Create topical authority without content sludge
The content strategy: a compact topic system that supports revenue pages, demonstrates expertise, and avoids low-value publishing.
Publish a small number of genuinely expert pieces that support your money pages, instead of a large number of interchangeable posts that support nothing.
You will have a 90-day editorial plan where every article has a buyer job, evidence source, and revenue-page connection.
Topical authority is not publishing every keyword in a cluster. It is demonstrating complete, useful expertise around the decisions your product helps buyers make.
IF YOU SKIP IT / Volume publishing dilutes the exact authority you are trying to build: 400 thin posts make the model less sure what you are an expert in, not more.
- Product-led content
- Editorial content that shows the product doing the job inside the how-to, at whichever of three tiers the topic honestly supports: the product is the answer, one route, or absent. Lesson 4 here; the money-page version is Module 03.
- Content strategy
- The decision about what to publish, in what order, for which revenue page. This module is the content strategy: a small topic system aimed at buyer decisions, not a calendar of keywords.
- Topical authority
- Being the source engines and buyers expect to answer a subject, earned by depth on one territory rather than breadth on many.
- Cluster
- A money page plus the supporting articles that answer the questions around its decision, all linking to each other.
- Information gain
- Material that adds something competitors do not provide, such as original data, a tested workflow, or expert evidence. It improves usefulness; citation and ranking effects still need measurement.
- Content decay
- The slow traffic loss of aging pages. The refresh queue in this module is the antidote.
Each phase feeds the next. Don't skip ahead, the artifact at the end is only trustworthy if every phase ran.
- 01Anchor
Choose the revenue page and buyer decision the cluster must strengthen.
- 02Research
Review search results, AI citations, communities, sales calls, and internal expertise for gaps.
- 03Differentiate
Commit to original data, a real workflow, a tool, a visual teardown, or expert synthesis.
- 04Brief
Specify answer sections, evidence, examples, visuals, internal links, reviewer, and conversion path.
- 05Produce
Capture expert input before drafting and retain source notes for fact checking.
- 06Distribute
Plan sales enablement, community participation, partner outreach, and newsletter reuse.
- 07Refresh
Update based on rankings, citations, conversions, product change, and evidence decay.
- ✓You now have
90-day editorial map
Choose a revenue-connected topic
The cluster must strengthen a product or use-case page.
Think about the professor everyone recommends for one subject. Nobody says 'she knows everything'; they say 'for supply chains, talk to her'. Topical authority works exactly like that reputation, and it is destroyed the same way: by trying to be known for everything at once.
So start from the money page instead of a keyword list. What must a buyer understand before that page makes sense? What breaks during implementation? What do they measure after? Those questions are your cluster. The four hundred posts covering everything else are the sludge this module's title warns about, and every one of them blurs what you are an expert in.
Every post competes alone. Nothing compounds, and the model cannot tell what you are an expert in.
Four expert pieces, each linking into the page that closes. Authority concentrates instead of scattering.
- 01
Start with one priority money page from Module 03.
- 02
List the prerequisite questions, implementation questions, mistakes, benchmarks, and objections around that decision.
- 03
Discard ideas that cannot naturally help the buyer reach or succeed with the product.
- Start a cluster sheet with the money page at the top. Brainstorm rows from three prompts: what must a buyer understand first, what goes wrong during implementation, what do they measure after.
- Check existing coverage with a site: search in Google (site:yourdomain.com topic) so you extend instead of duplicate.
- Cross off any row that cannot name the money page it supports and the sales objection it disarms.
Every topic on your list completes the sentence "this helps a buyer reach or succeed with [money page]" without stretching.
- PRO TIP
Ask sales: 'what do you wish every prospect had read before the first call?' That list is the cluster, pre-validated.
- WATCH OUT
Adjacent is not connected. An article can share keywords with your money page and still do nothing for the decision it owns.
OPTIONAL / AI CO-PILOTRun this lesson with AI, the prompt, sources and expected output+
Choose the business priority, interpret exceptions, protect confidential data, challenge weak evidence, and approve the final decision.
