System: MWMS
Document Type: Operating Framework
Authority Level: MCR Source Of Truth
Status: Active
Version: v1.0
Primary Location: MCR
Future Operational Destination: Sales Brain, Content Brain, AIBS Brain, Ads Brain, Automation Brain, Data Brain, Compliance Brain, Risk Brain, Customer Brain, Product Brain, HeadOffice Brain
Parent Page: Sales Brain
Owner: Martyn
Developer Boundary: Do Not Touch M’s Active Build Areas Unless Specifically Assigned
Source Of Truth: MCR
Last Reviewed: 2026-06-21
Source / Origin: AI Automations by Jack material covering Nano Banana image generation, personalised prospect images, logo-integrated visuals, website-derived lead magnets, personalised likeness models, synthetic voice, website walkthrough videos, prospect-specific scripts, candidate approval, asset storage, and outreach delivery
MWMS Classification: Sales Brain Operating Framework / Personalised Visual Sales Asset Standard / Synthetic Media Governance Framework / Prospect-Specific Asset Production System / Visual Outreach Approval Framework
Primary Brain: Sales Brain
Supporting Brains: Content Brain, AIBS Brain, Ads Brain, Automation Brain, Data Brain, Compliance Brain, Risk Brain, Customer Brain, Product Brain, HeadOffice Brain
Related Pages: Sales Brain Canon, MWMS AI Assisted Outreach And Sales Follow Up Automation Framework, MWMS Outbound Lead Enrichment And Cold Outreach Governance Framework, MWMS AIOS Lead Capture And Conversion Infrastructure Framework, MWMS Market Driven Social Content Production Framework, MWMS Prompt Architecture And Automation Output Reliability Framework, MWMS Client Communication Automation Framework, MWMS AI Tool Permission And Access Framework, MWMS AI Automation Security And Risk Checklist, MWMS Client Context Isolation And Privacy Boundary Standard, MWMS Source Visibility And Evidence Display Standard, MWMS AI Output Validation Standard
Purpose
The purpose of the MWMS Personalised Visual Sales Asset Production And Governance Framework is to define how MWMS creates, reviews, approves, stores, delivers, measures, and governs personalised visual and audiovisual assets used in sales, outreach, lead generation, client acquisition, follow-up, and commercial communication.
This framework covers assets such as:
personalised prospect images
company-specific graphics
logo-integrated visuals
custom comic-style lead magnets
website-derived visual assets
prospect-specific thumbnails
personalised presentation covers
AI-generated likeness images
synthetic spokesperson images
personalised short videos
website walkthrough videos
AI-generated voice messages
voice-note style assets
prospect-specific social visuals
diagnostic screenshots
visual proposal previews
sales enablement graphics
This framework exists because personalised visual assets can increase:
attention
relevance
response
memorability
perceived effort
sales conversation quality
demonstration value
However, these assets also create material risks involving:
brand misuse
logo misuse
false endorsement
identity confusion
synthetic likeness
voice cloning
privacy
copyright
deception
wrong-client assets
unsupported personalisation
inaccurate observations
automated sending
reputational harm
The purpose of this framework is therefore not simply to help MWMS create more eye-catching assets.
It is to ensure every personalised sales asset is:
relevant
source-backed
accurate
permission-aware
identity-safe
brand-safe
reviewed
approved
traceable
appropriate for the selected channel
measured against a defined commercial objective
Core Doctrine
The MWMS doctrine is:
Personalisation is not permission.
Generation is not approval.
Visual realism is not truth.
A personalised asset must not imply a relationship, endorsement, action, identity, or human effort that did not occur.
The complete personalised visual asset lifecycle is:
Define Commercial Objective
→ Identify Prospect Or Recipient
→ Confirm Source And Identity
→ Select Asset Type
→ Define Personalisation Evidence
→ Check Logo, Brand, Likeness, Voice, And Source Permissions
→ Create Prompt And Template Contract
→ Generate Candidate Assets
→ Validate Accuracy And Identity
→ Review Brand, Authenticity, And Disclosure Risk
→ Select Candidate
→ Obtain Human Approval
→ Approve Channel And Recipient
→ Deliver Or Attach Through Approved Workflow
→ Confirm Delivery
→ Track Response And Outcome
→ Retain, Refresh, Reject, Or Delete Asset
→ Improve Through Controlled Testing
An asset workflow is incomplete if it only produces an image or video.
