REVL Role Plan — How I'd Run It

★The plan on one screen

Section	The one thing to remember
1. B2B (the #1)	Don't automate to keep everyone warm like every other franchise. Track behaviour to triage, so human time goes on the genuinely keen. Build it inside the Franchising Code.
2. B2C	Behaviour-triggered member journeys off Mindbody (onboarding, churn-prevention, the review flywheel) + franchisee enablement through on-brand templates.
3. Deliverability	Win the Primary inbox (the phone notification). REVL has no DMARC + broken SPF, both fixable. The Yard just moved ahead.
4. Tech, AI & Data	A thin Supabase data layer unifies CRM + Mindbody + ads so you see the whole journey. Operationalise Claude across the whole team. API/MCP-first, skip Zapier.
5. Analytics	Build our own dashboards off that database, not tool-native reports. Claude queries it in plain English.
6. Paid	10 years' experience, available as upside. Deliberately not my focus.

My three sharpest assets (what most candidates can't bring)

A live audit of REVL's actual email setup — specific, verifiable, fixable. Not a CV, real work.
A measurable result: one automation made a studio the most-reviewed BFT in Australia.
A real AI system — 230,000+ messages of reps, not a few YouTube videos. This doc is the proof.

1B2B Lifecycle Marketing — the franchise pipeline

This is the #1 part of the role: own the end-to-end franchise-recruitment lifecycle. Here is how I'd build it, and why my version is different from how most franchises run it.

The reframe: this is a qualification system, not an automation system

Every other franchise builds workflows to keep everyone warm. That is the commodity version: push all leads through the same pipe and hope. What I'd build is a system that tells us who is genuinely keen and who isn't, so we spend real human time on the few that will convert, and let cheap automation hold the rest.

How every franchise sees it	My angle
Automation = efficiency	Tracking = triage
Keep everyone warm	Find the genuinely keen
Same pipe for all leads	Human time on hot leads, automation on cold
Email opens / clicks = "engaged"	Landing-page depth = real intent
Goal: more touches	Goal: invest effort where the reward is

Effort follows intent. High-intent leads get tailored time, sales calls, and dialed comms (highest reward, worth the effort). Low-intent leads sit in smart, low-cost workflows that don't burn anyone's time and re-trigger if they heat up. Nobody's dropped, nobody's over-served.

Why the tracking is the qualifier

The deep content is the filter: a landing page with real profit numbers, what a day running a studio looks like, the honest reality of ownership.
Someone genuinely keen will open the email, click through, and consume that. Someone who won't even do that hasn't got the motivation, and that is the signal.
You cannot triage on email data. You can on behavioural data.

The lead A vs lead B problem (the core argument)

Two leads, identical on email	What email sees	What the system sees
Lead A	Opened once, clicked once	Visited once, 30 seconds
Lead B	Opened once, clicked once	Visited 6 times, 30 min accumulated

Email tracks open-once / click-once, so A and B look the same. They are not the same lead and need completely different handling. Site and pixel depth is the only way to tell them apart.

The system, component by component

1. The two-page capture funnel — the profiling engine

Ad / form (Page 1)→ Profiling page (Page 2)→ Behavioural profile→ Matched workflow

Page 1: the ad/form landing (capture).
Page 2 is not a thank-you page. It is a strategically built profiling page — the linchpin of the whole system. Interactive tools and sections deliberately structured to surface why this person filled out the form (freedom? income? owning something? leaving their job?).
It tracks time-on-section, clicks, and scroll depth, so the page itself becomes a data-gathering instrument, not a dead end.

2. The behavioural profile — the user model

Fuse everything: form answers + page-2 behaviour + ongoing site visits + email behaviour + pixel data.
Two axes per lead: what motivates them (freedom vs income vs territory vs lifestyle) and how hot they are (visit frequency × depth).

3. Behaviour-triggered nurture — workflows triggering workflows

Entry workflow is set by the page-2 profile.
Branches on email behaviour. Example: opened "freedom of being your own boss" but ignored "territories available" → route into the freedom track, drop the logistics track.
Re-triggers on return behaviour: comes back, hits certain pages → flips to hot lead → fires a specific email tied to what they've already seen, or pushes them to sales.

