V0.2 tenant workspace

SITEMAP.XML operating spine

Org context: sitemap.xml

AI playbooks

Templates and launch playbooks

Approved scripts, response patterns, launch checklists, and reward-review playbooks that keep AI output practical and governed.

Playbooks

6

Approved or guarded playbooks available to operators.

Owner only

1

Playbooks that require owner-only approval.

Guarded

2

Launch or automation playbooks with explicit guardrails.

Matched reviews

6

Current review candidates matching these playbook areas.

Template lifecycle

Version state across approved, restricted, training, and draft playbooks.

Approved versions3
Training/restricted2
Drafts needing approval1

Testing framework

Each playbook defines outcome measures and risk guardrails.

Tracked metrics24
Channels covered15
Forbidden-claim rules22

Variable readiness

Template fields that need database-backed substitution before real sends.

Variables mapped30
Fallback paths6
Queued growth drafts0

Release controls

Operational gates before templates can move from draft to approved usage.

Release steps30
Audit event types18
Data sources referenced24

Approval workbench

What an operator should review before turning a template into a provider-backed action.

Owner

Approver must match the policy owner before the template can move lifecycle state.

Data

Every variable needs a source field, fallback value, and missing-data behavior.

Test

Each release needs a sample window, success metric, and rollback trigger.

Trial enquiry follow-up

Lead nurture · SMS + email

Approved draftv1.4

Convert a warm enquiry into a booked trial without pressure or invented offers.

Lifecycle

approved

Approved

2026-04-28

Template

Thanks for reaching out. The best next step is a trial class so we can place you in the right group and answer questions in person.

Checks

Confirm program and age group
Use real trial availability
No fabricated discounts
Escalate price objections to staff
Variables
leadNameprogramNameageGrouptrialWindowstaffOwner
Forbidden claims
Guaranteeing progressInventing discountsFalse scarcityMedical advice
Tests
Booked trial rateReply rateManual escalation rateOpt-out rate
Release steps
Owner approves copyMap variablesRun 20-lead sampleCompare booked-trial liftLock approved version
Data sources
Lead sourceProgram catalogueTrial availabilityStaff owner
Version history
v1.4 tightened price escalationv1.3 added trial-window variablev1.2 removed urgency language
Audit events
template.approvedlead_nurture.queuedmessage.drafted
Fallback

If the lead asks about price resistance, queue owner call instead of sending a discount.

Primary metric: booked trial within 7 days. Guardrail: opt-out below 2%.

After-hours AI receptionist handoff

AI receptionist · Phone / web

Guardedv0.9

Capture intent and context when staff are offline, then hand the conversation to a person.

Lifecycle

training

Approved

Needs owner sign-off

Template

I can help get the right person to follow up. Please share the member age, preferred class type, and best contact time.

Checks

Do not answer injury or medical advice
Escalate billing and cancellation
Record source channel
Create same-day task
Variables
contactNamememberAgepreferredClassbestContactTimesourceChannel
Forbidden claims
Child pickup permissionsBilling changesCancellation promisesInjury advice
Tests
Captured contact completenessEscalation accuracySame-day callback completionMissed call recovery
Release steps
Owner approves escalation matrixTrain fallback wordingRun transcript reviewMeasure sensitive-topic escape rateEnable limited hours
Data sources
Conversation transcriptChannel sourceContact fieldsEscalation owner
Version history
v0.9 added phone transcript supportv0.8 blocked billing/cancellation repliesv0.7 added age-group prompt
Audit events
ai_receptionist.training_reviewedhandoff.queuedsensitive_topic.escalated
Fallback

If intent or risk is unclear, ask one clarifying question then create a staff task.

Primary metric: complete enquiry captured. Guardrail: zero sensitive-topic auto-responses.

Urgent public review response

Reputation · Google / Facebook

Owner onlyv1.1

Respond quickly without exposing private member, child, billing, or incident information.

Lifecycle

restricted

Approved

2026-04-30

Template

Thanks for raising this. We take it seriously and would like to speak directly so we can understand and resolve it properly.

Checks

No personal details
No argument in public
Owner approval required
Log response outcome
Variables
reviewSourcesentimentownerNamecontactPath
Forbidden claims
Member namesIncident detailBilling dispute detailDefensive tone
Tests
Response timeOwner approval latencyFollow-up resolutionReputation recovery
Release steps
Owner approves tonePrivacy checkDraft responseRecord contact pathMonitor public outcome
Data sources
Review sourceSentimentOwner contact pathResolution status
Version history
v1.1 added child/privacy restrictionv1.0 approved neutral responsev0.9 removed defensive wording
Audit events
review_response.draftedowner.approvedpublic_reply.posted
Fallback

If the review references a child, injury, or dispute, draft a private-contact response only.

Primary metric: owner-approved response within same day. Guardrail: no privacy exposure.

Referral reward approval

Referrals · Internal task

Reviewv1.2

Make sure rewards are earned, correctly applied, and visible to billing.

Lifecycle

approved

Approved

2026-04-26

Template

Referral reward ready for review. Confirm referred member conversion, reward eligibility, and payer impact before applying.

Checks

Check conversion state
Check previous reward
Confirm discount/credit path
Notify advocate after approval
Variables
advocateNamereferredNamerewardTypepayerAccountconversionDate
Forbidden claims
Duplicate rewardsUnapproved account creditPublic recognition without consent
Tests
Reward approval timeDuplicate preventionBilling accuracyAdvocate thank-you sent
Release steps
Verify conversionCheck duplicate rewardPreview billing impactApprove creditNotify advocate
Data sources
Referral stateMembership conversionBilling accountReward ledger
Version history
v1.2 added payer impact checkv1.1 added duplicate preventionv1.0 approved internal task wording
Audit events
reward.reviewedbilling_credit.previewedadvocate.notified
Fallback

If payer impact is unclear, hold reward and queue billing review.