It must also govern:
who the asset represents
which company it references
what evidence supports the personalisation
whether logo or brand use is appropriate
whether a person’s likeness or voice is authorised
whether the asset may be sent
whether disclosure is required
what version was delivered
what commercial outcome occurred
Scope
This framework applies to:
cold outreach visuals
warm outreach visuals
sales follow-up visuals
lead magnets
prospecting value assets
diagnostic assets
personalised presentation assets
proposal-support visuals
client demonstration assets
social outreach assets
personalised email images
personalised LinkedIn assets
personalised WhatsApp assets where authorised
personalised video outreach
synthetic voice outreach
AI likeness models
AI-generated human imagery
company-specific visual examples
future AIBS visual personalisation products
This framework does not authorise unrestricted scraping, unauthorised logo use, deceptive synthetic media, voice cloning without permission, automatic external sending, or mass personalisation without compliance review.
Core Definition
A personalised visual sales asset is a visual or audiovisual communication asset modified or generated using recipient-specific information.
That information may include:
company name
company logo
website appearance
public business information
verified business event
recipient name
approved relationship context
known sales need
approved client data
MWMS Definition
An MWMS Personalised Visual Sales Asset is:
A source-linked, recipient-specific, permission-aware, human-reviewed visual or audiovisual commercial asset created to provide relevance, value, explanation, demonstration, or sales support without misrepresenting identity, endorsement, relationship, human effort, or business facts.
Strategic Principle
The objective is not to create impressive AI media.
The objective is to improve the quality of a commercial conversation.
Personalised visual assets should support:
relevance
value
clarity
demonstration
diagnosis
trust
decision-making
They should not rely solely on novelty.
The MWMS standard is:
Value Before Novelty
Evidence Before Personalisation
Permission Before Likeness
Accuracy Before Realism
Approval Before Delivery
Outcome Before Scale
The Personalised Visual Asset Operating Model
Every personalised visual sales asset system should operate across sixteen layers:
Commercial Objective Layer
Recipient And Identity Layer
Source And Evidence Layer
Asset Classification Layer
Brand And Logo Permission Layer
Likeness And Voice Permission Layer
Prompt And Template Contract Layer
Generation Layer
Candidate Selection Layer
Accuracy And Quality Review Layer
Authenticity And Disclosure Layer
Approval And Delivery Authority Layer
Storage Retention And Deletion Layer
Performance And Experimentation Layer
Failure Recovery Layer
Governance And Continuous Improvement Layer
- Commercial Objective Layer
Every asset must have a defined commercial purpose.
Possible purposes include:
start a sales conversation
demonstrate a capability
provide a useful observation
support a proposal
explain an opportunity
increase message relevance
create a memorable follow-up
show a possible future state
support a diagnostic
improve lead-magnet engagement
Commercial Objective Record
Campaign:
Prospect:
Offer:
Asset Purpose:
Expected Action:
Channel:
Success Metric:
Owner:
Rule
An asset should not be created merely because the technology can produce it.
- Recipient And Identity Layer
The system must identify:
prospect
company
contact
recipient
relationship
campaign
correct website
correct logo
correct email or channel
Identity status may be:
Verified
Probable
Unverified
Conflicting
Duplicate
Rule
A personalised asset must not be generated for external use where identity is materially uncertain.
Recipient Identity Record
Prospect ID:
Company:
Contact:
Role:
Website:
Domain:
Email:
Channel:
Relationship:
Identity Confidence:
Verification Source:
- Source And Evidence Layer
Every personalised element should have a source.
Possible source types include:
official company website
official logo source
approved CRM record
client-provided file
verified public announcement
approved sales conversation
current diagnostic
current proposal
approved form submission
Source Evidence Record
Personalisation Element:
Source:
Source URL Or Record:
Capture Date:
Evidence Type:
Confidence:
Current:
Safe To Use:
Human Review Required:
Rule
The system must not invent company facts, awards, business problems, or interests to make the asset feel personalised.
Evidence Classification
Verified Fact
Public Observation
Client-Provided Fact
Derived Signal
AI Inference
Unverified Claim
Rule
AI inference must not be presented as confirmed fact.
- Asset Classification Layer
MWMS recognises four primary personalised asset classes.
Prospecting Value Asset
An unsolicited but relevant asset intended to begin a commercial conversation.
Examples:
company-specific image
short diagnostic visual
custom comic
personalised video
Traditional Lead Magnet
An asset requested or accessed through a value exchange.
Examples:
guide
report
calculator
checklist
assessment
Diagnostic Asset
An asset demonstrating an observed problem, opportunity, gap, or possible improvement.
Examples:
website review
funnel visual
conversion observation
content gap example
Sales Enablement Asset
An asset supporting an existing sales conversation.
