Working Prototype
The Prototype
One Australian audience file, named at source and run end to end. Every figure is computed from the file, and every source opens to the page that proves it. Campaign segments, marketing materials, social uploads and Leadership Team reports are built from the same file. All currency figures AUD.
Each section below links to the duty it has to satisfy on the marketing compliance page, where the Spam Act, the Privacy Principles and the complaint response steps are set against the columns that evidence them.
End to End, In Motion
An Australian Consented Audience, Run End to End
Records
241
20 columns
May be sent to
193
84% of resolved people
Marathi at home
118
of the contactable audience
States covered
7
40 postcodes
Sources5 build · 7 set aside
Source codes
- LANPLANP — ABS Census — language used at home
- SALSAL — ABS QuickStats — suburb ancestry and birthplace
- APAP — Australia Post postcode reference
- ASGSASGS — ABS statistical areas and correspondences
- 1P1P — First-party opted-in list
- APP7APP7 — Australian Privacy Principle 7 — direct marketing
- SPAMSPAM — Spam Act 2003 — channel consent
Trust ladder
- Official statisticOfficial statistic — Enumerated or statutory figure
- Reference standardReference standard — Authoritative geography or coding list
- Primary record — first partyPrimary record — first party — Collected at opt-in by the operator
- RegulatorRegulator — Law or guidance that constrains use
- Licensed panelLicensed panel — Vendor panel — not a sendable list for ancestry
- Modelled estimateModelled estimate — Assigned by model, not declared by the person
Marathi used at home: 22,263 nationally, 9,753 New South Wales, 7,170 Victoria — the counts the state mix follows
- Validity
- Official statistic
- How to authenticate it
- A national census, not a sample: enumerated August 2021 under the Census and Statistics Act 1905. The standard Community Profiles fold Marathi into “Other Indo-Aryan”, so these counts come from TableBuilder: cross-tabulate Language used at home against State and the three figures reproduce exactly
- Cost
- Free — TableBuilder Basic costs nothing; it asks for a registration
Suburb-level ancestry and country of birth. Tarneit, for one: 56,370 people, Indian ancestry 15,199 (27.0%), India-born 16,231 (28.8%)
- Validity
- Official statistic
- How to authenticate it
- Each suburb has its own QuickStats page keyed by its SAL code, so any figure here can be re-read at the source in a few seconds
- Cost
- Free — the ABS publishes QuickStats at no charge
The locality-to-postcode list every suburb and postcode in this file was resolved against
- Validity
- Reference standard
- How to authenticate it
- Australia Post sets Australian postcodes, so its own file is the authority. The Standard Postcode File is the CSV every suburb and postcode here was matched against
- Cost
- Quote on request — Australia Post publishes no price for the CSV files, and the free PDF booklet is licensed for non-commercial use only, so a commercial audience file needs the licensed productprice page
The SA2 code each postcode resolves to, and the correspondence that makes postcode-to-SA2 many-to-many
- Validity
- Reference standard
- How to authenticate it
- The correspondence files themselves, under Creative Commons: CG_POSTCODE_2021_SA2_2021.xlsx is the postcode-to-SA2 table this file joins on, and it carries the ratio that makes the join many-to-many
- Cost
- Free — Creative Commons Attribution 4.0
Name, email, phone, mobile, age, the self-declared ancestry and language, the consent flag and the chosen channel — the only lawful route to person-level contact detail
- Validity
- Primary record — first party
- How to authenticate it
- Each record carries the consent timestamp, the form it came from and the purpose consented to, so any single person can be traced back to the moment they opted in
- Cost
- No fee for the data — the operator collects it. The platform that holds it publishes its price: Audience Republic Basic USD 98/mo billed annually at USD 1,028, Plus USD 163/mo billed annually at USD 1,713, entry tiers to 10,000 contacts; Enterprise on application
Sensitive information, which includes racial or ethnic origin, may be used for direct marketing only with the individual’s consent
- Validity
- Regulator
- How to authenticate it
- The regulator’s own guidelines on the Australian Privacy Principles — the text a court would be pointed at
- Cost
- Free — guidance, not a data purchase
