Silicon Valley GirlThe 2026 Business Playbook: Leverage AI Before Your Competitors Do | Epidemic Sound CEO
CHAPTERS
- 0:00 – 1:02
Epidemic Sound at scale: viral reach, value chains, and the cost of success
Marina introduces Oscar Höglund and frames the conversation around two themes: how content goes viral (and how Epidemic Sound influences it) and how to build a company without burning out. Oscar previews core ideas—understanding value chains for leverage, and the personal breaking point that forced him to redesign his life and leadership.
- •Oscar’s role and context: CEO/co-founder of Epidemic Sound; music behind billions of daily views
- •Core business lens: leverage comes from understanding value chains
- •Tease of AI-driven transformation in music + content workflows
- •The personal cost of scaling: near-quitting moment and rebuilding after it
- 1:02 – 3:58
What makes content go viral: emotion, human connection, and audio as storytelling power
Oscar explains that virality isn’t a hack; it’s about creating meaningful human connection, with music acting as emotional infrastructure. He argues that relying on famous songs can reduce a creator’s control because audience associations vary widely.
- •Virality rooted in sparking emotion and connection, not just trends
- •Music’s job is to shape feeling; it’s part of the narrative
- •Using well-known commercial music can “abdicate” storytelling control
- •Epidemic’s advantage: observing cross-platform patterns at massive volume
- 3:58 – 7:47
Trend insight: post-COVID “comfort music” and why genres surge
Using Epidemic’s dataset, Oscar shares an example of how global mood shows up in music choices. After COVID, creators and audiences gravitated toward warm, reassuring tracks—classical music spiked as “comfort food” for the ears.
- •Macro events influence audience emotion and content tone
- •“Comfort food” analogy applied to soundtracking and mood
- •Calming/warm/classical music rose as reassurance content grew
- •Data at scale reveals emotional cycles, not just aesthetic trends
- 7:47 – 11:09
The “CROSS” strategy: cognitive dissonance that pulls attention
Oscar tells the story behind “crosses”: people are drawn to surprising combinations (a mismatch between expectation and reality). He connects this to content and music—when audio and visuals create either intense harmony or deliberate contrast, engagement jumps.
- •Crosses = unexpected pairing that creates cognitive dissonance
- •Tabloid example: celebrity success vs hidden struggle draws attention
- •TV example: “anti-commercial” host fronting the most commercial show
- •In content: audio can create a second story that makes visuals pop
- 11:09 – 12:20
AI content revolution at Epidemic: democratizing “music supervision”
Marina probes how AI can help creators with music progression and editing decisions. Oscar describes a vision where solo creators get access to tools once reserved for expensive production teams—recommendations, sound effects, mastering guidance, and performance insights.
- •Goal: level the playing field between solo creators and big production teams
- •AI as unlock for guidance (not replacing human connection)
- •Data-driven recommendations: what works, where engagement happens, platform trends
- •Creator personalization improves as the system learns a channel’s direction
- 12:20 – 13:50
Creator workflow tools already shipping: AI Voice and adaptive track length
Oscar outlines practical tools Epidemic has begun releasing. A major breakthrough is adapting music to the edit (changing track length) instead of forcing creators to re-cut their story around a fixed track duration.
- •Tools released: AI Voice and other creator aids
- •Key innovation: adjust track length to fit a finished video edit
- •Reverses the traditional workflow (edit-to-track becomes track-to-edit)
- •Protects the story structure while making music fit naturally
- 13:50 – 20:18
A 2026-ready content framework: define your sound, then iterate with data
Oscar gives creators a repeatable approach: start with identity—values, tone, and the emotional signature you want—then use tools and feedback loops to refine. He emphasizes being a “ferocious learner” by studying creators who achieve the emotions you want to evoke.
- •Start with brand sound: values, core story, emotional intent
- •Use platform tools for suggestions and benchmarking against peers
- •Personalization improves through repeated interaction and feedback
- •“Ferocious learner” mindset: study models, reverse-engineer emotional craft
- 20:18 – 23:03
Sponsor segment: Pop.Store’s ECHO AI assistant for managing comments and DMs
Marina shares a tool recommendation aimed at another creator bottleneck: audience engagement. Pop.Store’s AI assistant scans comments/DMs, flags important messages, and helps reply faster to increase interaction and reach.
- •Creator pain point: time spent in comments and DMs
- •ECHO scans/triages messages across Instagram and Facebook
- •Claimed outcome: faster replies and higher engagement signals for algorithms
- •Positioning: an “all-in-one creator HQ” that reduces burnout
- 23:03 – 23:59
From feature → product → platform: how Epidemic Sound evolved
Oscar breaks down a scaling model: start with a narrow feature, expand into a must-have product, then open infrastructure into a platform. He explains Epidemic’s initial legal innovation (ownership model) and how software tooling turned music licensing into an integrated workflow.