Examples:
proposal visual
solution mock-up
custom workflow illustration
implementation preview
Rule
The asset class must be recorded because permission, delivery, claims, and follow-up rules differ.
- Brand And Logo Permission Layer
A logo or brand element may be used only where its use is:
relevant
accurate
non-deceptive
appropriately sourced
limited to the commercial context
unlikely to imply false endorsement
Brand elements may include:
logo
business name
brand colours
website screenshot
product image
public-facing visual identity
The system must not:
modify a logo in a damaging way
place a logo in offensive or misleading content
imply the company created the asset
imply the company endorsed MWMS
present a fake partnership
use the wrong logo
use an outdated brand identity without review
Logo Use Status
Approved
Conditionally Approved
Review Required
Restricted
Rejected
Rule
A publicly visible logo is not automatically unrestricted creative material.
- Likeness And Voice Permission Layer
Human likeness and voice require stronger controls.
Possible likeness subjects include:
MWMS owner
MWMS employee
client representative
prospect
public figure
synthetic fictional person
Likeness use must confirm:
identity owner
consent
permitted use
permitted channels
permitted duration
training data source
storage
retention
revocation
deletion
Voice use must confirm:
voice owner
recording source
consent
commercial-use permission
permitted scripts
disclosure requirements
revocation process
Rule
No voice clone or likeness model should be created or used without explicit authority from the person represented.
Synthetic Likeness Rule
A generated image must not imply that a real person:
held a sign
wore branded clothing
visited a location
endorsed a company
recorded a message
performed an action
unless that implication is authorised and not misleading.
- Prompt And Template Contract Layer
Every repeatable asset should use a defined visual prompt or media contract.
Visual Prompt Contract
Asset Type:
Scene:
Subject:
Recipient Or Brand:
Permitted Logo:
Required Text:
Composition:
Style:
Aspect Ratio:
Resolution:
Reference Asset:
Controlled Variables:
Fixed Elements:
Prohibited Elements:
Likeness Permission:
Voice Permission:
Disclosure Requirement:
Output Count:
Quality Criteria:
Prompt Version:
Template Version:
Rule
The prompt must separate fixed structure from recipient-specific variables.
Visual Variable Types
Company Name
Logo
Recipient Name
Business Type
Website Observation
Offer
CTA
Background
Text
Colour Palette
Industry Context
Rule
Only approved variables should be dynamically populated.
Reference Asset Rule
Reference images may be analysed for:
composition
layout
contrast
visual hierarchy
camera angle
text placement
subject position
style category
They should not be copied so closely that the output becomes a deceptive replica or infringement risk.
- Generation Layer
Generation may use:
text-to-image
image-to-image
reference-guided image generation
video rendering
voice synthesis
website capture
image compositing
template rendering
The generation workflow should record:
model
provider
prompt version
template version
inputs
reference assets
generation date
cost
output count
status
Generation Status
Queued
Running
Completed
Failed
Partial
Blocked
Rule
A successful API response means only that generation completed.
It does not mean the asset is accurate, permitted, approved, or ready to send.
- Candidate Selection Layer
Where possible, generate multiple candidates for review.
Candidate review should consider:
identity accuracy
logo accuracy
text accuracy
visual quality
brand fit
recipient relevance
professional suitability
authenticity risk
channel suitability
Candidate Status
Unreviewed
Shortlisted
Selected
Rejected
Regenerate
Blocked
Rule
The system should not automatically choose a candidate solely because it rendered successfully.
Selected Asset Record
Asset ID:
Candidate ID:
Selected By:
Selection Reason:
Rejected Candidates:
Required Corrections:
- Accuracy And Quality Review Layer
Every external asset should be checked for:
correct company
correct person
correct logo
correct spelling
correct website
correct business description
correct public observation
correct text
correct offer
correct CTA
visual defects
distorted faces
incorrect hands or objects
unreadable text
wrong background
wrong brand colours
unwanted logos
unintended implications
Quality Status
Pass
Pass With Conditions
Needs Revision
Reject
Rule
Visual quality does not override factual accuracy.
Text Accuracy Rule
Any generated text inside an image or video must be checked manually before use.
Logo Accuracy Rule
A distorted, altered, incomplete, or incorrect logo must not be used externally.
- Authenticity And Disclosure Layer
The system should determine whether disclosure is required.
