Every advisor I know has the same reaction to the word audit: a vague dread, followed by a mental inventory of where things might be missing. That dread is usually worse than the reality, because most people are preparing for a review that does not happen.
The review is narrower than you think
A reviewer is not going to read your entire book. They will pick a handful of files — often the ones with the largest face amounts, a replacement, or a complaint attached — and work through them properly. What they are testing is whether your process exists, not whether every record is immaculate.
Which means a practice with a consistent, modest process across two hundred files will come out better than one with an excellent file here and a gap there.
The question underneath every other question
Strip away the specifics and nearly everything reduces to: was this written down at the time, or afterwards?
A needs analysis dated the day of the meeting carries weight. The same document produced three months later, after a question was raised, carries almost none — and a reviewer can usually tell the difference. File metadata, document version history and the sheer tidiness of reconstructed notes all give it away.
This is the single highest-value habit in a practice: capture the reasoning while the client is in front of you, even roughly. Rough and contemporaneous beats polished and retrospective every time.
What they will ask to see
- The needs analysis. Not just that one exists, but that the recommendation follows from it. A file recommending $1,000,000 of term against an analysis showing a $300,000 need is a problem no amount of tidy paperwork fixes.
- The reason why. Why this product, why this carrier, why this amount. “The client asked for it” is not a reason; it is a description of what happened.
- What else you considered. Recommending from a shelf of one looks like a recommendation from a shelf of one.
- Consent. What the client agreed to, and when. If it changed, when it changed.
- Replacements. If you replaced existing coverage, the comparison and the client’s acknowledgement of what they gave up.
Where practices actually come unstuck
In my experience it is rarely the big things. It is the ordinary drift:
- A signed document that lives in an email folder rather than the client file, and cannot be found two years later.
- A needs analysis completed on paper, scanned badly, never indexed.
- Consent obtained verbally, recorded nowhere.
- Notes in three places — a CRM, a notebook, and your memory — where only the third is complete.
None of these is misconduct. All of them are difficult to defend, because the defence depends on you remembering accurately under pressure about a conversation you had years ago.
The practical fix
You do not need a compliance department. You need the evidence to be a by-product of the work rather than a separate task you do later — because a separate task is one you will skip in a busy week, and busy weeks are when the files that matter get written.
Concretely, that means: the analysis is completed in the meeting, the document is signed there, and it is filed against the household automatically rather than by you remembering to. If your system requires discipline to stay compliant, it will fail in exactly the weeks you most need it not to.
One thing worth doing this week
Pick your three largest cases from the last year and open the files as though you were the reviewer. Can you show, without reconstructing anything, what the client needed, what you recommended, why, and what they agreed to?
If yes, you are in better shape than you think. If no, you have learned something far more cheaply than you would have learned it later.