Something Digital 2026 · 25-26 August · Brisbane

Foundations of an AI Ready marketing team

Profiles hold the brand. Personas and voices speak for it. Brand Data stores the facts the business is made of, then connects those facts back to the brand.

Why context

Everyone on the same page

Brand goes wrong when people work from different versions of the truth. Context must be persistent, current and shared.

Persistent

The same for every brief.

Profiles, personas, voices, Brand Data.
Written once, then reused for years.
Never resets with a new campaign.

Current

True for this piece of work.

Campaign, audience, channel, offer.
Time-bound, drawn from the brand.
Specific work, not a new brand.

Shared

Reachable by everyone.

One record for EVERYONE, inc. AI.
No private versions that diverge.
Agreement is tangible, understood..

Miss any one of these and people invent their own context. That is how brand drifts.

Current team structure

Where most teams are today

The current structure

Who does the work

Owner or general manager

Budget and final say

Marketing manager

Plans, briefs and approves everything

Coordinator

Social, email, events

Designer

Artwork and assets

Agency

Web, ads, SEO

Where the brand lives

PDF guidelines

A shared drive

People’s heads

Modern marketing team

The SMB reality

AI-powered, still human-led

The people

Marketing manager

Strategy, brand, sign-off

Marketing coordinator

Content, social, email, web

Agency

Optional

Freelance support at peaks

The governance layer

Drafting

Repurposing

Brand checks

Monitoring

One shared brand record

Profiles

Personas

Voices

Brand Data

AI-native team

The next level

The AI-native team

The crew

Marketing engineer

Builds the workflows

Brand owner

Owns the record

AI agents

Do the repeatable work

Ultra-modern AI native tools and platforms

BrandHalo

Cursor

Resend

Amplitude

AI-RSE

Wired to one brand record

Read by every tool through APIs and MCP.

Existing business stack

Old team vs AI-enabled team

What changed

Before and after AI

The old team

The AI-enabled team

Guidelines in a PDF

Brand as live data

Knowledge in people’s heads

Context in every tool

Weeks per campaign

Days per campaign

One channel at a time

Every channel at once

Consistency by memory

Consistency by check

What an agent is

Everyone says agent

What an agent actually is

Level 1

Custom GPT

You ask

A prompt and some files

It writes back

Build effort

It answers, it does not act.

Level 2

Dedicated agent

A trigger fires

One job, with tools

Brand record

Website

The work is done

Build effort

One job, done without you.

Level 3

Multi-agent system

Something happens in the business

Orchestrator

Research

Draft

Check

APIs write to your systems

You approve the exceptions

Build effort

A whole process, running itself.

Same word, three very different things.

How the models connect

SD26 · Brisbane · 25-26 August 2026

Brand data model

How the four models connect

Profiles: The source of truth for strategy, visuals, messaging and compliance.
Personas: Audience portraits linked to a profile, with motivation and channels.
Voices: Named speakers with tone, topics and guardrails for each brand.
Brand Data: Collections define the model. Items hold the facts. Connections join them.

Organisation

Tenant

Brand Profile

Master, sub or campaign

Persona

Who you speak to

Voice

Who is speaking

Brand Data

Collections, items, connections

Connections

Item to profile, persona, voice, knowledge or another item

Profiles

01 · Profiles

The brand record

The source of truth for strategy, personality, visuals, messaging and compliance. Campaign profiles add mechanics under a parent brand.

Record

One profile per master, sub-brand or campaign

Name
Type: master / sub / campaign
Parent brand
Default for organisation
Status: draft / pending approval / approved
Version + display order

Compliance

Brand guidelines URL
Trademark status
Brand contact email
Notes
Frameworks (privacy, ads, AI, and more)
Applicability + scope

Brand essence

Tagline
Purpose
Mission
Vision
Values
Brand promise
Positioning (nested)
Rally cry
Positioning statement
Brand narrative
Challenge statement
Proof points
Tagline do / do not rules

Personality

Archetypes (name, weight 0-100)
Scenario guidance + triggers
Traits
Primary tone
Secondary tone
Writing style

Audience segments

Segment name
Description
Key needs
Demographics
Pains and fears
Hopes and goals
Desired feeling

Competitive landscape

Competitor name + website
Positioning
Differentiation
Threat: very low to very high
Quick overview
Our differentiators

