Niche Research and Displacement-Adjusted Market Selection

Overview

v1 selected niches on CPC and volume. v2 subtracts the things that destroy value: answers a machine gives away, compliance you cannot afford, reviewers you cannot recruit, and competitors with a decade of trust. The output is a ranked list of ten niches and a defensible reason that only three launch now.

2.1 What is high-CPM plus displacement risk?

What is it?

Two independent properties of a topic: how much advertisers pay for the audience, and how completely a generated answer can satisfy the query without a visit.

Why it matters

The most valuable-looking keywords are often the most displaced, because advertisers bid on early-funnel definitions that are exactly what summaries are good at. Value is realised only where a click is still needed.

How to do it

1. List 20 real topics from one niche, mixing definitions, comparisons, tools and current-data queries.
2. Estimate advertiser value from keyword tools and, better, from who is buying ads on the live results.
3. Estimate AI vulnerability: ask two assistants and judge whether their answer would end the user's task.
4. Plot on a 2×2 and keep only the high-value/low-vulnerability quadrant as launch candidates.

Tools needed

Keyword tool of your choice, live results pages, two AI assistants, spreadsheet.

Lab 2.1

Build the matrix for 20 topics in one Phase 1 niche. Highlight the five you would fund and the five you would never write.

Artifact

CPM vs displacement matrix (20 topics, 2x2 quadrants, funded/rejected)

Common mistakes

• Using CPC as a proxy for your own future RPM without any traffic-mix assumption.
• Judging vulnerability from one assistant on one day.
• Discarding a niche instead of discarding its displaced sub-topics.

Pro astuces

• Advertisers bidding on a query proves commercial value even where organic clicks are collapsing — that is a signal to build the tool, not the explainer.
• Keep a "displaced but valuable" list; those topics are lead magnets waiting to happen.

2.2 What is audience value?

What is it?

The economic and behavioural quality of the people behind the queries, independent of keyword metrics.

Why it matters

Two niches with identical CPC can differ tenfold in newsletter conversion and lifetime value. Audience value predicts owned-audience growth, which v2 treats as the real asset.

How to do it

1. Score the eight dimensions below from 1 to 5 using evidence, not vibes.
2. Evidence sources: existing communities, paid products in the space, job titles, active newsletters, forum activity.
3. Interview three people in the audience and record verbatim quotes about their recurring problems.
4. Reject niches scoring low on "likelihood to subscribe" and "recurring need" even if CPM is high.

Audience value dimensions
Budget            Advertiser value
Urgency           Trust requirement
Recurring need    Likelihood to subscribe
Decision complexity   Likelihood to return

Tools needed

Community platforms, competitor newsletters (subscribe to five), three interviews, spreadsheet.

Lab 2.2

Profile the buyer for each of the three Phase 1 niches. Note that all three share a buyer — the small-business operator — and write one sentence on how that reduces your research and newsletter costs.

Artifact

Audience value profile per niche, with three verbatim quotes each

Common mistakes

• Describing a demographic instead of a job and a budget.
• Choosing an audience you cannot reach or credibly speak to.
• Ignoring trust requirement, which drives reviewer cost later.

Pro astuces

• Subscribe to competitor newsletters for a month; their recurring segments reveal proven return triggers.
• If a paid product already exists for the audience, budget exists. That is stronger evidence than search volume.

2.3 What is a displacement-adjusted niche score?

What is it?

One number that adds durable upside and subtracts structural cost and risk.

Niche Score =
    Advertiser Value
  + Commercial Intent
  + Recurring Need
  + Tool Potential
  + Original Data Potential
  + Newsletter Potential
  + Partnership Potential
  - AI Answer Vulnerability
  - Compliance Burden
  - Reviewer Cost
  - Competition Strength

Why it matters

It forces the trade-off into the open. Health benefits scores high on advertiser value and catastrophically on reviewer cost and compliance; SMB cybersecurity scores lower on CPM but high on tool potential and partnerships.