Consent, sender identification and a working unsubscribe on every commercial email or SMS
- Validity
- Regulator
- How to authenticate it
- The consolidated text of the Act on the Federal Register of Legislation — the primary source, not a summary of it
- Cost
- Free — legislation, not a data purchase
Five million opted-in Australian profiles carrying name, email, phone, gender, date of birth and state — and no ethnicity, ancestry or language field at all
- Validity
- Licensed panel
- How to authenticate it
- A marketplace listing that states its own attribute set, so the absence of an ancestry field can be read off the page
- Cost
- Free sample on the listing, which carries no per-unit charge; the full data sets are quote on request
Over 20 million opted-in Australian consumers, filterable to postcode, suburb and state — with no ethnicity, language or country-of-birth filter offered
- Validity
- Licensed panel
- How to authenticate it
- The platform page lists its own filter set, which is where the missing ancestry dimension shows
- Cost
- A$0.43 a record plus GST, the vendor’s own published breakdown: base record 25.0c, phone 8.0c, email 10.0c
Consumer classification into 14 Groups and 52 Types, covering 99% of Australian households — a segment on the household, not an ethnicity on a named person
- Validity
- Modelled estimate
- How to authenticate it
- The vendor describes its own build and coverage; the classification is modelled, so it sits below an official statistic
- Cost
- Quote on request — no price, rate card or fee appears on the product page
Over 5,000 behavioural, demographic, geographic and lifestyle attributes used to enrich an existing customer record
- Validity
- Modelled estimate
- How to authenticate it
- An enrichment product: it appends to a record the buyer already holds, and the appended values are modelled
- Cost
- Quote on request — the enrichment page carries an enquiry form in place of a price
Over 60,000 interviews a year, carrying language used at home among its variables — a survey panel that sizes a segment, never a list to send to
- Validity
- Licensed panel
- How to authenticate it
- A continuous survey with a published sample size and method; it measures a population, and carries no contactable record
- Cost
- A subscription is quote on request, but the online store publishes off-the-shelf profiles built from it: A$1,529 including GST standard, A$4,950 including GST premium
54 Personas in 6 Communities, geo-projected down to mesh block — again an area classification, not a person attribute
- Validity
- Modelled estimate
- How to authenticate it
- Personas are assigned to areas by model, so a person inherits an area’s label rather than declaring their own
- Cost
- Quote on request — the product page routes to an enquiry rather than a rate card
The one identified national file of Australians — and s 91B of the Commonwealth Electoral Act 1918 shuts commercial use of it, which is a large part of why no lawful public person-level list exists
- Validity
- Official statistic
- How to authenticate it
- The AEC’s own electoral-roll access page sets out who may receive roll data and for what. A marketing organisation is not on that list
- Cost
- Not for sale — there is no commercial access path, at any price
What this buys — named campaign segments, channel kits and Leadership Team report specs are built from the same contactable audience below. Open Campaign segments to see the live counts.
Legal footing. The consent gate is what answers a complaint: which duty, which column, which test. Consent duties and the columns that evidence them
Reference Geography
Where These People Sit
Geography vs people
- Real referenceReal reference — Suburb, postcode, state and SA2 from ABS and Australia Post
- Fictional peopleFictional people — Names, emails and phones are synthetic — never a live contact list
Suburbs
47
real localities
Postcodes
40
Australia Post reference
ABS SA2 areas
38
ASGS 2021 codes
States and territories
7
all eight are in scope
What this buys — suburb-level out-of-home and geo-targeted social use these localities; see Marketing materials and Social activation.
Suburb, postcode, state and SA2 are real reference geography; the people are fictional.
Legal footing. Targeting a suburb uses reference geography, not a personal attribute — but the send still sits under the Spam Act and APP 7. The instruments a locality-targeted campaign has to satisfy
The Columns
Every Field, and What Stands Behind It
Source codes on each column