Visual assets

Logos (orientation, colour, role)
Logo files: SVG / EPS / PNG / JPG
Usage note, clear space, min size
Logo do / do not + co-brand notes
Primary / secondary colours
Neutral / tertiary colours
Colour specs: hex, RGB, CMYK, PMS
Gradients
Typography + source (Google / Adobe / custom)
Type hierarchy + do / do not
Adobe Fonts kit ID
Image style + imagery categories
Example images
Documents (guidelines, decks)

Messaging

Elevator pitch
Key messages
Boilerplate
Preferred vocabulary
Do not say
CTA guidance
Voice dimensions
Voice do / do not rules
Writing examples (off / on brand)
Phrase swaps (avoid → use)

Campaign (when type is campaign)

Lives on a campaign profile, always under a parent brand

Context / background
Goal: awareness / acquisition / retention
Objectives (1 primary, up to 5 more)
Ordered messages
KPIs (title + measurement)
Offer + CTA
Context and CTA page URLs
Start and end dates
Campaign assets (up to 50)
Channels

Personas

02 · Personas

Who you speak to

Audience portraits linked to a profile. Demographics, motivation, channels and a face.

Identity

Name
Age
Occupation
Location
Income range
Description

Portrait

Uploaded photo
Generated avatar style
Avatar background + hair
Avatar seed
AI headshot (Enterprise)

Motivation

Goals
Pain points
Buying behaviour

Reach

Preferred channels

Links and status

Brand profile
Active
Created / updated

Voices

03 · Voices

Who is speaking

Named speakers with tone, topics and guardrails. Linked to a profile so each brand can have more than one voice.

Speaker

Name
Job title
Department
Email
Bio
Photo or generated avatar

Voice character

Communication style
Tone
Personality traits

Content

Content focus
Preferred topics
Expertise areas

Guardrails

Do list
Do not list
Key messages

Links and status

Brand profile
Active
Created / updated

Brand Data

04 · Brand Data

Collections, items, connections

A graph you design: collections define the model, items hold the facts, connections join those facts to the brand.

Collection (the model)

You design the shape. Templates can start you off.

Name + slug
Description + icon
Status: active / archived
Source: manual / template / import
Optional brand profile

Field (a column on the model)

Key + label
Type: text, number, date, boolean
Type: select, multi-select, rich text
Type: URL, reference, asset
Required + sort order
Options (for selects)
Validation: min, max, length, pattern
Reference collection
Help text

Item (a row of facts)

Name
Field values (the data)
Brand profile
Group
Tags
Status: draft / active / archived
External ID (for imports)

Organisation helpers

Groups (nested categories)
Tags

Connection type + connection

Items can link to other items, or to a profile, persona, voice or knowledge article

Name + slug
Source collection
Target: item / profile / persona / voice / knowledge
Cardinality: one-to-many or many-to-many
Directional
Connection metadata

Worked examples

Same pattern, different collections

Same pattern, different collections. Locations use the Local Area Marketing pack. Products ship in Core. Healthcare, education and automotive packs follow the same shape.

Locations

Local Area Marketing · lam_location

Clinics, stores or sites. The Core pack uses a shorter location model; this is the full local marketing set.

Trading nametext
Addresstext
Suburbtext
Regiontext
Countrytext
Postcodetext
Phonetext
Emailtext
Websiteurl
Descriptionrich text
Primary categorytext
Store codetext
Hourstext
Launch datedate
Google Business Profileurl

Typical connections

Offers treatment / has device / can use offer
Has team member / works with partner
Item can also link to a brand profile

Products

Core business · product

Things the business sells. Same collection pattern as Locations: fields on the model, facts on each item.

Summarytext
Descriptionrich text
Product pageurl
Availabilityselect

Typical connections

Offered by brand (product → profile)
Targets persona (product → persona)
Referenced in knowledge (product → FAQ)

Foundations of an AI Ready marketing team

Why context

Persistent

Current

Shared

Current team structure

Modern marketing team

AI-native team

Old team vs AI-enabled team

What an agent is

How the models connect

Profiles

Record

Company information

Compliance

Brand essence

Personality

Audience segments

Competitive landscape

Visual assets

Messaging

Campaign (when type is campaign)

Personas

Identity

Portrait

Motivation

Reach

Links and status

Voices

Speaker

Voice character

Content

Guardrails

Links and status

Brand Data

Collection (the model)

Field (a column on the model)

Item (a row of facts)

Organisation helpers

Connection type + connection

Worked examples

Locations

Products