5 metrics every service manager should review every Monday morning

Why Monday and why these five

Most service managers we know start Monday by triaging emails for two hours, walking the floor for the first crisis, and then reacting to whatever fire is loudest. By Tuesday they're already behind. The week is over before it started.

The cure is fifteen minutes of structured review before the floor opens. Five numbers, every Monday, in the same order, in the same format. The discipline isn't about the metrics; it's about replacing reactive triage with a proactive frame for the week. Managers who do this consistently end up running their stores instead of the other way around.

Why these five? Each one is a leading indicator. Each one points at a specific lever the manager can pull this week. And each one is captureable from a modern queue tool, your DMS, or both — without an analyst, without exporting CSVs, and without an MBA.

Metric 1: Last week's queue throughput

What the number means

Three sub-numbers that go together: how many cars actually moved through your bays last week, what the median customer waited, and what percentage of customers left at or before the time you quoted them. Throughput tells you about capacity. Wait tells you about flow. On-time rate tells you about promises.

What good looks like

Signal	Good	OK	Bad
Cars served vs prior 4-week avg	Within ±5%	Within ±10%	>10% drop
P50 wait (express)	<30 min	30-45 min	>45 min
P90 wait (express)	<55 min	55-80 min	>80 min
On-time rate (delivered ≤ quoted)	>85%	70-85%	<70%

What to do if it's bad

• Throughput drop usually means a tech is out, a bay is down, or your appointment mix shifted. Check the schedule first.
• P50 wait spiking almost always means a parts staging or routing problem. See our throughput article.
• P90 spiking with P50 fine means one or two outlier customers — often a no-show that hosed the rotation, or a scope change that wasn't reflected in the queue. Walk those individual cases.
• On-time rate <70% means your advisors are quoting unrealistic times. See why a 30-minute transparent wait beats a 20-minute surprise.

Metric 2: Customer satisfaction signal

What the number means

Three sub-numbers: your rolling NPS or CSI for last week, the count of completed surveys (a low survey count makes any score noisy), and a list of every customer who left a rating below your threshold (typically below 4/5 or below 8/10). The list of low-raters is the most important part. Every one of those customers needs a callback this week.

What good looks like

Signal	Good	OK	Bad
Rolling 30-day CSI	>4.6 / 5	4.3-4.6	<4.3
Survey response rate	>22%	15-22%	<15%
Low-rating callouts handled within 48hr	100%	80-99%	<80%

What to do if it's bad

• CSI score drop — pull the actual customer comments before doing anything else. The score doesn't tell you why. The free-text answers do.
• Low survey count usually means your post-visit SMS isn't firing reliably. Audit the trigger.
• Low-raters not getting callbacks is the single biggest CSI killer over time. Assign one person (advisor, BDC, or yourself) to own the callback list every Monday and Wednesday. Track to closure.

Metric 3: Tech performance summary

What the number means

For each tech: how many cars they completed, their average completion time vs SRT, the variance (are they consistent?), and any come-back flags from the prior 30 days. This is not a public ranking — it's an internal management tool. Don't post it.

What good looks like

Signal	Good	OK	Bad
Tech avg vs SRT	Within ±10%	±10-25%	>25% over
Tech-week consistency (std dev)	<15%	15-25%	>25%
Come-back rate per tech	<3%	3-5%	>5%
Workload balance across techs	±10% of mean	±20%	>±20%

What to do if it's bad

• One tech drifting slow: private conversation. Could be a personal issue, a tool problem, a service-mix problem. Don't assume; ask.
• One tech with come-backs spiking: walk one of the come-back ROs with them. Identify whether it's a quality issue, a parts issue, or a diagnostic depth issue. Coach accordingly.
• Workload imbalance: means your routing is broken or one tech is gaming the system. See our article on data-driven routing.
• One tech consistently top of every metric: tell them. Out loud. In writing. Senior techs are the cheapest, easiest retention play in fixed ops.

Metric 4: Recalls captured + uncontacted

What the number means

Three sub-numbers: how many recalls you actually completed last week (in dollars and count), how many were declined (with reasons), and how many remain uncontacted from previous weeks. The third number is the queue your BDC or callback owner needs to work this week.

What good looks like

Signal	Good	OK	Bad
In-store capture rate	>50%	30-50%	<30%
Decline rate (when offered)	<15%	15-30%	>30%
Uncontacted older than 14 days	0	1-5	>5

What to do if it's bad

• Capture rate <30%: the recall isn't being surfaced at check-in. Fix the workflow — see our recall article for the playbook.
• Decline rate >30%: the script is wrong. Refresh advisors on the "factory pays us" frame. Role-play five times.
• Uncontacted backlog growing: assign a dedicated callback owner with a weekly cadence. The list cannot be everyone's job, because it then becomes nobody's job.

Metric 5: This week's appointment book + capacity forecast

What the number means

For each day this week: how many appointments are on the books, what hours are heaviest, how many tech-hours you have available, and what walk-in capacity remains. The point is to spot problems on Monday that would otherwise blow up on Friday.

What good looks like

Signal	Good	OK	Bad
Day with appointments >90% of capacity	0 days	1 day	>1 day
Tech callouts known by Monday	All of them	Most	Discovering on the day
Saturday appointments vs capacity	≤75%	75-90%	>90%
Parts orders for the week's known scope	Placed Mon	Placed Tue	Placed when needed

What to do if it's bad

• Saturday is overbooked: Monday is the moment to call a couple of the lower-priority appointments and offer a Tuesday slot with a $20 incentive or a free tire rotation. Customers usually take it; the alternative is a Saturday meltdown.
• Tech callout discovered on the day: push for advance notice as a cultural norm. The tech who calls out at 7 AM is a problem; the tech who calls out at 5 PM Sunday gives you 13 hours to redistribute.
• Parts not ordered: the parts manager doesn't see the appointment book unless you share it. Set a Monday-morning standing handoff.

The printable Monday-morning template

Print this. Tape it to the wall behind your desk. Fill in five numbers in fifteen minutes. Forward it to your GM. Repeat next Monday.

The discipline matters more than the data

If you do this for one week, nothing changes. If you do it every week for two months, you start to see patterns nobody else in the building sees. You catch the slow drift in P90 wait before it becomes a CSI emergency. You catch the recall backlog before it becomes a $200K hole. You catch the tech who's quietly disengaging before they hand in notice.

The five-metric ritual is the single highest-leverage habit a service manager can build. Most managers don't do it because they "don't have time." The truth is the opposite: they don't have time precisely because they don't do it. The fifteen minutes Monday saves five hours by Friday. The math holds every week.

Related reading: Bay throughput: the metric that separates good from great · Why tech performance data should be running your bay assignments · How dealerships are leaving $100K+ in warranty revenue on the table