Possible disclosures include:
AI-generated image
AI-generated video
synthetic voice
illustrative mock-up
concept visual
demonstration only
Disclosure may be required where:
the asset could reasonably be mistaken for a real photograph or recording
a real person’s likeness is used
a synthetic voice sounds like a real person
the asset implies a real event
law, platform policy, client policy, or ethical standards require it
Rule
MWMS must not intentionally create a false belief that a human personally recorded, photographed, or created something when that did not occur.
Human-Effort Honesty Rule
A message must not say:
“I made this personally for you”
“I recorded this for you”
“I took this photo”
unless that is true.
Permitted truthful alternatives may include:
“We created this concept for your business.”
“We generated this example using your public website information.”
“This is an illustrative mock-up.”
- Approval And Delivery Authority Layer
Approval must be separate from generation.
Possible authority levels include:
Generation Only
Internal Review
Approved For Internal Demonstration
Approved For Client Review
Approved For Named Recipient
Approved For Campaign
Rejected
Restricted
Before external delivery, confirm:
recipient
channel
asset class
source evidence
brand permission
likeness permission
voice permission
disclosure
message context
commercial objective
approval
suppression status
Rule
Asset approval does not automatically create message-send authority.
Delivery Channels
SMS link
Client Portal
Proposal
Presentation
Social Post
Landing Page
Delivery should comply with the relevant communication and outreach framework.
- Storage Retention And Deletion Layer
Generated assets may contain:
personal data
client data
prospect identity
logos
website screenshots
likeness data
voice data
training images
private campaign information
Storage records should define:
asset location
owner
client
campaign
access
sensitivity
retention period
deletion date
training-data status
model status
revocation status
Likeness Training Data Rule
Training images should be:
authorised
minimised
securely stored
access-controlled
deletable
not reused for unrelated purposes
Voice Training Data Rule
Voice samples should be governed as sensitive identity assets.
Rule
Generated likeness and voice systems require a deletion and revocation process.
- Performance And Experimentation Layer
Personalised assets should be measured.
Possible metrics include:
delivery rate
view rate
click rate
reply rate
positive reply rate
qualified conversation rate
booking rate
conversion rate
complaint rate
unsubscribe rate
asset approval rate
regeneration rate
cost per asset
cost per reply
cost per qualified conversation
cost per client
Test variables may include:
asset versus no asset
image versus plain text
video versus image
diagnostic versus novelty
short versus detailed
logo use versus no logo
personalised observation versus generic message
Rule
A personalised asset should not be declared effective from isolated examples.
Control Group Standard
Where practical, compare against a suitable control.
Scaling Rule
Do not scale until:
accuracy is acceptable
approval workload is manageable
complaints are low
positive reply quality is useful
conversion value exceeds production cost
brand and identity risk are controlled
- Failure Recovery Layer
Possible failures include:
wrong client
wrong logo
wrong person
bad likeness
bad voice
unreadable text
generation timeout
provider failure
asset-storage failure
broken link
wrong attachment
duplicate delivery
rejected candidate
permission uncertainty
disclosure uncertainty
Recovery actions may include:
regenerate
replace source
correct prompt
remove likeness
remove logo
switch to non-personalised asset
route to human review
block delivery
delete asset
Rule
Failure recovery must not bypass approval.
Retry Standard
Retries should define:
failure reason
maximum attempts
what changes
cost limit
duplicate-prevention rule
human-review threshold
- Governance And Continuous Improvement Layer
The system should learn from:
human rejections
accuracy corrections
response outcomes
complaints
brand concerns
likeness concerns
channel performance
asset cost
conversion quality
Learning may improve:
prompt templates
asset types
review criteria
personalisation evidence
channel choice
generation models
approval rules
disclosure language
Learning must not automatically:
expand likeness use
weaken permission
remove disclosure
increase campaign volume
reuse training data
change brand policy
Rule
Changes must be reviewed, versioned, tested, and approved.
Personalised Visual Asset Record
Asset ID:
Campaign ID:
Prospect ID:
Client ID:
Company:
Contact:
Asset Class:
Asset Type:
Commercial Objective:
Offer:
Channel:
Source Website:
Personalisation Evidence:
Evidence Status:
Logo Source:
Logo Permission Status:
Likeness Owner:
Likeness Permission Status:
Voice Owner:
Voice Permission Status:
Reference Asset:
Prompt Version:
Template Version:
Model:
Provider:
Generated Date:
Generation Cost:
Candidate Count:
Selected Candidate:
Accuracy Status:
Quality Status:
Disclosure Requirement:
Approval Status:
Approved By:
Approved Recipient:
Delivery Status:
Delivery Date:
Outcome:
Retention Status:
Deletion Date:
Last Updated:
Asset Approval Checklist
Before approval, confirm:
commercial purpose is defined
recipient is verified
company is verified
website is correct
personalisation evidence exists
facts are current
logo is correct
logo use is appropriate
likeness permission exists where required
voice permission exists where required
generated text is accurate
brand name is spelled correctly
visual does not imply false endorsement
visual does not imply a false event
asset does not contain another client’s data
CTA is correct
channel is appropriate
disclosure is defined
suppression is checked
human reviewer is recorded
delivery authority is separate
Rule
An attractive asset that fails any material identity, permission, or accuracy check is not approved.
