I’ve shown you three ways an AI rollout drains a business. This is the fourth state, and the only one that pays. Same fictional company, same £1.5M bet, same starting point. The difference is not the technology and not the budget. It is one thing: the friction in this firm is the productive kind.
The business we’re modelling
The same reference company runs through all four of these case studies, so you can see exactly what changes when the behaviour does.
The baseline
- Sector
- B2B logistics and supply chain
- Revenue
- £50M a year
- EBITDA
- 12%, so £6M
- AI investment
- £1.5M, spent over 12 months
- The promise on that spend
- 20% efficiency gain, around +£2M EBITDA by Year 2
Same firm. Same plan. Now give it a culture where people are aligned on the goal and free to challenge how you get there.
What Productive Friction looks like
This firm has high alignment and high agency at once. People agree on where they’re going, and they have the room to say so when the route is wrong. There is psychological safety, so the frontline speaks up. There is executive backing, so when they do, it counts.
In week two of the rollout, a frontline team challenges a flawed parameter in the system. They’re not being difficult. They can see something the planners couldn’t, and the culture lets them say it before it ships rather than after it breaks. That is friction working for the business instead of against it. The disagreement is the asset.
So the rollout goes live and stays stable in 90 days, instead of dual-running forever, or leaking out the side, or grinding through endless readiness workshops.
The maths
This is the only state where the £2M promise turns into something bigger.
The gain
- Rollout
- £1.5M, live and stable in 90 days
- Efficiency gain
- 20%, liberating £1M of human capacity, with no layoffs
- New business
- Freed capacity wins +£5M of new accounts
That freed £1M of capacity is the part most firms never reach. The people who would have spent the year double-checking spreadsheets or sitting in readiness workshops are instead selling, serving, and winning work. Put it on the P&L and the picture inverts.
The outcome
- Revenue
- £50M rises to £55M
- EBITDA by Year 2
- Climbs to £8.5M (Multiplied Output)
£6M of EBITDA climbs to £8.5M and the top line lifts to £55M. Compare that to the other three states, where the same £1.5M produced a £200k drag, a £630k profit hit, or a half-million EBITDA loss. The spend was identical every time. Productive Friction is what turns it from a cost into a return.
You’ve seen one version of this in the headlines
In its first month live, Klarna’s AI assistant handled the workload of around 700 customer-service agents. Resolution times fell from 11 minutes to under 2. The company put the profit improvement at roughly $40M a year. The technology mattered, but the result came from an organisation ready to actually run it, fast, at scale, without the hedging or the gridlock that sinks most rollouts.
At full scale
Klarna · the work of 700 agents, in month one
That is Productive Friction at full scale. The tool was good. The readiness to use it properly is what produced the number.
Source: Klarna, AI assistant handles two-thirds of customer service chats in its first month
The point
The same AI, the same money, run by an organisation that’s aligned and free to push back, doesn’t drain the business. It adds £5M to the top line and lifts EBITDA from £6M to £8.5M. The Friction Scan tells you which state you’re actually in, before you spend a penny more on the tools.
This analysis is conducted out of curiosity, utilising hypothetical numbers and applying The Unbiased Strategic Clarity Methodology™. It is intended as a simplistic educational case study demonstrating the business impact of assumed organisational dynamics for leaders navigating the current technology era.
