claims;
• consistent visible structured information;
• controlled monitoring of citations, referrals and branded demand;
• documented limitations.

It cannot guarantee inclusion, wording, ranking or a recommendation by an AI system.

6. Measurement and governance

Require a baseline, reporting dictionary and decision rules. Track page/query performance, indexed canonicals, qualified organic sessions, internal progression and leads where measurement is reliable.

AI visibility should be reported by observable platform evidence, referrals and controlled samples—not collapsed into an unexplained proprietary score.

Define who owns Search Console, analytics, tag management, content, source files, structured data and dashboards. Specify approval, deployment, rollback and exit procedures.

A 100-point RFP scorecard

Reject any proposal that guarantees rankings, treats every tool warning as equally important or proposes new pages before checking the existing portfolio.

What Crynet can help decide

Crynet can scope Web3 SEO across technical foundations, content consolidation, evidence and AI-search measurement while preserving valuable existing URLs.

Send the site, priority markets, Search Console access or exports, analytics baseline and current content inventory. We can return a decision map showing what to fix, consolidate, create and measure.