Twenty+ years, four company lifecycles and one new era: What building, selling and scaling companies taught me, and how AI changes everything

I started my career in 2004, and for the last twenty+ years I’ve been building and leading companies across almost every stage you can imagine. Two consumer startups I founded. One acquisition and multi-year integration into a major corporate. A high-growth SaaS scale-up that transformed from a product into a platform. And now, the inflection point that is reshaping the entire landscape: AI as infrastructure.

Different industries. Different customers. Different teams. Different pressures.

But the same underlying patterns, everywhere.

As I look across these chapters, the themes are surprisingly consistent. Companies evolve through predictable stages. Teams break in predictable ways. Leadership has to reinvent itself on a predictable cadence. And when you ignore these patterns, things go sideways quickly.

The only thing that doesn’t follow those old rules anymore is AI. AI rewrites everything. It forces a completely new model of thinking, building and scaling.

This is what the last twenty years taught me and why the next twenty are going to look nothing like the first.

Across every company I’ve been part of - founder, operator, or executive - I saw the same four versions play out. Different contexts, same lifecycle.

Version 1 — Startup

In my first consumer startup, scale meant user numbers, not revenue. No one cared about monetisation. You were optimising for love, not margins.

The second startup was similar: usage, virality, retention. It was about making something people flocked to, not building organisational muscles.

Speed mattered more than structure. Instinct mattered more than process.

Version 2 — Growth engine

As both products grew, the early “move fast” approach began to creak. Coordination cost increased. Priorities needed discipline.

And in the second startup, product-market fit demanded a kind of focus we simply hadn’t needed before.

After those chapters, during the acquisition and integration years, this stage became unmistakable.

The startup I joined had to fit into a corporate ecosystem with a different pace, different priorities and different incentives - the perfect stress test for any scaling model.

Growth makes everything heavier. And everything slower. Unless you redesign the system.

Version 3 — Platform

By the time I was leading a large SaaS scale-up, this pattern was second nature.

We moved from a single product to multiple surfaces and workflows. Cross-functional execution became the heartbeat of the organisation. Customer value came from flows, not features. This stage is where companies stop being products and become ecosystems.

Version 4 — AI-native organisation

This is where we are now. This is the version that didn’t exist earlier in my career but is unavoidable today.

AI isn’t a layer. It isn’t a feature. It isn’t an add-on.

AI is the new organisational infrastructure.

Everything above it - teams, products, workflows, leadership - must be rebuilt with that reality in mind.

In one high-growth chapter, we structured Customer Success around driving expansion revenue because it compensated for slow new business growth.

For a year, it worked brilliantly - until it didn’t. Customers began saying things like: “Whenever I speak to your team, they put a contract in front of me.”

Engagement fell. Adoption slipped. Retention wobbled.

It took less than twelve months for a good idea to become a structural liability. And it taught me one of the most important scaling truths: When the company evolves, the operating model must evolve with it. If it doesn’t, your customers will feel the break before you do.

This became one of the core principles across all my roles: You cannot design an organisation for the stage you are in. You must design it for the stage you’re entering.

I learned this the hard way, repeatedly. Here’s the cycle I wish I’d known earlier:

1. Sense the stage shift

You know it when launches slip, communication slows and customers say, “You’re not listening.”

2. Diagnose what no longer fits

What worked at 10 people won’t work at 50.

What worked with one product won’t work with three.

What worked with one customer segment won’t work with five.

3. Redesign the organisation

Not reactively. Deliberately.

Structure, ownership, communication, focus - the system beneath the work.

4. Stabilise, then rebuild again

When it feels smooth, start rethinking the next version.

There was a stage where launch quality was dropping, communication was breaking down and customers felt unheard.

We had talented people. We had great intentions. We had the wrong structure.

We rebuilt Customer Success entirely around delivery, adoption and professional services - the part of the system customers were actually experiencing. Time-to-value dropped. Adoption rose. Retention stabilised above 95% NRR and 115% GDR.