How to do it

1. Score each factor 1–10 with a written justification — a score without a sentence is noise.
2. Weight the subtractions at least as heavily as the additions.
3. Score all ten niches in one sitting so the scale stays consistent.
4. Sanity-check: would you personally fund the top three with your own money?

Tools needed

Displacement-adjusted niche scorer, your matrix from 2.1 and profiles from 2.2.

Lab 2.3

Score all ten academy niches. Publish the justification column. Then re-score the bottom niche as if you had unlimited reviewer budget and note how much of the score is really a resource constraint.

Artifact

Displacement-adjusted scorecard for 10 niches with written justifications

Common mistakes

• Scoring in isolation over several days, producing an inconsistent scale.
• Under-weighting reviewer cost, the most common cause of stalled YMYL sites.
• Scoring competition by domain metrics rather than by content quality and originality.

Pro astuces

• Tool potential and original data potential are the two factors most predictive of survival. Break ties with them.
• Record the score date. Re-score annually; markets and displacement both move.

2.4 What is competitor and AI-surface research?

What is it?

Mapping who currently owns attention across every surface, including the assistants that never show a results page.

Why it matters

Your real competitors may be a forum, a vendor's documentation and a newsletter — not the sites in the top ten. And the brands assistants cite repeatedly are the ones you must out-originate.

How to do it

1. For ten priority queries, record: organic top five, AI summary sources, forum threads, videos, newsletters, competitor tools, comparison pages, data assets, communities.
2. Note which competitors own a named tool or dataset — those are the durable positions.
3. Record which brands get cited by assistants, and what format their cited page uses.
4. Identify one gap you can own with original data within 90 days.

AI-surface competitor audit
Query | Platform | Brands cited | Format cited | Their tool/data asset |
Gap we can own | Evidence needed | Date

Tools needed

Two assistants, browser, competitor newsletters, and the AI-surface tracking sheet.

Lab 2.4

Audit ten queries in one niche and identify the single dataset nobody maintains. That dataset is a candidate for one of your two data assets.

Artifact

AI-surface competitor audit with one 90-day data opportunity

Common mistakes

• Benchmarking only against publishers, ignoring vendors and communities.
• Copying a competitor's tool rather than finding the adjacent one nobody built.
• Not dating the audit; citations change and undated notes mislead.

Pro astuces

• If assistants cite the same three brands everywhere, look for the sub-topic where they cite nobody.
• Vendor documentation is often the citation winner — beat it with independence and comparison, not with a rewrite.

2.5 What is compliance pre-screening?

What is it?

Nine questions asked before you commit, so legal exposure is a choice rather than a surprise.

Compliance pre-screen
Does this involve insurance leads?    Legal advice?
Mortgage leads?                       Tax advice?
Loan brokering?                       SMS/phone consent?
Health advice?                        EU/UK personal data?
                                      Affiliate disclosures?

Why it matters

Lead generation in regulated verticals can require licensing; consent rules govern phone and SMS follow-up; EU/UK data brings its own obligations. Discovering this after building a lead funnel wastes the funnel.

How to do it

1. Answer all nine questions per niche with yes/no/unsure.
2. For every yes, write the specific question you will take to counsel (Course 4 builds the sheet).
3. Decide whether the niche can launch with advertising and affiliate revenue only, deferring lead-gen.
4. Record the decision and a review date in the risk register.

Tools needed

Legal counsel question sheet, risk register, FTC endorsement guidance for disclosure obligations.

Lab 2.5

Pre-screen all ten niches. Produce a table showing which can launch immediately, which need counsel first, and which need counsel only before lead-gen.

Artifact

Compliance pre-screen for 10 niches with launch conditions

Common mistakes

• Assuming an affiliate model avoids all regulatory questions — disclosure obligations still apply.
• Copying a competitor's disclaimers; you do not know whether theirs are compliant.
• Answering "unsure" and moving on without logging the question.

Pro astuces

• Sequence revenue: display and affiliate first, lead-gen after counsel. It de-risks launch without delaying it.
• Bundle all your counsel questions into one paid hour rather than drip-feeding.

2.6 Why launch 3 sites first?

What is it?

