What I write about
The Cloud
My work on what organisations choose to own, what they choose to rent, and what they are prepared to depend on.
My take
Cloud sits in the middle of my career arc. I came to it as a founder, after building for the web and ecommerce, and re-engineered my own platform before moving more than 400 clients to the cloud. That taught me that the migration is usually the most visible part and rarely the most consequential. The lasting choices sit in the economics and operating model that follow: what the organisation commits to, which capabilities it chooses not to own, and how much freedom it retains to change course.
I don’t think there is one correct cloud posture. Some workloads justify deep commitment because the economics and capability are compelling. Others need portability, sovereign control, or a credible route out. That is why I treat cloud financial management as an operating discipline, not a procurement exercise revived when a contract comes up for renewal.
The question I keep coming back to is whether the organisation still understands the trade it has made. Cloud can replace fixed infrastructure with on-demand capability and give an organisation room to move, but the trade now reaches far beyond the monthly bill, into commitments, concentration, sovereignty, energy, and technical debt. The work gathered here follows those choices from the first migration business case to the infrastructure an organisation ultimately depends on.
Latest writing (29)
Measuring AI value: A strategic framework for Boards and business leaders
Measuring AI value needs more than total cost of ownership. A strategic framework for Boards, built on what the cloud business case taught me.
Amazon Bedrock launches Custom Model Import
Custom Model Import lets an organisation bring the models it trained on its own data into Amazon Bedrock's managed, secured environment.
AWS launch Meta Llama 3 on Amazon Bedrock
Meta's Llama 3 reached Amazon Bedrock in April 2024: open-weight models with better reasoning, code generation and instruction following, as a managed service.
AWS launch Anthropic Claude 3 Opus on Amazon Bedrock
Claude 3 Opus landed on Amazon Bedrock, making it the first managed service to offer all three Claude 3 models in general availability.
The Executive's Guide to the Digital Workplace
My contribution to the AWS Executive's Guide to the Digital Workplace: what leaders need to weigh when driving workplace digitalisation.
Driving organisational transformation in a downturn
A downturn is a forcing function for transformation, not a reason to pull back. Organisations that sustain the investment emerge with structural advantages.
The CFO is central to any cloud transformation
No cloud transformation succeeds without the CFO at its centre: my interview with CFO Magazine in the Netherlands on why finance leads the move.
IDC Multicloud Events
Cloud strategy, financial management and migration approaches, in one conversation: my fireside chat for IDC Multicloud Events with Cloudreach and Intel.
Why now is not the time to take your foot off the gas
The pandemic accelerated digital transformation for many Boards. A return to normality is not a reason to slow down; the momentum is the advantage.
Accelerating innovation with Cloud Financial Management
Cloud spend becomes innovation only when it is tied to business initiatives: my podcast conversation with Nationwide's Joe Daly on what the discipline changed.
Planning a cloud migration? Here's what you should consider
A cloud migration case that stops at TCO flatters itself. Price the whole move: double-running, decommissioning, and the people who carry it.
Retail Week
Retailers are becoming technology platforms: the Amazon viewpoint I gave Retail Week for its analysis of how, and what it takes.
No articles match that filter. Clear the filter.
The ideas that underpin my writing
Ideas I’ve named and matured writing about Cloud: what each one means, and where it started.
Adaptive Localisation
A strategy of running different AI approaches in different markets, tuned to each market's regulatory, cultural, or competitive conditions, deliberately trading consistency for regional advantage. For a Board, choosing this stance means accepting real complexity costs and being ready to answer why the organisation treats one market's rules differently to another's.
AI Skills Paradox
AI simultaneously threatens to automate certain roles while creating acute talent shortages in others, requiring organisations to prepare workers for both displacement and new opportunities at once.
Cultural Sovereignty
The often-overlooked dimension where American, European and Chinese AI cultures embody different worldviews, disruption, deliberation and harmony, shaping architecture, governance and stakeholder engagement beyond regulation.
Board Briefings
For directors who want to go deeper, my Board Briefings bring the related writing, evidence, and practical questions together in one place.
AI Infrastructure
Compute, energy, and the plumbing beneath enterprise AI
AI Risk
Seeing and pricing the risks AI actually creates
Operating AI
Building the Centre of Excellence that scales adoption
AI Business Cases
Investment cases that survive contact with the Board
The Board in the Machine
Signals
My early reads on infrastructure and technologies that are changing what organisations can build, buy, and depend on, while the picture is still forming.
Post-Quantum Cryptography
Quantum-resistant cryptography standards are published and migration targets set. Data stolen today may be decrypted later, which makes the transition a Board question now.
Quantum Computing
Quantum error correction has crossed important experimental thresholds. The Board question is how to time readiness against evidence, not vendor roadmaps alone.
Agentic AI
Generative models are being given goals, tools, and the authority to act. The Board question is where to transfer agency, and under what limits.
Questions Boards ask about The Cloud
How do we tell whether moving an application to the cloud makes sense?
By pricing the whole journey, not the destination: alongside run-cost comparisons, the assessment must include the migration bubble (the temporary double-running of old and new), the decommissioning cost of what you leave, and the resource plan for the people who carry the move. TCO models that stop at the target-state bill systematically flatter the case; that discipline became the Cloud Business Case vocabulary.
Is technical debt putting a ceiling on our AI ambitions?
Agents are only as capable as the systems they can reach: decades-old data lineage, undocumented logic layers, and brittle integrations cap what any agentic deployment can safely do. If the honest audit shows agents working around the estate rather than through it, the debt is the strategy constraint, and no amount of model capability buys past it.
What hybrid hosting strategies could balance cost, compliance, and control?
For most organisations, the practical answer is workload-tiered placement: sovereignty-critical systems in environments that meet their residency, operational-control, and continuity requirements, and cost-sensitive workloads where the economics run best. The design work is in the classification, and in keeping the boundary movable as those requirements change.
How do we quantify the true cost of AI sovereignty versus dependency?
The comparison has to price more than compute: energy, any premium for stronger sovereignty controls, and the legal, continuity, and long-term commercial exposure created by infrastructure governed elsewhere all belong in the same model. The comparison is incomplete until both the visible premium and the dependency exposure are on the page.
Should we own our compute or rent it, and can we secure the energy?
Securing a long-term energy contract can reduce price volatility, but it does not by itself secure grid access, planning approval, or resilience, and it does not ensure the compute remains useful for as long as the contract runs. Direct ownership makes sense where demand is large and predictable, control or sovereignty is material, and the organisation can keep expensive capacity well used. Cloud makes more sense where demand is variable, access to changing capability matters, or flexibility is worth more than ownership. The practical answer may be a portfolio: own the capacity whose control is strategic, and source the rest through cloud providers. The Board should be able to say which workloads belong in each category, and what trade it is making.