Why it matters now
Budgets are rising quickly, and few companies feel mature. That combination makes it easy to spend for the sake of being seen to act. I see two opposite mistakes: spending widely out of fear of falling behind, and waiting for the dust to settle. The dust is not going to settle. Learn quickly and cheaply, and spend heavily only where the evidence supports it.
- Companies plan to roughly double their AI spending in 2026, from about 0.8% to about 1.7% of revenue.BCG, AI Radar 2026 ↗
- 92% of companies plan to increase their AI investment, but only about 1% of leaders call their company mature.McKinsey, Superagency in the Workplace, 2025 ↗
- Over 80% of leaders expect their AI investments to pay back within 2–3 years.Wharton and GBK Collective, 2025 AI Adoption Report ↗
Common mistakes
- Spending driven by fear of missing out rather than by expected value.
- Overpromising timelines to the board or the organisation.
- Waiting for certainty while others learn.
- Spreading investment thinly across many small bets.
What to do instead
- Learn fast and cheaply. Get people using AI and running small experiments now. Learning is the cheap part.
- Fund in stages. Release money in steps tied to evidence: a working pilot, measured value, readiness to scale.
- Concentrate. Put most of the investment into a few areas where the value is largest.
- Set realistic horizons. Agree payback expectations up front, and be honest about how long real change takes.
- Review the portfolio every quarter. Move money from what is not working to what is.
Your first step this week
Sort your AI spending into two piles: learning and adoption, and scaling. Check that every item in the scaling pile is tied to evidence, not hope.
Related guides
- What should our AI strategy be, and where do we focus?
- Why aren’t we seeing a return on AI, and how do we measure it?
- Should we build, buy or partner for AI?
All AI leadership principles and guides · 100 AI use cases by function