Validation Plan
Enter hub password to continue.
That's not right. Try again.
The Validation Plan: Every Cheapest Test, Sequenced
2026-07-14 · every cheapest test the panel proposed, in the order that kills the business fastest and cheapest if it's going to die
What this is
The adversarial panel gave the Menopause Practice Growth plan 15 hard findings, and for each one it named the cheapest real-world test to prove or kill it. Scattered across three model reviews, that's a to-do list you can't act on. So I pulled every one of those tests and consolidated them into a tight set, each with a pre-committed pass/fail line, in the order that gets you an answer for the least time and money. That comes to five experiments still worth running, plus one the business has already closed (lead sourcing is done, since the directory is already scraped) and one folded into the interviews (the competitive question, now asked of clinicians directly instead of by mystery-shopping).
The sequencing rule is simple and it matters: desk tests first. Anything you can settle alone at your own desk (the market-size count on data you already own, the delivery-hours check, the tool-cost math) runs before the tests that need other humans (the interviews and the outreach A/B). A desk test can kill this in an afternoon, and there's no sense spending two weeks interviewing clinicians about a business that a spreadsheet would have stopped on day one.
Where things actually stand
Before the plan, be honest about what's already answered. Two of the panel's scariest findings are largely defused, not because the panel was wrong to raise them, but because the business has moved since it wrote them. Here's the real scoreboard.
| Panel finding | Status now | Why |
|---|---|---|
| Market size / SAM (panel called 1,500-4,000 fiction) | Largely defused | You now sell worldwide and to in-person clinics, not just US telehealth, and you already own resolved data on 4,851 practices. The panel's US-telehealth framing was too narrow. Confirm density on the data you have. |
| Delivery hours for one person (panel called it physically impossible) | Largely defused, pending confirmation | Your own estimate is about 1.5 hours per client per week even top tier with video. If that holds you carry 15 to 20 clients solo. The panel assumed 8 to 12. One real build confirms which number is true. |
| Lead pipeline (panel called scraping toxic, sourcing unscalable) | Answered | The directory is already scraped and the data is in hand, so acquisition-side sourcing is done. The only residual is compliant outreach (anti-spam rules by country), and that's already handled in the outreach A/B page, no need to re-derive it. |
| Competitive substitution (generalists, platforms like Midi/Alloy) | Moved into the interviews | You're not mystery-shopping. The question (do generalists already cover this at the same price, would clinicians join Midi/Alloy instead of hiring you) now gets asked of clinicians directly in the discovery interviews, so it still gets tested. |
| Willingness to pay, and at what price | Open | This is the one that decides the business. Every willingness-to-pay number in the plan is a guess. Only past spending, named in real dollars, settles it. |
| Is the evidence / regulatory wedge a purchase driver | Open | You believe fact-checked scripts and regulatory fluency are a moat. No buyer has paid for that yet. Until one does, it's a story, not a barrier. |
So the short version: market size, delivery hours, and lead sourcing are largely handled. Market size by your worldwide-plus-in-person scope and the 4,851 practices you already hold, delivery by your roughly 1.5-hour-per-client estimate that one real build will confirm, and sourcing because the directory is already scraped. The competitive question isn't gone, it just moved into the interviews, where you ask clinicians directly instead of mystery-shopping. What's genuinely still open is narrower and sharper: will enough clinicians pay, and at what price, and is the evidence wedge a real purchase driver. The plan below spends its cheapest, fastest tests confirming the defused ones, and its real energy on the open ones.
The five experiments, in order
Each card names the assumption it kills, exactly what to do, what it costs (mostly your time), the pass/fail line the panel set, the honest status right now, and the instrument that's already built to run it.
SAM density desk-check
PARTLY ANSWERED FATALKills this assumptionThat the addressable market is 1,500 to 4,000 funded buyers. The panel showed the plan's own ICP filters can collapse that to a few hundred, and the first 17 enriched prospects came back thin (2 HOT, 10 WARM).
Do thisApply the full six-dimension funded-buyer rubric (independent decision control, cash-pay or membership, menopause-branded, telehealth or in-person, visible marketing gap, reachability) to a multi-state sample of the data you already own. Count only HOT plus WARM true buyers, then extrapolate. This is a desk exercise on your 4,851 resolved practices, and you can largely automate the rubric scoring. It is not a re-scrape.
CostAn afternoon, mostly scoring you can script. No new data collection.
