"Grow into the future, or shrink into the past": Accenture Song's Ndidi Oteh on AI, relevance, and the new creative pecking order
AI has moved the bar for "good" creative work. That's the point. The new bar to beat is excellence - and that's still human territory.
Accenture Song's global CEO, Ndidi Oteh, put it plainly at SXSW Sydney: "AI makes a lot of things get to good. But it doesn't make a lot of things excellent." The edge now belongs to those who experiment, upskill, and build the technical foundations to operate at scale.
Translation for creatives and marketers: stop waiting for the cavalry. You are it.
Competing with machines - and winning
Oteh is bullish on human creativity in an AI-saturated market, even as she acknowledges a tougher competitive field. AI will flood the market with passable work. Excellence will differentiate. Context, taste, judgment, and lived experience still matter.
She gave a sharp example: a client compared her answer to Claude. The outputs matched on facts. What she added was lived context - what went wrong on similar projects, how leaders behaved, what actually changed outcomes. That's the human advantage.
The relevance playbook
Oteh's operating system for staying relevant cuts across four dimensions: culture, individual connection, technology as a workbench, and industry relevance. It's practical, not theoretical.
- Run the business: Ideas are cheap without delivery. Get clear on the workflow, talent, and tech that take concepts to market.
- Authentically reinvent: Your customers have changed. So must your product, experience, and brand behavior.
- Stay human-centric: AI automates tasks. It won't automate imagination, purpose, or trust. Design around those.
Stop worshipping nostalgia
The industry loves the way things were. That's a trap. Oteh's take: pride in past work should not slow you down. Be first to build what's next. The teams that matter in the next two years will use AI to do something we haven't seen before - and they'll have the architecture to scale it.
From good to great: what changes now
- "Good" is democratized: Anyone can ship decent copy, images, and edits. Quality alone is no longer a moat.
- Excellence is curated: Taste, insight, and context separate brands that stick from content that scrolls by.
- Curation is a job: As Manon Dave put it, being a curator is as important as being a creator - and often harder.
Tasks vs outcomes
Much of the fear around AI is because teams fixate on tasks, not outcomes. Writing a cheque isn't the goal - accessing funds is. Same with marketing: the goal is growth, loyalty, and signal in the market, not the volume of assets produced.
Expect fewer roles centered on repetitive tasks. Expect higher demand for roles that define outcomes, architect systems, and direct AI toward measurable results.
"Become the cavalry" - what to do next
- Adopt AI as your workbench: Integrate models into research, concepting, iteration, and production. Keep humans in the loop for taste and judgment.
- Build for scale: Don't stop at pilots. Define your technical architecture, data flows, model governance, and review layers. Otherwise, it's "just an ad."
- Train for relevance: Upskill in prompt design, agent workflows, and creative QA. If you lead, make it safe to experiment and ship.
- Curate aggressively: Set brand taste. Codify references, constraints, and examples of "excellent" so teams and models align.
- Measure outcomes: Tie AI-assisted work to CAC, LTV, brand lift, and speed-to-market. Kill vanity metrics.
Leadership notes
Oteh steps in after a period of significant growth, with Accenture Song consolidated and scaled under prior leadership. The mandate now: move faster, build inclusively, and use AI to create work that couldn't exist before - not just produce more of what already exists.
A simple checklist for creative teams
- Define "excellent" for your brand with examples - and what "good" looks like so you don't settle.
- Map the creative supply chain. Mark tasks for automation, augmentation, and human-only judgment.
- Stand up an AI QA process: originality, bias checks, IP safety, and brand voice alignment.
- Create a live library of curated prompts, references, and do/don't rules.
- Ship one AI-enabled experiment per quarter that impacts a real KPI.
The hard truth
Some jobs will shrink. That's real. But new roles are emerging around curation, orchestration, and system-level thinking. The ones who learn, build, and lead will have more options - and more influence.
As Oteh puts it: "You either grow into the future, or shrink into the past." The market will reward those who choose growth.
Resources
Your membership also unlocks:
