APS doesn't disqualify. It sequences.
The job of this rubric is to tell the team — and the ad system — which buyer is ready now and which buyer is ready later. Not to filter buyers in or out. That distinction matters operationally (don't waste rep time on the wrong tier today) and matters for Meta strategy (long-cycle buyers are exactly who the offline conversion API needs to surface back to the ad optimiser).
"We don't really have bad customers as such — most people who inquire are buyers, it's just a matter of when and what they're buying."
Two empirical findings drive everything below:
- Channel of arrival is the strongest close-rate predictor. Website form ~90% close > phone (same-day) > ShopperFly > Messenger. Messenger has the volume; the warmer channels do the converting.
- Power type (10-amp vs 32-amp) is the single most decisive qualifying question. Combined with seat count and timing, three quick-replies in a Messenger opener pre-tier 80%+ of leads before a rep touches them.
Four channels. Wildly different conversion.
| Tier | Channel | Volume | Close rate | Treatment |
|---|---|---|---|---|
| C1 · Hottest | Website inquiry form | 1–2 / day | ~90% | Top of rep queue. No qualifying friction. |
| C2 · Hot | Phone call inbound | "Heaps" / day | Same / next day | Pick up immediately. Treat as ready to quote. |
| C3 · Warm | ShopperFly (on-site chat) | Lower than Messenger | Higher than FB | Done first daily — already on the website signals real intent. |
| C4 · Volume | Facebook Messenger | ~3K leads/mo | 21.1% conv | Bulk lane. Tier inside this channel (see §03). |
The website inquiry form is currently hard to find on the APS site. If even a small share of Messenger-driven traffic converted via the website form, total close rates would lift dramatically without more ad spend. This is the single highest-leverage on-site change.
Three quick-replies do most of the work.
| Question | Why it matters | Where it goes |
|---|---|---|
| Power type — 10-amp (plug-and-play) or 32-amp (hardwired)? | Binary. 10-amp = budget social spa, plug into a regular socket. 32-amp = top-of-range, requires an electrician — and some properties physically can't support it without a $6K meter upgrade. | Messenger opener quick-reply |
| Seat count — 4 / 5 / 6+? | Shortlist of spa models and price band falls out of this. | Messenger opener quick-reply |
| Timing — ready now / 1–6 months / 6+ months / unsure? | Sets buyer-readiness tier (R1–R4 below). | Messenger opener quick-reply |
These quick-replies are compatible with Meta Lead Ads instant form or a Messenger flow opener with quick replies. Either way, the answers should land on the lead record so reps and the lead system can route accordingly.
Everyone qualifies. Sequence them.
Ready · close in 0–30 days
Qualifies if all of: concrete spec signal (seat count, power type, or specific model interest); property is ready (slab installed, deck built, electrician engaged, spa zone prepared); engages 3+ exchanges or provides phone/email proactively.
Action: rep priority response, full quote, push for deposit close.
Soon · 30–180 day close · active nurture
Qualifies if: has spec direction (size or power type) but property timeline is 1–6 months; researching actively — comparing models, asking maintenance/warranty; interest-free finance interest signalled.
Action: quote sent, deposit pursued ("secures price + stock"), follow-up cadence with quiet-day price-rise nudges.
Researching · 6–24 month close · long-cycle
Qualifies if: property is being built or planned (not just renovated); curious / asking general questions; no power type or size answer yet; long-cycle property investment, not impulse.
Action: educational content path, monthly value email, low-touch from rep until property gets closer.
Critically: do not disqualify. This is exactly Meta's invisible 84% — the cohort offline conversions need to surface back into the ad optimiser.
Aware-only
Qualifies if: single message, no engagement, no spec direction, no property situation; generic enquiry, no return reply.
Action: automated reply with model gallery + pop-up nurture flow. No rep minutes until they re-engage.
Things to design around — not filter out.
| Friction | Cause | What APS does today | What we can build |
|---|---|---|---|
| Property not ready (#1) | Build delay, slab/deck/fence pending | Long-cycle nurture; deposit holds price + stock | R3 nurture flow; CAPI keeps these in Meta optimisation despite long cycle |
| 32-amp infeasible | $6K+ meter upgrade required | Pivot to 10-amp model | Surface 10-amp options early; "check your power" question in opener |
| Freight quote shock | Far-state $900–$2,200; even local $400–550 surprises | Conversation-first WA/TAS/NT; upfront NSW/VIC/QLD | Geography-aware quote display; freight app already in build |
| No interest-free finance interest | Customer flat refuses finance | Push price/timing instead | Hum already in place — quote-flow integration only |
| "Just looking" | True early research | Light nurture | R3 / R4 flow |
Genuine non-fits · rare but real
- Sub-budget intent — under-$2,000 inflatable spas. APS entry is ~$3,999. Polite redirect.
