Sample audit: field sales team at a dealer-services company
An illustrative example based on a real engagement. Names and figures are changed.
This is the kind of report the Operations & AI audit ends with. It says where time and money go today, which fixes are worth doing, in what order, and what to leave alone.
The short version
The team sells finance and insurance products to car dealerships. Reps spend their days on the road, and every visit has to be logged through a long form on a phone.
So visits get logged late, from memory, or not at all. Managers can't see what's open, promises to dealers slip, and reps prepare for each visit by digging through old entries.
The biggest win is not AI. It's a shorter form. After that, summaries and briefs written by AI save the most time, and voice is worth a field test once the basics work.
Scope, people and systems
Scope
- How a dealer visit is booked, recorded and followed up, from the calendar invite to the next visit
- How managers see what the team is doing
- Out of scope: commissions, product pricing and dealer onboarding
People interviewed
- Four field reps, from new starters to the longest-serving
- Two regional sales managers
- The sales director
- The person who looks after the CRM
- Plus a day riding along on dealer visits
Systems
- The team's own field-sales CRM, running on AWS
- Outlook, for booking visits
- A weekly spreadsheet the managers build by hand
- A sample of recent visit records
What we found
Impact figures are rough estimates from interviews and the record sample. They show scale, not a promise.
- Finding 1
The visit form is too long for a phone
Over twenty fields per visit, including five goal dropdowns. Reps fill it in between dealers or at night, from memory. Managers only use about eight of the fields.
~3 hours per rep per week on visit admin
- Finding 2
Many visits are never logged
When the day runs long, the form is the first thing to go. In the sample, a large share of visits had no record, or one added days later.
Roughly 1 in 4 visits missing or late
- Finding 3
Goal statuses are out of date
Goals only change when a rep remembers to update them. Managers can't tell what is really open, so they ask in calls and messages instead.
~half a day per manager per week chasing updates
- Finding 4
Promises to dealers get lost
Follow-ups agreed in the meeting live in the rep's head or a notebook. Some never happen, and dealers notice.
Several dropped follow-ups per rep per month
- Finding 5
Prep means digging through old entries
Before each visit, reps scroll back through past records to work out what was said and what is still open.
~15 minutes of prep per visit
- Finding 6
Visits are booked twice
Visits are booked in Outlook, then entered again in the CRM. The two drift apart, and the manager's view is only as good as the CRM side.
~1 hour per rep per week, plus missed clashes
What to fix, in what order
Ranked by impact against effort. Each fix is labelled with the kind of work it needs, because not all of it is AI.
- Order: 1Process changeCut the visit form to the fields people actually use
Agree the short list with managers. Everything else becomes optional or goes.
- Impact
- High
- Effort
- Low
- Order: 2Custom buildA phone-first visit screen on the existing CRM
Big buttons, one hand, saves as you go. Same data underneath.
- Impact
- Medium
- Effort
- Medium
- Order: 3IntegrationBook visits in the CRM and send them to Outlook
One place to book. Calendars stay in sync and clashes show up.
- Impact
- Medium
- Effort
- Low
- Order: 4AI agentAutomatic visit summary and a brief before the next visit
AI writes what was covered and agreed, then pulls open goals and promises into a short brief.
- Impact
- High
- Effort
- Medium
- Order: 5AutomationWeekly manager summary and a needs-attention list
Replaces the hand-built spreadsheet. Visits with no record stand out.
- Impact
- Medium
- Effort
- Low
- Order: 6AI agentVoice debrief that drafts the record and suggests goal updates
Reps talk it through in the car. Nothing is saved until they confirm. Field test first.
- Impact
- High
- Effort
- High
- Order: Not recommendedLeave it aloneReplace the CRM with an off-the-shelf sales tool
It would mean moving every record and retraining the team, to fix problems the steps above already solve.
- Impact
- Low
- Effort
- High
Three phases, each useful on its own
You can stop after any phase and keep what it delivered.
- Phase 1
Make the record easy
- Shorter visit form, agreed with managers
- Phone-first visit screen
- Booking in the CRM, synced to Outlook
Delivers: Every visit is booked and logged in one place, on the phone, the same day.
- Phase 2
Let the record do the work
- Automatic summary of every visit
- A brief before the next visit
- Weekly manager summary and needs-attention list
Delivers: Reps walk in prepared. Managers see what's open without chasing anyone.
- Phase 3
Test voice, then decide
- Voice debrief with a few reps on real visits
- Suggested goal updates and follow-ups, each one confirmed by the rep
- Roll out only if the field test shows it saves time
Delivers: A clear yes or no on voice, based on real visits rather than a demo.
Not everything needs fixing
- The CRM itself
- It holds years of dealer history, the team knows it, and it can be extended. Replacing it would cost more than the problems it causes.
- Outlook for everything else
- Reps keep their email and calendar as they are. Only visit booking moves.
- Emails to dealers
- Dealers value the personal relationship. AI can draft a follow-up, but the rep sends it. Nothing goes out automatically.
- AI that predicts which dealers to visit
- Tempting, but the visit records aren't reliable enough yet to learn from. Worth a look once Phases 1 and 2 have been running a while.
Want this for your business?
Start with a free 30-minute call. Tell me what's slowing you down, and I'll tell you honestly whether an audit is worth it.
Book a meeting