Why Creatives Struggle to Niche Down: Chris Do's ICP Framework, Tested

Direct Answer

Niching down feels dangerous because it turns an abstract possibility into a visible choice. A broad creative can imagine serving everyone; a specialist must name one buyer, one costly situation, and one result. That means accepting that some people will not see themselves in the message.

Chris Do's live coaching session makes the case that this constraint is useful. A focused niche sharpens the language, proof, offer, referrals, outreach, and delivery system around a repeatable customer. But the practical version needs one addition: a niche is a hypothesis to validate, not an identity to defend forever.

Watch Chris Do's Live Coaching Session

Video credit: Chris Do and The Futur. Watch the original video on YouTube. The article below preserves the coaching framework, labels creator anecdotes as such, and separates them from independently supported market-research guidance.

First, the "97%" Claim Is Not Evidence

The video title says 97% of creatives stay broke, but neither the supplied transcript nor the linked session provides a dataset, survey, denominator, or definition of "broke." It should not be repeated as a statistic. The session's actual argument is qualitative: most people Chris Do observes struggle to choose, and he believes specialists usually perform better than generalists.

The U.S. Small Business Administration's current market-research guidance supports the underlying discipline, not the number: evaluate demand, market size, economic indicators, buyer location, market saturation, pricing, and competition before committing. Focus is a strategic choice; evidence decides whether the choice is commercially sound.

Why Choosing a Niche Feels So Hard

1. Opportunity-cost fear. Saying "Texas home-service firms" sounds like losing every SaaS, education, health, and retail project.
2. Identity confusion. Creatives mistake a market focus for a permanent description of everything they are capable of doing.
3. Evidence avoidance. Staying broad delays the moment when buyers can reject a precise offer.
4. Novelty addiction. A new niche feels energizing before the slow work of interviews, proof, delivery, and referrals begins.
5. Premature abstraction. "Coaches," "small businesses," and "people who need design" are categories, not usable buying contexts.
6. Confusing a failed offer with a failed market. The caller tried education for a few months and found many coaches could not afford his automation. That may reject the segment, the price, the problem, the channel, or the offer. It does not identify which one without more evidence.

The fix is to stop asking, "What niche am I?" and ask, "Which buyer-problem pair deserves the next fixed period of focused learning?" That question is reversible, testable, and much less dramatic.

Five Ways to Niche Without Shrinking Your Identity

You can combine two or three dimensions when each one changes the work. "We help $2M-$5M Texas HVAC companies respond to inbound leads within five minutes" is more operational than "AI automation for blue-collar men." Industry, revenue, location, and lead-response pain affect the buying process. Age and gender may not. Use protected or demographic traits only when they are genuinely relevant, lawful, and respectful; firmographics, triggers, and workflows are usually better B2B constraints.

Rebuilding the Live Workshop Example

In the session, Chris Do forces choices across demographic, career, industry, revenue, and geography dimensions. The group lands near older owners of blue-collar businesses in Texas, earning $2M-$5M, who know technology could improve responsiveness. That is a useful draft, but it is not yet an ICP.

A testable niche statement might become: "We help owner-led HVAC firms in Texas with $2M-$5M revenue reduce missed inbound leads by connecting their existing phone, form, CRM, and scheduling workflow, with staff approval before outbound follow-up." It is still a hypothesis. Interviews and a paid pilot decide whether it deserves investment.

Score a Niche Before You Commit

Score each candidate from 0 to 5. Do not average away a fatal zero: a market with no budget, no access, or no deliverable outcome is not rescued by personal enthusiasm.

Chris Do suggests looking backward at your best clients: they paid well and on time, made decisions quickly, were easy to work with, and exist in meaningful numbers. That is a good seed list. Add two columns: why they bought now and why your delivery worked for them. Those answers reveal the trigger and operating fit that generic demographics miss.

Is the Market Difficult Because It Is Valuable or Because It Is Bad?

"Hard" is not a moat by itself. A good difficult niche rewards accumulated knowledge. A bad difficult niche consumes effort without increasing the probability of a sale, a successful delivery, or a referral.

The Dip, Corrected: Persistence Is Not the Whole Lesson

Chris uses Seth Godin's The Dip to explain the stage where early novelty disappears, effort rises, results stall, and people restart elsewhere. That pattern is real. But Godin's own Q&A about the book adds the crucial decision: ask whether the endeavor can respond to courage, effort, and investment. The lesson is strategic commitment and strategic quitting, not endurance at any price.

Set the decision before the discomfort arrives: "I will run 15 interviews, make 30 relevant approaches, present five offers, and attempt two paid pilots within 45 days. I will continue if the pain repeats and at least one buyer pays. I will revise if access or wording fails. I will leave if budget and urgency remain absent." That is persistence with instrumentation.

