Cut through dashboard noise. Here are the 12 KPIs that actually matter for growing service businesses, with benchmarks and where to find the data.
The Only 12 Metrics a 10-50 Person Company Needs to Track
Your dashboard has 47 metrics. You look at 3 of them. The other 44 are noise creating the illusion of measurement without the benefit of insight.
I know because I've seen your dashboard. Maybe it's a Databox board with 30 widgets. Maybe it's a Google Sheet with 14 tabs. Maybe it's three different tools that nobody opens after setup week.
The result is the same: you don't actually know how your business is running. You have data. You don't have visibility.
The "Track Everything" Trap
Somewhere along the way, founders got the message that more data equals better decisions. It doesn't.
More data equals more noise. More dashboards equals more time staring at screens. More metrics equals more opportunities to cherry-pick the numbers that confirm what you already believe.
I worked with a 28-person agency last year that tracked 53 metrics across four dashboards. The founder logged into their reporting tool maybe twice a month. When I asked which metrics drove their decisions, the answer was honest: "Revenue and whether clients seem happy."
53 metrics. Two of them used. The other 51 were theater.
Here's the uncomfortable truth: the businesses that make the best operational decisions track fewer metrics, not more. They just track the right ones. And they check them on a schedule.
You don't need a $500/month BI tool. You need 12 numbers, checked weekly, that tell you whether the business is healthy or bleeding.
The 4-Category Framework
Every service business between 10 and 50 people runs on four engines: delivery, money, pipeline, and people. If any one of these is broken, you'll feel it. But you won't always know which one is the problem until you're measuring.
Here are the 12 metrics that cover all four. Nothing else.
Category 1: Delivery Health (4 Metrics)
This is the work. The thing clients pay you for. If delivery is broken, nothing else matters because clients leave.
Metric 1: On-Time Delivery Rate
What it tells you: Whether you keep promises.
Formula:
On-Time Delivery Rate = (Projects Delivered On or Before Deadline / Total Projects Delivered) × 100
Where to get the data: Your project management tool. Compare original deadline to actual completion date. If you don't track deadlines in your PM tool, that's your first problem.
What "good" looks like: >90%. Below 85% means clients are noticing. Below 75% means they're talking about it with their colleagues.
When it's bad, check:
- Are scopes accurate? If every project runs over, your estimating is broken, not your team.
- Are handoffs clean? Delays often happen in the gaps between people, not in the work itself.
- Is the team overloaded? Check the capacity buffer metric below.
Most companies don't track this formally. They have a vague sense that "we're usually on time." That vague sense is almost always wrong. Measure it for one month and you'll be surprised.
Metric 2: Client Satisfaction Score
What it tells you: Whether clients would hire you again.
Formula: Ask them. At the end of every project or quarterly for retainer clients. One question: "On a scale of 1-10, how likely are you to recommend us?" (NPS) or "How satisfied are you with our work?" (CSAT).
Where to get the data: A simple post-project survey. Typeform, Google Forms, or even a question in your project wrap-up email. Don't overthink the tool.
What "good" looks like: >8/10 average or NPS >40. Below 7/10 means you have a service quality problem.
When it's bad, check:
- Is it the work quality or the experience? Sometimes the deliverable is great but the process was painful. Those are different problems.
- Is one team member or project type dragging the score down? Don't average away the problem. Segment by team, service line, and client size.
Opinionated take: I prefer simple CSAT over NPS for companies under 50 people. NPS was designed for consumer brands with millions of customers. You have 20-40 clients. Just ask if they're happy.
Metric 3: Utilization Rate
What it tells you: Whether your team's time is generating revenue.
Formula:
Utilization Rate = (Billable Hours / Available Hours) × 100
Where to get the data: Your time tracking system. If you don't have one, stop reading and go set one up. Harvest, Toggl, or Clockify. Pick one. Mandate it. Available hours is NOT 40 hours per week. It's 32-35 hours after meetings, admin, and breaks.
What "good" looks like: 70-80% for the team overall. Individual targets vary by role. Junior staff should be 80-85%. Directors should be 30-40%. If your Creative Director is at 80%, they're not directing anything.
