Retainer Proof
Generates a branded, client-ready monthly care-report PDF for small WordPress agencies from free public signals, with nothing installed on the client site.
One product, one founder, thirty days. Everything below is built to be executed inside the contest window — the official rules still decide what counts.
Decide in 20 seconds
The facts that decide whether this is worth thirty days of yours.
Owner-operator of a 1-10 person WordPress design or dev agency in the US, UK, Canada or Australia selling monthly care plans across 15-60 small-business client sites.
Operational simplicity 5.7/10
Why this could work in 30 days: one founder can ship the MVP in 5 days; the plan targets a first sale on day 10; 35 sales at $29 clears $1,000 of collected revenue; strongest signal is speed to first revenue at 8.7/10.
Score and projections come from the generated report and are not guarantees. The $100K Hackathon is run by Ship or Die — read the official rules before you build.
Your first 24 hours
12 concrete actions. Tick them off as you go — nothing is saved or sent.
Product opportunity and positioning
Target customer
Owner-operator of a 1-10 person WordPress design or dev agency in the US, UK, Canada or Australia selling monthly care plans across 15-60 small-business client sites.
Why now
The buyer already charges monthly and already has a renewal conversation on the calendar, so this is ammunition for an existing budget line rather than a new purchase.
Problem
The agency promises ongoing maintenance but has no client-visible proof of work; assembling a report by hand takes 20-40 minutes per client, so most agencies send nothing and absorb doubt emails and quiet cancellations at renewal.
Solution
The agency adds a client domain and brand colours once; a scheduled pipeline pulls public signals, diffs them against last month's snapshot, and emails a one-page white-label PDF titled Maintenance Report for [Month] with changes, grades and a plain-English summary.
1. Product opportunity and positioning
The buyer is the owner-operator of a 1-10 person WordPress agency in the US, UK, Canada or Australia selling monthly care plans across roughly 15-60 small-business sites. That owner already bills a monthly maintenance fee per client. What they cannot produce is the artifact: a client-visible, forwardable page that says what changed on the site this month and why the retainer is still worth paying. Assembling that by hand - uptime screenshots, plugin screens, a PageSpeed grab - takes 20-40 minutes per client and gets skipped. The cost of skipping is quiet: "what am I paying for?" emails and non-renewals 30 days later.
Retainer Proof sells into an existing budget line, not a new one. It is not observability for engineers; it is proof-of-work for a business owner. The wedge is the passive check: nothing is installed on the client's site, no credentials are requested, no client cooperation is needed. TLS certificate expiry, DNS and nameserver state, HTTP status and latency samples, Core Web Vitals and a WordPress version fingerprint are all publicly observable. Diffed month over month against a stored snapshot, they become a one-page PDF titled Maintenance Report for [Month], carrying the agency's logo, brand colour, grade badges and an editable "what we did this month" paragraph.
Differentiation against the realistic alternative - Canva, a screenshot collage, or nothing - is that the output is generated rather than assembled, and the diff is the part a human will not do. Differentiation against full observability platforms is the buyer: those are priced and sold to engineering teams and require an agent. The JEV score of 67.4 (band B) reflects that tension honestly: speed-to-first-revenue scored 8.71 because the buyer is identifiable and already paying for the outcome, while market demand scored 5.86 because the category is unproven. That means the first 30 days must be measured in delivered reports and forwarded reports, not impressions.
The narrow wedge is deliberate. Do not add malware scanning, uptime alerting, client dashboards or a template marketplace. Each of those moves the product away from the one sentence that sells it: this is the page your client sees for the money they already send you.
2. Smallest sellable MVP
The sellable unit is one branded PDF for one domain for one month. Everything else is convenience. Ship: an onboarding form (client name, domain, logo, brand colour, monthly send day); collector jobs for TLS expiry, DNS and nameserver state, three daily HTTP status and latency samples, and a PageSpeed Insights mobile score; a WordPress fingerprint pass over public HTML and asset query strings; a snapshot diff engine; a React-PDF one-pager with grade badges and an editable copy block; Stripe Checkout with magic-link accounts and plan limits enforced by domain count; and a scheduled monthly send through Resend with a preview link before dispatch.