- LANPLANP — ABS Census — language used at home
- SALSAL — ABS QuickStats — suburb ancestry and birthplace
- APAP — Australia Post postcode reference
- ASGSASGS — ABS statistical areas and correspondences
- 1P1P — First-party opted-in list
- APP7APP7 — Australian Privacy Principle 7 — direct marketing
- SPAMSPAM — Spam Act 2003 — channel consent
Column types
- texttext — Free-text or name field
- numbernumber — Numeric measure such as age
- flagflag — Controlled Y/N or E/M code
- datedate — Calendar date on the consent event
- codecode — Stable key or geographic code
Legal mark
- §Scale icon — Sensitive information or a statutory obligation on the column
| Column | Type | Source | What it is | Where it comes from |
|---|---|---|---|---|
| record_id | code | Stable key for the record | Generated surrogate key — the natural key a warehouse hashes on | |
| first_name | text | 1P | Given name | Fictional. Marathi, wider Indian and general given-name pools |
| last_name | text | 1P | Family name | Fictional. Marathi and wider Indian family-name pools |
| text | 1P | Email address | Fictional, on the reserved example.com domain (RFC 2606) | |
| phone | text | 1P | Landline number | Fictional, inside the (0x) 5550 xxxx drama range |
| mobile | text | 1P | Mobile number | Fictional, inside the 0491 57x xxx drama range |
| age | number | 1P | Age in years | Weighted to the 25–44 band that dominates recent India-born arrivals |
| state | code | LANP | State or territory | Mix follows the ABS Census 2021 counts of Marathi used at home |
| suburb | text | SALAP | Suburb / locality | Real locality, resolved against the Australia Post postcode reference |
| postcode | code | AP | Postcode | Real postcode for that locality |
| sa2_code_2021 | code | ASGS | ABS statistical area level 2 | Real ASGS 2021 code the postcode resolves to — the geography join key |
| sa2_name_2021 | text | ASGS | SA2 name | Real ASGS 2021 name for that code |
| ethnicity_nationality | text | SALAPP7 | Ancestry, on ABS categories | Self-declared by the person at opt-in, on ABS Census ancestry categories. A bought-in ethnicity append cannot be used for marketing however it was collectedSensitive information — s 6, Privacy Act 1988 (Cth). APP 7.4 needs the person’s consent |
| marathi_speaking | flag | LANP | Marathi used at home | Self-declared at opt-in, on the ABS Census variable Language used at home (LANP)Travels with the same consent as the ancestry column |
| consented_for_marketing | flag | 1PAPP7 | Current marketing consent | Derived state — the current position of the consent event historyAPP 7.4 — sensitive information may be used for direct marketing only with consent |
| contact_preference | flag | 1PSPAM | Channel the person chose — E email, M mobile | Channel consent, enforced at send time by the preference joinSpam Act 2003 (Cth) — consent, sender identification and a working unsubscribe |
| consent_timestamp | date | 1P | When the current consent state was set | The date the latest consent event was recorded |
| consent_source | text | 1PSPAM | Where the consent came from | Checkout opt-in, waitlist, newsletter, ballot, box office or a permission-based partner list |
| consent_purpose | text | What the person consented to | The stated purpose the consent was given for | |
| data_status | flag | Real extract or synthetic sample | Required column — travels to every downstream table and tile |
What this buys — every campaign segment and Leadership Team report below names the columns it consumes from this dictionary.
Legal footing. Two columns on this list carry a consent requirement that the others do not. Why ancestry and language are treated as sensitive information
Warehouse
Data Mart Tables and Their Joins
Table kinds
- DimensionDimension — Slowly changing descriptive attributes
- CorrespondenceCorrespondence — Many-to-many geography bridge with a ratio
- FactFact — Event at a point in time — consent, send, sale
- Serving viewServing view — The only object a campaign may send from
Key badges
- pkpk — Surrogate primary key
- nknk — Natural business key
- fkfk — Foreign key to another table
dim_person
dimOne row per person per change — history kept, one row current
- pkperson_sk
- nkrecord_id
- first_name
- last_name
- age
- ethnicity_nationality
- marathi_speaking
- fkgeo_sk
- valid_from
- valid_to
- is_current
dim_geography
dimOne row per suburb and postcode
- pkgeo_sk
- nkpostcode
- suburb
- state
- sa2_code_2021
- sa2_name_2021
dim_contact_point