When it's bad, check: Our full breakdown of utilization benchmarks by role and agency type covers what to do when the numbers are off. The short version: cut non-billable meetings before you cut people.
Metric 4: Rework/Revision Rate
What it tells you: Whether you're getting it right the first time.
Formula:
Rework Rate = (Projects Requiring Unplanned Revisions / Total Projects) × 100
Where to get the data: Track revision rounds in your project management tool. The key word is "unplanned." Scoped revisions are normal. The third round of edits that wasn't in the SOW is rework.
What "good" looks like: <15%. Below 10% is excellent. Above 20% means something is systematically broken in your discovery, scoping, or approval process.
When it's bad, check:
- Is your discovery process thorough enough? Rework usually starts at kickoff, not at delivery.
- Are approvals happening at checkpoints or only at the end? The later you get feedback, the more expensive it is to act on.
- Is there a specific client or team member pattern? Sometimes one account manager's projects have 30% rework while everyone else is at 10%.
Category 2: Financial Health (3 Metrics)
Delivery metrics tell you if clients are happy. Financial metrics tell you if you'll survive.
Metric 5: Gross Margin by Service Line
What it tells you: Which services make money and which ones just keep people busy.
Formula:
Gross Margin = ((Revenue - Direct Costs) / Revenue) × 100
Direct costs = labor, contractors, tools, and materials directly tied to delivering that service. Not rent. Not your salary. Not the office snack budget.
Where to get the data: Your accounting software plus project-level time tracking. You need both. Revenue from QuickBooks/Xero, hours from your time tracker, then calculate the labor cost per project.
What "good" looks like: 50-70% for professional services. Below 40% means the service line is a commodity or you're underpricing it. Above 70% is excellent leverage.
When it's bad, check:
- Are you tracking direct costs accurately? Most companies undercount because they don't include the true loaded cost of labor (salary + benefits + taxes + overhead).
- Is one service line subsidizing another? This happens constantly. Your strategy work at 65% margin is funding your execution work at 25% margin. That's fine if it's intentional. It's not fine if you don't know it.
This is the metric most small businesses skip. They know total revenue. They know total profit. But they don't know which services are printing money and which are bleeding it. If you run an operations audit, margin by service line is always one of the biggest revelations.
Metric 6: Revenue Per Employee
What it tells you: How efficiently your company converts headcount into revenue. This is the single best proxy for operational efficiency.
Formula:
Revenue Per Employee = Annual Revenue / Total Headcount (including contractor FTEs)
Where to get the data: Your P&L and headcount list. Use trailing 12 months to smooth out seasonality. Count contractors who work dedicated hours as fractional FTEs.
What "good" looks like: $150K-$250K for service businesses. Below $120K and you're either overstaffed, underpriced, or operationally inefficient (likely all three). Above $250K and you're running a tight operation.
When it's bad, check: Read our complete breakdown of revenue per employee benchmarks. The short version: don't hire to solve an efficiency problem. Fix utilization, processes, and pricing first.
Metric 7: Cash Runway
What it tells you: How long you can survive if revenue stops tomorrow. This metric exists to prevent surprises.
Formula:
Cash Runway (months) = Current Cash Balance / Average Monthly Burn Rate
Burn rate = total monthly expenses minus revenue. If you're profitable, your "burn rate" is negative and runway is theoretically infinite. Track it anyway. Profitable companies run out of cash all the time because revenue on an invoice isn't cash in the bank.
Where to get the data: Your bank balance and your monthly expense total. Check it weekly.
What "good" looks like: >3 months for stable businesses. >6 months if you're growing fast or in a volatile market. Below 2 months is a fire alarm.
When it's bad, check: Our cash flow forecasting guide walks through the 13-week cash flow model that gives you week-by-week visibility. Don't wait until runway is short to start watching it.
Category 3: Pipeline Health (3 Metrics)
Delivery and finances tell you about today. Pipeline tells you about 90 days from now. This is where most companies get blindsided.
Metric 8: Close Rate by Source
What it tells you: Where your good clients come from - not just any leads, but the ones that actually close.
Formula:
Close Rate = (Deals Won from Source / Total Opportunities from Source) × 100
Where to get the data: Your CRM. If you don't have one, a spreadsheet with columns for Lead Source, Deal Amount, and Won/Lost works. The key is tracking the source of every deal.