Personalised Image Standard
A personalised image should define:
subject
brand element
message
recipient relevance
background
style
aspect ratio
channel
approval
Images must not:
misrepresent a real event
depict a person in an offensive or compromising way
use an unauthorised likeness
alter a logo misleadingly
contain false claims
imply endorsement
Rule
The image should support the message rather than replace the offer.
Personalised Video Standard
A personalised video should define:
recipient
business observation
script
visual source
website capture
voice
speaker identity
duration
CTA
disclosure
approval
Video scripts should be:
brief
specific
accurate
useful
human-quality
non-manipulative
Videos must not:
claim a full audit occurred where only public-page analysis occurred
claim the sender personally recorded the video where synthetic production was used
use a cloned voice without authority
include unverified claims
show confidential information
Rule
A personalised video is a sales asset, not proof that a human performed bespoke consulting work.
Synthetic Voice Standard
Synthetic voice may be used only where:
the voice owner approved it
the script is approved
the use case is approved
the channel is approved
the recording is not deceptive
disclosure is applied where required
The system must record:
voice owner
consent date
approved purpose
approved channels
revocation status
Rule
Voice cloning is an identity-sensitive capability.
It must not be treated as an ordinary formatting tool.
Website Capture Standard
Website captures may support:
diagnostics
personalised videos
proposal visuals
demonstrations
Website captures should record:
URL
capture date
page
viewport
source owner
public or authenticated status
Capture must not expose:
private account data
personal customer data
internal dashboards
restricted content
Rule
Public website capture does not authorise unrelated reuse or misleading alteration.
Personalised Lead Magnet Standard
A personalised lead magnet should be classified correctly.
Traditional Lead Magnet
Requested by the user through a value exchange.
Prospecting Value Asset
Sent to begin a conversation.
Diagnostic Asset
Shows a relevant finding or opportunity.
Sales Enablement Asset
Supports an existing sales process.
The system should not call every custom image a lead magnet.
Rule
The asset type must match the real acquisition mechanism.
Personalised Asset Message Standard
The message accompanying the asset should explain:
why it is relevant
what the asset represents
what is illustrative
what useful point it demonstrates
what next step is proposed
It should not rely on:
shock
confusion
fake familiarity
false compliments
false urgency
creepy personal data
Rule
The recipient should understand the value of the asset without being misled about how it was created.
Multi-Channel Asset Standard
The same asset may be considered for multiple channels.
Before reuse, confirm:
format
size
recipient
permission
message context
platform rules
frequency
previous delivery
suppression
Rule
An asset created once must not be sprayed across every available channel.
Reference Pattern Extraction Standard
MWMS may analyse successful visual examples to identify:
hook structure
layout
contrast
subject placement
text hierarchy
emotion
camera angle
background
brand integration
The resulting asset must remain sufficiently original.
Rule
Extract the pattern.
Do not clone the creative identity.
Failure Modes
Failure Mode 1: Public Logo Treated As Unrestricted
The workflow uses any logo found online without review.
Correction
Record the source, context, and permission status.
Failure Mode 2: Wrong Logo
An asset uses another company’s brand.
Correction
Validate domain, company, and logo before generation.
Failure Mode 3: False Endorsement
The asset implies the prospect supports MWMS.
Correction
Use clear commercial context and avoid partnership-style presentation.
Failure Mode 4: Synthetic Image Presented As Real
A generated likeness is treated as a genuine photograph.
Correction
Apply authenticity review and disclosure where required.
Failure Mode 5: Voice Clone Without Permission
A person’s voice is reproduced without explicit authority.
Correction
Block creation and use until permission exists.
Failure Mode 6: Personalisation Evidence Is Invented
The asset references an award, interest, or problem that was not verified.
Correction
Require a source-backed personalisation record.
Failure Mode 7: Prospect Is Depicted Doing Something False
The image shows the recipient using, holding, endorsing, or saying something they never did.
Correction
Avoid unauthorised real-person depictions.
Failure Mode 8: Wrong Client Asset Is Sent
A personalised asset is attached to another prospect’s message.
