Collect signals from websites, content, sources and brand mentions.
Research by FunkyMEDIA
Evidence behind our AI Search and Brand Mentions work.
Research is not a separate service. It is the evidence layer behind the way FunkyMEDIA analyses brands, external sources and changing search behaviour.
We test assumptions before turning them into recommendations.
RESEARCH AT A GLANCE
10 yearsproprietary data
600,000+articles analysed
38,509outgoing references
4,315unique target domains
Long-term observation helps us separate one-off patterns from recurring signals.
Why we research
Because visibility claims are easy. Evidence is harder.
AI Search and Brand Mentions are still developing fields. Terminology changes quickly, platforms behave differently and simple cause-and-effect claims are often unreliable.
Our research gives us a stronger basis for deciding what deserves attention, what needs testing and what should remain only a hypothesis.
Research #1
How often do the same domains recur across independent sources?
We analysed outgoing references across an 11-source sample to observe recurrence and source overlap.
38,509outgoing links / references
23,375source pages / documents
15,931unique target URLs
4,315unique target domains
2,697domains in 2+ sources
→
918in 5+ sources
→
200in 8+ sources
→
87in 9+ sources
→
11in 10+ sources
→
2across all 11 sources
OBSERVED PATTERN
Repeated presence across independent sources is rarer than simple one-off presence.
We use recurrence as one descriptive signal when studying how brands and references appear across different source environments.
Research #2
Large-scale AI-assisted content experiment.
March 2023 — June 2024
600,000+
articles analysed
~198,000
visits observed on the record day
This observation is not presented as universal causation or as a publishing-volume recommendation. The experiment is used to study patterns at scale.
How evidence enters client work
Research is useful only when it changes a decision.
Look for recurrence, differences, context and relationships.
Separate measured facts from FunkyMEDIA interpretation and working hypotheses.
Use the evidence to prioritise audits, strategy, implementation and measurement.
Research limitations
We do not turn observations into universal rules.
Observed signal ≠ guaranteed outcome
Patterns can inform decisions without proving that the same result will occur for every brand.
Sample ≠ the entire web
Research samples describe the analysed material, not every possible source or market.
Correlation ≠ causation
Where causal evidence is unavailable, we keep the interpretation conservative.
Missing data ≠ negative evidence
An absent signal is not automatically treated as proof that something is weak or wrong.
Public Research Lab
The deeper research lives on FunkyMEDIA.space.
FunkyMEDIA.eu stays focused on the agency and its services. Detailed datasets, experiments, case studies and research materials belong in the Research Lab.
From evidence to action
Research should make the next decision clearer.
If you want to know what the evidence means for your brand, start with an audit or review our Services.