Does the retention curve flatten?
MapView is pre-revenue. The next 8 months are not about profit — they are about producing measured evidence of a durable, cheap-to-acquire, retaining user base, packaging it into a deck, and raising. The retention curve is the scoreboard; revenue is a secondary signal.
Retention curve — flattening vs decaying target
Investigator → Guardian, with the handoff target
Proof-point status
Now
Current gate
Next decision
Days to deck-ready
Top three risks
1. P2 is structurally thin by M6
The Investigator→Guardian handoff needs a 3–6 month buying window; Investigators acquired in M2 largely won't have purchased by deck time. The hinge of the thesis has the fewest observations exactly when it's needed. Mitigation: a Guardian seed cohort recruited directly.
2. The curve decays instead of flattens
If week-24 retention falls below half of week-12 (the flattening test — defined once on /proof), the value prop isn't earning return visits and Gate 3 is a no-go regardless of anything else.
3. Acquisition is costlier than modelled
If cost-per-retaining-user stays above $50 (the Gate 2 kill threshold — see /model) and organic share doesn't rise, the "cheap to acquire" half of the pitch fails even if retention holds.
The moat is impartiality
Consumer positioning — verbatim
Short
Domain and REA show you what's for sale, and only the good bits. MapView shows you what a property is really like: the good and the bad. So you can decide for yourself.
Long
For everyday Australian property buyers and owners who are making the biggest financial decision of their lives inside an industry built to sell to them, MapView shows you what a property is really like — the good and the bad — so you can decide for yourself. [Opening clause verbatim; body reconstructed from the approved argument — confirm against source.]
Investor framing investor only
MapView is the independent information layer for Australian property — the impartial source of truth outside the sales machine.
This is the investor narrative, not the consumer voice — the consumer voice is the "good and the bad" line above.
The moat
Every incumbent information source is sales-aligned: agents are paid by vendors, REA and Domain are paid by agents. The buyer has the most at stake and the least information. Incumbents cannot close that gap without cannibalising the vendor relationship that pays them. Impartiality is a category-defining moat, not a feature gap.
Hooks
Buyer hook
"What aren't they telling you about this property?"
Leads with Site Alerts, not MVPS. Easements, flood, bushfire, heritage are emotionally concrete and PR-able.
Owner hook
"What is your neighbour about to build?"
Permit alerts — builds the durable Guardian relationship that flattens the curve.
Tone of voice — verbatim
- Confident, not arrogant
- Plain-spoken, not technical
- On the buyer's side, always
- Cheeky challenger, not a whinger
- Trustworthy through transparency
- Added (not in source): never alarmist — fear opens the door, facts close the sale
Wrong vs right framing
Rejected
"Comprehensive property data app" — utility framing: low emotion, low virality, low retention.
"Property credit score / passport" — cold, implies a single transaction. (The MVPS lives inside the product; it is not the brand.)
Approved
"Independent information layer / impartial source of truth." Category-defining, emotionally resonant (fairness), retention-aligned (the impartial source you keep returning to).
Editorial rule — the fear/impartiality line
Operating rule: every risk surfaced must be paired with the property's genuine strengths, and MapView never publishes a negative-only view of a property. The hook opens with risk; the report must close with a balanced, evidenced picture. Testable standard: each property page/report carries ≥1 strength callout for every risk callout (a ≥1:1 ratio, audited per release). This is a content standard, not a sentiment.
PR legal standard
- Verification threshold: any risk claim (flood, bushfire, easement, heritage) must cite the authoritative source (council overlay, state dataset), be date-stamped, and be independently checkable before publication.
- De-identification default: property-specific stories must not name the address or make the property identifiable without written evidence and legal review. Naming a live listing carries defamation and misleading-conduct exposure.
- Aggregate-vs-property weighting: aggregate and suburb-level stories carry far less risk — weight the PR calendar toward them. Property-specific stories are the exception, not the rule.
- Legal review trigger: any story naming a specific listing/address, or making a comparative claim ("the listing didn't mention it"), requires a contemporaneous listing screenshot and triggers legal review before publish.
- Sign-off: a named legal/exec owner signs off property-specific stories; the marketing owner signs off aggregate stories.
Investigator → Guardian
Two archetypes, one person at two life stages. Acquire the Investigator, engineer the conversion to Guardian — the handoff is what flattens the curve.
Tier priorities
Tier 1
First-home buyers and burnt-once upgraders — emotionally engaged, data-hungry, high activation, long buying window. The primary Investigator pool.
Tier 2
Individual investors (repeat research, durable) and settled owners (the Guardian pool — harder to acquire, durable once acquired).
The handoff
An Investigator buys, and in that moment becomes a Guardian: home_registered fires, and they shift from "what aren't they telling me?" to "what is my neighbour about to build?". Retention stops being time-boxed to the buying window and becomes indefinite. The conversion rate (investigator_to_guardian_conversion) is the single most important number in the pitch — and the hardest to observe in 8 months (see /proof → P2 timing).
Cut list — and why
Never sell to the sales side — moat breach
Agents, buyer's agents, vendors and listing portals are removed entirely as segments and partners. Selling to the people who sell property makes MapView a vendor-aligned source — it destroys the impartiality moat and the entire positioning.