Correction
Validate asset ID, prospect ID, recipient, and campaign before delivery.
Failure Mode 9: Unreadable Generated Text
Text inside an image is distorted or incorrect.
Correction
Require manual text review or add text deterministically after generation.
Failure Mode 10: Logo Is Distorted
The generated model recreates the logo inaccurately.
Correction
Use a controlled composition method or reject the asset.
Failure Mode 11: Novelty Replaces Value
The asset is visually interesting but commercially irrelevant.
Correction
Require a defined value proposition and next action.
Failure Mode 12: Every Prospect Receives The Same Gimmick
Only the logo changes.
Correction
Use real evidence-based relevance and test whether the asset adds value.
Failure Mode 13: Generated Asset Automatically Sends
The workflow moves directly from generation to external delivery.
Correction
Insert human approval and delivery authority.
Failure Mode 14: Likeness Training Data Is Exposed
Personal training images are publicly hosted or broadly accessible.
Correction
Use secure storage, limited access, retention, and deletion controls.
Failure Mode 15: Synthetic Voice Implies Personal Recording
The recipient believes the sender recorded the message specifically.
Correction
Use truthful wording and disclosure where needed.
Failure Mode 16: Website Analysis Is Presented As Full Audit
A limited public-page review is overstated.
Correction
Describe the asset as an observation, example, or preliminary diagnostic.
Failure Mode 17: Asset Is Reused After Permission Revocation
A likeness or voice continues to be used.
Correction
Maintain revocation state and block future generation and delivery.
Failure Mode 18: Reference Asset Is Copied Too Closely
The output reproduces another creator’s distinctive work.
Correction
Extract general patterns and create original execution.
Failure Mode 19: Multiple Candidates Are Never Reviewed
The first successful render is used.
Correction
Use candidate status and human selection.
Failure Mode 20: Metrics Reward Attention Only
The system optimises opens or views despite poor sales quality.
Correction
Measure qualified conversations, conversion, complaints, and cost.
Failure Mode 21: Asset Link Expires
The recipient receives a broken image or video.
Correction
Validate hosting, access, and link duration before delivery.
Failure Mode 22: Duplicate Asset Delivery
Retries or multi-channel workflows send repeated assets.
Correction
Use asset, campaign, recipient, and channel IDs.
Failure Mode 23: Wrong Disclosure
An asset is labelled inaccurately or no disclosure is provided where needed.
Correction
Define disclosure status before approval.
Failure Mode 24: Prompt Injection Through Website Content
External webpage instructions influence the asset workflow.
Correction
Treat scraped content as untrusted data.
Governance Responsibilities
Sales Brain
Owns:
commercial objective
prospect relevance
offer alignment
asset class
message purpose
sales follow-up
outcome
Content Brain
Owns:
visual quality guidance
brand language
human-quality messaging
creative structure
asset originality
AIBS Brain
Owns:
client-facing productisation
AIOS use cases
client packaging
service scope
value demonstration
Ads Brain
Supports:
creative testing
hook analysis
visual performance
campaign adaptation
Automation Brain
Owns:
generation workflow
status checks
candidate routing
storage
delivery orchestration
failure handling
duplicate prevention
Data Brain
Owns:
asset records
prospect linkage
evidence records
version history
permission states
retention
outcome data
Compliance And Risk Brains
Own:
brand risk
logo risk
likeness permission
voice permission
privacy
copyright escalation
disclosure
platform risk
regulated claims
Customer Brain
Supports:
recipient experience
trust
clarity
non-manipulative communication
Product Brain
Supports:
review interfaces
approval queues
asset preview
campaign controls
HeadOffice
Owns:
cross-Brain authority
high-risk synthetic media
policy exceptions
strategic review
Future AI Employee Ideas
Personalised Asset Strategist
Primary Brain: Sales Brain
Purpose:
Selects the appropriate asset type, commercial objective, and recipient-specific value.
Visual Personalisation Evidence Analyst
Primary Brain: Research Brain / Sales Brain
Purpose:
Finds and validates safe business-relevant personalisation evidence.
Visual Prompt Contract Designer
Primary Brain: Content Brain / Prompting Framework
Purpose:
Creates governed visual prompt and template contracts.
Brand And Logo Use Reviewer
Primary Brain: Risk Brain / Content Brain
Purpose:
Reviews logo accuracy, brand use, false endorsement risk, and visual context.
Likeness And Voice Permission Steward
Primary Brain: Compliance Brain / Risk Brain
Purpose:
Maintains consent, permitted use, revocation, retention, and deletion records.
