An audit is not an inspection of technicians. It is an inspection of decisions, most of which were made in an office.
Why it matters
Operations rarely lack data; they lack a moment where somebody compares intention with outcome. An audit is that moment. It finds the stops that should not be on that day, the windows nobody needed, the territory boundary that stopped matching the customer base two acquisitions ago.
It also produces a baseline, which is the only way a later improvement can be proved rather than claimed.
How you notice it slipping
The last audit was the implementation. Improvements are described in adjectives rather than numbers. Nobody kept a before, so every after is an opinion.
What to do
The output should fit on one page: three or four causes, each with the evidence that found it and the change it implies. An audit that ends in a spreadsheet nobody opens had the same effect as not running one.
Audit a small set properly rather than everything superficially: two territories, four weeks, planned against actual. Write down what you expect to find first — the gap between expectation and result is usually the most useful thing the audit produces.