Clean, classify, compare, calculate, and draft rows from supplied evidence. AI may surface patterns; it does not own strategy.
Ahrefs / Semrush, plus page/query baseline from GSC.
| Revenue page | Buyer job | Supporting questions | Current coverage | Pipeline fit | Cluster decision |
|---|---|---|---|---|---|
| /agency-reporting/ | Prove reporting value | KPIs, benchmarks, QA | Partial | High | Build |
CLAUDE / CHATGPT PROMPTAnalyze the evidence without outsourcing the decision+
Remove personal or confidential data. Attach the named exports, explain every column, and tell the model when the dataset was collected.
You are assisting a human B2B SaaS SEO and AEO operator with Module 04: Create topical authority without content sludge. Current lesson: Choose a revenue-connected topic Objective: The cluster must strengthen a product or use-case page. Required artifact: 90-day editorial map BUSINESS CONTEXT I WILL PROVIDE - Product, category, target market, pricing model, sales motion, and geography - The priority customer segment and the commercial outcome for this 90-day cycle - Topic universe exported from Ahrefs / Semrush - Page/query baseline exported from GSC - Definitions for any internal fields, stages, scores, and abbreviations TASK 1. Start with one priority money page from Module 03. 2. List the prerequisite questions, implementation questions, mistakes, benchmarks, and objections around that decision. 3. Discard ideas that cannot naturally help the buyer reach or succeed with the product. REQUIRED OUTPUT Return a table using these exact columns: Revenue page | Buyer job | Supporting questions | Current coverage | Pipeline fit | Cluster decision. For every recommendation, cite the source row, URL, call note, or data point that supports it. Add a confidence column in your analysis: High, Medium, or Low. List missing evidence separately instead of guessing. Finish with a section named HUMAN DECISIONS REQUIRED. RULES - Do not invent search volume, revenue, customer statements, product capabilities, or competitor facts. - Do not treat correlation as causation. - Preserve contradictory evidence and explain why it conflicts. - Do not make the final priority or publishing decision. Prepare the evidence for a human owner. - Use this example only as a format reference, not as evidence: For agency reporting software: client reporting KPIs, dashboard templates, reporting automation, and data-quality checks.
Every topic supports a revenue decision. A named human owner must verify this before the lesson is complete.
For agency reporting software: client reporting KPIs, dashboard templates, reporting automation, and data-quality checks.
Find the information gap
A page needs a reason to exist beyond rewriting the current results.
Before writing anything, answer one uncomfortable question: why does the internet need this page from you? If your article would be the eleventh version of the same list, it does not, and no optimization changes that.
The fix is information gain, which is a fancy phrase for 'something on the page that exists nowhere else'. Original data. A tested workflow. A template that opens and works. Here is the good news: you are sitting on it. Your support tickets, your product database, your onboarding times are all numbers nobody else can publish. When a model has ten interchangeable listicles and one page with real figures, the figures win the citation. Be the figures.
- 01
Review the top results and cited AI sources for recurring sections and unsupported claims.
- 02
Interview a subject-matter expert for practical constraints, exceptions, and examples.
- 03
Choose an information gain: original data, workflow, template, test, or expert synthesis.
- Read the current top five results for the topic and write down the sections they all share. That shared outline is the floor, not the target.
- Ask the same question in ChatGPT, Perplexity, and Gemini, and note which sources get cited. Those pages are the incumbents you must out-inform.
- Book 30 minutes with your subject-matter expert, record it, and ask for exceptions, real numbers, and 'what everyone gets wrong'. Pick your information gain from that recording.
Every brief names its information gain in one sentence, and at least one piece in the quarter is built on data only you have.
- PRO TIP
Your support tickets and product database are original data nobody else can publish. Aggregated and anonymized, they are instant information gain.
- WATCH OUT
Longer is not gain. The eleventh 3,000-word version of the same list loses to 800 words with one number nobody else has.
OPTIONAL / AI CO-PILOTRun this lesson with AI, the prompt, sources and expected output+
Choose the business priority, interpret exceptions, protect confidential data, challenge weak evidence, and approve the final decision.
