Agencies are not obsolete. They are often excellent at long-horizon brand work. The problem is that many performance-driven teams now face a different operational reality, and the classic agency model is not optimized for that reality.
The real conversation is not agency versus marketplace. It is task-model fit.
What agencies still do better
Let us be fair first. Strong agencies deliver value in areas that crowdsourced systems do not replace:
- integrated brand strategy
- cross-channel campaign coherence
- complex production orchestration
- long-term stewardship
If you need a flagship brand platform, a high-stakes TV spot, or enterprise stakeholder alignment, agencies can be the right tool.
Where the model strains
Now the pressure point: high-volume, performance-focused digital creative.
That operating mode demands:
- lots of concept variation
- rapid cycle time
- frequent iteration
- lower cost per asset
The agency structure, retainers, account layers, review loops, and finite creative bandwidth, can make this expensive and slow at scale.
This is not a talent problem. It is a system design problem.
The economics mismatch
For many growth teams, unit economics are the deciding factor.
If an engagement produces 5 to 10 assets for $10K to $20K, cost per asset remains high. That can be acceptable for hero creative, but difficult for high-frequency performance testing.
A campaign model that yields 20 to 60 submissions at lower total budget shifts the math dramatically.
Agency operating profile
- $15,000/month retainer
- 3-4 week cycle per major set
- 5-10 assets typical output window
- Selection by internal team review
- Pre-market validation usually separate or absent
Cost per asset
$1,500 to $3,000
Swayze campaign profile
- $5,000 budget + fees
- 1-2 week cycle depending on setup
- 20-60 submissions from creators
- Selection through community voting signal
- Validation integrated in process
Cost per submission
$83 to $250
Illustrative comparison, exact economics vary by campaign scale and participation.
Quality tradeoffs, honestly
Agency output often has stronger strategic coherence because a concentrated team internalizes your brand deeply.
Crowdsourced output tends to be:
- more diverse
- more surprising
- occasionally inconsistent
That inconsistency is not a bug in all cases. It can be the price of discovery.
For performance channels, discovery often matters more than polished uniformity.
The "and" strategy beats the "or" debate
High-performing brands increasingly use both models:
- agencies for strategic architecture and major brand campaigns
- creator marketplaces for high-velocity performance creative and pre-market selection
This is not hedge behavior. It is specialization.
Trying to force one model to do every job usually increases cost and lowers speed.
A decision framework for brand teams
Ask four questions before choosing channel:
- Is the objective strategic cohesion or rapid creative throughput?
- How many fresh concepts are needed this cycle?
- How fast must learning loops run?
- Is pre-launch validation required before spend scales?
If answers point to high volume and high speed, campaign marketplaces become more compelling.
Why old assumptions persist
Many teams still anchor to the "hero ad" era where one big concept could carry a quarter. Today, platform fragmentation and feed dynamics reward continuous testing and faster adaptation.
System design must follow media reality.
A model built for infrequent big swings will struggle in a world of persistent micro-optimization.
Final thought
The case against agencies is not that they are bad. It is that they are sometimes the wrong instrument for a specific class of modern performance problems.
The winning stack is modular: strategy where strategy is needed, scale where scale is needed, and validation where risk is highest.
Add a high-velocity creative lane
Use Swayze campaigns for diverse concept volume and community validation, while keeping your agency for long-horizon brand work.