Explicitly out of scope for all 30 days: no WordPress plugin, no site login or credentials, no client-facing dashboard, no malware or security-audit claims, no uptime paging, no CRM or invoicing, no marketplace or affiliate program. Every one of those would add a week and remove none of the objection the buyer actually has.
3. Technical implementation plan
Next.js on Vercel, Supabase Postgres for accounts, domains and monthly snapshots, Vercel Cron for the monthly run, Node fetch collectors, PageSpeed Insights free API, React-PDF for the artifact, Stripe Checkout and Billing Portal, Resend for delivery. Build order is schema, onboarding, collectors, diff engine, PDF renderer, billing, deploy - five working days.
The thing most likely to break is fingerprinting. Minified, CDN-cached or hardened HTML hides the WordPress core version and asset query strings, so a report comes back thin or wrong. Mitigation: fall back to feed generator tags, readme files and asset path patterns, and when a signal cannot be observed, print "not detected this month" rather than guessing. A wrong number in a client-facing PDF is a refund and a lost agency. Collectors must fail loudly to the founder and softly to the reader.
4. Pricing and monetization
Three tiers ship on day five: Single Report at $29 one-time, Care Plan at $99 per month for up to 15 client domains, Agency at $199 per month for up to 60 domains with direct-to-client sending and per-client copy blocks. The $29 tier exists to make the first yes trivial and to prove the artifact actually gets forwarded; it is a qualification step, not a business. A realistic 30-day revenue blend lands near $95-$120 ARPU because most collected revenue is a monthly plan charged once inside the window. The upsell path is mechanical: $29 buyer forwards one report, then moves to $99; $99 customer asks to send directly to clients, then moves to $199.
5. First-customer acquisition
One channel only: direct outreach to agency owners whose own site advertises a monthly care or maintenance plan. Build a 200-row list from X, LinkedIn and public agency directories, capturing owner name, the exact care-plan claim they make, and one client domain. Run that agency's own domain through the pipeline and attach the finished PDF to the first message. The angle: "This is what your client sees for their $X a month. Yours, white-labelled, in 60 seconds." Follow up at day 3 and day 7 with the same PDF and one question: want the next one automated? Week-one target: 120 qualified rows, 40 personalised sends with attachments, 5 replies, 2 sample reports delivered.
6. 30-day execution roadmap
Days 1-5 are build, with the product accepting payments on day 5. Days 6-7 are sample generation and the first 40 sends. Days 8-14 are outreach waves, first $29 conversions and the first $99 close; the first sale is targeted for day 10. Days 15-21 automate the scheduled monthly send, ship the $199 Agency tier and audit report accuracy against real client sites. Days 22-30 are a fourth and fifth outreach wave, community posts of the sample artifact, an upsell pass over every paying account, and reconciliation of collected revenue before the 16 December close.
7. Revenue and profit scenarios (projections only)
These are hypotheses, not verified revenue. Conservative: 4 customers at $95 blended ARPU for $380 gross, $12 processing fees, no ad spend, $25 operating cost. Base: 12 customers at $105 for $1,260 gross, one $29 refund, $43 fees, no ad spend, $60 operating cost. Upside: 30 customers at $118 for $3,540 gross, $70 refunds, $122 fees, $180 paid marketing, $145 operating cost. Hackathon profit excludes hosting and tooling by rule; business profit includes them. Processing fees assume Stripe at roughly 3.2-3.4% blended.
8. Risks and validation experiments
The three tests that matter. First, fingerprint reliability: run 30 real agency-client domains through the collector and count how many yield a usable core-version or asset-version diff; below 60% usable, the diff becomes the weaker signal and the report must lean on certificate, DNS and PageSpeed instead. Second, conversion from artifact to payment: send 15 sample PDFs and measure how many reply and how many pay $29; under 1 payment from 15 sends, rewrite the message, not the product. Third, forwardability: after the first five reports are delivered, ask each buyer directly whether they forwarded it to the client; if none did, the artifact is wrong regardless of accuracy. Kill criteria: if 40 personalised sends with attached PDFs produce zero replies by day 12, stop outreach and re-examine whether the sample PDF itself reads as an ad rather than as proof.