Clean, classify, compare, calculate, and draft rows from supplied evidence. AI may surface patterns; it does not own strategy.
GSC, plus content crawl from Screaming Frog.
| Topic | SERP pattern | AI source pattern | Missing evidence | Information gain | Expert |
|---|---|---|---|---|---|
| Agency KPI benchmarks | Generic lists | Review sites | Real ranges | Customer dataset | Data lead |
CLAUDE / CHATGPT PROMPTAnalyze the evidence without outsourcing the decision+
Remove personal or confidential data. Attach the named exports, explain every column, and tell the model when the dataset was collected.
You are assisting a human B2B SaaS SEO and AEO operator with Module 04: Create topical authority without content sludge. Current lesson: Find the information gap Objective: A page needs a reason to exist beyond rewriting the current results. Required artifact: 90-day editorial map BUSINESS CONTEXT I WILL PROVIDE - Product, category, target market, pricing model, sales motion, and geography - The priority customer segment and the commercial outcome for this 90-day cycle - Page/query baseline exported from GSC - Content crawl exported from Screaming Frog - Definitions for any internal fields, stages, scores, and abbreviations TASK 1. Review the top results and cited AI sources for recurring sections and unsupported claims. 2. Interview a subject-matter expert for practical constraints, exceptions, and examples. 3. Choose an information gain: original data, workflow, template, test, or expert synthesis. REQUIRED OUTPUT Return a table using these exact columns: Topic | SERP pattern | AI source pattern | Missing evidence | Information gain | Expert. For every recommendation, cite the source row, URL, call note, or data point that supports it. Add a confidence column in your analysis: High, Medium, or Low. List missing evidence separately instead of guessing. Finish with a section named HUMAN DECISIONS REQUIRED. RULES - Do not invent search volume, revenue, customer statements, product capabilities, or competitor facts. - Do not treat correlation as causation. - Preserve contradictory evidence and explain why it conflicts. - Do not make the final priority or publishing decision. Prepare the evidence for a human owner. - Use this example only as a format reference, not as evidence: Instead of another KPI list, publish anonymized benchmark ranges by agency size and client model.
Each brief names a defensible information gain. A named human owner must verify this before the lesson is complete.
Instead of another KPI list, publish anonymized benchmark ranges by agency size and client model.
Create evidence-led briefs
A brief should direct research and decisions, not dictate a word count.
A brief that says '1,500 words on reporting KPIs, mention the keyword five times' produces exactly the mediocrity it describes. The writer did their job. The brief was the problem.
Treat the brief as a research assignment instead: who is reading, what situation they are in, what they must be able to do afterward, and which evidence you will show them. List the sources before the outline. A brief that requires a real dashboard screenshot and a review from your analytics lead starts the writer from substance; a brief that requires nothing gets you another confident post with nothing inside.
- 01
Name the reader, situation, question, and action the page must enable.
- 02
List required first-party and external sources before outlining.
- 03
Specify answer-first sections, examples, visual proof, internal links, and claims requiring review.
- Use a fixed brief template: reader, situation, question, action enabled, required evidence, internal link targets, reviewer, pass condition.
- Attach the evidence before the outline: the SME recording, the data export, the screenshots. A brief with empty evidence fields is not ready to assign.
- Name the reviewer in the brief and get their yes before writing starts, because expert review time is the scarcest input you have.
Each brief names the reader, the required evidence, and a reviewer, and none of them specifies a word count as the goal.
- PRO TIP
Paste three verbatim buyer quotes from Module 01 into every brief. The writer's tone problem solves itself.
- WATCH OUT
The moment a brief specifies a word count as the goal, quality becomes optional. Specify the questions to answer instead.
OPTIONAL / AI CO-PILOTRun this lesson with AI, the prompt, sources and expected output+
Choose the business priority, interpret exceptions, protect confidential data, challenge weak evidence, and approve the final decision.
Clean, classify, compare, calculate, and draft rows from supplied evidence. AI may surface patterns; it does not own strategy.