dimOne row per person per channel
- pkcontact_sk
- fkperson_sk
- channel
- address
- is_preferred
dim_campaign
dimOne row per campaign
- pkcampaign_sk
- nkcampaign_id
- name
- market
- channel
xref_postcode_sa2
xrefOne row per postcode and SA2 pair, with its share
- fkpostcode
- fksa2_code_2021
- ratio
fact_consent_event
factOne row per consent event — granted or withdrawn
- pkconsent_event_sk
- fkperson_sk
- fkdate_sk
- channel
- purpose
- action
- source
fact_campaign_send
factOne row per person per send
- pksend_sk
- fkperson_sk
- fkcampaign_sk
- fkdate_sk
- channel
- delivered
- opened
- clicked
fact_ticket_sale
factOne row per ticket transaction
- pksale_sk
- fkperson_sk
- fkcampaign_sk
- fkdate_sk
- tickets
- amount_aud
vw_marketable_audience
viewOne row per contactable person, per channel they chose
- fkperson_sk
- channel
- address
- state
- sa2_code_2021
- marathi_speaking
The joins
Cardinality
- 1:11:1 — One row maps to exactly one row
- 1:M1:M — One parent, many children
- M:MM:M — Many-to-many — needs a correspondence with a ratio
| From | To | On | Card. | |
|---|---|---|---|---|
| dim_person | dim_geography | dim_person.geo_sk = dim_geography.geo_sk | 1:M | |
| dim_geography | xref_postcode_sa2 | dim_geography.postcode = xref_postcode_sa2.postcodeA postcode can span several SA2s and an SA2 several postcodes, so the correspondence carries a ratio rather than a single code | M:M | |
| dim_person | dim_contact_point | dim_person.person_sk = dim_contact_point.person_sk | 1:M | |
| dim_person | fact_consent_event | dim_person.person_sk = fact_consent_event.person_skConsent is an event with a timestamp, a purpose and a source. The Y/N column on the file is the current position of this history, not the record of it | 1:M | |
| dim_person | fact_campaign_send | dim_person.person_sk = fact_campaign_send.person_sk | 1:M | |
| dim_campaign | fact_campaign_send | dim_campaign.campaign_sk = fact_campaign_send.campaign_sk | 1:M | |
| dim_person | fact_ticket_sale | dim_person.person_sk = fact_ticket_sale.person_sk | 1:M | |
| dim_campaign | fact_ticket_sale | dim_campaign.campaign_sk = fact_ticket_sale.campaign_sk | 1:M | |
| fact_consent_event | vw_marketable_audience | latest GRANT per person and channel, with no later WITHDRAWThe view is the only object a campaign may send from | 1:1 | |
| dim_contact_point | vw_marketable_audience | dim_contact_point.person_sk = vw_marketable_audience.person_sk and is_preferred | 1:1 |
The tests that guard it
Test types
- uniqueunique — One row per key
- not_nullnot_null — No orphan surrogate or foreign key
- relationshipsrelationships — Referential integrity to the parent dimension
- accepted_valuesaccepted_values — Controlled vocabulary only
- expression_is_trueexpression_is_true — Business rule enforced by a test
unique
dim_person.person_sk, dim_geography.geo_sk
one row per key
not_null
every surrogate and foreign key
no orphan rows
relationships
fact_*.person_sk → dim_person.person_sk
referential integrity
accepted_values
channel in (E, M) · action in (GRANT, WITHDRAW) · state in the eight codes
no free text in a controlled column
expression_is_true
vw_marketable_audience has no person whose latest event is WITHDRAW
consent is enforced by a test, not by convention
Provenance chips used on this page
Provenance
- Actual spendActual spend — Money already spent on this programme
- Published pricePublished price — The price the vendor lists
- QuotedQuoted — A written quote held on file
- CalculatedCalculated — Worked out from the figures beside it
- Official statisticOfficial statistic — An official statistic or statutory record
- Illustrative — from sample dataIllustrative — from sample data — Computed from the downloadable sample files
What this buys — Leadership Team reports read from these tables. Campaign send and ticket sale facts fill once the pilot transacts; audience size and reach resolve from today's file.
Legal footing. The consent event table is the record a regulator asks for; the tests keep the serving view honest. Where the consent evidence lives, and the tests that hold it
Commercial Activation
Campaigns, Materials, Social and Leadership Reports
Named segments, asset kits, ad-platform uploads and Leadership Team report specs — each figure below is computed from the same consented file as the walkthrough above. Revenue and engagement figures resolve once the pilot transacts; audience size and reach resolve today. See also the data ecosystem and investment pages for the wider monetisation sequence.