The lesson was clear: People succeed when systems support them. People struggle when systems fail them.

And across every company I’ve worked in - from tiny startups to global corporates - that has been universally true.

Leadership has to evolve at the same pace as the company - or faster. Across twenty years, this ladder has held up everywhere:

Player: You execute. You deliver. You hold the details.

Coach: You develop others. You multiply output through people.

Architect: You design the systems that determine how work flows.

Strategic Operator: You orchestrate the entire organisation. Your value is clarity, focus and alignment.

If you stay too long on one rung, the company outgrows you. If you climb consistently, you become the leverage the organisation needs.

Every chapter of my career had a messy middle. Different contexts, same feeling:

• too big to be small
• too small to be big
• too fast to be stable
• too complex for the old model to hold

One of the most intense periods came during a quarter with major launches stacked back-to-back, dependencies everywhere, expectations sky-high and very little slack in the system.

It wasn’t a people problem. It was a systemic one.

That messy middle taught me that chaos is almost always a leadership lag indicator:

The system hasn’t caught up with the ambition.

This pattern showed up in both startups, in the corporate integration, and again in the high-growth scale-up. Different companies, same truth.

Across every company I’ve built or led, I eventually learned the same lesson:

Features grow products. Platforms grow companies.

This pattern first emerged in the consumer startups, user behaviour always revealed the limitations of feature-led thinking.

It showed up again during the acquisition and integration. Where incentives, systems and teams were misaligned because the organisation wasn’t built as a platform.

And it was unmistakable in the scale-up years. Where platform-level thinking became the only way to deliver speed, consistency and value across multiple teams.

Two patterns that forced the shift

1. Sales selling the future, not the present In moments of pressure, roadmap promises replaced value articulation. A trap I’ve now seen in every company under stress.

2. Customer Success drowning in feature requests Ninety percent weren’t features they “needed.” They were symptoms of system gaps: adoption, onboarding, clarity. Across every chapter, the truth was the same:

Customers ask for features when the system around the product fails them. Platform thinking solves this. Feature thinking cannot.

For nearly two decades, the same scaling principles applied everywhere. And then AI arrived.

We experimented with chat interfaces, lightweight prototypes and quick LLM layers. And none of them stuck. Not because the tech wasn’t good - but because the approach was wrong.

AI-as-feature is cosmetic. AI-as-infrastructure is transformational.

The tipping point was realising that AI needed to:

• be invisible
• improve workflows by default
• sit beneath the product
• reshape organisational design
• change day-to-day work
• become part of the foundational architecture, not the interface layer

Once that clicked, everything changed. AI shifts companies from:

• human-performed tasks, to
• human-directed systems

It rewrites the operating model.

Here’s the framework that now underpins my thinking:

Layer 1 — Cognitive automation

Summaries, extraction, triage, interpretation.

Layer 2 — Workflow Intelligence

Automated handoffs and coordinated processes.

Layer 3 — Organisational Intelligence

Prediction, prioritisation, scenario modelling.

Layer 4 — Leadership Leverage

Real-time clarity, strategic augmentation and systemic optimisation.

AI isn’t another tool. AI is how work will happen. AI is how companies will scale. AI is the new infrastructure.

This is the shift that ties everything together:

Old model: Scale = headcount × productivity

AI-native model: Scale = systems × leverage × intelligence

This is the structural break between the last twenty years and the next twenty. Companies that design around AI infrastructure will move faster, operate leaner and outperform competitors who try to bolt AI onto old models.

Looking across the last two decades, every chapter revealed the same patterns:

• Startup velocity
• Growth strain
• Platform necessity
• Cultural foundations
• Systemic clarity
• Leadership reinvention

And now, a new era: AI isn’t the next wave. AI is the new foundation.

The work ahead, for me and for companies everywhere, is to build the organisational, technical and cultural systems that operate natively in this new environment.

This isn’t a closing chapter. It’s a widening of scope.

A recognition that the lessons across twenty years of building, selling and scaling companies now meet the reality of the AI era.

I’m excited for what comes next.