Screaming Frog, plus expert brief inputs from CRM / Gong.
| Section | Question | Evidence | Visual | Internal link | Reviewer | Pass condition |
|---|---|---|---|---|---|---|
| Benchmark table | What is normal? | Anonymized data | Distribution chart | /agency-reporting/ | Data lead | Method disclosed |
CLAUDE / CHATGPT PROMPTAnalyze the evidence without outsourcing the decision+
Remove personal or confidential data. Attach the named exports, explain every column, and tell the model when the dataset was collected.
You are assisting a human B2B SaaS SEO and AEO operator with Module 04: Create topical authority without content sludge. Current lesson: Create evidence-led briefs Objective: A brief should direct research and decisions, not dictate a word count. Required artifact: 90-day editorial map BUSINESS CONTEXT I WILL PROVIDE - Product, category, target market, pricing model, sales motion, and geography - The priority customer segment and the commercial outcome for this 90-day cycle - Content crawl exported from Screaming Frog - Expert brief inputs exported from CRM / Gong - Definitions for any internal fields, stages, scores, and abbreviations TASK 1. Name the reader, situation, question, and action the page must enable. 2. List required first-party and external sources before outlining. 3. Specify answer-first sections, examples, visual proof, internal links, and claims requiring review. REQUIRED OUTPUT Return a table using these exact columns: Section | Question | Evidence | Visual | Internal link | Reviewer | Pass condition. For every recommendation, cite the source row, URL, call note, or data point that supports it. Add a confidence column in your analysis: High, Medium, or Low. List missing evidence separately instead of guessing. Finish with a section named HUMAN DECISIONS REQUIRED. RULES - Do not invent search volume, revenue, customer statements, product capabilities, or competitor facts. - Do not treat correlation as causation. - Preserve contradictory evidence and explain why it conflicts. - Do not make the final priority or publishing decision. Prepare the evidence for a human owner. - Use this example only as a format reference, not as evidence: The brief requires one real dashboard, a downloadable QA checklist, and review by the analytics lead.
Experts and source material are assigned. A named human owner must verify this before the lesson is complete.
The brief requires one real dashboard, a downloadable QA checklist, and review by the analytics lead.
Write with the product in the answer
Product-led content shows the product doing the job inside the how-to, at whichever tier the topic honestly supports.
Read most SaaS blogs and the product appears in one of two places: nowhere, or in a closing paragraph that begins 'tools like ours'. Both are the same mistake from opposite directions. A guide that never shows the product doing the job has left the strongest evidence it owns on the table, and a guide that forces the product into a topic it does not belong in reads as a sales deck, and buyers know it in one paragraph.
So tier every topic before the brief goes out. Tier one, the product is the answer: the article opens inside it, with the real screens and the actual clicks. Tier two, the product is one route: the framework comes first, the product is one honest implementation, and the alternatives get named. Tier three, the product does not belong: it stays out of the body and the money page gets one link, or none. The test is simple. Could a reader finish the task with the product open? Yes is tier one, partly is tier two, no is tier three, and the answer decides what the writer is allowed to do.
Could a reader finish this task with the product open?
“How to automate multi-client reporting”
PRODUCT SHOWS UP ASThe article opens inside the product. Real screens, the actual clicks, the output.
THE BRIEF NEEDSA captured workflow in the brief, and the product named in the first 70 words.
“How to build a client reporting KPI framework”
PRODUCT SHOWS UP ASThe framework first. The product as one honest implementation, with the alternatives named.
THE BRIEF NEEDSScreens for the part it does, and a plain sentence about the part it does not.
“How to run a client kickoff call”
PRODUCT SHOWS UP ASNothing in the body. One link to the money page, or none.
THE BRIEF NEEDSA reviewer who will not add it back.
- 01
Before the brief is assigned, give the topic a tier: the product is the answer, the product is one route, or the product does not belong.
- 02
For the first two tiers, put the real workflow on the page: the screens, the clicks, and the output, with the product named where it does the work rather than in a closing pitch.
- 03
For the third tier, keep the product out of the body and link the money page once, if at all. Forcing it is what turns a guide into a sales deck.
- Add a Tier column to the Module 04 brief template, filled by the person who owns the topic, not the writer. Tier one, two, or three, plus one line on why.