Pass / fail lineIf HOT plus WARM per state comes in under roughly 15 after the full filter, the 1,500 to 4,000 SAM is fiction. But score it against your real scope: worldwide, and in-person clinics included, not just US telehealth.
Where it standsLargely defused before you start. Your decision to sell worldwide and to in-person providers materially enlarges the pool beyond the panel's US-telehealth framing, and you own the data. This desk-check confirms it on numbers instead of belief.
Instrument: Your 4,851-practice ICP audit (the business plan explains the beachhead)
Delivery time-box
PARTLY ANSWERED FATALKills this assumptionThat one person can deliver the flagship Get Known tier at the client count needed to net $6,500 a month. All three models independently said the labor per client is too high, estimating 8 to 12 hours each.
Do thisMock-deliver Get Known for one fictional client for a full week, or better, track every minute on your next one or two real client builds: research a script, verify the claims against peer-reviewed sources, edit a short video, post to the platforms, draft the report. Log only the minutes you personally spend that Claude did not. Multiply by the number of clients you'd need to hit the income target.
CostOne week of honest timekeeping on work you're doing anyway.
Pass / fail lineIf weekly hours exceed roughly 30 to 35 before any sales time, the model is unworkable at that client count. Watch video editing specifically, it's the one task that creeps past the estimate.
Where it standsLargely defused, pending confirmation. Your own estimate is about 1.5 hours per client per week even top tier with video, which puts you at 15 to 20 clients of capacity, not the panel's ceiling. This test confirms that estimate before you bet the business on it. The full time-audit template is on the discovery page.
Instrument: Time-audit template, in the discovery playbook
Margin and tool-cost desk calc
OPEN SERIOUSKills this assumptionThat the low tiers ($300 Get Found, $799 Get Chosen) throw off enough margin to matter, and that the income target survives after software and self-employment tax.
Do thisPrice the fully-automated stack for the $300 tier: the actual software licenses per location (local-SEO tooling, citation management, whatever the automation really needs). Subtract that plus your effective tax rate from the price to find real cash margin per account. Then write the exact monthly revenue you need for $6,500 post-tax at your tax rate, and the minimum client mix in plain numbers.
CostA half-day spreadsheet. No interviews.
Pass / fail lineIf the $300 tier's cash margin after tools and tax won't compound toward the target without an unmanageable client count, that tier is a loss leader, not a revenue line. Hold the flagship price: if a prospect balks at $2,250, downsell to $799 rather than discount the labor-heavy video tier.
Where it standsOpen, but it's arithmetic you can finish today. It pairs with the delivery time-box: hours and margin together tell you the real minimum client count.
Instrument: Pricing and tier logic in the business plan
Willingness-to-pay discovery interviews
OPEN FATALKills this assumptionThat enough clinicians will actually pay a retainer, and sustain it. Two models called this the finding that decides the business, and every number behind it in the plan is a guess. This is also where the competitive question now lives, so it covers two panel findings at once.
Do thisRun 15 structured Mom-Test interviews with owners across telehealth, in-person, and a couple of international practices, weighted toward the 816 no-website beachhead. Ask the three money questions the panel wrote: what do you spend monthly on marketing today, is patient acquisition a funded top-3 priority this quarter, and would you pay around $799 a month before any results. Reconstruct real past purchases, not hypotheticals. Add the two competitive questions the panel wanted tested, straight to the clinician instead of by mystery-shopping: who else have you considered or gotten a quote from, and what did they charge, and have you thought about joining a platform like Midi or Alloy instead of hiring anyone, why or why not.
CostThe biggest one: 15 conversations over two to three weeks, five to eight a week, plus the outreach to book them (plan on contacting 100 to 150 people).
Pass / fail lineIf fewer than roughly 30 percent report existing spend at or above $300 a month AND a scarce calendar AND willingness to pay before results, the ICP is not a funded buyer. And get the price number: you need 6 or more owners naming a real monthly figure at or above about $500, backed by past spend. On the competitive read: if owners tell you a generalist already covers fact-checking at the same price, or that they'd rather join a platform than hire anyone, the wedge is decorative.
Where it standsGenuinely open, and the single most important thing on this page. The full playbook (sample, scripts, guide, rubric) is ready to run, and it now carries the competitive-substitution question too.
