RevParPro · Decision memo
Prepared 2026-08-20 · answers Founding Checkout $108 Bounce Risk · four independent seats consulted, one cross-vendor · every quantity below carries its source
The note this memo answers says the printed $9/month line "is ink and cannot change." A live
query against RevParPro production on 2026-08-20 says otherwise: mail_campaigns holds one row
(dev-frozen-check, status draft), mail_pieces holds only the dev demo
piece, plan_products.gm_named_list_monthly.price_cents is NULL, and
rpp_accounts holds one row with status comped. No campaign minted, no per-piece codes,
no final PDF, and the envelope still carries the placeholder 123 Sample Street, which blocks
printing outright.
Nothing has been printed and nobody has ever paid RevParPro anything. Every option is still open at zero cost, and there is no conversion, churn, or refund data anywhere in the system. Every conversion number anyone cites, in this memo or elsewhere, is a guess.
Sources: live PostgREST query against RPP prod, 2026-08-20; RPP Direct Mail Campaign.md
lines 140 and 337.
Charge $9 a month. Take the card at the ask, charge nothing that day, and take the first $9 on the day the July report actually ships. Kill the annual condition entirely.
Recorded terms: .agents/prompts/rpp-paid-compset-funnel-agent-2.md
(gm_named_list_monthly 900 cents/month, rpp_membership_standard 1500 cents/month,
rpp_monitored_hotel_addon 900 cents/month per hotel beyond the first, and the continuity rule).
Comptroller cadence: June 2026 first populated 2026-07-27, July 2026 first appeared 2026-08-10
(sift_ingest_periods, research/07). Live reserve copy: funnel brief section 4.5.
Marginal cost: every figure computed from free Texas public tax receipts,
supabase/functions/_shared/compset-engine.ts buildMonthlyAudit lines 592 to 899.
It costs the metric you said you care about, in order to protect an amount that is small, and it does not actually buy the churn defense it was hired for.
| Charge | Gross | Card fee | Net | Fee % |
|---|---|---|---|---|
| One month at $9 | $9.00 | $0.561 | $8.439 | 6.23% |
| One year prepaid at $108 | $108.00 | $3.432 | $104.568 | 3.18% |
| Twelve monthly charges | $108.00 | $6.732 | $101.268 | 6.23% |
| Annual advantage, per member per year | n/a | n/a | $3.30 | n/a |
Applies Stripe's published US standard rate of 2.9% plus $0.30 per successful charge. unverified RevParPro has no Stripe account, so this is the published rate, not a contracted one. Postage for the whole 100-piece drop is $78.00 (100 at $0.78, USPS Notice 123 metered First-Class 1 oz, effective 2026-07-12), so $3.30 a member is noise against the cost of the drop itself.
The founding discount costs $6.00 a month gross against the recorded $15.00 standard, which nets to $5.826 a month, or $69.91 a year, per member, forgone forever.
| Founding members | Forgone per year | Forgone over five years |
|---|---|---|
| 3 | $209.74 | $1,048.68 |
| 5 | $349.56 | $1,747.80 |
| 10 | $699.12 | $3,495.60 |
| 50 | $3,495.60 | $17,478.00 |
Arithmetic on the recorded $15.00 standard minus the $9.00 founding rate, net of the assumed published card rate above. Independently recomputed this session, not taken from a single source.
Annual-only beats monthly on year-one cash only if the $108 wall barely dents signups. Writing k for the share of would-be signups who still sign up at $108, and m for the months an average monthly founder pays:
| If a monthly founder pays | Annual-only must keep this share of signups to win |
|---|---|
| 12 months | 96.8% |
| 9 months | 72.6% |
| 6 months | 48.4% |
| 3 months | 24.2% |
Solves k × $104.568 > m × $8.439, so k > 0.0807m. Verified arithmetic this session. Neither k nor m exists anywhere unverified, because RevParPro has never charged anyone. The point is not the exact number: the condition only wins if the $108 ask costs you almost nothing, and your own stated worry is that it costs you a great deal.
| Retained | $9 monthly | $108 prepaid | Difference |
|---|---|---|---|
| 12 months | $101.27 | $104.57 | $3.30 |
| 24 months | $202.54 | $209.14 | $6.60 |
| 36 months | $303.80 | $313.70 | $9.90 |
| 60 months | $506.34 | $522.84 | $16.50 |
Net of the assumed published card rate. Not a forecast: no retention curve exists, so read this as contribution per retained period, not lifetime value. At three years the cadence is worth $9.90. The cadence choice is a conversion decision, not a revenue decision.