- For tier one and two, capture the workflow yourself before the writer starts: the screens, the clicks, the output, in the order a reader would do them. Store them in the brief so the writer describes what happened rather than what the feature page claims.
- For tier three, put the money page in the internal link targets once and strike the product from the outline. If a reviewer adds it back, the tier was wrong, not the writer.
Every brief carries a tier before writing starts, tier-one and tier-two pieces show real screens and real output, and no tier-three piece mentions the product in its body.
- PRO TIP
The tier-one opening is the same shape as a Module 07 answer passage: who it is for, what the product does, in the first 70 words, then the depth. Write it once and reuse the structure.
- WATCH OUT
If the product only partly does the thing, say which part. A tier-two piece that implies tier one is the editorial version of the Module 11 product-fit trap, and it costs every other claim on the site.
OPTIONAL / AI CO-PILOTRun this lesson with AI, the prompt, sources and expected output+
Choose the business priority, interpret exceptions, protect confidential data, challenge weak evidence, and approve the final decision.
Clean, classify, compare, calculate, and draft rows from supplied evidence. AI may surface patterns; it does not own strategy.
CRM / Gong, plus topic universe from Ahrefs / Semrush.
| Topic | Tier | Reader task | Product shows up as | Screens captured | Money page link | Reviewer |
|---|---|---|---|---|---|---|
| How to automate multi-client reporting | 1 | Ship a scheduled report | The workflow, opening screen onward | 6 | /agency-reporting/ | Product lead |
CLAUDE / CHATGPT PROMPTAnalyze the evidence without outsourcing the decision+
Remove personal or confidential data. Attach the named exports, explain every column, and tell the model when the dataset was collected.
You are assisting a human B2B SaaS SEO and AEO operator with Module 04: Create topical authority without content sludge. Current lesson: Write with the product in the answer Objective: Product-led content shows the product doing the job inside the how-to, at whichever tier the topic honestly supports. Required artifact: 90-day editorial map BUSINESS CONTEXT I WILL PROVIDE - Product, category, target market, pricing model, sales motion, and geography - The priority customer segment and the commercial outcome for this 90-day cycle - Expert brief inputs exported from CRM / Gong - Topic universe exported from Ahrefs / Semrush - Definitions for any internal fields, stages, scores, and abbreviations TASK 1. Before the brief is assigned, give the topic a tier: the product is the answer, the product is one route, or the product does not belong. 2. For the first two tiers, put the real workflow on the page: the screens, the clicks, and the output, with the product named where it does the work rather than in a closing pitch. 3. For the third tier, keep the product out of the body and link the money page once, if at all. Forcing it is what turns a guide into a sales deck. REQUIRED OUTPUT Return a table using these exact columns: Topic | Tier | Reader task | Product shows up as | Screens captured | Money page link | Reviewer. For every recommendation, cite the source row, URL, call note, or data point that supports it. Add a confidence column in your analysis: High, Medium, or Low. List missing evidence separately instead of guessing. Finish with a section named HUMAN DECISIONS REQUIRED. RULES - Do not invent search volume, revenue, customer statements, product capabilities, or competitor facts. - Do not treat correlation as causation. - Preserve contradictory evidence and explain why it conflicts. - Do not make the final priority or publishing decision. Prepare the evidence for a human owner. - Use this example only as a format reference, not as evidence: “How to automate multi-client reporting” is tier one and opens inside the product. “How to run a client kickoff call” is tier three and never mentions it.
The 90-day cadence matches real capacity. A named human owner must verify this before the lesson is complete.
“How to automate multi-client reporting” is tier one and opens inside the product. “How to run a client kickoff call” is tier three and never mentions it.
Publish at the quality cadence
One defensible piece beats five interchangeable posts.
Teams pick a publishing schedule first,'two posts a week', then discover their experts have four spare hours a month, and quality quietly drops to fit the calendar. Flip the order. Count the real hours available for research, expert review, and design, and let the schedule fall out of the counting.