What "good" looks like: Varies wildly by source. Referrals typically close at 40-60%. Cold outbound might be 5-15%. Inbound from content might be 15-30%. The specific numbers matter less than the comparison between sources.
When it's bad, check:
- Are you investing in the sources that actually close? Most companies spend the most on the channels with the lowest close rates because those channels are the most visible.
- Is the definition of an "opportunity" consistent? If one salesperson logs every conversation as an opportunity and another only logs qualified prospects, the data is useless.
The insight that matters: Don't optimize for lead volume. Optimize for close rate by source. 10 referrals that close at 50% are worth more than 100 cold leads that close at 3%.
Metric 9: Average Time-to-Close
What it tells you: How long your sales cycle takes from first contact to signed contract.
Formula:
Avg Time-to-Close = Sum of Days from First Contact to Close / Number of Closed Deals
Where to get the data: Your CRM. Track the date each opportunity was created and the date it was marked won or lost.
What "good" looks like: For B2B service businesses: 14-30 days is fast, 30-60 is normal, 60-90 is slow, 90+ means something is broken. The benchmark depends on deal size. A $5K project should close faster than a $150K engagement.
When it's bad, check:
- Is there a specific stage where deals stall? Track time per stage, not just total time. If proposals sit for 3 weeks, the problem is in your proposal process, not your prospecting.
- Are you qualifying early enough? Long sales cycles often mean you're spending time on prospects who were never going to buy.
Metric 10: Pipeline Coverage Ratio
What it tells you: Whether you have enough in the pipeline to hit your revenue target.
Formula:
Pipeline Coverage Ratio = Total Pipeline Value / Revenue Target for Period
Where to get the data: CRM pipeline value (weighted by probability) compared to your quarterly or monthly revenue target.
What "good" looks like: 3x is safe. Meaning if your quarterly target is $300K, you need $900K in pipeline. 2x-3x is normal. Below 2x is an alarm. Below 1.5x means you're almost certainly going to miss your number.
When it's bad, check:
- Is your pipeline real? Stale opportunities that haven't been updated in 60+ days aren't pipeline. They're wishful thinking. Clean your CRM monthly.
- Is your win rate accurate? If you're closing 20% of pipeline but calculating coverage assuming 33%, you're lying to yourself.
A rule I enforce with every client: If pipeline coverage drops below 2x, the founder's calendar shifts. Cancel the internal projects. Stop the rebranding exercise. Go sell. Nothing else matters if you don't have enough pipeline.
Category 4: Team Health (2 Metrics)
You can have perfect delivery, strong finances, and a full pipeline. If your team is maxed out or leaving, it all falls apart in 6 months.
Metric 11: Capacity Buffer
What it tells you: Whether you have room to absorb a new project, a sick team member, or an emergency without everything breaking.
Formula:
Capacity Buffer = (1 - Team Utilization Rate) × 100
Or more practically: how many hours per week does your team have available beyond current commitments?
Where to get the data: Your time tracking system. Subtract current booked/utilized hours from available hours.
What "good" looks like: 15-20% buffer. A 10-person team at 80% utilization has 20% buffer. That's roughly 64 hours per week of breathing room across the team. Enough to handle a new client onboarding or cover when someone's out.
When it's bad:
- Below 10% buffer: You're one emergency away from missed deadlines. Don't take on new work until you either hire or offload something.
- Above 30% buffer: You're overstaffed or under-selling. Either trim headcount or invest harder in sales.
This metric is the bridge between delivery health and team health. It tells you whether growth is possible without burning people out. If you're reading the agency operations playbook and wondering how to plan capacity, this is where you start.
Metric 12: Voluntary Turnover Rate
What it tells you: Whether people want to stay.
Formula:
Annual Voluntary Turnover = (Voluntary Departures in 12 Months / Average Headcount) × 100
Only count voluntary departures - people who chose to leave. Terminations and layoffs are a different metric.
Where to get the data: HR records. For a 10-50 person company, you probably know every departure off the top of your head.
What "good" looks like: <15% annually for service businesses. Below 10% is excellent. Above 20% means you have a retention problem that is actively costing you money.
