Walk into most service businesses doing $1M a year and ask the founder what their gross margin was last month. You'll get a pause. Then a guess. Then a promise to check with the bookkeeper. Meanwhile, the same founder can rattle off their Instagram follower count, their MRR, and how many leads came through the site last week. That's the problem. Not a lack of data. A lack of the right data, looked at often enough to matter.
Below are the five service business KPIs worth tracking every single week. Not monthly. Not quarterly. Weekly. And before we get to them, a quick word on the metrics you should probably stop obsessing over.
The vanity metrics quietly costing you money
Revenue is the loudest vanity metric of them all. A $2M year with 18% margins is a worse business than a $900K year with 42% margins, but the first founder gets the LinkedIn applause. Followers, impressions, website traffic, and even total lead volume fall into the same trap. They feel like progress. They rarely predict it.
The tell is simple. If a number can go up while your bank account goes down, it's probably a vanity metric. Track it if you want, but don't run your week around it.
The five below are different. Each one connects directly to cash, retention, or growth capacity. Watch them weekly and you'll spot problems 60 to 90 days before they show up in your P&L.
1. Customer LTV (by cohort, not average)
Lifetime value is one of the most misused numbers in small business metrics. Most founders calculate it as average revenue per customer times average lifespan. That gives you one blended number that hides everything interesting.
Instead, calculate LTV by acquisition cohort. Group customers by the quarter they signed on. Then track how much each cohort has paid you to date and how many are still active. You'll see patterns like this: customers acquired through referrals in Q1 have an average LTV of $18,400 and 71% are still active. Customers acquired through paid search in Q2 have an LTV of $6,200 and only 34% are still active.
Now you know something useful. Referrals aren't just cheaper to acquire. They stick around three times longer. That's a hiring decision, a marketing budget decision, and a service delivery decision, all sitting in one metric.
Weekly action: Update your cohort table every Monday. If a recent cohort is trending 20% below the last one at the same age, dig in this week, not next quarter.
2. Gross margin per service line
If you offer more than one service, blended gross margin lies to you. A design agency I know was running at 38% blended margin and thought that was fine. When they broke it out by service line, they found their brand identity work was 61%, their web dev was 44%, and their ongoing social media retainers were losing 4% every month. They'd been subsidizing a bad product with two good ones for 18 months.
Calculate it this way for each service line: revenue minus direct labor cost minus direct software and contractor cost, divided by revenue. Direct labor means the fully loaded cost of the people doing the work, including payroll taxes and benefits. Don't skip that part. Founders who count only base salary consistently overstate margin by 15 to 20 points.
Once you can see margin by service line, three decisions get easier. Which service to raise prices on. Which one to kill. Which one to hire aggressively into. Most agency KPIs dashboards skip this entirely and just show revenue by service. That tells you nothing about health.
Weekly action: Run margin by service line every Friday. Flag any line that dropped more than three points week over week.
3. DSO (days sales outstanding)
This one is boring, and that's exactly why it's neglected. DSO is the average number of days between invoicing a client and getting paid. In a healthy service business, it should sit between 15 and 30 days. Anything over 45 and you have a working capital problem you probably haven't named yet.
Here's the math. If you invoice $200K a month and your DSO is 55 days, you have roughly $360K locked up in receivables at any given moment. If you got that down to 25 days, you'd free up about $195K in cash. That's a hire. Or three. Or a full quarter of runway.
DSO also tells you something about your client base. Clients who pay slowly usually complain more, scope-creep more, and churn faster. There's a strong correlation between payment behavior and every other kind of behavior. When you tighten payment terms and enforce them, you don't just improve cash flow. You filter out clients who were going to be a problem anyway.
Weekly action: Every Monday, pull an aging report. Anything over 30 days gets a call that day, not an email next week. Anything over 60 days gets a partial work stoppage until it's resolved.
4. Retention cohort (not blended churn)
Churn is the sibling of LTV, and it has the same problem. A blended monthly churn rate of 4% sounds fine until you realize your Q1 cohort churns at 1% and your Q3 cohort churns at 9%. Something changed. You need to know what.
Build a simple retention cohort table. Rows are the month a customer signed on. Columns are month 1, month 2, month 3, and so on. Each cell shows what percentage of that cohort is still active. This one table will teach you more about your business than any dashboard your accountant can build.
What to look for:
- A cliff at month 3 usually means an onboarding or expectation-setting problem
- A cliff at month 12 usually means your contracts are annual and you're losing renewals you thought were automatic
- Newer cohorts retaining worse than older ones is a warning that your product or delivery has slipped, or your acquisition channel has changed
You don't need fancy software. A Google Sheet updated weekly is enough. What matters is looking at it every week and asking why the numbers moved.
Weekly action: Refresh the cohort table every Wednesday. Any cohort that dropped a client that week gets a five-minute conversation about why.
5. Lead-to-consult rate
Most founders track leads. Fewer track consults booked. Almost none track the ratio between the two, week by week. That ratio is one of the fastest early warning systems in your business.
If your lead-to-consult rate drops from 34% to 22% over three weeks, something in your top-of-funnel changed. Maybe a new lead source is bringing in unqualified traffic. Maybe your intake form got broken. Maybe the person responding to inbounds is 48 hours slow. You'll spot the problem now, not two months from now when revenue tanks.
The reverse is also useful. If your rate climbs from 30% to 45%, figure out what's working and pour more into it. Most growth in service businesses doesn't come from finding new channels. It comes from noticing what's already working and doing more of it, faster than anyone else does.
One note on definitions. A lead is anyone who raised their hand. A consult is a scheduled meeting on the calendar. Don't count discovery calls that got rescheduled twice or leads who ghosted after one email. Be strict. The number is only useful if the definition is clean.
Weekly action: Every Friday, count leads received and consults booked from that week's cohort. Track the rolling four-week average. Investigate any two-week swing greater than 10 points.
How to actually run the weekly review
Five KPIs. Thirty minutes. Same time every week. That's the whole system.
The mistake most operators make is treating weekly business metrics like a report card. They look at the numbers, feel good or bad, and move on. That's not a review. That's a mood.
A real review has three parts. First, what changed and why. Not just the direction of the number, but the actual cause. Second, what you're going to do about it this week. Not next quarter. This week. Third, what you'd expect to see next week if your action worked. That last part is what turns a review into a feedback loop.
The other thing worth saying: someone has to own each of these numbers. Not the founder for all five. If you're the person on the hook for LTV, margin, DSO, retention, and lead-to-consult, you're also the person who won't get to any of them because you're delivering client work. Assign each metric to a person. Give them the authority to fix what's broken. Then review together on Friday.
Most of the operators we work with find that once they instrument this properly, their biggest constraint isn't visibility. It's capacity. They can see the problems now. They just don't have the hours to fix them, because they're still processing invoices and editing videos and cleaning inboxes at 9pm. That's a different conversation, and it usually leads to the same answer: hire before you're ready, but only into roles the numbers tell you are worth it.
Start with the five metrics. Give it four weeks. You'll be running a different business by the end of the month.