Age, gender, household
Age bands, sex/gender, household type, children, life stage, and main shopper.
age_band
sex
household_type
children
Audience modelling
Target groups that fit Swedish media planning and synthetic agents
Swedish media products tend to expose planning variables such as age, gender, geography, household context, income, education, interests, and media behaviour. This page translates that market language into an auditable schema for RBM agents.
Core fields
Gender, age, education, income
Planning overlay
Interests, media, shopping, mobility
ABC handling
Derived bridge, not a source truth
Recommended schema
Canonical agent dimensions
These are the fields the dashboard should treat as first-class filters. They map cleanly to public Swedish statistics, media-kit conventions, and the current human layer fields.
Education, income, work
Education, income decile, employment, occupation, industry, and work flexibility.
education
income_decile
occupation
telework
Planning tiers
Transparent affluence labels derived from income, education, tenure, and role.
sv_socioeconomic_tier
is_abc1
source_social_grade
Category affinity
Practical advertising interests such as retail, mobility, culture, sport, finance, travel, and sustainability.
interests
shopping_role
spending_category
Channel overlays
Keep ORVESTO, MMS, Ocast, TV, radio, print, digital, and OOH metrics as channel overlays.
media_touchpoints
reach_context
frequency
Mobility context
Trips, mode, region, anchor type, daypart, exposure opportunity, and route context.
mode
daypart
region
impressions
Market evidence
What Swedish media surfaces expose.
Audience panels and media kits
Age, gender, geography, household, education, income, and interest categories
appear repeatedly across Swedish publisher and planning surfaces.
OOH and mobility
Outdoor planning needs audience variables to remain tied to trips, daypart,
panel inventory, and opportunity-to-see logic.
Compatibility labels
ABC-like labels are useful for buyers, but should be generated from transparent
agent attributes rather than treated as canonical Swedish ground truth.
Implementation rule
Separate source, derivation, and activation.
Store raw agent attributes, derived planning labels, and channel-specific overlays as separate columns. This keeps the model explainable when a campaign asks for broad adults 25-54, high-income urban households, C+ decision makers, or interest-led segments.
source fields
derived labels
channel overlays
campaign filters
Evidence landscape
Swedish media references for target-group variables
Each source should be treated according to its role: broad cross-media profiling, channel overlay, first-party publisher activation, OOH reach modelling, or public canonical base.
ORVESTO / Mediafacts
- Observed targeting variables
- Media use, sociodemographics, consumption, interests, habits, spending, internet use, stores; Mediafacts reach/profile runs by gender, age and geography.
- Schema implication
- Use as the broad Swedish cross-media profiling reference, subject to licence and exact dictionary access.
MMS TV / Total Video
- Observed targeting variables
- Target groups by all/men/women and age intervals; total-video modules include gender with multiple age groups.
- Schema implication
- Implement MMS as a channel overlay rather than the whole synthetic-agent taxonomy.
TV4 Digital
- Observed targeting variables
- Age, gender, geography, interests, life stages, residential situations, devices and program contexts.
- Schema implication
- Maps cleanly to age/gender/geo/lifestage/housing/interests plus media-context fields.
Schibsted / Admarket
- Observed targeting variables
- Gender, age bands 18-25 to 66+, geography, demographics and interests.
- Schema implication
- Use as a digital publisher overlay with first-party-data provenance.
Bauer Media audio
- Observed targeting variables
- Channel profiles by age/gender/interests; digital audio adds geography, device, time, context, vertical, weather and mood.
- Schema implication
- Use as audio overlay; do not mix channel listener profiles with person-level truth.
Outdoor Impact
- Observed targeting variables
- OOH reach/frequency/VAC from inventory, traffic flows and Ipsos travel/background studies; can break down VAC by target groups.
- Schema implication
- Keep VAC native; use target groups only as a reach overlay.
Publisher kits / Ocast
- Observed targeting variables
- Public sales profiles use gender, age, urbanity, education, high income, high-spender indices and category interests.
- Schema implication
- Treat sales-language personas separately from measured segment definitions.
SCB public standards
- Observed targeting variables
- SUN education, SSYK occupation, income registers and alternatives to discontinued SEI.
- Schema implication
- Use SCB-compatible dimensions as the public, auditable canonical base.
ABC / C+ bridge
Recommended compatibility layer
Use an explicit derived tier until a paid Swedish media codebook confirms a different market-specific definition. The label should be recalculable and never overwrite the source fields.
Agent field
Purpose
Suggested values
sv_socioeconomic_tier
Primary Swedish planning tier
Affluent, comfortable, mainstream, constrained
is_abc1
Buyer-facing ABC1 proxy
true / false / unknown
source_social_grade
Only when a licensed source supplies it
A, B, C1, C2, D, E, null
Calibration
How to make the filters trustworthy.
Compare marginal distributions to SCB and survey sources, then validate segment behaviour against trips, impressions, and campaign-facing reach profiles. Every high-value segment should have a source note and a freshness date.
SCB demographic margins
Human layer field coverage
RBM impression filters
Channel overlay checks