9. First 24 hours
Hours 1-3: register the domain, create the Vercel project and Supabase database, wire Stripe in test mode, deploy a skeleton. Hours 4-6: build the first 50 rows of the agency list with owner names and care-plan claims quoted verbatim. Hours 7-9: write the collectors and run them against five known WordPress sites to confirm TLS, DNS, HTTP sampling and core-version fingerprint all return values. Hours 10-12: draft the outreach message, render one sample PDF for a self-owned domain, and check that the one-pager reads cleanly on a phone. Output by midnight: a deployed skeleton, 50 qualified rows, a verified collector, and one sample PDF good enough to attach.
Smallest sellable MVP
A single branded one-page PDF for one client domain, generated from public signals and diffed against last month, delivered within minutes of a $29 payment.
Must-have features
- Onboarding form: client name, domain, logo, brand colour, monthly send day
- Collectors for TLS expiry, DNS/nameserver state, three daily HTTP status and latency samples
- PageSpeed Insights mobile score capture
- WordPress fingerprint pass over public HTML and asset version query strings
- Month-over-month snapshot diff (cert renewed, core version change, score delta, downtime minutes)
- React-PDF one-page renderer with logo, grade badges and editable copy block
- Stripe Checkout plus magic-link account with plan limits by domain count
- Scheduled monthly send via Resend with preview link before dispatch
Explicitly out of scope
- No WordPress plugin, agent or code install on client sites
- No client-facing dashboard or login for the agency's customers
- No malware scanning or security-audit claims
- No uptime alerting, on-call paging or incident response
- No CRM, invoicing or helpdesk features
- No template marketplace or affiliate program in the first 30 days
Demo flow (what the buyer sees)
- Agency owner signs up and enters client name, domain, logo, brand colour and send day
- Pipeline runs collectors immediately and stores a first snapshot
- Report renders as a one-page branded PDF with grades and editable copy block
- Owner previews it via emailed link, edits the 'what we did this month' paragraph
- Owner pays $29 or subscribes, then forwards the PDF or schedules the next send
Technical implementation plan
Stack
Architecture
A Next.js app on Vercel holds the onboarding form and account pages. Supabase stores accounts, domains, per-domain brand settings and one snapshot row per month. A Vercel Cron job wakes monthly per domain, runs the collectors, writes a new snapshot, diffs it against the previous one, renders a React-PDF, stores it, and emails a preview link to the agency. Stripe webhooks set plan state and domain limits; Resend handles both preview and final delivery.
Build steps
- Create Supabase schema: accounts, domains, snapshots, reports, subscriptions
- Build onboarding form with logo upload and brand colour picker
- Write collectors for TLS expiry, DNS/NS, HTTP status and latency sampling
- Add PageSpeed Insights call and WordPress fingerprint parser
- Build snapshot diff engine producing a change list per domain
- Build React-PDF one-page renderer with grade badges and copy block
- Integrate Stripe Checkout, webhooks and magic-link auth with plan limits
- Wire Vercel Cron monthly run and Resend preview/final send
- Deploy to Vercel, run 10 real domains end to end, fix thin reports
AI components
- No LLM required for v1; the plain-English summary block is generated from deterministic templates over the diff fields, with an optional LLM rewrite of the agency's own copy if added later
Pricing and monetization
Single Report
Entry offerFor Agency owner testing the artifact with one client before committing
- One white-labelled branded PDF for one domain
- Month-over-month change summary
- Full-detail HTML version for editing
Care Plan
For Agency running up to 15 client care plans
- Up to 15 client domains
- Automated monthly branded PDFs
- Snapshot diffing and change log
- Scheduled send with preview before dispatch
- Agency logo and colour white-labelling
Agency
For Agency running 16-60 client care plans
- Up to 60 client domains
- Direct-to-client sending from agency domain
- Custom report copy blocks per client
- CSV export of all monthly grades
Model: Freemium-free sample to paid one-off report, then monthly subscription priced by client-domain count.
First revenue offer: Free branded sample report of their own agency domain in the first message, then $29 for a report on one real client domain delivered within 30 minutes of payment.