Asset Candidate Reviewer
Primary Brain: Content Brain / Sales Brain
Purpose:
Scores generated candidates for accuracy, quality, relevance, and commercial suitability.
Synthetic Media Authenticity Reviewer
Primary Brain: Risk Brain / SIT Brain
Purpose:
Checks deception risk, disclosure, false event implications, and human-effort honesty.
Personalised Asset Delivery Safety Agent
Primary Brain: Automation Brain / Sales Brain
Purpose:
Checks asset, recipient, channel, approval, suppression, and duplicate-delivery state.
Personalised Asset Performance Analyst
Primary Brain: Experimentation Brain / Sales Brain
Purpose:
Measures qualified conversations, conversion, complaints, and cost.
Drift Protection
This framework protects MWMS from:
tool-driven visual gimmicks
unauthorised logo use
false endorsements
wrong-client assets
synthetic likeness abuse
voice cloning without consent
creepy personalisation
invented evidence
automatic sending
deceptive authenticity
copying competitor creatives
public exposure of training data
attention metrics without sales value
mass personalisation without governance
Drift Signals
Watch for:
“The logo is public, so we can use it.”
“It looks real enough.”
“No one will know it is AI.”
“The voice sounds exactly like him.”
“Just send the first image.”
“Every prospect should get one.”
“The website says enough.”
“We can say we audited them.”
“The asset will get attention.”
“Use the same concept for everyone.”
“More personalised means more effective.”
“We do not need a control group.”
“The image generated successfully, so it is ready.”
“Just host the training images publicly.”
“The prospect will think I recorded it.”
Rule
When these signals appear, return to evidence, permission, identity, authenticity, approval, commercial value, and outcome measurement.
Implementation Boundary
This framework defines the operating standard.
It does not authorise immediate development of:
image-generation automations
video-generation pipelines
voice-cloning systems
likeness models
asset approval interfaces
personalised outreach campaigns
logo ingestion systems
website-capture workflows
Before implementation, the responsible Brain must define:
commercial objective
asset class
recipient type
source evidence
logo rules
likeness rules
voice rules
prompt contract
generation provider
candidate count
review criteria
approval state
delivery channel
disclosure
storage
retention
deletion
performance metrics
scale limits
Rule
No personalised visual sales system should be built without a defined identity, permission, authenticity, approval, and delivery boundary.
Strategic Summary
Personalised visual sales assets can create strong commercial value when they combine:
verified prospect context
useful insight
clear visual communication
appropriate branding
human-quality messaging
controlled production
human review
truthful delivery
measurable sales outcomes
The durable opportunity is not any single image or video model.
The durable opportunity is the governed system that converts approved recipient data into a relevant asset without crossing identity, brand, privacy, or authenticity boundaries.
The strongest starting use cases are:
personalised diagnostic image
company-specific concept mock-up
short website observation video
proposal-support visual
approved prospecting value asset
These should begin with:
small volume
verified prospects
source-backed personalisation
human candidate selection
human send approval
clear measurement
Final Standard
The MWMS final standard is:
Every personalised visual sales asset must have a defined commercial objective, verified recipient, source-backed personalisation, appropriate brand use, explicit likeness and voice authority where required, a governed prompt contract, human quality review, authenticity review, delivery approval, retention controls, and measurable outcome.
A valid asset system must define:
asset ID
campaign
prospect
company
contact
asset class
asset type
commercial objective
source evidence
logo source
brand permission
likeness owner
likeness permission
voice owner
voice permission
prompt version
template version
model
candidate count
selected candidate
accuracy status
quality status
disclosure
approval
recipient
channel
delivery status
outcome
retention
deletion
AI may generate the asset.
AI may not independently authorise identity use, brand use, likeness use, voice use, disclosure, or external delivery.
That is the MWMS Personalised Visual Sales Asset Production And Governance Framework.
MWMS System Change Log
Version: v1.0
Date: 2026-06-21
Author: HeadOffice
Change
Created the MWMS Personalised Visual Sales Asset Production And Governance Framework using the AI Automations by Jack material covering:
Nano Banana image generation
image-to-image automation
logo-integrated visuals
website-specific image prompts
personalised prospect images
custom comic-style assets
website screenshots
personalised lead magnets
AI likeness model training
multiple generated candidates
human approve-or-deny stages
prospect-specific backgrounds
synthetic voice
website walkthrough videos
prospect-specific scripts
asset storage
email-draft integration
The framework converts the course’s visual personalisation tactics into a governed sixteen-layer sales asset system.