4. The sales dashboard — behaviour-driven handoff

Behaviour fires qualified leads to a sales dashboard with context: "viewed the franchise page 6× in 3 days."
Sales ring armed with the profile (what this person cares about, what they've already seen), not a cold "you filled a form."

5. The Facebook / competitive-auction reality

The pixel does two jobs: retargeting REVL's own ads, and exposing the truth that the moment someone is in-market, Meta floods them with franchise ads — REVL's and every competitor's.
So the creative and the nurture have to feel like they understand this specific person's pain, or they get out-resonated in a crowded auction. The profiling isn't a nice-to-have, it's how you win the attention war.

6. The AI angle — and it's the "work smarter with AI" the brief asks for

Old constraint: you pre-built generic sequences because bespoke emails were too slow to make.
Now: design and ship specific, on-brand emails fast through Claude → far more dialed, per-segment and even per-person, without the old labour cost.
That's a concrete answer to "help us work smarter through AI." Not a buzzword, an actual capability shift.

The guardrails that keep it credible

It still lives inside the Franchising Code

The dialed, personalised emails are exactly where compliance bites. Any financial figure in a nurture email must match the formal disclosure document. The information statement still has to land within 7 days of enquiry, before any other document, regardless of behaviour. There's a 14-day disclosure window before any signing or payment, plus a separate 14-day cooling-off after. The clever system runs inside fixed legal gates.

The tracking has a consent layer

Pixel and behavioural profiling needs the consent basics handled (Privacy Act), and the email side needs consent + unsubscribe (Spam Act). Not a blocker, but it's built in from the start, not bolted on.

Connects to the tech stack

The database and behavioural-profiling infrastructure is part of the tech-stack conversation (Section 3). HubSpot is fine — it has the automations and API to link Claude into. The system is stack-agnostic enough that "what they already have" isn't a blocker, it's a starting point.

2B2C Email & CRM Support

Two jobs sit inside this: the member lifecycle (journeys triggered off Mindbody behaviour) and franchisee enablement (templates + support so ~50 studios send on-brand without me touching every email). Plus a win I can already prove.

The proven win to lead with

They told me they couldn't get REVL's brand image out of Claude. I did, with no trouble. On top of that I rebuilt one of their emails to render properly in dark mode and sent it live. So "a genuine eye for brand" and "work smarter with AI" aren't claims here, they're already demonstrated.

2.1 The member lifecycle — journeys off Mindbody behaviour

The shift is the same engine as the B2B triage idea, pointed at members: stop sending campaign blasts to everyone, start running journeys triggered by what people actually do in-studio (trials, first visit, attendance, package expiry). The Mindbody API + a developer account is what unlocks it.

Journey	Trigger	Why it matters
Onboarding	New member joins	First 30/60/90 days is the highest-churn window. Get it right or lose them early.
Trial → member	Trial booked / attended	Convert while motivation peaks, not a week later.
Review flywheel	~30 days in, going well	"How's it going? If you've got a sec, leave us a review." Google ranks businesses with consistent, recent reviews (it's how they know you're active). Reviews compound into local SEO, member acquisition, and franchise credibility.
Referrals	Happy-member moments	Cheapest acquisition there is. Triggered off positive signal, not blasted.
Churn prevention	Attendance drops off	Re-engage before they cancel. REVL's sub-5% attrition claim is a marketing asset worth protecting, not just an ops stat.
Win-back	Lapsed / cancelled	One good automated attempt, then suppress so the list stays clean.

The cancellation / suspension save flow (already built)

Not a one-click cancel. An assessment-style form that asks why someone's cancelling or suspending, then serves a tailored offer or a check-in based on their answers (pause instead of quit, downgrade, a hold). It turns the exit moment into a save moment. This is a flow I've built before, so it's a working asset, not a concept.

Proof point: I made a studio the most-reviewed BFT in Australia

One automated message did it. A casual, personal check-in at ~30 days, not a corporate ask. The gist: "you've been with us a month, how have you found it, anything we can do better, and if you've got a sec a quick review would genuinely help us." That single automation drove a studio to the most reviews of any BFT in the country.