Primary metric: eligible reward applied accurately. Guardrail: no duplicate credit.

Landing page launch

Landing pages · Website

Checklistv0.8

Launch a conversion page only after offer, audience, tracking, and source-of-truth checks pass.

Lifecycle

draft

Approved

Not approved

Template

Landing page ready for approval. Confirm audience, program details, CTA, tracking source, and follow-up owner.

Checks

Program copy matches live timetable
CTA maps to lead source
Analytics tag present
Rollback owner known
Variables
audienceprogramNameofferctatrackingSourcerollbackOwner
Forbidden claims
Outdated timetableUnapproved pricingMissing consent languageUnsupported guarantee
Tests
Page conversionLead source accuracyCTA completionQuality of booked trials
Release steps
Confirm offerConfirm timetableMap CTA sourceSet rollback ownerLaunch limited traffic
Data sources
Program timetableOffer recordLanding sourceLead intake
Version history
v0.8 added rollback ownerv0.7 added analytics sourcev0.6 blocked unapproved pricing
Audit events
page.reviewedpage.publishedlead_source.captured
Fallback

If tracking or pricing is missing, keep page in draft and queue owner review.

Primary metric: qualified lead conversion. Guardrail: all leads retain source context.

Campaign follow-up closeout

Campaigns · Email / SMS

Reviewv1.0

Close campaign follow-up loops so no interested lead or member is left in limbo.

Lifecycle

approved

Approved

2026-04-29

Template

Campaign follow-up review: identify uncontacted responders, conversion-ready leads, and messages that need staff approval.

Checks

Consent retained
Uncontacted responders listed
Owner assigned
Outcome recorded
Variables
campaignNameaudiencerespondersownernextMessage
Forbidden claims
Messaging unsubscribed contactsUnowned follow-upSending after offer expires
Tests
Responder contact rateConversion-ready handoffsUnsubscribe rateFollow-up completion
Release steps
Segment respondersVerify consentAssign ownerQueue approved follow-upClose outcomes
Data sources
Campaign respondersConsent stateLead statusOwner task
Version history
v1.0 approved closeout checklistv0.9 added consent gatev0.8 added uncontacted responder report
Audit events
campaign.followup_reviewedmessage.queuedoutcome.recorded
Fallback

If consent or owner is missing, pause follow-up and create a campaign review task.

Primary metric: responders contacted with outcome recorded. Guardrail: no consent violations.

Playbook use map

How these templates connect back into current AI/growth decisions.

Lead nurture

Front desk lead

0 linked

AI receptionist

Owner

3 linked

Reputation

Owner

3 linked

Referrals

Front desk

0 linked

Landing pages

Growth owner

0 linked

Campaigns

Marketing owner

0 linked

What becomes durable next

The next database-backed layer for this area.

Template versions

Persist approved templates with owner, version, channel, and last approval.

Experiment variants

Attach A/B versions to campaigns, landing pages, and lead nurture paths.

Outcome scoring

Measure booked trials, conversion, reply quality, review outcome, and reward impact.

Queued drafts

0 growth drafts currently available for replay inspection.

Channel matrix

Which channels each playbook is allowed to touch before provider execution exists.

Trial enquiry follow-up

SMSEmailStaff task

After-hours AI receptionist handoff

Web chatPhone transcriptStaff task

Urgent public review response

Google draftFacebook draftOwner task

Referral reward approval

Internal taskBilling noteMember message

Landing page launch

Website pageLanding pageCampaign CTA

Campaign follow-up closeout

EmailSMSStaff taskCampaign report

Measurement plan

How each template proves it is working without creating unsafe automation.

Trial enquiry follow-up

Primary metric: booked trial within 7 days. Guardrail: opt-out below 2%.

After-hours AI receptionist handoff

Primary metric: complete enquiry captured. Guardrail: zero sensitive-topic auto-responses.

Urgent public review response

Primary metric: owner-approved response within same day. Guardrail: no privacy exposure.

Referral reward approval

Primary metric: eligible reward applied accurately. Guardrail: no duplicate credit.

Landing page launch

Primary metric: qualified lead conversion. Guardrail: all leads retain source context.

Campaign follow-up closeout

Primary metric: responders contacted with outcome recorded. Guardrail: no consent violations.

Version registry

Current version, lifecycle, approval state, and latest change notes for each playbook.

Trial enquiry follow-up

Front desk lead · 2026-04-28

v1.4approved

v1.4 tightened price escalation

After-hours AI receptionist handoff

Owner · Needs owner sign-off

v0.9training

v0.9 added phone transcript support

Urgent public review response

Owner · 2026-04-30

v1.1restricted

v1.1 added child/privacy restriction

Referral reward approval

Front desk · 2026-04-26

v1.2approved

v1.2 added payer impact check

Landing page launch

Growth owner · Not approved

v0.8draft

v0.8 added rollback owner

Campaign follow-up closeout

Marketing owner · 2026-04-29

v1.0approved

v1.0 approved closeout checklist

Database shape preview

Tables this should graduate into when we persist templates properly.

ai_playbooks

Template title, owner, area, lifecycle, current version, allowed channels, and policy scope.

ai_playbook_versions

Versioned template body, variables, forbidden claims, release steps, and approval metadata.

ai_playbook_experiments

Variant, sample window, metrics, rollback trigger, and result summary.

ai_playbook_audit_events

Every approval, release, provider handoff, rollback, and outcome update.