Four defensible pieces in 90 days, each supporting a money page and disarming a sales objection, will beat thirty generic posts on every number you report to a board. And when capacity slips, cut scope, never quality. Three excellent pieces compound. Six rushed ones subtract.
- 01
Estimate research, expert, design, writing, and review capacity honestly.
- 02
Sequence content so supporting articles link into the priority money page as they publish.
- 03
Refresh based on performance and changed evidence, not an arbitrary date.
- Build a capacity row per person: realistic hours per week for research, writing, design, and expert review. Sum it before promising anything.
- Sequence the map so each piece publishes in the same week its internal links into the money page go live.
- Set a quarterly refresh pass: re-verify stats, update dates, and re-check the pages your cluster cites.
The 90-day map fits capacity your experts confirmed, and every ship date has a named internal-link target waiting for it.
- PRO TIP
A refresh of a page that already ranks usually beats a new post on effort-to-result. Check Search Console for pages sliding on impressions and start there.
- WATCH OUT
When capacity slips, cut scope, not quality. Three excellent pieces this quarter compounds; six rushed ones subtracts.
OPTIONAL / AI CO-PILOTRun this lesson with AI, the prompt, sources and expected output+
Choose the business priority, interpret exceptions, protect confidential data, challenge weak evidence, and approve the final decision.
Clean, classify, compare, calculate, and draft rows from supplied evidence. AI may surface patterns; it does not own strategy.
Ahrefs / Semrush, plus page/query baseline from GSC.
| Asset | Research hrs | Expert hrs | Design hrs | Review hrs | Ship week | Owner |
|---|---|---|---|---|---|---|
| KPI benchmark report | 12 | 3 | 8 | 2 | 5 | Content lead |
CLAUDE / CHATGPT PROMPTAnalyze the evidence without outsourcing the decision+
Remove personal or confidential data. Attach the named exports, explain every column, and tell the model when the dataset was collected.
You are assisting a human B2B SaaS SEO and AEO operator with Module 04: Create topical authority without content sludge. Current lesson: Publish at the quality cadence Objective: One defensible piece beats five interchangeable posts. Required artifact: 90-day editorial map BUSINESS CONTEXT I WILL PROVIDE - Product, category, target market, pricing model, sales motion, and geography - The priority customer segment and the commercial outcome for this 90-day cycle - Topic universe exported from Ahrefs / Semrush - Page/query baseline exported from GSC - Definitions for any internal fields, stages, scores, and abbreviations TASK 1. Estimate research, expert, design, writing, and review capacity honestly. 2. Sequence content so supporting articles link into the priority money page as they publish. 3. Refresh based on performance and changed evidence, not an arbitrary date. REQUIRED OUTPUT Return a table using these exact columns: Asset | Research hrs | Expert hrs | Design hrs | Review hrs | Ship week | Owner. For every recommendation, cite the source row, URL, call note, or data point that supports it. Add a confidence column in your analysis: High, Medium, or Low. List missing evidence separately instead of guessing. Finish with a section named HUMAN DECISIONS REQUIRED. RULES - Do not invent search volume, revenue, customer statements, product capabilities, or competitor facts. - Do not treat correlation as causation. - Preserve contradictory evidence and explain why it conflicts. - Do not make the final priority or publishing decision. Prepare the evidence for a human owner. - Use this example only as a format reference, not as evidence: Ship four expert pieces in 90 days, each supporting one page and one sales objection, instead of 30 generic posts.
Every brief carries a product tier before writing starts. A named human owner must verify this before the lesson is complete.
Ship four expert pieces in 90 days, each supporting one page and one sales objection, instead of 30 generic posts.
The workspace shows a 90-day editorial plan for the demo company. Note that it contains four pieces, not forty. Look at what each row carries: a buyer job, a named information gain, and the money page it supports. If a topic idea cannot fill those three columns, it does not ship.
Topical authority portfolio
Plan expert content around revenue pages, information gain, internal links, and a production capacity the team can sustain.
Cluster coverage and evidence depth
Coverage is useful only when the assets add evidence and move buyers toward a decision.
Check each item only after the artifact meets the standard. Progress saves on this device; use the academy-home backup to move it elsewhere.