Drafted as a clean template (casual, on-brand, no fluff):

Hey [First name], you've hit your first month with us. Massive effort. Just wanted to check in: how have you found it so far? If there's anything we could do to make it better, hit reply and let me know. And if you've got a spare 30 seconds, a quick Google review would genuinely mean a lot. They really do help us out. [review link]
Cheers, [Name]

Why it's strong for REVL: it's a real, measurable B2C lifecycle result, it feeds the review flywheel (Google ranks active, consistently-reviewed businesses), and the proof points at member acquisition and franchise credibility. Decide in the room whether to name the specific studio/client or keep it anonymous.

Why it ranks you: the local 3-pack

Google's local map pack shows three businesses, and that's where the clicks go. The studios that win those slots have the highest review counts, the highest ratings, and get reviews consistently and recently (recency tells Google you're active and open).

REVL Greenslopes example: search that area and REVL Greenslopes is missing from the top three. The studios that are showing have fewer reviews and aren't as consistent, so REVL is losing visibility it could easily win. A simple 30-day review automation closes exactly that gap. (Eyeball check, so reconfirm the live ranking before quoting it in the room.)

2.2 Franchisee enablement — templates + support

HQ builds on-brand, conversion-tested templates in the CRM. Managers send and schedule them locally, inside guardrails. Local flavour, central control of brand + deliverability.
A real support system around it: feature webinars, written docs, and a dedicated help channel (Slack) so a manager can reach whoever owns the CRM, fast, with an automation issue or a question.
This is the JD's "tools, templates, and technical support" bullet word for word. My Knowsi angle fits exactly: build tools simple enough that a non-marketer studio manager actually uses them.

2.3 The sending architecture (the fork to have a POV on)

Unknown until I ask: do B2C campaigns send from one master account, per-location sub-accounts, or templates pushed to managers? My POV either way:

However it works today, HQ should own brand, deliverability, and compliance centrally, and franchisees should localise on-brand templates inside guardrails. You never let 50 studios send raw, uncontrolled email. That's a brand-drift and sender-reputation disaster (see Section 3).

2.4 Questions to confirm with her (genuine unknowns)

What CRM, and how's it split? Is there a master account that provisions sub-accounts to franchisees?
B2C campaigns: sent from one account, per-location sub-accounts, or templates managers send themselves?
What automations have they run? Just Facebook leads in, or journeys off in-studio behaviour (purchases, trials, bookings)?
How deep do they want to go on automation?
Are they on-selling a CRM to franchisees (GHL? HubSpot?), or running it all centrally?
What level of franchisee support: webinars, an open help channel, ongoing feature support?

3Deliverability & DNS — getting into the Primary inbox

This sits under both B2B and B2C. It's the most concrete, verifiable thing I bring, because I checked REVL's live setup and it's fixable.

The whole game in one sentence: a Primary-inbox email triggers a notification on someone's phone. Promotions gets no notification. Spam is invisible. That phone notification is what we're fighting for, and it's decided by things most senders never touch.

3.1 Why their image-only emails hurt them

REVL sends image-heavy emails. Inbox filters can't read what's inside an image, so they can't tell what the email is, which pushes it toward Promotions or Spam.
Real HTML text is machine-readable: filters classify it correctly, and it can actually be tested and optimised before send.
To find out from her: what are their open and click rates, and are low opens really a content problem or a placement problem (landing in Promotions/Spam)?

3.2 My pre-send process

Design the email, then run it through deliverability scoring (subject line and body), fix the spam triggers it flags, and confirm Primary placement before it goes out.
To find out from her: do they use any service that scores deliverability before send? Do they test different subjects and messaging?

The principle that sounds senior

Authentication (SPF/DKIM/DMARC) decides inbox vs spam. Content and engagement decide Primary vs Promotions. Two different levers, two different fixes.

3.3 The three records, explained plainly

Record	Stands for	What it does
SPF	Sender Policy Framework	A public list of which servers are allowed to send email for your domain. Stops randoms sending as you.
DKIM	DomainKeys Identified Mail	A cryptographic signature on each email proving it really came from you and wasn't tampered with in transit.
DMARC	Domain-based Message Authentication, Reporting & Conformance	The policy that tells inboxes what to do when SPF/DKIM fail (`none` = let it through, `quarantine` = spam folder, `reject` = block), and emails you reports on who's sending as you.

