“What’s the best practice for this?” It’s probably the most common question asked in business meetings, strategy sessions, and project planning discussions. And increasingly, it’s the first thing we ask AI tools when we’re looking for solutions.
But here’s the uncomfortable truth: if everything AI generates is based on “best practice,” we’re systematically eliminating the possibility of breakthrough innovation.
The best practice trap
By definition, best practice is what we used to do. It’s the accumulated wisdom of past successes, codified into repeatable processes and approaches. Best practices are incredibly valuable—they help us avoid reinventing the wheel, reduce risk, and build on proven foundations.
But they’re also, by their very nature, backward-looking.
When we ask an AI tool to “write a marketing strategy using best practices,” we’re essentially asking it to synthesise what has worked before. The AI scours its training data, identifies patterns from successful campaigns, and produces something that looks professional, comprehensive, and safe.
What we get is competent. What we don’t get is groundbreaking.
The innovation paradox
Innovation, by definition, requires doing something that hasn’t been done before—or at least, doing it in a way that’s meaningfully different. The most successful companies, products, and campaigns often succeed precisely because they broke best practice.
Consider some of the most innovative moments in recent business history:
- Netflix didn’t follow the best practice of physical video rental stores
- Airbnb ignored the best practice of professional hospitality management
- Tesla abandoned the best practice of traditional automotive manufacturing and sales
- Dollar Shave Club threw out the best practice of premium razor marketing
Each of these companies succeeded by identifying where best practice was actually holding their industries back, then doing something radically different.
The AI amplification effect
AI tools are incredibly good at pattern recognition and synthesis. They can analyse thousands of examples and distil them into coherent, well-structured outputs. This makes them excellent for:
- Creating solid, professional-looking strategies
- Generating comprehensive project plans
- Producing well-researched content
- Developing technically sound solutions
But this same strength becomes a weakness when innovation is the goal. AI tools are trained on what has already been done, thought, and published. They’re optimised to produce outputs that feel familiar and credible—which is the opposite of what breakthrough innovation requires.
The AI slop problem
There’s another, more insidious issue emerging: AI slop. This is the term for the flood of AI-generated content that’s increasingly filling the internet—bland, generic, technically correct but creatively bankrupt material that all starts to sound eerily similar.
We’re seeing it everywhere:
- Blog posts that hit all the SEO keywords but say nothing new
- Marketing copy that’s grammatically perfect but emotionally flat
- Business strategies that cover all the bases but lack genuine insight
- Product descriptions that are comprehensive but completely forgettable
The problem isn’t just that this content is mediocre—it’s that it’s becoming the new baseline. When everyone uses AI to generate “best practice” content, the internet becomes an echo chamber of recycled ideas, each iteration slightly more generic than the last.
The model collapse risk
Here’s where it gets really concerning: as AI-generated content proliferates, future AI models are increasingly being trained on data that includes this synthetic content. This creates a feedback loop that researchers call “model collapse.”
Imagine a photocopier making copies of copies. Each generation loses a bit of fidelity, and eventually, you end up with something that barely resembles the original. The same thing can happen with AI models trained on increasingly synthetic data—they start to lose the diversity and creativity that made the original training data valuable.
While this technical challenge will likely be solved through better data curation and training techniques, it illustrates a deeper problem: when we rely too heavily on AI to generate our ideas, we risk creating a world where genuine innovation becomes increasingly rare.
The more we ask AI for “best practice” solutions, the more we contribute to a homogenisation of thinking that makes breakthrough innovation not just harder, but potentially impossible.
Where expertise becomes essential
This is where human expertise and structured innovation processes become not just valuable, but essential. True innovation requires:
1. Understanding the “why” behind best practices
Before you can break rules effectively, you need to understand why they exist. An expert doesn’t just know what the best practices are—they understand the underlying principles, constraints, and assumptions that created them.
When those assumptions change (new technology, shifting customer behaviour, regulatory changes), the best practices built on them may no longer be optimal.
2. Identifying constraint opportunities
Innovation often happens at the intersection of constraints. The best practices in your industry might exist because of limitations that no longer apply:
- Technical constraints that new technology has solved
- Economic constraints that new business models have overcome
- Regulatory constraints that have been updated or removed
- Customer behaviour constraints that have shifted
3. Systematic experimentation
Rather than asking “What’s the best practice?”, innovative organisations ask “What if we tried…?” They create structured processes for:
- Hypothesis generation: What assumptions could we challenge?
- Rapid prototyping: How can we test new approaches quickly and cheaply?
- Learning loops: How do we capture insights from both successes and failures?
- Scaling decisions: When do we double down on something new vs. returning to proven approaches?
A framework for innovation beyond best practice
At Versantus, we’ve developed a process that balances the safety of best practice with the potential of breakthrough thinking:
Phase 1: Map the landscape
- Understand current best practices and why they exist
- Identify the assumptions and constraints that shaped them
- Map where those constraints might be changing
Phase 2: Challenge assumptions
- What if the opposite were true?
- What if we removed this constraint entirely?
- What if we optimised for a completely different outcome?
Phase 3: Rapid experimentation
- Create small, low-risk tests of new approaches
- Use AI tools to rapidly prototype and iterate
- Measure both traditional metrics and innovation indicators
Phase 4: Intelligent scaling
- Know when to persist with something new vs. returning to proven approaches
- Build new best practices from successful innovations
- Share learnings to elevate the entire organisation
The role of AI in innovation
This doesn’t mean AI has no place in innovation. When used thoughtfully, AI can be a powerful innovation accelerator:
Use AI for rapid iteration: Generate multiple variations of an idea quickly, then apply human judgment to identify the most promising directions.
Use AI for research synthesis: Let AI help you understand the current landscape faster, so you can spend more time thinking about what comes next.
Use AI for scenario planning: Explore “what if” scenarios and their implications more thoroughly than would be practical manually.
Use AI for execution: Once you’ve identified an innovative approach, use AI to help implement it efficiently.
The expertise imperative
The organisations that will thrive in the AI era won’t be those that follow best practices most efficiently. They’ll be those that can systematically identify where best practices are holding them back and create new approaches that others will eventually copy.
This requires a different kind of expertise—not just knowing what works, but understanding why it works, when it might stop working, and how to create something better.
It requires people who can ask the right questions, not just generate the right answers.
Beyond the comfort zone
Best practices are comfortable. They’re safe. They’re defensible in board meetings and strategy reviews. “We followed industry best practice” is rarely a career-limiting statement.
But comfort is the enemy of innovation.
The most successful organisations we work with have learned to be systematically uncomfortable. They use best practices as a starting point, not an ending point. They ask “What if we’re wrong about this?” as often as they ask “What’s the best way to do this?”
They understand that in a world where AI can generate competent, best-practice solutions instantly, the real competitive advantage lies in the ability to think beyond what’s already been done.
The bottom line
Be careful what you wish for when you ask for best practice. You might get exactly what you asked for—and nothing more.
As AI-generated content floods the internet and models risk being trained on increasingly synthetic data, the organisations that stand out will be those that resist the pull of algorithmic mediocrity. They’ll be the ones that understand the difference between competent execution and genuine innovation.
The future belongs to organisations that can balance the efficiency of proven approaches with the courage to venture into uncharted territory. That requires human expertise, structured innovation processes, and the wisdom to know when to follow the rules and when to break them.
AI can help you execute better. But it can’t help you imagine what doesn’t exist yet.
That’s still a uniquely human superpower—and it’s becoming more valuable every day.
—
Ready to move beyond best practice and unlock breakthrough innovation? Contact our strategy team to discuss how we can help you identify and test new approaches that could transform your industry.