Cut — renters
Excluded from the launch strategy: they do not convert to Guardians on a property they don't own, so they contribute no P2 and thin the retention cohort. Revisit only if a renter→buyer handoff can be shown.
Moved — conveyancers, brokers, accountants
Not consumer segments — they move to /b2b as the non-selling side: low-cost consumer-acquisition channels (co-branded reports) and an upside signal, never a revenue business in this window.
One loop, two ignition sources
Not a funnel. A flywheel with two things that light it: paid social (buys Investigators fast) and data PR (earned reach, creates Guardian demand).
Readouts derived from /model. Drag the scrubber, click a stage, or use the scenario toggle.
A closed loop: Share re-enters Acquire. Ignition (Paid Social + Data PR) lights it from outside and fades as organic share grows; Convert to Guardian is the engine that stops the curve decaying. Revolutions = cumulative registrations ÷ 1,000; velocity ∝ cumulative consistent users; leakage and conversion follow the scenario's consistent-user and purchase-conversion rates.
Ignition 1 — Paid Social
Buys Investigators fast and validates hooks. Funded by investment, not by $9.95 report margin — the v1 CAC-ceiling-from-gross-profit constraint no longer governs channel selection (see /model). The measure is cost-per-retaining-user, not cost-per-registration.
Ignition 2 — Data PR
Earned reach that creates Guardian demand directly ("what is your neighbour about to build?"). Slower to spin, but it produces the owners the flywheel needs — and the credibility the deck needs.
Three compounding loops
Shareable property pages
A property page is a shareable object passed between partners, family and advisors — each share re-enters Acquire with zero marginal cost.
Investigator → Guardian handoff
The same person, two life stages. Converting buyers into owners converts a time-boxed relationship into a durable one — the retention flywheel's core.
Lifecycle emails
Watchlists and permit alerts that pull users back — the cheapest re-entry into Engage & Retain.
Five-stage lifecycle
Acquire → Activate → Engage & Retain → Convert to Guardian → Share — drawn above as a closed loop. The full event taxonomy (stage → events → what it proves) lives once on /measurement; it is not duplicated here.
What we have to prove to raise
Seven proof points. P0 is proven by P1 + P2; P3 proves affordability; B2B proves it scales cheaply; the $ are bonuses.
| ID | Proof point | Question | Target | Status | Satisfying evidence |
|---|
Retention curves
The numeric definition of "flattening"
P2 timing risk — the primary structural risk
Cohort table
| Reg. cohort | Registrations | Consistent-User rate (target) | Consistent Users | Cost / retaining user | Read at |
|---|
Eight months, three gates, two clocks
Gates — pre-committed, numeric
Two clocks — growth and raise in parallel
Growth clock
Raise clock
Kill criteria (pre-committed)
| Initiative | Kill / hold threshold |
|---|
M1 dependency — instrumentation before spend
Instrumentation must be complete before any paid spend (M2). The instrumented app build must be released before M2 — iOS App Store review adds 1–7 days, Play Console hours–days. Schedule the instrumented release by M1 week 3 to protect the M2 start. Web registration + analytics must also be instrumented in M1. Any App Store / distribution delay pushes M2 — and the whole two-clock timeline — right.
Four-tier measurement hierarchy
Tier 1 is the hero — the retention curve. Tier 4 is vanity and is never led with.
| Tier | Purpose | Metrics |
|---|
Locked metric definitions
Locked — reproduce verbatim, do not redefine. Change control below.
Event taxonomy — the instrumentation spec
| Lifecycle stage | Event | Meaning | What it proves |
|---|
Instrumentation checklist
Definition versioning (change control)
Definitions are locked. Any change requires a definition change request: a named owner proposes it, it is logged with date + reason, it needs growth-owner sign-off, and every historical cohort is re-computed under both the old and new definitions for the transition window. The taxonomy carries a version (e.g. metrics v1.0) and a definition is never silently redefined.
The non-selling side
In this window B2B is not a revenue business. It is (a) a low-cost consumer-acquisition channel via co-branded reports, (b) an upside signal for the pitch, (c) validation that the dataset has commercial value.
Moat guardrail — who MapView will never sell to
Agents, buyer's agents, vendors, and listing portals. These are the sales side. Selling to them makes MapView vendor-aligned and destroys the impartiality moat. This list is binding and must not be eroded by any commercial opportunity.
Targets (non-selling side)
| Target | Why | Cycle |
|---|
Co-branded reports — the tracking chain
A co-branded report must carry a trackable link measuring Report → Registration → User Retained. A partner is only valuable if their reports produce retaining users — not clicks. Chain:
Pipeline stages
| Stage | Definition | This window |
|---|
Budget, cohorts & cost-per-retaining-user
The cohort-size requirement drives the budget — not the reverse. All figures are modelled proposals for approval.
Inputs
Budget envelope (8 months)
Cost-per-retaining-user by source
Organic share timing
Secondary — revenue is a signal, not the scoreboard
The $9.95 report is a probe, not the business — proof point $1 (people will pay for the data). Consumer data is free; near-zero marginal cost-to-serve is the point. The paid-social channel is funded by investment, not report margin.