Most conversations about AI start in the same place. They start with the technology, then look for an industry to point it at.
We start at the other end, with the person who has spent years inside an industry and knows exactly which part of it is broken or just can be better. AI is the multiplier. Domain expertise is what you multiply.
That is straightforward to put on a website. It is harder to defend in front of 6,000 people who build businesses for a living. So on 9 and 10 September we took it to Ideas Fest at Champneys Tring and argued it out on stage.
Craig Donaldson, our Executive Chair, took the business-building half. He was the founding chief executive of Metro Bank and took it to a billion-pound plus listing, so the position comes with receipts. His case was the unfashionable one: revenue discipline, the true cost of acquiring a customer, and the difference between money you have earned and money somebody has given you.
Glen Robinson took the technology half. Glen is Managing Director of Transform and C4VC, our Centre for Value Creation, and spent years as National Technology Officer at Microsoft UK advising governments and large companies on technology strategy and regulation. His thread was Blockbuster and Netflix.
The familiar version is that Blockbuster missed streaming. Glen's version is less comfortable. Blockbuster saw it. They had the chance to buy Netflix in 2000 and turned it down. What they were missing was not foresight. It was the technical capability, the operating model, and the conviction to act on something they already knew.
Seeing it was never the hard part. Blockbuster knew. Most incumbents know. The distance between knowing and acting is made up of their operating model, which makes it a build and decision making problem rather than an insight problem.
The technology is the cheapest part of the stack. Models are commoditising quickly. What is not commoditising is knowing which problem inside an industry is worth solving. That knowledge sits with people who have spent their career in it, and it cannot be scraped, fine-tuned out of a base model, or hired in a hurry.
Revenue teaches you things capital cannot. £1 of repeatable revenue is worth 10x equity capital, because revenue proves something works and capital only proves somebody believed you.
Speed without discipline gets expensive. The agentic age rewards moving early, but punishes moving early without knowing what a customer wants and the cost of winning and keeping them.
Craig and Glen were arguing one position from two ends, and that position is the Tramlines.
Tramlines works across three operating divisions and one investment platform, applying the same model at different scale. Build backs domain experts with capital, technology and operating support across a 24-month accelerator. Scale takes established SMEs whose growth is tied to the next hire and rebuilds the operating model with AI to break this link. Transform supports large institutions in innovating from within and migrating their operating model to ultimately replace their pre-AI one. Invest opens all three to investors through one platform.
What sits underneath all four is C4VC, which is Glen's point made concrete. The thing Blockbuster lacked was not an idea. It was a team, a platform, and an operating model capable of acting. That is what a centre for value creation is for.
Expertise stays at the front. Technology sits behind it.