“Adults 25-54, household income $50K+” sounds like targeting. It isn’t. It describes nearly half the adult population. Most of them will never need a PI lawyer.
Demographic targeting is how agencies waste CTV budgets.
The Demographic Targeting Trap
Traditional TV buying used demographics because nothing better was available. You bought “Adults 25-54” on programs that indexed well for that demo.
CTV can do the same thing. But why would you?
Adults 25-54, HHI $50K+ in Philadelphia DMA:
- Population: ~1.5 million adults
- Potential PI clients in next year: ~15,000 (1%)
- Your targeting efficiency: 1%
You’re paying to reach 99 people who will never need you for every 1 who might.
Why Demographics Fail for PI
Option A
Option B
Demographics Can't Predict
Accidents are random. A 28-year-old making $60K is no more likely to be injured than a 52-year-old making $80K. Workplace injuries happen at any age. Demographics don’t predict injury.
Cases Come From Everywhere
Your last 100 cases probably span age 22 to 68, income $30K to $200K. No demographic filter would have captured them all.
Wide Targeting = Wide Waste
Who You Reach With 'Adults 25-54'
- − People who never drive
- − People who work from home
- − People with no injury exposure
- − People who already have attorneys
- − People outside your service area
All at the same CPM as potentially qualified households.
Behavioral Targeting: The Alternative
Behavioral targeting identifies what people DO, not who they ARE:
Predictive Behaviors for PI
| Behavior | Why It Matters |
|---|---|
| Recent auto loan | New vehicle, unfamiliar handling |
| Motorcycle ownership | Higher injury severity risk |
| Rideshare usage | Frequent passenger exposure |
| Workers’ comp research | Employment injury concern |
| Medical facility visits | Current or recent injury |
| Construction employment | Workplace injury risk |
| Physical therapy interest | Existing injury treatment |
| Active sports household | Recreational injury potential |
These behaviors predict injury likelihood better than age or income.
Intent Behaviors
Even more valuable, people actively considering legal help:
| Behavior | Signal Strength |
|---|---|
| Legal services research | Very high |
| Personal injury content consumption | High |
| Competitor website visits | High |
| Insurance claims research | Medium |
| Medical malpractice content | Medium-high (if relevant) |
Intent behaviors indicate someone is already in consideration mode.
The Efficiency Difference
Option A
Option B
Demographic Approach
Audience: Adults 25-54, $50K+ in DMA. Population: 1.5M households. Relevance: ~1-2%. Effective CPM: $40 ÷ 1.5% = $2,667 per relevant household.
Behavioral Approach
Audience: Auto loans + motorcycle + workers comp in DMA. Population: 150K households. Relevance: ~15-25%. Effective CPM: $45 ÷ 20% = $225 per relevant household.
Same base CPM. 10x difference in cost to reach relevant households. That’s not optimization. That’s a different business model.
Why Firms Still Use Demographics
Legacy Thinking
“That’s how TV buying has always worked.” True for broadcast. CTV enables better.
Simplicity
Demographic targeting is easy. Behavioral targeting requires strategy.
Reach Fears
“But I want to reach everyone!” You can’t afford to reach everyone effectively. Better to reach fewer relevant households more frequently than spray across everyone and hope.
Platform Defaults
Some platforms default to demographic targeting. You have to intentionally build behavioral audiences.
Building Behavioral Audiences
Geography (Required)
Start with your service area: DMA for broad coverage, zip codes for precision, radius around office, exclusions for areas you don’t serve.
Behavioral Signals
Add relevant behaviors as OR conditions: Auto loan holders, motorcycle owners, workers’ comp researchers, rideshare users, medical facility visitors. Each signal adds relevance without over-narrowing.
Intent (When Available)
Add high-priority segments: Legal services in-market, PI content consumers, competitor site visitors. These get priority impression delivery.
Demographic Guardrails Only
Use demographics as guardrails, not targeting: Age 25-64 (reasonable PI range), HHI $35K+ (ensures economic damages potential). This excludes clearly irrelevant without over-restricting.
The Audience Sizing Balance
Too narrow = insufficient reach:
- Can’t build frequency
- Statistical noise in data
- Optimization impossible
Too broad = demographic waste:
- Paying for irrelevant impressions
- Low efficiency
- High effective CPM
Target: Audience representing 15-30% of DMA households with behavioral relevance.
First-Party Data: The Ultimate Behavioral Signal
Your own data beats any third-party behavioral segment:
Client lookalikes:
- Based on people who actually became clients
- Specific to your practice and market
- Unavailable to competitors
If your clients share characteristics, lookalike targeting finds more households like them.
For building exclusive audiences, see Exclusive CTV Audiences.
Testing Behavioral vs. Demographic
Run a comparison:
Test A: Demographic targeting (Adults 25-54, $50K+) Test B: Behavioral targeting (layered signals)
Measure:
- Verified visit rate
- Cost per verified visit
- Conversion rate
- Cost per lead
Behavioral will likely outperform significantly, proving the concept with your data.
What This Means for Budget
Behavioral targeting means:
- Smaller raw audience
- Higher relevance per impression
- Better frequency against interested households
- Lower cost per result
- Potentially lower total impressions needed
You might need fewer impressions to achieve better results.
For budget implications, see CTV Budget by Market Size.
The Taqtics Approach
Our Behavioral Approach
- + No demographic-only campaigns
- + Layered behavioral signals standard
- + First-party integration when possible
- + Intent targeting included
- + Continuous optimization toward conversion
Demographics are guardrails. Behaviors are targeting.
For the complete targeting framework, see the CTV advertising guide.
References
Principles derived from campaign performance data and industry targeting best practices.