Expected ACV: $99
Upsell path: $29 single-report buyers move to $99 Care Plan once they forward one report to a client; $99 accounts move to $199 Agency when they ask to send directly to clients or exceed 15 domains.
First-customer acquisition
Direct personalised outreach to WordPress agency owners who publicly advertise a monthly care or maintenance plan.
Where to find them
- X search for 'WordPress care plan' and 'monthly maintenance plan' in bios
- LinkedIn agency owner profiles listing maintenance retainers
- Public WordPress agency directories and 'hire us' pages with care-plan pricing
- WordPress agency Slack and Discord communities
- Local business directories listing WordPress maintenance providers in US/UK/CA/AU
Exact outreach motions
- Send 40 personalised emails with the agency's own branded sample PDF attached and their name in the first line
- Follow up on day 3 with the same PDF and one question: want the next one automated?
- Follow up on day 7 with a one-line case of what changed on their own domain this month
- Offer $29 single report to anyone who replies, delivered within 30 minutes
- Post the same sample PDF as a build-in-public artifact in two WordPress agency communities
Message template
Hi {first_name} - I saw {agency} sells monthly care plans. I ran your own site through a tool I built and attached the report your clients would get: certificate status, DNS, uptime samples, PageSpeed score and WordPress version changes, all from public signals with nothing installed. Want the same white-labelled for one of your client sites? It takes 60 seconds to set up.30-day execution roadmap
Week 1 · Build & validate · 7 days
Infrastructure and list start
- Create Vercel project, Supabase database and Stripe test keys
- Write schema for accounts, domains and snapshots
- Start agency list with first 50 rows
Deliverable: Deployed skeleton plus 50-row qualified agency list
Domain capture and branding
- Build onboarding form with logo upload and colour picker
- Store brand settings per domain
- Enforce send-day selection
Deliverable: Working onboarding form saving domains to Supabase
Public signal capture
- Write TLS, DNS/NS and HTTP status/latency collectors
- Add PageSpeed Insights call
- Run against five real WordPress sites
Deliverable: Collector run storing a full snapshot for five domains
Turning signals into an artifact
- Build snapshot diff engine with change list
- Add WordPress fingerprint parser with fallbacks
- Render one-page React-PDF with badges and copy block
Deliverable: Sample PDF rendered for own test domain
Billing and deploy
- Wire Stripe Checkout for $29, $99 and $199
- Add magic-link auth and plan limits
- Deploy to production with cron config
Deliverable: Live product accepting real payments
Arm the outreach
- Run 25 target agency domains through pipeline
- Finish list to 120 qualified rows
- Fix the three worst report renderings
Deliverable: 25 branded sample PDFs plus 120-row list
First 40 sends
- Send 40 personalised emails with attached PDF
- Post sample PDF in two WordPress agency Slacks
- Log replies in a simple spreadsheet
Deliverable: 40 sends dispatched, first replies logged
Week 2 · Launch & first sale · 7 days
Volume and follow-up
- Send 40 more personalised emails
- Run day-3 follow-up on wave 1
- Reply to every inbound within two hours
Deliverable: 80 total sends, follow-up cadence running
Turn replies into paid reports
- Offer $29 single report to every replier
- Deliver each purchased report within 30 minutes
- Fix any report accuracy complaint same day
Deliverable: Three delivery-ready paid reports queued
Close first paying customer
- Close first paid order
- Ask for one-line testimonial on the report
- Record the exact objection that nearly stopped them
Deliverable: First revenue collected in Stripe
Extend the list
- Send 40 more emails to fresh rows
- Grow list from 120 to 160
- Post sample artifact in a third community
Deliverable: 160-row list, 120 total sends
Single report to subscription
- Email every $29 buyer with $99 Care Plan offer
- Attach their own report re-rendered for a second domain
- Offer to set it up for them in the call-by-email
Deliverable: First $99 subscription started
Fix broken reports
- Audit every report delivered for wrong or missing signals
- Add 'not detected this month' fallback copy
- Patch fingerprint parser for cached sites
Deliverable: Zero unresolved accuracy complaints
Week two checkpoint
- Count sends, replies, payments, refunds
