The information / advice boundary

When to use it

Reach for this when someone says any of these — they are the phrases the skill itself triggers on:

• “are agents giving advice”
• “where's the line between information and advice”
• “should agents recommend a product”
• “what would you do”

How it works

The method, in the order the skill runs it. The full procedure — tables, worked examples and the edge cases — is in the skill itself.

1. Get the definition before looking at anything

   Ask compliance for the boundary as it applies to your products, and get it as testable criteria rather than principles. Typically the distinguishing features are whether the statement is personal to the customer's circumstances and whether it constitutes a recommendation to act. General factual information about how a product works usually sits on the safe side.

2. The four zones

   Classify what agents actually said, because the middle two are where the work is.

3. Where to look

   The high-yield conversations: Direct requests, Customers describing their circumstances and asking an open question.

4. Look for over-caution too

   An audit that only counts crossings will drive agents further into evasiveness, which is its own harm and generates its own complaints: Reading terms verbatim instead of explaining them, Customers contacting repeatedly because they never got a usable answer.

5. The fix is usually a script, not training

   Where agents cross the line, look at what they had to work with.

Related skills

• Collections conductCheck customers in financial difficulty were treated fairly
• Chargeback disputesCheck chargebacks are handled to the scheme’s deadlines
• Financial promotionsFind promotional wording in replies that needs approval
• Scam and fraud signalsCatch scams customers describe to support before systems see them