If an earnings stream relies partly on customers finding it too difficult to act, AI can change the effort involved. The useful executive question is where that dependence exists and what you can do about it.
Three different pieces of evidence
Complaint research suggests that assistance can change outcomes. Subscription and broadband evidence describes existing costs of inaction. Together they justify investigating exposure. They do not establish a single causal chain from AI to a universal loss of profit.
A real dispute, with clear limits
The Building Trust guide describes residents organising documents, interrogating explanations and pursuing a structural-warranty dispute with AI assistance alongside professional and collective work. It is a practical example of a better-prepared customer.
The accepted settlement reported in the account is one result in one dispute. It should not be read as a typical claim, money already received or a sum attributable entirely to AI. The full story explains the evidence and the work involved.
Read the Building Trust case ↗
Open the insurance white paper ↗
What does it mean for the business?
Separate three questions. Where are customers entitled to a better resolution? Where does the business earn revenue from a decision being deferred? Where could a clearer, fairer offer earn continued custom? These questions may involve different owners and different economics.
Allocate time to investigate the customer journey, check the evidence behind decisions and improve the proposition. Finance can test exposure; Customer and Legal can examine resolution; Revenue and Product can develop a stronger reason to stay.
Try a sensitivity calculation
The defaults below are deliberately hypothetical. They are not derived from the studies above. Select one revenue stream and test low, middle and high assumptions using your own data. The output is gross contribution exposed under those assumptions, not a forecast of net profit.
One thing to do: map one point of abandonment
Bring Finance, Customer and Revenue together for 45 minutes. Pick one cancellation, renewal, switching or complaint journey. Identify the point where customers give up, who benefits financially and what a customer with well-organised evidence could do differently.
Leave with one friction map, one sensitivity range and one improvement to test. Measure resolution and customer value alongside the financial effect. A legitimate, clear offer should have a reason to survive when acting becomes easier.
Turn the exposure into an offer-design experiment →
video · 4 min
AI lawyers are here – can you trust artificial intelligence for legal help?
University of Sheffield · 2025-08-27
A short introduction to the opportunity and limits of AI-supported legal help, connecting customer empowerment with responsibility for accurate decisions.
Listen or watch for: What can a better-prepared customer now question, and where will they still need expert support?
Perspective: University and legal-support practitioner perspectives, including student tests; not evidence of a measured increase in successful legal claims.