- Kill or rewrite the message based on reply rate
- Ship landing page with real sample PDF
Deliverable: Week-two metrics sheet and live landing page
Week 3 · Repeat & systematize · 7 days
Scheduled monthly send
- Build Vercel Cron monthly run per domain
- Add Resend preview link before dispatch
- Test full cycle on two accounts
Deliverable: Automated monthly send working end to end
Fill to 200
- Send 40 more emails
- Take list from 160 to 200 rows
- Add a second client-domain example to the attachment
Deliverable: 200-row list complete, 160 total sends
Day-7 re-contact
- Run day-7 follow-up on all non-repliers
- Send a one-line change summary from their own domain
- Ask every replier who did not buy one question why
Deliverable: Follow-up sent, objection list updated
Ship $199
- Add direct-to-client sending from agency domain
- Add per-client custom copy blocks
- Add CSV export of monthly grades
Deliverable: Agency tier live and purchasable
Landing page rewrite
- Rewrite hero around the forwardable artifact
- Add three sample report screenshots
- Add pricing table with domain limits
Deliverable: Rewritten landing page live
Did they forward it
- Email every paying customer asking if they forwarded the report
- Offer free copy block edit for anyone who did not
- Log answers as forward-rate data
Deliverable: Forward-rate number per customer
Week three checkpoint
- Reconcile Stripe revenue and fees
- Process any refunds immediately
- Decide whether to push $199 or deepen $99
Deliverable: Week-three revenue reconciliation
Week 4 · Scale what works · 7 days
LinkedIn channel
- Send 30 LinkedIn messages to agency owners
- Attach one anonymised sample PDF
- Book any call offered, cap at 15 minutes
Deliverable: 30 LinkedIn touches, first calls booked
Build-in-public post
- Publish a thread with the sample report and the diff logic
- Reply to every comment with the tool link
- Post the same in one agency newsletter or community
Deliverable: Public post with sample artifact attached
Cut setup friction
- Reduce onboarding to a single screen
- Auto-detect WordPress version on domain entry
- Remove the manual send-day step until first send
Deliverable: One-screen onboarding shipped
Push $99 to $199
- Email all Care Plan accounts about the 15-domain ceiling
- Show them their current domain count
- Offer a one-month upgrade at the same price
Deliverable: Upgrade offers sent to all subscribers
Stop day-one drop-off
- Contact anyone who has not generated a report in 14 days
- Generate and send their report for them
- Ask what stopped them sending it
Deliverable: Reactivation attempted on every dormant account
Warm leads only
- Re-contact every replier who never bought
- Offer a free second-domain report
- Close any pending $29 orders
Deliverable: Warm-lead conversions closed or marked dead
Week four checkpoint
- Count paying accounts and monthly recurring total
- Check refunds and chargebacks
- Fix the top two support questions in-product
Deliverable: Week-four metrics and MRR snapshot
Days 29-30 · Consolidate · 2 days
Final revenue push
- Send final offer to all non-converted leads
- Confirm next monthly send is scheduled for every account
- Collect any outstanding invoices
Deliverable: All accounts scheduled, pipeline cleared
Lock the number
- Reconcile Stripe payouts against TrustMRR records
- Document refunds, fees and ad spend
- Write post-mortem with what broke in the collectors
Deliverable: Final revenue, fee and refund ledger
Revenue and profit scenarios
How the numbers are defined
Hackathon profit — Collected revenue − refunds − chargebacks − payment processing fees − paid marketing spend.
Business profit — Hackathon profit − operating costs (hosting, domains, tooling, AI/API delivery costs).
The contest ranks entrants on hackathon profit only. Hosting, tooling and AI/API delivery costs are not deducted by the organisers, but each sale still has to cover them — so we show both. Definitions verified on 2026-11-15 against the official rules.
Stated assumptions
- Base blended ARPU is about $105 because most collected revenue is a monthly plan charged once inside the 30-day window (Base).
- Payment processing assumes Stripe at roughly 3.2-3.4% blended including per-transaction fees (all scenarios).
- Paid marketing is $0 in Conservative and Base; outreach is organic and manual, so ad_spend stays at zero (Conservative/Base).
- Upside spends $180 on sponsored community placements and boosted posts only after the first five paying customers exist (Upside).
