Telstra shut down its muru-D accelerator in 2024. Westpac wound back its Reinventure fund. ANZ’s innovation lab underwent yet another restructure. NAB’s innovation team has been reorganised three times in five years.

These aren’t isolated failures. They’re symptoms of a fundamental mismatch between how large Australian corporations approach innovation and what actually produces results.

The Standard Playbook

The corporate innovation lab playbook has been remarkably consistent across organisations and industries for the past decade. It goes roughly like this:

1. CEO reads a book about disruption or attends a conference
2. Board approves an innovation budget (typically $5-15 million per year)
3. Company hires a Chief Innovation Officer or equivalent
4. A physical space is created—open plan, standing desks, whiteboards, exposed brick if possible
5. Teams are assembled to work on “breakthrough” ideas
6. Demo days are held, executives attend, enthusiasm is expressed
7. Nothing reaches production or market
8. Budget gets cut in the next cost review
9. Lab quietly disappears

The cycle typically takes 18-36 months. We’ve watched it repeat across Australian banking, telecommunications, mining, insurance, and retail for over a decade now.

Why It Doesn’t Work

Structural isolation. Innovation labs are typically set up as separate units, physically and organisationally removed from the core business. This feels right—give innovators space to think freely, away from corporate bureaucracy. In practice, isolation means disconnection from the customers, data, distribution channels, and regulatory knowledge that a corporate innovator should be able to access as advantages.

A startup competing with a corporate innovation lab’s project has to build everything from scratch. The corporate lab should have unfair advantages—existing customer relationships, proprietary data, regulatory expertise, distribution at scale. But isolation strips away these advantages, leaving the lab competing like a startup without a startup’s speed or equity incentives.

Wrong success metrics. Most corporate innovation labs measure activity rather than outcomes. Patents filed. Prototypes built. Hackathons held. Startup partnerships formed. These are easy to report to the board. They’re also largely meaningless.

The only metric that matters is whether the innovation lab produced something the company deployed at scale and that generated measurable business value. By that standard, the failure rate across Australian corporate innovation programs is above 85%.

Talent mismatch. Corporate innovation labs tend to attract two types of people: corporate employees who want a break from business-as-usual, and external hires who’ve never worked in a large organisation. Neither group has the combination of entrepreneurial instinct and corporate navigation skills needed to push innovations through a large organisation’s immune system.

The best corporate innovators are people who deeply understand how the parent company works—its decision processes, power structures, risk appetite—and can work within those constraints while pushing boundaries. These people are rare and tend to be more valuable in operational roles, which is where they usually end up.

What Might Work Instead

A few Australian organisations have found approaches that produce better results than the traditional innovation lab model.

Embedded innovation teams. Instead of a separate lab, place small innovation-focused teams within business units, working on problems those units actually face. Suncorp’s approach to insurance product innovation, where development teams sit within the insurance division rather than in a separate innovation function, has produced faster time-to-market than their previous centralised model.

Problem-first mandates. Rather than “go innovate,” give teams specific business problems to solve. “Reduce claims processing time by 50%” is a mandate that can produce innovation. “Explore the future of insurance” is a mandate that produces slide decks.

External partnerships with accountability. Working with external partners—whether startups, consultancies, or custom AI development firms—can inject capability that internal teams lack. But only if the partnership has clear deliverables tied to business outcomes, not just exploration.

Venture building vs venture funding. Several Australian corporates have shifted from funding external startups (which rarely generates strategic value for the parent) to building new ventures internally with startup-like governance. The venture has its own P&L, its own targets, and its own timeline for demonstrating viability.

The Cultural Problem Nobody Wants to Discuss

Underneath the structural issues lies a cultural one. Most large Australian organisations are fundamentally risk-averse. Their incentive structures reward consistency and punish failure. Promotions go to people who deliver reliable results, not to people who try ambitious things that might not work.

Innovation requires tolerance for failure. Not reckless failure—disciplined experimentation where some experiments don’t produce results. But when a failed experiment goes on someone’s performance review as a negative, the rational response is to stop experimenting.

Until incentive structures change, corporate innovation will remain largely performative. The lab will produce demos, the board will nod approvingly, and the core business will continue operating exactly as it did before.

That’s not cynicism. It’s pattern recognition based on a decade of watching this cycle repeat across Australian industry. The organisations that break the pattern will be the ones that treat innovation not as a separate activity but as an integrated capability—measured by outcomes, embedded in operations, and supported by governance structures that tolerate controlled failure.

Most aren’t there yet. Whether they get there before the next disruption cycle is an open question.