3.4 What REVL has and hasn't (checked live, 22 June 2026)

Record	Status	So what
SPF on `revl.com.au`	Broken	Split into two records. The HubSpot include sits in its own orphaned TXT with no `v=spf1` prefix, so HubSpot mail isn't SPF-authorised. Re-confirmed live today.
DMARC on `revl.com.au`	Missing	No record at all. No spoofing protection, and it fails the 2024 Google/Yahoo bulk-sender rules. Re-confirmed live today. The top fix.
DKIM	Confirm	Selector-specific, so confirm per sending platform once inside. Each ESP (HubSpot, SendGrid, ACR) needs its own DKIM set and aligned.
Campaign sender	ACR / acrmailer	Live campaigns send via ACR (the "acr.fit" domain), which also has no DMARC, so campaign mail has zero alignment.

Bulk-sender rules (5,000+/day to Gmail): SPF or DKIM and a DMARC record, complaint rate under 0.3% (target under 0.1%), one-click unsubscribe on every send. REVL currently fails the DMARC requirement.

3.5 The category — live check, today

Brand	DMARC policy	Read
REVL	none	Exposed. Config-only fix away from leading the pack.
BFT	none	Also exposed.
Fitstop	p=none	Published but not enforcing.
The Yard	p=quarantine	Just moved up (was p=none in June). A competitor is now actively enforcing.
Chemist Warehouse, THE ICONIC	p=reject	Where mature senders sit. The destination.

The angle

The fitness category was asleep on this. The Yard just woke up. REVL and BFT are still on nothing, and REVL can leapfrog both with a config change. Extra: a DMARC policy at quarantine or reject is the gate for BIMI (your verified logo showing in the inbox). The Yard can pursue that now. REVL can't until DMARC is fixed.

3.6 The tooling stack

Job	Tool	Cost
Score an email + predict Primary vs Promotions before send	GlockApps (Inbox Insight, rates 1-100 across 70+ seed inboxes, checks SPF/DKIM/DMARC + blacklists)	~$59/mo
Fast free spam-score + fix list per email	Mail-Tester (does not predict the Gmail tab)	free
Track engagement depth (read-time, client, device, dark mode) from the HTML, any ESP	Litmus Email Analytics	~$99-199/mo
True conversions (the metric that matters)	UTM tags → GA4 + Mindbody	free
Ongoing reputation	Google Postmaster Tools + DMARC reporting (dmarcian / EasyDMARC)	free

Principle: auth gets you to the inbox, content and engagement decide the tab, and post-Apple-MPP opens are unreliable, so measure clicks, read-time and conversions.

4Tech Stack, AI & Data

The JD asks to own the stack, own integrations, and use AI to scale. My take goes a layer deeper: the real opportunity isn't squeezing more out of the tools they already have, it's building a thin data + AI layer on top so the team finally understands what's actually happening, and so anyone can produce specialist-quality work.

4.1 The core idea: own your data layer

Every tool in the stack is good at its own job and blind to the rest. The CRM knows comms, Mindbody knows bookings, the ad platforms know clicks, the landing pages know visits. Nothing sees the whole journey, so you can't actually answer why and how people convert.

The fix: a central database (Supabase) that reads from everything via API. The CRM and Mindbody keep doing their jobs. The database is a unifying insight layer on top, not a replacement.
What it unlocks: full-journey tracking (lead source, which ad, what they engaged with, how they converted), one customer profile any team can call, and custom insights beyond what any single tool's dashboard can show.
Why it's safe, not key-person risk: it's documented, version-controlled in Git, and the team is trained on it (see 4.4). The build and the enablement are the same project.

The example that proves the whole idea

A franchise lead lands in the CRM. Without a unified database, you see a name and an email. With one, you might see they've done 100 sessions at a studio, that their favourite class is Sweat, that they're already a power-member. That completely changes how sales handles them, and you'd never know it otherwise. One database, and a B2B lead and a B2C member become the same person.

The grown-up caveat to say out loud

Owning the database means owning responsibility for securing the member and lead data in it (Privacy Act). Built secure and access-controlled from day one, not bolted on later.

4.2 Integrations: API / MCP-first, skip Zapier

Most platforms now expose APIs, and increasingly MCPs, so we connect to them directly.
Zapier is expensive and priced per-task, so not worth it at any real volume. Cheaper alternative is Pabbly (more webhooks, far cheaper), but with APIs and MCPs available we often don't need middleware at all. Web-scraping is the last-resort fallback if a platform is closed.
The position: lean, modern, direct integration, not paying a monthly toll on every automation.

4.3 Landing pages as data instruments

Built conversion-optimised and with tracking baked in (source, ad, on-page behaviour) — the same engine as the B2B profiling funnel in Section 1.
Tracking links + custom-domain URLs mean an email click lands on a specific page and the whole path is captured, feeding the customer profile.
That profile then triggers things: someone who keeps returning to a promo page gets flagged or fired into an automation.
Webflow note: I build landing pages in code today. The JD says Webflow "or similar CMS" — same thinking, I'd pick the tool up fast.

4.4 AI, operationalised across the whole team

This is the real version of "help us work smarter through AI." Not me using AI well in a corner. The whole marketing team levelled up.

A business Claude setup where everyone has a pre-built shared folder, is trained on Claude Code, and knows enough Git/GitHub to pull and stay in sync.
My role: maintain those folders so the messaging, branding, SOPs and skills are always current and correct.
The goal: cut workload across the team so people spend their time planning and executing, not grinding.

Everyone has the same Claude model and gets completely different results. That's not prompting. It's file architecture and engineering.

A pre-built skill system and shared file structure encode the criteria for the work, so the system knows what a quality landing page actually requires.
Which means anyone on the marketing team can build an LP to the same standard as the specialist, because they're working off the same structure. Capability stops living in one person's head.
Hooks strip the AI tells so nothing ships sounding like a robot. Commands keep projects and chats clean.

Why the architecture is the moat

AI has a ceiling, and it's set by whoever builds the machine, not by the model. As models get smarter, great, but a smarter model won't fix bad work, bad filing, or bad architecture. Knowing how to set it up is the skill that compounds.

My credibility on this is reps, not theory: nine months deep in it, 230,000+ messages through Claude, a system tuned by studying the right people and adapting it over and over. It's not something you learn from a few YouTube videos. You learn it by building it, breaking it, and fixing it enough times to know why it works.

5Analytics & Reporting

This is the payoff of the data layer in Section 4, so I'd reframe the brief here. It's not "track performance across CRM channels" using whatever dashboards the tools give you. It's build our own.

The problem with tool-native dashboards: people arrive through the CRM, through Mindbody, through ads, through landing pages. Trying to understand performance across all those silos is possible but painful, and you only ever see what each tool chooses to show you.
The model: one central database that feeds off everything, then a dashboard built for each team, department, or person, showing exactly what they need to make good calls on performance, conversions and profitability.
Claude queries it directly. Anyone can ask the database a question in plain English and get the answer, rather than waiting on a report or learning a BI tool.

The line: "I wouldn't position this as reporting off CRM dashboards. I'd build our own database that everything feeds into, then surface team-specific dashboards from it. That's the difference between guessing from fragments and actually understanding the full picture."

Measurement principles still apply on top: post-Apple-MPP, opens are unreliable, so lead on clicks, conversions and (for members) Mindbody activity as the real signal of "engaged".

6Paid Advertising

The JD lists this as preferred, not essential, and I'd treat it honestly.

I've run paid for around 10 years (Meta, Google), so it's covered if it's useful to the role.
But I'm deliberately not pitching it as my focus. My strength and my preference is the CRM, lifecycle, automation and AI side. That's where I do my best work and where this role's biggest gaps are.
Happy for paid to be upside, not the centre of the job.

It also connects back: the landing pages, tracking and data layer in Section 4 make whatever paid does run far more measurable, because we'd actually see what each ad produces all the way through to a join.