Instrument: The 15-Interview Discovery Playbook
Evidence-wedge value test (the A/B is the test)
OPEN SERIOUSKills this assumptionThat buyers will choose and pay a premium for evidence-validation and regulatory fluency over a generalist agency at the same price. All three models attacked this as process theater until a buyer pays for it.
Do thisThis is exactly what the outreach A/B test measures. Angle A leads with the evidence, regulatory, and research-backed wedge. Angle B leads with filling calendars and a fixed-price pilot. Split the same class of prospects and measure reply rate, discovery calls booked, and price questions. Reinforce it in interviews: ask 5 to 10 prospects whether fact-checked scripts would actually change their hiring decision or make them pay more. You can also offer 3 free scripts and watch whether they accept or insist on rewriting the clinical claims themselves.
CostRuns on top of the interviews and the outreach you're already sending. No extra week.
Pass / fail lineIf Angle B wins, or Angle A only pulls polite non-buyers, the wedge is not the purchase driver. In interviews, if more than 3 of 5 say fact-checking isn't a decisive factor, the moat gives you no pricing power.
Where it standsOpen. The A/B is the cheapest test for this finding, and it's a companion page that may be publishing alongside this one.
Instrument: The outreach A/B test (Angle A is the wedge)
Already closed, no action needed
The panel raised one more test that time has overtaken. It stays on the record so nothing looks skipped, but there's nothing for you to run.
Lead pipeline: sourcing already done
DONEAssumption testedThat there's no fast, compliant, repeatable way to feed the business with leads. The panel doubted the sourcing could scale.
Why it's closedThe directory is already scraped and the 4,851 practices are resolved and in hand. Acquisition-side sourcing is finished, not a test to run. The one live thread is compliant outreach of the data you hold (anti-spam rules differ by country), and that's already covered.
Instrument: Compliant-outreach guidance lives in the outreach A/B page
The two to three week calendar
Front-load the desk kills so a cheap test can stop you before the expensive ones start. Then run the interviews and the A/B together, because both need the same two weeks and the same outreach.
| When | Focus | What runs |
|---|---|---|
| Week 0 | Desk kills and setup | Run the SAM density desk-check on your owned data, start the delivery time-box on your next real build, and finish the margin and tool-cost calc. Then pull the interview list, send the first 10 outreaches, and set up the two A/B angles. |
| Weeks 1 to 2 | Run the human-dependent tests | Run the 15 interviews, five to eight a week, so patterns stay fresh, and work the competitive questions into each one. Send the outreach A/B in batches and log reply, discovery, and price questions by angle. End every interview by asking for a referral. |
| End of week 2 to 3 | Synthesize and decide | Roll the interviews into an opportunity map, fill the decision rubric, and read the desk results together. Then hit the single gate below. Three weeks instead of a year to know. |
The single go / refine / kill gate
At the end, one decision, and it comes down to the two questions that actually decide this business now: (1) will enough clinicians pay, and at what price, and (2) is your real per-client delivery time low enough to make that price a living. Everything above feeds these two. Decide the lines now, before a single warm conversation talks you into whichever answer you were hoping for.
GO — build it
The desk tests survive: SAM density holds under your worldwide, in-person scope, the delivery time-box lands near 1.5 to 2 hours per client, and the tier margins clear after tools and tax. AND the interviews come back with 6 or more owners naming a real monthly number at or above roughly $500 backed by past spend, most raising the acquisition pain unprompted, and at least 3 putting a paid pilot or a genuine sales conversation on the calendar. That combination means your target is about 9 to 10 clients, comfortably inside your capacity. Start selling.
REFINE — narrow and retest
The pain is real but the price anchors low, most owners land well under $500. That's the dangerous middle: a low price means you'd need 20, 30, 40 clients, and 40 clients at 1.5 hours each is 60 hours a week, straight back through the delivery wall. Or the time-box drifts toward 4-plus hours per client, usually on video. Don't proceed and don't quit. Narrow the segment (toward the 816 no-website beachhead, or in-person local SEO where value is easiest to prove), cap the video in the package, reshape the offer, and test the price again.
KILL — stop, and be glad it was three weeks
Owners don't raise the patient-acquisition pain on their own and won't spend real money on it. OR the delivery time-box confirms 30 to 35-plus hours a week at the client count you'd actually need, and no price you can charge closes that gap. OR the interviews say plainly that a generalist agency already does this at the same price and the evidence wedge changes nobody's decision. Any one of those, on its own, is a stop. The honest move then is to stop, and you'll have spent three weeks finding out instead of a year.