$9 a month is not a credibility problem. It is category-correct at the bottom rung, and the close analogs all price monthly, discount annually, and trial without a card.
| Product | What it is | Monthly | Annual | Trial |
|---|---|---|---|---|
| Texas Bar Tab | Texas public receipts, named businesses | $9.95 | $99.95 | free 30-day |
| BoozeReports | Texas public receipts, named bars | $12.00 | $99.00 | 7-day, no card |
| Source Strategies Silver | Texas hotel tax data, the incumbent | n/a | $349.00 | 1-week, no card |
| Source Strategies Gold | same, deeper tier | n/a | $849.00 | 1-week, no card |
| Source Strategies Platinum | same, top tier | n/a | $1,249.00 | 1-week, no card |
| AirDNA Research | short-term-rental analog | $125.00 | $400.00 | free tier, account |
| Lighthouse Starter | rate shopping and benchmarking | 99€ | not stated | free dashboard |
| STR / CoStar entry reporting | the thing GMs actually have | n/a | from $750 | participation-gated |
| RevParPro founding | Texas hotel tax data, named peers, monthly | $9.00 | none at the door | free June report, no card |
All competitor prices from
~/agents-runtime/docs/rpp-mailer-funnel-brief-archive/research/06-analog-products.md,
vendor pages read 2026-08-18, independently re-read against that archive this session. AirDNA figures from a
Wayback snapshot of airdna.co/pricing captured 2026-08-09. STR entry price is a HotelMinder third-party
listing, not a CoStar page.
The reading that matters. Two products selling the identical shape of thing, Texas public filings cleaned up and delivered monthly, sit at $9.95 and $12.00 a month. RevParPro at $9.00 sits inside the established band, not below it. Cutting to $1 or free to remove friction would tell the buyer you do not believe $9 is worth it, and would convert a customer into a trialer. Meanwhile every close analog offers annual as a discount and trials with no card. Requiring $108 up front with no trial would be the most aggressive ask in the entire category, from the vendor with the least reputation in it.
The fine line under the founding card currently argues against the offer it sits under.
Cancelling is the single act that destroys the founding rate, so the current line advertises the one thing that costs the member $69.91 a year. Stated plainly instead, that sentence is the churn defense the annual condition was invented to provide, it is carried in the member's head every month, and it costs zero conversions and zero dollars to enforce. All three internal seats flagged this line independently.
Leave it. It is already correct under this model. The mismatch everyone has been trying to solve
exists only because the annual condition was bolted on after the line was written: remove the condition and
the paper is right as drafted. Two seats proposed
Founding: $9/mo for life at 49 characters, and it is a better offer line, but it prints an
uncapped lifetime promise from a company that has never charged anyone, on an artifact that has already been
physically proofed through the envelope windows. Not worth reopening for two characters. Revisit "for life"
for the 500 to 1,000 piece batch, once the first cohort has produced evidence.
Character counts measured this session, not estimated. Current line verified in
artifacts/direct-mail/rpp-best-western-austin-sample-v06.html line 773.
Four seats, briefed from one packet with verified facts and assumptions separated, answering independently without seeing each other.
Unanimous, four out of four: price at $9 a month, monthly cadence, and delete the annual condition. No seat defended the $108 ask. Because fast agreement is a warning sign, note that the seats disagreed sharply on when money moves and on what to print, which is recorded below rather than smoothed away.
Advisor · offers and friction
agrees $9 monthly, charged at checkout, plus an unconditional monthly refund.
dissents Would charge the $9 on day one, not on delivery, and would print "for life." Argues the constraint is proof, not price: a GM at a hotel doing $200,000 a month in receipts is not stopped by $9, he is stopped by not knowing you will exist next month, and a refund answers that where a lower price does not.
CFO · unit economics
agrees $9 monthly, card on file now, first charge on delivery. Supplied the break-even and the forgone-margin math above.
flags The strongest counter to "for life": it is an uncapped promise from a company with no pricing history, and a stated term would keep repricing power. His mitigation, adopted here, is the scope limit: the lock covers the base membership and first hotel only.