When it's bad, check:
- The cost is real: Replacing a mid-level employee costs 50-150% of their annual salary when you factor in recruiting, onboarding, lost productivity, and the work that doesn't get done during the transition. At a 25-person company with $80K average salaries, every departure costs you $40K-$120K.
- Conduct exit interviews. Not the HR checkbox kind. Actual conversations. The patterns in why people leave are almost always fixable.
- Is it compensation, growth, management, or workload? Each requires a different fix. Don't throw money at a management problem.
How to Build This Dashboard
You don't need fancy tools. You need consistency.
Tier 1: Google Sheets (Free, Start Here)
Build one spreadsheet with 12 rows and 52 columns (one per week). Enter numbers every Monday. Add conditional formatting: green for on-target, yellow for warning, red for alarm.
This is genuinely enough for most companies under 30 people. Don't let anyone sell you a $300/month dashboard tool until you've actually maintained a spreadsheet for 3 months.
Your 12-Metric Dashboard (Google Sheet)
Row 1: On-Time Delivery Rate Target: >90%
Row 2: Client Satisfaction Score Target: >8/10
Row 3: Utilization Rate Target: 70-80%
Row 4: Rework Rate Target: <15%
Row 5: Gross Margin (Service Line A) Target: >50%
Row 6: Gross Margin (Service Line B) Target: >50%
Row 7: Revenue Per Employee Target: >$150K
Row 8: Cash Runway (months) Target: >3 months
Row 9: Close Rate by Source Target: varies
Row 10: Avg Time-to-Close (days) Target: <45 days
Row 11: Pipeline Coverage Ratio Target: >3x
Row 12: Capacity Buffer Target: 15-20%
Row 13: Voluntary Turnover (annualized) Target: <15%
Yes, that's 13 rows. Gross margin gets split by service line because the average is meaningless. One service at 65% and another at 25% is not a business with "45% margins." It's a business with one profitable service and one that needs to be fixed or killed.
Tier 2: Databox or Geckoboard ($50-$200/month)
When your data lives in multiple tools (CRM, PM tool, accounting, time tracker), pulling it manually into a spreadsheet gets tedious after a few months. That's when a dashboard tool earns its cost.
Databox connects to HubSpot, QuickBooks, Google Sheets, and 70+ other tools. It pulls data automatically and displays it on a wall TV or browser dashboard.
Geckoboard is simpler and more visual. Good for status-board style displays.
When to upgrade: When you catch yourself skipping the Monday review because the manual data entry takes too long. The review matters more than the tool.
Tier 3: Custom n8n + Google Sheets (Power Users)
If you want automation without the monthly SaaS fee, build workflows that pull data from your tools into a central Google Sheet automatically.
n8n can pull CRM data from HubSpot, time data from Harvest, financial data from QuickBooks, and dump it all into a formatted sheet every Sunday night. Monday morning, your 12 numbers are waiting for you.
This requires setup time (4-8 hours) but costs nothing to run and updates automatically. It's what we build for most of our clients.
The Weekly Review Ritual
The dashboard doesn't matter if nobody looks at it. Here's the ritual that makes it work.
When: Every Monday. 15 minutes. Not negotiable.
Who: You (the founder/CEO) and anyone who owns a number on the dashboard. For most companies, that's you, your ops lead, and your sales lead. Three people.
What:
Monday Metrics Review - 15 Minutes
MINUTE 0-2: Open dashboard. Note any red or yellow indicators.
MINUTE 2-5: DELIVERY CHECK
- On-time delivery rate this week?
- Any client satisfaction issues flagged?
- Utilization: over or under target?
- Rework trending up or down?
MINUTE 5-8: FINANCIAL CHECK
- Margins by service line: any surprises?
- RPE trending up or down?
- Cash runway: still comfortable?
MINUTE 8-11: PIPELINE CHECK
- Coverage ratio: above or below 3x?
- Close rate: any source performing unusually well or poorly?
- Time-to-close: any deals stuck?
MINUTE 11-13: TEAM CHECK
- Capacity buffer: room for new work?
- Any turnover signals? (This one you feel more than measure.)
MINUTE 13-15: DECISIONS
- Based on what we just saw, do we need to change anything?