- Refunds stay low because the first artifact is delivered within minutes; one $29 refund in Base, roughly 2% in Upside (Base/Upside).
- Operating cost covers Vercel, Supabase, Resend, a domain and PageSpeed overage; it is not deducted from hackathon profit but is from business profit (all scenarios).
- No sale is below cost: a single report costs under $0.10 in compute and API calls.
- Monthly subscriptions are counted once at first charge inside the contest window; renewals after 16 December are not counted.
- Conversion from 200 personalised sends is assumed at 3-5% paid in Base and Upside, 2% in Conservative (all).
Conservative
4 customers- Refunds
- $0
- Chargebacks
- $0
- Payment fees
- $12
- Paid marketing
- $0
- Hackathon profit
- $368
- Operating costs
- $25
- Business profit
- $343
Base case
12 customers- Refunds
- $29
- Chargebacks
- $0
- Payment fees
- $43
- Paid marketing
- $0
- Hackathon profit
- $1,188
- Operating costs
- $60
- Business profit
- $1,128
Optimistic
30 customers- Refunds
- $70
- Chargebacks
- $0
- Payment fees
- $122
- Paid marketing
- $180
- Hackathon profit
- $3,168
- Operating costs
- $145
- Business profit
- $3,023
Model your own numbers
Runs in your browser · nothing is saved- Collected revenue
- $348
- Payment fees (3.9%)
- $14
- Paid marketing
- $0
- Hackathon profit
- $334
- Business profit
- $274
Hypothetical arithmetic, not a forecast. The fee rate is a flat 3.9% assumption, refunds and chargebacks are left at zero here, and none of this is tracked data.
All figures are scenario projections built from stated assumptions, not verified revenue, customers or traction.
Main risks and validation experiments
Risk 1Fingerprinting WordPress core and plugin versions from public HTML fails on cached, minified or hardened sites, producing thin or wrong reports and refunds.
Layered fallbacks (feed generator tag, readme files, asset path patterns) and an explicit 'not detected this month' state instead of a guess.
Run the collector against 30 real agency-client domains on day 6 and count how many return a usable core-version or asset-version diff.
If fewer than 60% of domains yield a usable diff, demote version fingerprinting and rebuild the report around certificate, DNS, latency and PageSpeed only.
Risk 2Agencies conclude they can make the same PDF in Canva and never convert past the $29 single report.
Lead with the month-over-month diff, which is the part a human will not do by hand, and show it in the attached sample.
After the first five $29 deliveries, ask each buyer directly whether they forwarded the report to their client and whether they would pay monthly for it.
If zero of the first five $29 buyers forward the report or accept a $99 offer by day 16, the artifact is not doing the job and the concept needs rework.
Risk 3The cold outreach volume needed for 40 customers in 30 days exceeds what one founder can send without deliverability damage.
Cap at 40 personalised sends a day, use a warmed domain, and prioritise attachment-led messages over volume.
Send 40 messages on day 7 and measure reply rate; anything under 5% means the message, not the volume, is the problem.
If 40 personalised sends with attached PDFs produce zero replies by day 12, stop outreach and re-examine the sample artifact.
Risk 4PageSpeed Insights free quota or rate limits throttle batch runs when dozens of domains render in the same hour.
Stagger cron runs across the month by send day rather than rendering everything on the 1st, and cache the last score.
Simulate 60 domain renders in one hour on day 15 and log every rate-limit response.
If more than 10% of renders fail on quota, move to per-domain staggered scheduling before onboarding any 16-60 domain account.