Added standards covering:
- commercial objective
- recipient identity
- source-backed personalisation
- asset classification
- brand and logo permission
- likeness permission
- voice permission
- visual prompt contracts
- fixed and dynamic variables
- reference asset analysis
- candidate generation
- human candidate selection
- visual accuracy
- text accuracy
- logo accuracy
- authenticity
- disclosure
- approval states
- channel authority
- storage
- retention
- deletion
- performance testing
- controlled scaling
Added explicit distinctions between:
- Traditional Lead Magnet
- Prospecting Value Asset
- Diagnostic Asset
- Sales Enablement Asset
Added explicit doctrine that:
- personalisation is not permission
- generation is not approval
- visual realism is not truth
- public logo visibility does not create unrestricted creative authority
- synthetic likeness must not imply a false action or endorsement
- cloned voice requires explicit authority
- AI-generated assets must not be presented as personally recorded or photographed when that is false
- website observations must not be presented as a complete audit
- asset approval and message-send authority are separate
Change Impact Declaration
This is a new Sales Brain framework.
It creates the governing layer for personalised images, videos, synthetic voice, logo-integrated assets, website-derived visuals, and other recipient-specific sales media.
The framework supports but does not replace:
- MWMS AI Assisted Outreach And Sales Follow Up Automation Framework
- MWMS Outbound Lead Enrichment And Cold Outreach Governance Framework
- MWMS AIOS Lead Capture And Conversion Infrastructure Framework
- MWMS Market Driven Social Content Production Framework
- MWMS Prompt Architecture And Automation Output Reliability Framework
The framework does not authorise:
- unrestricted logo use
- voice cloning without consent
- likeness generation without authority
- automatic external sending
- misleading synthetic media
- false endorsement
- mass personalised outreach
- public exposure of training data
Pages Created
- MWMS Personalised Visual Sales Asset Production And Governance Framework
Pages Updated
- None as part of this page creation
Pages Deprecated
- None
Standalone Pages Not Created
The following standalone pages were not created because their durable intelligence is governed within this framework:
- MWMS Nano Banana Framework
- MWMS Flux Image Framework
- MWMS Personalised Image Outreach Framework
- MWMS Synthetic Voice Sales Video Framework
- MWMS AI Comic Lead Magnet Framework
- MWMS Logo Insertion Automation Framework
- MWMS AI Thumbnail Recreation Framework
- MWMS Personalised Loom Automation Framework
- MWMS Website Screenshot Sales Asset Framework
- MWMS AI Likeness Model Governance Framework
Registries Requiring Update
- MCR Page Registry
- Sales Brain Page Registry
- MCR Copy Map
- MWMS Course Absorption Decision Registry
Canon Version Update Required
No immediate Sales Brain Canon version change is required unless the Canon must record the addition of a new active framework.
The next scheduled Sales Brain Canon alignment should recognise:
- personalised visual sales assets
- asset classification
- logo and brand permission
- likeness and voice authority
- authenticity and disclosure
- asset approval
- outcome measurement
Change Log Entry Required
Yes.
The new v1.0 framework must be recorded in:
- MWMS System Change Log
- MCR Page Registry
- Sales Brain Page Registry
- MCR Copy Map
- MWMS Course Absorption Decision Registry
Strategic Absorption Result
The AI Automations by Jack material concerning Nano Banana, personalised images, logo-integrated graphics, AI likeness models, synthetic voice, website walkthrough videos, and prospect-specific sales assets has been absorbed into one governed MWMS framework.
The absorption preserves:
- dynamic visual personalisation
- reusable prompt templates
- prospect-specific evidence
- company branding
- multiple generated candidates
- human approval
- website capture
- personalised video
- synthetic media production
- asset delivery
- performance measurement
The absorption rejects:
- tool hype as architecture
- public logo visibility as unrestricted permission
- synthetic images being treated as genuine photographs
- cloned voice without explicit consent
- false claims of personal recording
- personalisation without evidence
- automatic external delivery
- wrong-client asset risk
- unreviewed generated text
- public exposure of likeness training data
- novelty as a substitute for offer quality
- scaling without qualified-response evidence
The resulting v1.0 framework establishes that MWMS personalised visual sales assets must be:
- objective-led
- identity-aware
- evidence-backed
- permission-aware
- brand-safe
- likeness-controlled
- voice-controlled
- prompt-governed
- candidate-reviewed
- accuracy-checked
- authenticity-reviewed
- approval-gated
- delivery-controlled
- retention-managed
- outcome-measured
END OF FULL FILE OUTPUT