A blog post written by:
Nils Vold, Partner at Verdane
After 25 years in tech, I’ve been through enough technology shifts to recognise the pattern: Every decade or so, a technology arrives that the business press calls revolutionary, and most companies treat as a tool.
The internet was going to change everything. For most incumbents, it became a website.
Mobile was going to change everything. For most, it became an app.
SaaS was going to change everything. For most, it became a billing change.
Cloud was going to change everything. For most, it became a hosting decision.
The companies that actually won in each wave weren’t the ones with the best version of the tool. They were the ones whose CEOs treated the technology as an enabler, and rebuilt their business model around it before someone else did.
AI fits the same pattern, with one key difference.
Every previous cognitive technology, be it writing, print, the database, and the search engine, extended human thinking by storing or retrieving it. AI is the first technology that reasons, decides and predicts. That isn’t a productivity tool. It’s a new factor of production.
Tool or enabler. That’s the whole question.
Walk into almost any company today and you’ll hear the same story: we’re more efficient because of AI. Developers ship faster. Marketers produce more content. Support agents close tickets quicker.
Now ask where that efficiency went. Almost nobody can point to a cost line that’s structurally lower or a revenue line that’s structurally higher. The productivity is real, but it dissolves into the existing organisation as slack – more meetings, marginally more output of the same work.
Parkinson’s Law, the old observation that work expands to fill the time available, eats the gain. Give people two hours back and the work spreads out to take two hours longer.
That’s what AI-as-a-tool looks like. Bolt it on to existing processes, get a productivity bump, watch competitors do the same, and watch the value get competed away to customers within 18 months. The upside is capped at single-digit margin improvement that doesn’t last.
AI-as-an-enabler is a different game altogether. It means rebuilding the business model around what the technology makes possible, exactly what the cloud winners did to the on-prem incumbents a decade ago.
License revenue became ARR.
Reactive support became proactive customer success.
Quarterly releases became continuous delivery with DevOps.
Sales motions, comp plans, org charts, KPIs – all rewritten and supporting a different business model.
The companies that did redesigned their business model took a two-three year P&L hit and transformation. The companies that didn’t, got acquired at low valuations or quietly lost momentum.
What actually has to change with AI
Product architecture
Software will need to moves from “humans do the work, software helps” to “software does the work, humans supervise”. AI agents and automated workflows will also fundamentally change the product experience and customer expectations.
Go-to-market
AI changes customer behaviour and thus go-to-market motions. Customer discovery is increasingly shifting towards LLM-driven interfaces, while seat-based pricing is weakening in favour of outcome-based models as software reduces the amount of human labour required. At the same time, AI-powered processes are improving targeting, pipeline quality and efficiency, driving both revenue uplift and cost savings.
Organisation design
AI breaks the historical link between revenue growth and headcount growth. Companies need smaller, more senior teams focused on oversight and escalation handling, while support, onboarding and operations can scale customers without costs and headcount growing proportionally.
Data flywheels
Proprietary feedback loops are becoming the moat. Companies improving products through customer interactions and workflow data build compounding advantages over time, while companies relying solely on foundation models become dependent on technology they do not control.
Why it’s worth the pain.
The real opportunity with AI is much larger than productivity gains, and it is still not discussed enough.
The companies that rebuilt around cloud captured profit pools that didn’t exist before. For example, they transformed into vertical SaaS categories worth tens of billions. Infrastructure layers worth hundreds of billions. Customer relationships that compounded for fifteen years. The redesigners didn’t just defend their businesses; they captured value the incumbents could not yet see. The same pattern repeated across the internet, mobile, SaaS and cloud. The largest value creation came from companies that redesigned their business model around what the technology made possible, not from companies that simply applied the technology to an existing model.
For AI, the profit pools will be even larger and form even faster. The companies that redesign their business model around AI now will be positioned to scale revenue, operations, and decision making at a fundamentally different level than before, allowing them to reach and capture those emerging profit pools ahead of the market. The companies that bolt AI onto existing processes will see the productivity gains competed away as AI becomes standard across the market.
It comes down to the CEO, supportive board and owners who understands the transformation
Most transformations fail at the same moment when somebody has to accept a short term pain, shrink a function, reprice a product, restructure a team, or take the margin hit this quarter.
The only thing that gets a company through that moment is a CEO willing to make the business model redesign, take some risky bets, backed by a board that will hold the line through the J-curve.
AI is a technology, not a strategy. Strategy is what leadership choose to with it.
The technology is commodifying fast. Organisational courage at the top is the scarce input. Leadership teams willing to redesign their operating model around AI will capture the next generation of value creation. The rest will use the same technology to compete harder for lower-margin businesses.