Named segments
6
Material kits
5
Social destinations
3
Leadership reports
4
Campaign segments
Named Audiences Ready to Send
Channel
- EE — Email preferred
- MM — Mobile preferred
- BothBoth — Either channel, respecting preference at send
Filter: marketable AND marathi_speaking = 'Y' AND contact_preference = 'E'
Breakdown
Filter: marketable AND age between 25 and 44 AND state in (VIC, NSW)
Breakdown
Filter: marketable AND contact_preference = 'M'
Filter: marketable AND consent_source contains 'box office'
Filter: marketable AND consent_source contains 'ballot'
Filter: marketable AND consent_source contains 'waitlist'
Legal footing. A segment is a filter over the serving view, so a withdrawal removes a person from every segment at once. Each segment sends only from the consented view
Marketing materials
Asset Kit Per Channel
Channel kit
- EmailEmail — Nurture and on-sale templates
- SMSSMS — Short-window mobile creative
- Out-of-homeOut-of-home — Suburb-keyed poster and transit
- PartnerPartner — Co-brand with venues and promoters
- WebWeb — Event landing variants
Subject, preview and body personalised to ancestry and language flags, with unsubscribe and sender identity
Outcome: Repeat opens and click-to-purchase on language-led tours
Segment: Marathi at home — email preferred · 77 people
160-character on-sale alert with stop path; mobile only where preference is M
Outcome: Same-day ticket conversion on release windows
Segment: SMS — on-sale day · 66 people
Poster and transit creative keyed to real suburbs and postcodes in the file
Outcome: Awareness in high-density India-origin localities ahead of on-sale
Segment: 25–44 metro Victoria and New South Wales · 86 people
Co-branded landing and email for permission-based partner lists
Outcome: Trusted introduction through venues and community promoters
Segment: Box-office re-engagement · 25 people
Geo and language variants of the event page, fed by SA2 and language flags
Outcome: Higher landing-to-checkout rate for the right city and language cut
Segment: 25–44 metro Victoria and New South Wales · 86 people
Legal footing. Three duties live in the creative rather than the data, and every template has to carry them. Sender identity, the unsubscribe and the SMS sender ID
Leadership Team reports
Report Specs the Warehouse Already Supports
Figure status
- Available nowAvailable now — Computed from today’s consented file
- Resolves laterResolves later — Needs campaign send and ticket sale facts from the pilot
- Grain
- One row per contactable person per preferred channel
- Cadence
- Weekly during the pilot; monthly thereafter
- Decision it drives
- Whether the Australian proof market has enough consented reach to proceed past gate G1
- Tables
- vw_marketable_audiencedim_persondim_geography
- Columns
- person_skchannelstatesa2_code_2021marathi_speaking
Available now
Marketable people
193
Email reachable
127
Mobile reachable
66
States covered
7
Marathi at home
118
Resolves once the pilot transacts
- Campaign open and click rates
- Ticket sell-through by segment
- Grain
- One row per person per send, joined to ticket sales
- Cadence
- Per campaign within 72 hours of send; weekly roll-up for Leadership
- Decision it drives
- Which segments to scale, pause or redesign before the next on-sale
- Tables
- dim_campaignfact_campaign_sendfact_ticket_salevw_marketable_audience
- Columns
- campaign_idchanneldeliveredopenedclickedticketsamount_aud
Available now
Named segments ready
6
Largest segment
25–44 metro Victoria and New South Wales (86)
Resolves once the pilot transacts
- Delivered / opened / clicked
- Tickets and amount_aud per campaign
- Cost per ticket by channel
- Grain
- One row per consent event; serving view is the sendable cut
- Cadence
- Daily automated test; weekly Leadership summary
- Decision it drives
- Go / no-go on each send; remediation if WITHDRAW events are not suppressed
- Tables
- fact_consent_eventdim_personvw_marketable_audience
- Columns
- actionpurposesourcechannelperson_sk
Available now
Consent rate
84%
Suppressed
38
Consent sources
6
Resolves once the pilot transacts
- Grant / withdraw event volume over time
- Unsubscribe rate after each send
- Grain
- One row per suburb / SA2 with audience and, later, sales
- Cadence
- Monthly; before each venue commitment
- Decision it drives
- Where to book the next three to five pilot events
- Tables
- dim_geographyxref_postcode_sa2vw_marketable_audiencefact_ticket_sale
- Columns
- suburbpostcodesa2_code_2021ratioticketsamount_aud
Available now
Suburbs
47
Postcodes
40
SA2 areas
38
Resolves once the pilot transacts
- Tickets and revenue by suburb and SA2
- Sell-through versus audience density
Legal footing. Consent health is a Leadership Team report because the downside is measured in penalty units. The exposure these reports keep visible
Social and platform activation
Hashed Uploads to Ad Platforms
Email matchable
127
Mobile matchable
66
Match keys
Uploadable records from this file: 193 (email 127 + mobile 66, respecting preference).
APP 7.4 consent required for ancestry-derived targeting; Spam Act unsubscribe must remain available on owned channels; Meta policy limits apply to sensitive attributes.
Ancestry and language stay on owned sends. The match file carries only hashed contact points with current marketing consent.
Uploadable records from this file: 193 (email 127 + mobile 66, respecting preference).
Upload only records with current marketing consent. Ancestry and language fields stay off the match file — they segment owned sends, not the ad platform upload.
Ancestry and language stay on owned sends. The match file carries only hashed contact points with current marketing consent.
Uploadable records from this file: 193 (email 127 + mobile 66, respecting preference).
Same consent gate as Meta and Google. Platform sensitive-category rules further restrict ethnicity-based lookalikes.
Ancestry and language stay on owned sends. The match file carries only hashed contact points with current marketing consent.
Legal footing. Hashed contact points with current consent go up. Ancestry and language stay on owned sends. What may be uploaded to an ad platform, and what may not