- Service / parts only — chemicals, jets, repairs. Route to Shopify or service team (matches the Shopify online rep cohort: 568 orders YTD, AOV $570).
- Existing customer post-sale — service ticket, not a sales lead.
- Bot / spam — gibberish names, identical templates across ads. Block.
- Duplicate within 30 days — merge to existing thread, don't create new lead.
Ad-level decisions outweigh lead-level ones.
The match-rate distribution across active ads is severe enough that ad-level moves are short-term highest-leverage.
| Tier | Definition | Examples | Action |
|---|---|---|---|
| A · High | ≥3% match rate, ≥200 leads | 120236972101480626 (4.3% · $425K) · 6821446035364 (4.7% · $168K) · 6604755269564 (3.2% · $186K) | Scale spend 30–50% |
| B · Average | 1.5–3% match rate | 6584368907964 (2.4%) · 6595192671164 (1.9%) | Hold |
| C · Below average | 0.5–1.5% match rate | 6610246573564 (1.3% · 15K leads — top volume but mediocre) | Audit creative; cut budget unless brand value justifies |
| D · Wasted spend | <0.5% match rate, ≥200 leads | 6594394517364 · 6568240892764 · 6612701437764 (0% match across 700+ leads) | Pause immediately |
These moves do not need Hayden — they are ad-account work. Pausing Tier D and scaling Tier A is the fastest match-rate lift available, with no creative or website dependency.
Manual rep-tag predictive value
Rep tagging is sparingly used today (<1% of leads tagged), but where it is used it is predictive:
- "Order placed" tag → 2.8% match rate (2× baseline)
- "Quote requested" → 1.9%
- "Appointment booked" → 5.6% (n=18, small but striking)
Recommendation: standardise tagging at first reply, quote, and close. The lead system shifts from an ingest log to a forecasting tool.
Rep workflow with Hayden's actual cadence.
| Touchpoint | Rep action (current) | What changes |
|---|---|---|
| New inquiry (any channel) | Hayden replies straight away — done by ~10am daily | Auto-tier on opener answers (after we add quick-replies) |
| Channel prioritisation | Implicit: ShopperFly first, then FB | Make explicit — surface website-form + phone above Messenger |
| Quote sent | Manual: PDF emailed to self, then to customer | Phase 02: quote-automation tool replaces this |
| Deposit close | "Secures price and secures stock" — Hayden's actual script | Capture deposit as CRM stage trigger |
| Quote-aging follow-up | Manual price-rise nudge on quiet days | Phase 02: automated daily report of Quotes >7d / >14d / >30d |
| Order placed | RE manual entry by rep (Shopify auto-flows) | CAPI sends back to Meta — closes the attribution loop |
What Hayden does not need
- A speed-to-lead automation: he already replies same-day.
- A new finance product: Hum-equivalent already in place.
- A "qualification gauntlet" before he sees the lead: he wants signal that helps him spec-match, not gates that block leads.
What this report drives.
- Messenger ad opener with 3 quick-replies (Power / Size / Timing) — ad-account work, instant deployment.
- Lead-system tier auto-classification — driven by the 3 opener answers; flows to rep priority queue.
- Channel routing — website form + phone calls surface above Messenger in the daily inbox.
- Ad-account moves — pause Tier D, scale Tier A. No waiting, no Hayden dependency.
- Quote-automation tool — Hayden's biggest time pain. Replaces the manual PDF email loop.
Two of these compound directly: the opener answers feed the tier classifier feeds the rep queue feeds the quote tool. Each link in that chain is a separate buildable.
How we'll know it worked.
| Validation | Method | Target |
|---|---|---|
| R1 close rate | Track 30-day close rate for opener-classified R1 leads | ≥25% |
| R4 close rate | Same for R4 leads | <0.5% (proves auto-flow doesn't hide buyers) |
| Channel close rate | Compare website-form vs FB-Messenger close at matched horizons | Form ≥80% / Messenger ≥20% |
| Tier A vs D removal impact | Pause D, scale A. Measure overall match rate over 4 weeks | Average match rate 1.4% → ≥2.5% |
| Hayden time-to-first-reply | Ensure tier auto-classification doesn't slow him | No degradation from current ~same-day |
What we still need.
| Gap | Owner | Resolution |
|---|---|---|
| Conversation content review pending | Engineering | Extend Meta MCP to fetch message bodies, OR pull a Page Inbox export. Validates whether language patterns of an R1 lead match the rubric. |
| PSID capture at quote time | Hayden + Entangle | Follow-up call — log FB user ID against customer record. Turns name-matching into deterministic attribution. |
| Property-readiness signal beyond opener | Hayden + Entangle | Mid-conversation "what's your install timeline" tag in lead system. |
| Spot-check 30 random matches per rep | Hayden | Confirms the 36.2% Messenger-attribution figure isn't name-collision noise. Locks the number once >90% confirm. |