Between 2003 and 2018, I was part of the senior leadership team at IG Group that grew the company from a £300m valuation to £2.7bn. I oversaw an engineering department that expanded from 4 people to 250. We built the first online trading platform, the first mobile trading platform, and the world's first Apple Watch trading app. None of these assets — the technology, the engineering culture, the customer relationships they created — appeared on IG's balance sheet.
That experience is why I started Opagio. And it is directly relevant to every company investing in AI today.
Every significant AI investment creates or enhances the same categories of intangible assets that drove IG's transformation. The challenge is that most companies cannot see them, do not measure them, and therefore cannot manage or value them. When 90% of firms report zero measurable AI productivity impact (NBER, February 2026), the question is not whether AI is creating value — it is whether anyone is looking in the right place.
The Corrado-Hulten-Sichel framework identifies the major categories of intangible capital investment. When I map AI investment against this framework, six categories emerge consistently. I can illustrate each one from direct experience, because I built every one of them at IG Group — before AI made the process faster, cheaper, and more scalable.
At IG Group, we built proprietary trading platforms that could handle real-time streaming prices and execute trades in under a second. The technology was a genuine competitive moat — competitors could not replicate 15 years of accumulated engineering decisions, architecture choices, and performance optimisation.
AI accelerates technology capital creation dramatically. Companies using AI code assistants are building software 30-55% faster according to multiple studies. But the intangible asset is not the speed — it is what the speed enables. Faster development cycles mean more rapid iteration, earlier product-market fit, and deeper technical moats built in less time.
Technology capital created by AI includes proprietary models, custom training data pipelines, fine-tuned systems, API integrations, and automated workflows. Under IAS 38, most of this investment is expensed immediately. It does not appear on the balance sheet, but it drives enterprise value.
What to measure: Number of proprietary AI systems deployed, custom model performance relative to generic alternatives, integration depth (how many business processes depend on AI systems), and replacement cost — what it would cost a competitor to replicate your AI technology stack from scratch.
IG Group generated enormous volumes of trading data — customer behaviour patterns, market data, execution analytics. This data made our platforms better with every trade. We used it to improve execution, personalise the user experience, and identify opportunities for new products. The data itself was an asset of immense value that never appeared in the financial statements.
AI transforms data from a passive byproduct into an active asset. Every AI system ingests data, learns from it, and creates new data through its outputs. The training data, the fine-tuning datasets, the feedback loops — these are compounding assets. They become more valuable over time, and they create switching costs that make it increasingly difficult for competitors to catch up.
A professional services firm that fine-tunes an AI system on five years of client engagement data has created a data asset that no competitor can replicate. The AI's recommendations improve with each engagement, and the accumulated data represents a genuine competitive moat — even though the balance sheet records it as zero.
What to measure: Unique data volume and quality, data refresh rate (how quickly your data asset grows), competitive exclusivity (is this data available to anyone else?), and the performance differential between AI systems trained on your data versus generic alternatives.
This is the intangible asset I care about most deeply, because I watched it create extraordinary value at IG Group and saw how invisible it remained to outside observers.
Organisational capital encompasses the management practices, processes, and institutional knowledge that determine how effectively a company operates. At IG, we deployed agile methodologies before "agile" became a named methodology. We evolved from weekly Friday-night deployments to continuous deployment with zero downtime. We built career pathways, competency frameworks, and an engineering culture that attracted and retained exceptional talent.
AI is transforming organisational capital in ways most companies do not yet recognise. When a company deploys AI across its operations, it is not simply automating tasks — it is encoding institutional knowledge into systems. The prompts, the workflows, the decision trees, the quality standards embedded in AI systems — these represent organisational capital in a new form.
Companies that invest most heavily in AI-driven organisational capital — automated workflows, AI-augmented decision-making, knowledge management systems — often appear less productive in the short term as teams learn new ways of working. This is the J-curve effect that the Solow Paradox describes. The value appears later, and in categories that traditional metrics cannot capture.
What to measure: Process automation depth (percentage of workflows that incorporate AI), knowledge codification rate (how much institutional knowledge is captured in AI systems versus individual expertise), and decision quality improvements — measured by outcomes, not by volume.
At IG, our mobile trading apps became massive customer acquisition platforms. The user experience was exceptional — real-time streaming prices honoured upon receipt were famous in City trading circles. Customers did not just use IG; they recommended it. The brand trust and customer evangelism this created was an intangible asset worth billions.
AI enables a step-change in customer relationship value. Personalised recommendations, proactive support, intelligent onboarding, and conversational interfaces all deepen customer relationships in ways that create measurable switching costs. The AI learns each customer's preferences, history, and needs — building a relationship asset that becomes more valuable with every interaction.
Customer relationships enhanced by AI create a double lock-in: the customer benefits from increasingly personalised service, and the company benefits from increasingly accurate predictions about customer needs. This mutual value creation is precisely the kind of intangible asset that traditional accounting cannot capture.
What to measure: Customer acquisition cost trends (AI should reduce CAC over time), retention rate improvements, net revenue retention (expansion revenue within existing accounts), and the lifetime value differential between AI-served and traditionally-served customers.
When IG launched the world's first Apple Watch trading app, Apple featured our logo at their launch event. That was not a marketing spend — it was a brand asset created through technology excellence. The recognition positioned IG as an innovation leader in financial services, attracting both customers and talent.