A deliberate constraint: Phase 1 is SMB Finance and Banking, B2B SaaS and AI Tools, and SMB Cybersecurity. The remaining seven niches are expansion exercises unlocked by the month 9 gate.

Why it matters

Ten simultaneous sites means ten newsletters, ten reviewer relationships and ten refresh backlogs, funded by one person's attention. Three sites with a shared buyer share research, sponsors, tooling and SOPs — and can each reach a real MVP.

How to do it

1. Confirm the three against your own scorecard — if your scores disagree, document why before overriding.
2. Write the launch order and the staggered start dates (do not start three sites in the same week).
3. Define shared infrastructure: one analytics account structure, one component library, one SOP set.
4. Write the expansion trigger: which gate result unlocks site four.

Tools needed

Your scorecard, the Course 0 budget model, and a calendar.

Lab 2.6

Write a one-page launch order memo: three sites, stagger dates, shared infrastructure, and the numeric condition that unlocks Phase 2.

Artifact

Phase 1/Phase 2 launch order memo

Common mistakes

• Launching all three the same week and starving each of attention.
• Treating Phase 2 as a promise rather than a conditional option.
• Choosing three niches with different buyers, losing every shared advantage.

Pro astuces

• Stagger by six to eight weeks; site two benefits from every mistake site one made.
• Build the shared component library during site one — it is your compounding asset.

Scenario assessment: the seductive high-CPM niche

A student's scorecard puts Health Benefits and HSA first on advertiser value and commercial intent. It also scores 9/10 on AI answer vulnerability (most queries are definitional), 8/10 on compliance burden and 9/10 on reviewer cost. They have no medical or benefits reviewer, and a 12-month budget that assumes 30 pages.

Final project: the ten-niche displacement scorecard

Create a displacement-adjusted scorecard for all ten niches and justify why only three launch first.

Deliverable
1. Scores for all 11 factors across 10 niches, with justifications
2. CPM vs displacement matrix for the top 3
3. Audience value profile for the top 3 (shared buyer analysis)
4. Compliance pre-screen with launch conditions for all 10
5. Launch order memo with stagger dates and the Phase 2 trigger

Pass standard: a reader who disagrees with your top three can see exactly which factor would have to change their mind.

Course 2 checklist

• 20 topics classified on CPM and AI vulnerability
• Three real audience interviews completed with verbatim quotes
• All ten niches scored in one sitting with written justifications
• Subtraction factors weighted at least as heavily as additions
• AI-surface audit dated and covering ten queries
• One unmaintained dataset identified as a data-asset candidate
• Nine-question compliance pre-screen completed per niche
• Counsel questions logged for every "yes" answer
• Phase 1 confirmed with stagger dates
• Numeric Phase 2 expansion trigger written down

AI-agent prompt for Course 2

Act as a market analyst for an independent publisher operating in an AI-mediated search environment.

I will paste: a list of candidate niches and my constraints (capital, hours, reviewer access, jurisdiction).

Do this:
1. For each niche, score 1-10 with a one-sentence justification: advertiser value, commercial intent, recurring need, tool potential, original data potential, newsletter potential, partnership potential.
2. Then score and SUBTRACT: AI answer vulnerability, compliance burden, reviewer cost, competition strength.
3. Output the ranked table with the net score and the deciding factor for each rank.
4. For the top three, produce a CPM vs AI-vulnerability matrix of 20 topics each, marking which topics to fund and which to never write.
5. Run the nine-question compliance pre-screen per niche and list the exact questions to take to counsel for every "yes".
6. Recommend a Phase 1 of three niches that share one buyer, with stagger dates and a numeric trigger for Phase 2.

Rules:
- Do not state CPC, RPM, licensing requirements or platform terms as facts. Output them as claim rows needing a primary source, with an owner and a 90-day expiry.
- Prefer tool and original-data opportunities over explainer topics.
- Flag any niche where launch should be advertising/affiliate only until counsel review.
Output as tables.

Primary sources

• Creating helpful, reliable, people-first content
• FTC endorsement, influencer and review guidance