CRO · the buyer
agrees $9 monthly, and brought the category price anchors that settle whether $9 reads as cheap.
dissents Would charge $9 at checkout and would keep a $90 annual toggle available from day one, never as default. Argues a recurring charge on the hotel P&L is the introduction to the management company you actually want, while an annual prepay hides on a personal card.
Cross-vendor contrarian, on Gemini
agrees $9 monthly, month to month.
dissents Says take no card at all for this drop: billing does not exist, the build is formally blocked, and arguing billing cadence for a system with no billing is the wrong fight. Would ship the card-less reserve that is already live and build behind it.
Two seats wanted the $9 charged at checkout. One wanted no card at all. The deciding facts were that the live shipped copy already promises "You pay $9 then, or not at all," so cohort one will have read that sentence before reaching any checkout, and that capturing a card with no immediate charge is a materially smaller build than metered subscription billing. Card on file now, first charge on delivery, satisfies the stated goal of cards down, keeps the promise already made in print and on screen, and puts the first dollar next to the first delivered unit of value. It is a better answer than any single seat gave.
| Unknown | Why it matters |
|---|---|
| Every conversion, churn, and refund rate | No RevParPro customer has ever paid anything. There is no data, and nothing in this memo forecasts any. |
| The real card processing rate | No Stripe account exists, so all fee figures use the published US rate unverified. Every fee number moves if the real rate differs. |
| Whose money it is | Assumed to be a personal card or a small expense line unverified. If procurement is involved, the ranking inverts and one $108 invoice can be easier than a recurring $9 charge. |
| Involuntary churn on $9 cards | Twelve small charges are twelve chances for a card to fail. Nothing in RevParPro measures this. If it runs high, the annual charge stops being a conversion tax and becomes a retention mechanism. |
| Cost per piece beyond postage | Paper, toner, and envelope carry no dollar figure anywhere. Every acquisition-cost number is a postage-only floor of $0.78 a piece and the true figure is higher by an unknown amount. |
| Whether a GM can name 4 to 12 competitors from memory on a phone | Untested, flagged in the August panel. The whole set-picker step depends on it. |
Delivery. Three of four seats landed here independently. The monthly issue generator is unbuilt
work, and the paid checkout plan is formally blocked at Phase 2, with Agent 3 returning
NEEDS_CONTEXT because migration 910 cannot represent the required upload lifecycle.
A model that charges next to every delivery punishes a slipped delivery every month instead of once a year. If the July issue is late, a stranger who just paid an unknown vendor gets nothing, cancels, and under the lock rule that cancel permanently destroys his founding rate. That converts a one-week engineering slip into a permanently lost customer, in two markets where hotel GMs talk to each other.
The refund guarantee is the mitigation: it turns a missed issue from a betrayal into a $9 apology. But the actual fix is that the July issue ships, by hand if it has to.
Sources: RPP Direct Mail Campaign.md "Latest 2026-08-20" and decision line 204;
funnel brief section 12.
| # | Action | Owner | Blocks |
|---|---|---|---|
| 1 | Correct the recorded commercial terms in .agents/prompts/rpp-paid-compset-funnel-agent-2.md: strike the annual condition, write the lock scope (base plus first hotel only), the break rule (voluntary cancel or unpaid, recovered card failure does not break it), and the token gate. The record is what the webhook gets built against, and it currently disagrees with the stated intent. | Woz | the webhook |
| 2 | Replace the founding card fine line on production. Text only, costs nothing, and it is the entire churn defense. | Woz | nothing |
| 3 | Answer one dated question before print: can a card be placed on file, authorization only with no charge, by mail date? That is a far smaller build than metered billing, and it is what makes the drop measure willingness to pay rather than intent. | Woz | the drop's success metric |
| 4 | Supply the real return address. This blocks printing outright and has nothing to do with pricing. | Ace | the entire drop |
| 5 | Commit to a July issue ship date, by hand if necessary. The pricing model above is only safe if delivery is reliable. | Woz | cohort retention |
The execution-mode policy at ~/agents-runtime/scripts/cos-execution-mode.py classifies
this decision as reconcile-first, because the recorded terms and the stated intent conflict. That
is why item 1 is first: reconcile the record before anything is written against it.