AI investments create brand value in two distinct ways. First, companies that deploy AI effectively build a reputation for innovation and operational excellence that attracts customers, talent, and investors. Second, AI-generated content, personalised experiences, and intelligent interactions become part of the brand experience itself.
What to measure: Brand perception surveys (before and after AI deployment), talent acquisition metrics (application volume, offer acceptance rates), media sentiment, and — critically — whether your AI claims are substantive or performative. The SEC is increasingly scrutinising AI-washing, and hollow claims destroy brand value faster than genuine capability builds it.
This is where the compounding power of intangible assets becomes most visible. At IG, every new client made the platform more liquid, which attracted more clients. Every new market we offered attracted new segments, who then demanded additional markets. The network effects were powerful, self-reinforcing, and — of course — invisible on the balance sheet.
AI amplifies network effects in several ways. AI-powered marketplaces learn from every transaction, improving matching and recommendations. AI platforms become more valuable as more users contribute data and feedback. And AI-generated insights create sharing incentives that drive organic growth.
The most valuable AI investments are those that create self-reinforcing loops: more data leads to better AI, which leads to better outcomes, which attracts more users, which generates more data. This flywheel effect is the most powerful intangible asset a company can build — and the hardest to replicate.
What to measure: User growth rate and its relationship to AI capability improvements, data network effects (does more data measurably improve AI performance?), and platform stickiness — the extent to which users become more engaged over time as AI learns their preferences.
The fact that these six assets are invisible to traditional accounting does not mean they are unmeasurable. For each asset, the framework requires three inputs:
| Asset Category | Investment Proxy | Output Metric | Replacement Cost |
|---|---|---|---|
| Technology Capital | AI R&D + infrastructure spend | Proprietary system performance vs. generic | Cost to rebuild from scratch |
| Data Assets | Data collection + curation cost | AI performance with your data vs. public data | Time and cost to recreate dataset |
| Organisational Capital | Process automation investment | Decision quality and speed improvements | Time to train new team to equivalent capability |
| Customer Relationships | AI-enhanced CRM + personalisation spend | LTV/CAC improvement, net revenue retention | Customer re-acquisition cost |
| Brand & Reputation | Innovation investment with market impact | Talent and customer acquisition premium | Brand rebuilding timeline and cost |
| Network Effects | Platform AI investment | User growth compounding rate | Competitive moat durability |
Replacement cost is a valuation concept from IAS 38 and the relief-from-royalty method. It asks: what would it cost a third party to recreate this asset from nothing? For AI-created intangible assets, replacement costs are often significantly higher than the original investment — because they include the data accumulated over time, the iterations that improved the model, and the institutional knowledge embedded in the system.
The AI investment cycle is following the same pattern I observed at IG Group. Companies are spending heavily on AI — global AI infrastructure investment exceeded $100 billion in 2025. CFOs are asking where the return is. Boards are scrutinising AI budgets. And the measurement frameworks being used to evaluate these investments are fundamentally inadequate.
If you measure AI ROI using traditional metrics — revenue per employee, cost reduction, task automation rates — you will miss most of the value being created. These metrics capture the first-order effects (task efficiency) while ignoring the second-order effects (intangible asset creation) that ultimately drive enterprise value.
Every AI project creates or enhances at least one of the six intangible asset types. Identify which categories each project contributes to.
Before measuring improvement, you need to know your starting position. Use the metrics in the table above as a starting framework.
Intangible assets compound. Data becomes more valuable. Network effects strengthen. Organisational knowledge deepens. Measure trajectories, not snapshots.
If a competitor wanted to replicate your AI capability from nothing, what would it cost them? This is the minimum value of the intangible assets you have created.
At IG Group, the City questioned our technology cost base for years. But the revenue growth that followed — driven entirely by the intangible assets we had built — vindicated every pound invested. The companies that measure their AI-created intangible assets today will be the ones that can demonstrate their true enterprise value to investors, acquirers, and lenders tomorrow.
The Opagio Growth Platform provides the tools to measure, track, and value these six intangible asset categories — turning invisible AI value creation into structured, defensible evidence that CFOs, boards, and investors can act on.
Ivan Gowan is the Founder and CEO of Opagio. He spent 15 years at IG Group (LSE: IGG), overseeing engineering growth from 4 to 250 people during the company's rise from £300m to £2.7bn. He built IG's first online (2003) and mobile (2004) trading platforms and launched the world's first Apple Watch trading app. MSc Edinburgh with neural networks research (2001).
National accounts now recognise R&D and software as capital assets. But the majority of intangible investment — organisational know-how, proprietary data, trained workforces, customer networks — still falls outside the measurement boundary. These are the seven categories that distort our understanding of productivity.
Read more →
Only 29% of executives can measure AI ROI confidently. The problem is not that AI fails to deliver value — it is that traditional ROI frameworks were not designed for investments that create intangible assets. Here is the 4-layer framework that connects AI spending to board-level outcomes.
Read more →
Global AI investment has surpassed $500 billion annually, yet total factor productivity growth remains stubbornly flat. This is Solow's paradox reborn: you can see AI everywhere except in the productivity statistics. History tells us why — and what needs to change.
Read more →Get the latest insights on intangible asset growth and productivity delivered to your inbox.
Book a free consultation to see how the Opagio Growth Platform can help your business.