John Rawls proposed a thought experiment in 1971: design the rules for a just society from behind a "veil of ignorance," where you don't know what position you'll occupy. You don't know if you'll be rich or poor, powerful or powerless, healthy or sick.

Behind the veil, you design fair rules because any unfair rule might apply to you.

Now apply this to your support policy. If you didn't know whether you'd be the customer or the company, would you design the same refund policy? The same response time SLA? The same chatbot that blocks human access?

The Refund Policy Test

Your current refund policy: "Refunds within 30 days. After 30 days, no refunds."

Behind the veil, you don't know if you're the company (who loses revenue on refunds) or the customer (who might discover a problem on day 31). Would you design a hard 30-day cutoff?

Probably not. You'd probably design something like: "Full refunds within 30 days, no questions asked. After 30 days, prorated refunds for legitimate issues (billing errors, broken features, service failures). After 90 days, case-by-case review for exceptional circumstances."

This policy protects the customer (who might discover a problem late) while protecting the company (from abuse by customers who use the product for months and then request full refunds). Behind the veil, it feels fair regardless of which side you're on.

The SLA Test

Your current SLA: "We respond within 24 hours."

Behind the veil: if you're the customer whose business depends on your product, and the product goes down on a Friday afternoon, is a 24-hour SLA acceptable? You'll be down all weekend. Your customers will be affected. You'll lose money.

A Rawlsian SLA might look like: "4 hours during business hours for all issues. 1 hour for urgent issues (outages, security, billing errors). 15 minutes for critical issues affecting multiple customers."

This is more expensive to deliver. But behind the veil, if there's a chance you'd be the customer with the outage, you'd want the faster response. The 24-hour blanket SLA only feels acceptable when you're sure you're the company, not the customer.

The Chatbot Test

Your current setup: all incoming messages go to a chatbot. The chatbot handles what it can. Reaching a human requires navigating the chatbot first.

Behind the veil: if you're the customer with a complex, emotional, or urgent issue, is the chatbot acceptable? If your billing was wrong by $500 and your rent is due, would you want to negotiate with a bot?

A Rawlsian chatbot design: the chatbot is the first interaction, but a visible "talk to a human" button is always available. The chatbot handles simple issues instantly (giving the customer a faster experience). Complex issues get routed to humans without friction.

This is what most customers would design if they had the choice. It's also what most support leaders would design if they imagined themselves as the customer.

Why Companies Don't Do This

Companies design support policies from the company's perspective, not the customer's. The CFO sees refund rates and tightens the policy. The VP of Ops sees support costs and delays the SLA. The product team sees chatbot deflection rates and removes the human escalation button.

Each decision is rational from the company's perspective. But behind the veil, each decision creates a policy you wouldn't want applied to yourself.

The veil of ignorance doesn't mean "always side with the customer." It means "design policies that are fair regardless of which side you're on." Sometimes that means the company's interests dominate (a customer who used the product for 11 months and wants a full refund isn't treated the same as one who used it for 1 day). Sometimes the customer's interests dominate (a billing error should be fixed immediately regardless of how long the customer has been a subscriber).

Applying This Practically

For every support policy, ask the Rawlsian question: "If I didn't know whether I'd be the customer or the company, would I design this policy?"

If the answer is yes, the policy is fair.

If the answer is "only if I'm the company," the policy favors you at the customer's expense. That doesn't mean it's wrong (companies need to be profitable), but it means the policy has a fairness cost that should be acknowledged and weighed against the business benefit.

If the answer is "only if I'm the customer," the policy is too generous. It might feel good but it's unsustainable. A 365-day full refund policy is customer-friendly and company-fatal.

The sweet spot is the policy that both sides would design for themselves. That's usually: generous for small amounts, reasonable for medium amounts, case-by-case for large amounts, and always accessible to a human for complex situations.

That's what fair support looks like. And if your current policies don't pass the veil test, it's worth asking why.