Sales team audit: a checklist you can complete in one evening
What to check in the CRM, the calls and the closed leads, how to convert the holes into money, and the order to fix them in. All you need is a one-month CRM export.
A sales team audit sounds like a month-long project with consultants and a slide deck. In practice, 80% of the value comes from a check the owner or head of sales can run in one evening with a CRM export. Here's that checklist — with an explanation of what exactly to look for in each item and what to do with what you find.
Contents
Why audit if "everything's visible anyway"
A manager sees sales through the reps' reports — and reports are written by the people being checked. An audit exists to look at the facts once without the retelling: how many enquiries were actually handled, how many calls were actually made, whether the refusal reasons are confirmed by the data.
The typical result of a first audit is 10–20% of "junk" that isn't junk: leads closed without a single dial attempt, "unqualified" ones nobody ever qualified, and deals that died simply because nobody set the next step.
Step 0: the data you need
A CRM export for the last 30 days: leads with creation date, first-response date, closing date, refusal reason and call count; deals with last-activity date and next step. Plus access to call recordings. If your CRM can't produce some of this — that is already the audit's first finding.
Block 1: inbound leads
1. First-response speed. Count how many enquiries were answered later than an hour in. The conversion gap between a five-minute call-back and an hour is measured in tens of percent — this is the most expensive hole of all.
2. Enquiries outside the CRM. Reconcile enquiries from ad accounts and messengers against CRM cards. An Instagram or WhatsApp message a rep scrolled past doesn't exist in any report — but you paid for it.
3. Leads with no task. How many cards live without a scheduled action? A lead without a task is a lead someone will remember by accident — or never.
Block 2: calls and conversations
4. "No answer" after one attempt. A large share of contacts picks up on the second to fourth attempt at different times of day. If the "no answer" status is set after a single dial, you are throwing paid enquiries away.
5. Calls under 10 seconds. Formally a call happened; in reality it didn't. Count their share among "handled" leads.
6. Five conversations that ended in "too expensive". Listen to the recordings: was there any justification of the price, or did the rep just agree and hang up? It's the fastest way to see the real quality of objection handling — and at volume, this is exactly what an AI quality-control layer listens to across every call, not five samples.
Block 3: closures and refusal reasons
7. Ten random "unqualified" leads. Open and verify: are they really unqualified? Usually a share of these closures doesn't hold up — no calls were made, there's no duplicate, the number is valid.
8. Deals without a next step. The customer said "let me think", the rep said "sure" — and no next-contact date exists. Formally the deal is alive; in reality it's lost.
9. Silence after the proposal. How many quotes went out with no follow-up call within two or three days? One such call recovers a noticeable share of "hanging" deals.
An automatic audit of the funnel and closure validity: disputed leads go back to work with a flag. Launch in 2 weeks.
Get a quote in 30 sec →Turning findings into money
Count in currency, not percentages — otherwise the conversation stays abstract. The formula is simple: disputed closures × your conversion rate × average deal size, plus the ad cost of the enquiries themselves.
Example: 200 leads a month, 60 closed as junk, and the check shows 15 closures with no real dial attempt. At 15% conversion and a €900 average deal, that's ≈2 deals × €900 = €1,800 a month of potential losses. Even recovering half pays back the automation in the first month. The full method with the formula is in our article on AI automation cost.
What to fix first
First-touch speed. An automatic task on every new lead with a hard deadline and an alert on breach. The effect is visible within the first week.
A dial-attempt rule. Minimum three attempts at different times — only then is a "no answer" status allowed.
A mandatory next step. A deal cannot remain without a next-contact date — the system enforces it, not the rep's memory.
Closure verification. An automatic check of every refusal reason against the actual call, chat and duplicate data. Disputed cases go back to work with a flag.
Base revival. Regular re-contact of refusals and "let me think" leads — manually, or with AI phone calls when the volume is large.
The first three items are CRM settings closed in a few days. The full breakdown of where enquiries leak is in why you lose leads: 9 leak points.
Frequently asked questions
How often should a sales team be audited?
One deep audit at the start of changes, then a monthly express check with this same checklist on a fresh export. Once control is automated, closure verification and response-speed checks run daily without you.
Can we run the audit without a consultant?
Yes — this checklist is designed for the owner or head of sales to complete alone in one evening. A consultant is needed not to check, but to implement the changes when nobody inside has the time and the authority.
What if the CRM doesn't have the data to check?
That is the audit's main finding: you're managing the team blind. Start with basic discipline — mandatory fields, call logging, all channels feeding the CRM — and repeat the audit in a month.