Hour-by-hour: the first 24 hours
Hours 1-3
- Register the domain and create the Vercel project
- Create Supabase project and write the accounts/domains/snapshots schema
- Wire Stripe in test mode with $29, $99 and $199 prices
Output: Deployed skeleton with empty schema and test-mode pricing
Hours 4-6
- Search X and LinkedIn for agencies advertising monthly care plans
- Record owner name, agency URL and the exact care-plan claim
- Build the first 50 rows of the list
Output: 50-row qualified agency list with quoted claims
Hours 7-9
- Write TLS expiry, DNS/NS and HTTP status/latency collectors
- Add WordPress core version fingerprint with feed and readme fallbacks
- Run against five known WordPress sites
Output: Verified collector returning a full snapshot for five domains
Hours 10-12
- Render one sample PDF for a self-owned domain
- Draft the outreach message and subject line
- Check the one-pager renders cleanly on a phone screen
Output: One attachable sample PDF and a final outreach message
Full 30-day launch report
Read the full generated report
Retainer Proof - 30-Day Launch Report
Concept version 1 | JEV 67.4/100 (band B) | Window: Mon 16 Nov 2026 13:00 UTC to Wed 16 Dec 2026 13:00 UTC | Solo founder, no funding, no team.
1. Product opportunity and positioning
The buyer is the owner-operator of a 1-10 person WordPress agency in the US, UK, Canada or Australia, selling monthly care plans across 15-60 small-business client sites. That owner already bills a monthly maintenance fee. What they cannot produce is the artifact: a client-visible, forwardable page showing what changed on the site this month and why the retainer is still worth paying. Doing it by hand - uptime screenshots, plugin screens, a PageSpeed grab - takes 20-40 minutes per client and gets skipped. The cost is quiet: "what am I paying for?" emails and non-renewals 30 days later.
Retainer Proof sells into an existing budget line, not a new one. It is not observability for engineers; it is proof-of-work for a business owner. The check is passive: TLS certificate expiry, DNS and nameserver state, HTTP status and latency samples, Core Web Vitals and a WordPress version fingerprint are all publicly observable. Nothing is installed on the client site, no credentials are requested, no client cooperation is needed. Diffed month over month against a stored snapshot, those signals become a one-page PDF titled Maintenance Report for [Month], carrying the agency's logo, brand colour, grade badges and an editable "what we did this month" paragraph.
The JEV score reflects the trade honestly: speed-to-first-revenue at 8.71 because the buyer is identifiable and already paying for the outcome; market demand at 5.86 because the category is unproven. The response is to measure the first 30 days in delivered and forwarded reports, not impressions. The narrow positioning is the asset - this is the page your client sees for the money they already send you. Do not broaden it.
2. Smallest sellable MVP
The sellable unit is one branded PDF for one domain for one month. Ship: an onboarding form (client name, domain, logo, brand colour, send day); collectors for TLS expiry, DNS and nameserver state, three daily HTTP status and latency samples, and a PageSpeed Insights mobile score; a WordPress fingerprint pass over public HTML and asset query strings; a snapshot diff engine; a React-PDF one-pager with grade badges and an editable copy block; Stripe Checkout with magic-link accounts and domain-count limits; and a scheduled monthly send through Resend with a preview link. Explicitly out: no WordPress plugin, no site login, no client dashboard, no malware claims, no uptime paging, no CRM or invoicing, no marketplace. Each of those would add a week and remove none of the buyer's actual objection.
3. Technical implementation plan
Next.js on Vercel, Supabase Postgres for accounts, domains and monthly snapshots, Vercel Cron for the monthly run, Node fetch collectors, PageSpeed Insights free API, React-PDF, Stripe Checkout and Billing Portal, Resend for delivery. Build order over five working days: schema, onboarding, collectors, diff engine, PDF renderer, billing, deploy.
The failure mode is fingerprinting. Minified, CDN-cached or hardened HTML hides the WordPress core version and asset query strings, so a report comes back thin or wrong. Fall back to feed generator tags, readme files and asset path patterns, and print "not detected this month" instead of guessing. Collectors should fail loudly to the founder and softly to the reader. A wrong number in a client-facing PDF is a refund plus a lost agency.
4. Pricing and monetization
$29 one-time Single Report, $99 per month Care Plan (15 domains), $199 per month Agency (60 domains, direct-to-client send, per-client copy blocks). The $29 tier makes the first yes trivial and proves the artifact gets forwarded; it is a qualification step, not a business. A realistic 30-day blend sits near $95-120 ARPU because most collected revenue is a monthly plan charged once inside the window. Upsell is mechanical: $29 buyer forwards one report, moves to $99; $99 customer asks to send directly, moves to $199.
