What you’ll ship

• A master CSV with clusters, intent, targets, and owners
• One Frase document per page with outline and entity checklist
• Assigned briefs with clear acceptance criteria

Guardrails

• One job per URL
• Descriptive internal links to hub and siblings
• Evidence links where claims are made

Helpful references

• Helpful content
• Structured data
• Crawlable links

Accounts

• Ahrefs or Semrush (keyword export)
• Frase (research, briefs, editor)
• Optional: a clustering tool for large datasets

Decisions up front

• Intent taxonomy: informational, commercial, transactional, navigational
• Stages: TOFU, MOFU, BOFU
• Page types: guide, comparison, solution, template, FAQ

Acceptance criteria

• Evidence links for claims
• FAQ only if Q&A appears on page
• Routes to hub and two siblings

Suggested fields to export

• keyword
• search_volume
• kd (keyword difficulty)
• cpc_usd
• parent_topic
• serp_features
• top_serp_urls (if available)

Tips

• Run separate exports by market if you plan international pages
• Use consistent locale and device settings
• Keep raw exports unchanged for audit trails

Normalization checklist

• Lowercase keywords and strip extra spaces
• Unify spelling where appropriate
• Standardize locale (en-us vs en-gb)

Deduping tips

• Strip punctuation before comparing
• Remove near-duplicates like “guide” vs “guides” if intent is same
• Keep branded terms separate

Mark exclusions

• Misaligned industries or buyer roles
• Low-value or irrelevant queries
• Queries you can’t serve credibly

What good clusters look like

• Shared SERP across top results
• One dominant job to be done
• Clear page type (guide, comparison, solution)

Optional tool

For large lists, you can use a clustering tool to speed up grouping and intent tags. Export the results to merge into your master sheet.

Simple scoring model

• Demand: normalize search_volume
• Difficulty: invert kd to reward easier wins
• Value: normalize cpc_usd or pipeline fit
• Fit: 1–5 based on product alignment
• Freshness: 1–5 if timely

Example weights

• Demand 30%
• Difficulty 20%
• Value 20%
• Fit 20%
• Freshness 10%

Result

Pick the top 10 clusters and assign owners. Re-score quarterly or when the SERP shifts.

Brief header (paste into the doc)

Head term: <cluster_name or top keyword>
Intent: informational / commercial / transactional
Stage: TOFU / MOFU / BOFU
URL target: /<slug>/
Primary CTA: <one action>
Sources to cite: [link, link, link]

Pull SERP and outline

• Scan recurring H2/H3 patterns
• Lift entities and common definitions
• Add 3–6 FAQs that appear on page

Use Google’s helpful content guidance when choosing what to keep from the SERP.

What to include

• Outcome-focused headline
• Outline with verbs in headings
• Entity checklist and definitions
• 3–6 FAQs that appear on page
• Primary CTA that fits the stage

Routing plan

• Link to hub near the top
• Two sibling links near the end
• Breadcrumbs and crawlable anchors

Quality bar

• Every claim has a reputable source
• No duplicate paragraphs across pages
• Natural tone with specific examples

Hand-off

• Share the Frase doc and the cluster row
• Confirm acceptance criteria
• Set a mid-way review checkpoint

Measure what matters

• Cost per shipped page
• Time to publish
• Impressions, clicks, and page routes after publish

Readability

• Short intro that sets the job of the page
• Scannable H2s and clear steps
• Tables for quick comparisons

Integrity

• Citations for claims, link the exact page
• FAQ appears on page if you use FAQ schema
• Stable slugs and descriptive anchors

Technical

• Article or BlogPosting schema as relevant
• Validate with the Rich Results Test
• Links crawlable per Search Central

Sheets formulas

• Normalize scores with MIN–MAX scaling
• Auto-generate slugs from titles
• Flag overdue items with conditional formatting

Exports

• Export fresh keywords quarterly
• Append to master sheet with dates
• Re-score and re-prioritize top clusters

Light scripts

• Batch-format titles and slugs
• Clean punctuation and spaces
• Create doc names from cluster_name

Standard Operating Procedure (SOP)

1. Export keywords from Ahrefs or Semrush with volume, KD, CPC
2. Clean and normalize; dedupe near duplicates
3. Cluster and assign intent and stage
4. Construct the master CSV with routing and owners
5. Pick top 10 clusters by priority_score
6. Create one Frase doc per target URL
7. Generate brief from SERP in Frase
8. Assign to writer with acceptance criteria
9. QA content, add schema if appropriate, route links
10. Publish and measure; refresh quarterly

CSV template (headers)

keyword,cluster_id,cluster_name,intent,stage,search_volume,kd,cpc_usd,source,top_serp_urls,notes,url_target,title_draft,slug_draft,h2_outline,entities_to_cover,questions_to_answer,internal_links_from,internal_links_to,owner,status,due_date,priority_score

If your browser previews CSV files, right-click the button and “Save link as…” to download.

Do I need clustering for every page

No. Use clustering when you have overlapping queries and want to avoid duplicate targets. For unique or branded topics, you can go straight to Frase.

Can one page serve multiple intents

Pick a primary intent. If the SERP shows mixed jobs, split into siblings and route between them with descriptive anchors.

What score should I aim for in Frase

Treat scores as coverage checks, not targets. Favor clarity, examples, citations, and correct routing.

Where should I add schema

Use Article or BlogPosting for most guides. Add FAQPage only if the Q&A appears on the page. Validate using the Rich Results Test.