- 0-3 action items. Assign owners. Due by Friday.
- If the answer is "everything looks fine," the meeting is over early.
The most important part is the last two minutes. Metrics without decisions are a spectator sport. The purpose of the review isn't to admire numbers. It's to decide: do we need to change anything this week?
Most weeks, the answer is no. Everything is on track. Meeting takes 8 minutes. Good.
Some weeks, one number is off. You spend 5 minutes diagnosing it and assign someone to fix it. Meeting takes 15 minutes.
Occasionally, multiple numbers are off at once. That's a sign of a systemic issue, not a bad week. Extend the meeting or schedule a deep-dive. But that should happen once a quarter, not once a week.
What Not to Track
Just as important as what you measure is what you ignore. Here's what you should not put on your KPI dashboard:
Vanity metrics. Website traffic, social media followers, email list size. These are marketing metrics, not operations metrics. Track them in your marketing team's tools if you want, but they don't belong in your operational dashboard.
Lagging indicators you can't act on. Annual revenue is useful for year-end reviews but useless for weekly decision-making. By the time annual revenue looks bad, you've had 9 months of warning signs in these 12 metrics.
Per-person performance metrics. Your operational dashboard is about the business, not individual performance reviews. Utilization by person is a management conversation, not a dashboard widget.
Anything that requires more than 5 minutes to calculate. If it takes 30 minutes to compute a metric, you'll stop computing it by month two. The best metrics are the ones you can pull in under a minute each.
What These 12 Metrics Look Like Together
Here's a snapshot of a healthy 30-person service business:
DELIVERY HEALTH
On-Time Delivery: 92% ✓ (target: >90%)
Client Satisfaction: 8.4 ✓ (target: >8.0)
Utilization: 76% ✓ (target: 70-80%)
Rework Rate: 11% ✓ (target: <15%)
FINANCIAL HEALTH
Gross Margin (Consulting): 62% ✓ (target: >50%)
Gross Margin (Execution): 41% ⚠ (target: >50%)
Revenue Per Employee: $188K ✓ (target: >$150K)
Cash Runway: 4.2 months ✓ (target: >3)
PIPELINE HEALTH
Close Rate (Referrals): 48% ✓
Close Rate (Inbound): 22% ✓
Avg Time-to-Close: 34 days ✓ (target: <45)
Pipeline Coverage: 2.8x ⚠ (target: >3x)
TEAM HEALTH
Capacity Buffer: 24% ✓ (target: 15-20%)
Voluntary Turnover: 12% ✓ (target: <15%)
Two yellow flags. The execution service line is running thin margins - that's a pricing or scoping problem worth investigating. Pipeline coverage is slightly below 3x - not a crisis, but the founder should spend an extra 5 hours on BD this week.
Everything else is green. This business is healthy. The Monday review for this week takes 10 minutes. No major decisions needed.
Now compare that to knowing "revenue is $5.6M and we have 30 people." Same company, zero insight. That's the difference between data and visibility.
12 Metrics. 15 Minutes a Week.
That's operational visibility. Not 47 widgets on a dashboard you never open. Not a BI tool that takes 3 months to implement. Not a 90-minute weekly leadership meeting reviewing slides.
Twelve numbers. Four categories. Fifteen minutes every Monday.
You'll catch delivery problems before clients complain. You'll see margin erosion before it hits your bank account. You'll spot pipeline gaps 90 days before revenue drops. You'll know when your team is running hot before someone quits.
The companies that operate well don't have more data. They have the right data, reviewed on a schedule, with the discipline to act on what it says.
Start this Monday. Pull the 12 numbers. It won't take 15 minutes because you'll have to hunt for some of them. That's the point. The hunting tells you where your measurement gaps are.
By week four, you'll have the habit. By week eight, you'll have trend data. By week twelve, you'll wonder how you ran the business without it.
Need Help Setting This Up?
If you're looking at this list and realizing you can't pull half these numbers because the systems aren't in place, that's a common starting point. Most companies we work with start with an operations audit to figure out where the gaps are, then we build the tracking infrastructure alongside the operational fixes.
Book a free 30-minute call and we'll figure out which of these 12 metrics you're missing - and what it's costing you not to track them.
Related Reading