5. First-customer acquisition
One channel: direct outreach to agency owners whose own site advertises a monthly care or maintenance plan. Build a 200-row list from X, LinkedIn and public agency directories, capturing owner name, the exact care-plan claim, and one client domain. Run that agency's own domain through the pipeline and attach the finished PDF. Angle: "This is what your client sees for their $X a month. Yours, white-labelled, in 60 seconds." Follow up at day 3 and day 7 with the same PDF and one question: want the next one automated? Week one: 120 rows, 40 personalised sends, 5 replies, 2 sample reports delivered.
6. 30-day execution roadmap
Days 1-5 are build, with the product accepting real payments on day 5. Days 6-7 generate 25 agency-specific sample PDFs and dispatch the first 40 sends. Days 8-14 are outreach waves, $29 conversions and the first $99 close, with the first sale targeted for day 10. Days 15-21 automate the monthly send, ship the $199 Agency tier, and audit report accuracy against real client sites. Days 22-30 add a LinkedIn wave, a build-in-public post, a one-screen onboarding, an upgrade pass over every paying account, and full reconciliation of collected revenue before the 16 December close. Highest-risk days are 9-10: if replies arrive but payments do not, the problem is the report, not the pitch.
7. Revenue and profit scenarios (projections only)
These are hypotheses, not verified revenue. Conservative: 4 customers, $95 ARPU, $380 gross; $12 processing fees; no ad spend; $25 operating cost. Base: 12 customers, $105 ARPU, $1,260 gross; one $29 refund; $43 fees; no ad spend; $60 operating cost. Upside: 30 customers, $118 ARPU, $3,540 gross; $70 refunds; $122 fees; $180 paid marketing; $145 operating cost. Fees assume Stripe at roughly 3.2-3.4% blended. Hackathon profit deducts refunds, chargebacks, processing fees and paid marketing only; business profit additionally deducts hosting, tooling and API delivery cost. No sale is priced below its delivery cost.
8. Risks and validation experiments
Test one, fingerprint reliability: run 30 real agency-client domains through the collector on day 6 and count usable diffs. Under 60% usable, demote version fingerprinting and rebuild the report around certificate, DNS, latency and PageSpeed.
Test two, artifact-to-payment conversion: send 15 sample PDFs and count replies and $29 payments. Under one payment from 15 sends, rewrite the message before touching the product.
Test three, forwardability: after the first five deliveries, ask each buyer whether they forwarded the report. If none did, the artifact is wrong regardless of accuracy, and the concept needs rework before more outreach.
Secondary risks with cheap probes: outreach deliverability (cap at 40 sends a day and watch reply rate), PageSpeed quota (simulate 60 renders in one hour on day 15), and accuracy complaints (audit every delivered report on day 13 and patch before scaling).
9. First 24 hours
Hours 1-3: register the domain, create the Vercel project and Supabase database, wire Stripe in test mode, deploy a skeleton. Hours 4-6: build the first 50 agency rows with owner names and quoted care-plan claims. Hours 7-9: write the collectors and run them against five known WordPress sites to confirm TLS, DNS, HTTP sampling and core-version fingerprint all return values. Hours 10-12: render one sample PDF for a self-owned domain, draft the outreach message, and check the one-pager on a phone. By midnight: a deployed skeleton, 50 qualified rows, a verified collector and one attachable sample PDF.
This brief was generated from public data about established startups, then narrowed and scored for a solo 30-day launch. Revenue and profit figures are scenario models, not verified results, guarantees or evidence of existing customers. Validate demand yourself before you spend a month on it — and check the contest rules for what counts toward your score.
Related ideas
All ideas →LogSlim Audit Pack
A browser-local tool that turns one 24-hour log sample into a paste-ready exclusion config bundle and a projected monthly savings report for teams overpaying for Datadog, CloudWatch or Elastic log volume.
Cloud Bill Autopsy
Cloud Bill Autopsy connects to a developer's read-only DigitalOcean account and delivers a PDF report listing every idle, oversized and orphaned resource, the monthly and annual dollar waste, and the exact fix for each.
PitchSheet
PitchSheet turns your Google Sheet into a live, shareable proposal with auto-updating pricing and view analytics, so freelancers and agencies never send stale numbers again.