- •Early stage = feature: a “vitamin,” not a necessity
- •Legal innovation: shift from rights fragmentation/representation to ownership
- •Pay artists upfront, acquire rights, and indemnify creators globally
- •Product stage: workflow tooling + creator integration; millions of users and major YouTubers
- 23:59 – 27:51
From platform moment to “Antidote”: opening internal infrastructure like AWS
Oscar describes Epidemic’s “Amazon moment”: internal data, tools, and expertise became valuable enough to externalize. He outlines a platform vision (Antidote) to connect storytellers, artists, and IP using Epidemic’s unique distribution and performance insights.
- •Analogy: AWS emerged from Amazon’s internal scaling infrastructure
- •Epidemic’s edge: data on how music travels online and affects outcomes
- •Platform ambition: connect creators and rights holders via shared infrastructure
- •Repositioning from catalog provider to ecosystem enabler
- 27:51 – 29:30
Value chain leverage: how to scale distribution by targeting the real decision-makers
Oscar gives a concrete go-to-market lesson: map the ecosystem to find leverage points. Instead of chasing thousands of individual editors or dozens of production companies, Epidemic sold to a handful of broadcasters who controlled budgets—and distribution cascaded downstream automatically.
- •Value chain mapping reveals who truly controls access and revenue
- •Sweden example: 5,000 editors → 50 production companies → 4 broadcasters
- •Close a few high-leverage deals, and adoption propagates downstream
- •Leverage definition: small input producing outsized output
- 29:30 – 31:10
Hiring A-players: build rooms where you’re not the smartest person
Marina highlights the team’s impressive pedigree, and Oscar shares his hiring philosophy. He seeks people smarter than himself and values “cross” profiles—unexpected combinations that create strength and originality.
- •Principle: hire people smarter than you; otherwise “we’re doomed”
- •Use hiring to multiply judgment and capability, not validate the CEO
- •Look for “cross” talent—unusual blends of skills/backgrounds
- •Team quality as a strategic advantage for platform-scale ambitions
- 31:10 – 33:04
Why hustle fails (as a lifestyle): the real success stack—talent, grit, and being a net giver
Oscar adds nuance: success does require relentlessness, but not nonstop chaos or dishonest “hustle.” He proposes a four-part model emphasizing grit and reputation with gatekeepers—winning by being someone others want to help.
- •Talent matters least; it’s only the ticket to play (relative edge)
- •Grit/relentlessness is chosen and often required for years
- •Industries have gatekeepers—your reputation determines access and lift-off
- •Operate as a “net provider”: give more than you take to compound goodwill
- 33:04 – 34:33
Network strategy in practice: ‘How can I help?’ and high-quality introductions
Oscar makes “giving” concrete through a repeatable networking tactic: proactive helpfulness and thoughtful intros. He explains how well-crafted introductions create durable positive-sum networks where others endorse you later in unexpected ways.
- •Default opener: “How can I help?” to prime generosity and reciprocity
- •Introductions as leverage: do fewer, but do them exceptionally well
- •Use brief research (even with ChatGPT) to craft compelling two-way context
- •Compounding effect: many indirect advocates across industries over time
- 34:33 – 35:30
Staying sane as CEO: discipline, guardrails, and the breaking point that forced change
Oscar answers Marina’s core question with one word: discipline. He shares strict boundaries (no early meetings, unreachable family time, no weekend work) and the painful moment when constant travel nearly ended his marriage—leading to a redesign of his role and priorities.
- •Discipline as the “unsexy” foundation for longevity and performance
- •Guardrails: no meetings before 9am, protected 6–9pm family time, no weekends
- •Travel constrained to weekdays; real summer vacations prioritized
- •Breaking point: spouse said life was better when he wasn’t there; he nearly quit and restructured everything
- 35:30 – 38:27
Top AI tools and the ‘Three WHY’ framework for building with AI
Oscar shares his primary AI stack—ChatGPT and Google/Gemini—focusing on deep usage so tools learn his context. He closes with a practical analytical method from consulting: ask “why” three times to reach root causes, then use AI to accelerate solutions.
- •Primary tools: ChatGPT + Google/Gemini suites as daily assistants
- •View: invest time so tools understand you and deliver compounding value
- •‘Three WHY’ = minimum bar for real analysis and root-cause clarity
- •AI advantage: speed from diagnosis to execution once the true problem is identified