Verify Market Size From the Bottom Up

A model can produce a plausible market story in seconds. It cannot turn an unsourced count into demand. For U.S. business niches, the Census Bureau's current small-business data guide points operators to NAICS classifications, County Business Patterns, the Economic Census, and Census Business Builder for establishment counts, employee ranges, payroll, locations, customer characteristics, and competitor density. Use equivalent official sources in other countries.

Start with a bottom-up serviceable market:

• Qualified accounts = establishments in the chosen industry, geography, and size band.
• Reachable accounts = qualified accounts you can identify and reach through lawful channels.
• Annual opportunities = reachable accounts x realistic share experiencing the trigger this year.
• Realistic wins = annual opportunities x observed meeting rate x observed close rate.
• Gross-profit opportunity = realistic wins x collected gross profit per client.

If 2,000 accounts exist but only 200 experience the trigger annually, 80 are reachable, 12 enter a serious sales process, and two buy, your working market is two clients at the current system's performance. That is more useful than a billion-dollar TAM slide. Improve access, trigger detection, proof, or conversion and recalculate.

A Better AI Prompt for Niche Research

Chris recommends giving Claude a detailed background, asking it to define the ICP, checking that enough customers exist, and using human judgment on the result. Keep that spirit, but make the model expose uncertainty instead of pretending it already knows current market size.

You are a market-research assistant, not a market oracle.

My capabilities:
[skills, proof, delivery capacity, geography]

My current offer:
[problem, deliverable, price range, time to value]

Candidate clients from my history:
[3-10 clients, including revenue, speed, fit, outcome, and why they bought]

Candidate niche:
[industry, company size, buyer role, trigger, geography]

First, ask up to 10 clarifying questions. Then produce:
1. A one-sentence niche hypothesis.
2. The costly problem and the buyer's likely plain-language description.
3. Qualification and disqualification criteria.
4. Assumptions that require evidence.
5. Primary public datasets or registries to verify account count and market trend.
6. Fifteen neutral interview questions that do not pitch or lead the buyer.
7. Three small paid-pilot offers with measurable acceptance criteria.
8. A score from 0-5 for pain, budget, reachability, proof advantage,
   repeatability, market depth, economics, and founder fit.
9. The strongest case against choosing this niche.
10. Continue, revise, and stop thresholds for a 30-day test.

Do not invent market counts, budgets, quotes, or sources.
Mark unknowns clearly. Separate evidence, inference, and hypothesis.

After the model responds, verify every count and market claim. More importantly, talk to buyers. AI can organize what you know and reveal missing questions; it cannot substitute for observing the purchasing reality.

A 30-Day Niche Validation Sprint

The numbers above are experiment targets, not universal conversion benchmarks. A niche with enterprise procurement or a small addressable population may need a longer window. The principle is to predefine enough evidence to distinguish market learning from endless activity.

Continue, revise, or stop

• Continue when pain repeats, the buyer has authority and budget, access improves, and at least one paid engagement validates the outcome.
• Revise when the problem is real but the buyer, trigger, price, channel, or deliverable is wrong.
• Stop when urgency, budget, access, or ethical delivery remains absent after a fair evidence window.

Keep the niche statement on your homepage, outreach, case-study pipeline, and content calendar for the duration of the test. Do not rewrite the positioning after every quiet day. Consistency is what lets the market produce interpretable feedback.

Video Chapters

The supplied transcript ends at approximately 12:38, although the submitted chapter list extends to 18:00. The navigation above follows the content available in the transcript rather than inventing later sections.

Bottom Line

Niching down is not a vow to remain small. It is a way to concentrate learning until your market can understand, trust, refer, and buy a repeatable offer. Chris Do's forcing function is valuable because it breaks indecision. The evidence layer makes it safe: verify the buyers, budget, pain, access, and economics before confusing stubbornness with strategy.

Choose one buyer-problem pair for the next test, not for the rest of your life. Keep what improves with focused effort. Revise what the market teaches you. Quit what cannot become viable.

Sources and Useful Links

• The Futur: Why 97% of Creatives Stay Broke
• The Futur official website
• The Futur: Find Your Niche - Ikigai Workshop
• The Futur: Finding Your Niche - Business Growth Strategy
• Chris Do: four client profiles to identify
• Seth Godin: The Dip and knowing when to quit
• U.S. Small Business Administration: market research and competitive analysis
• U.S. Census Bureau: data for customers, competitors, and small-business planning
• Census Business Builder: locate and analyze customers and markets
• JQ AI SYSTEMS: Zero-Click Marketing for Founders
• JQ AI SYSTEMS: Productized Services, Demand, and the Honest Math