15 KPIs for a Coaching Business (and the Number Behind Each)
Most coaches can tell you how last month felt. Busy, or quiet, or stressful in a way that did not match the bank balance. Far fewer can say how many people asked about coaching, how many got on a call, and how many who finished a package came back for a second one. Those three numbers change what you do on Monday. The feeling does not.
This article covers the coaching business KPIs that matter for solo and small-team coaches running a real practice: somewhere between three and fifty clients, a mix of one to one work, packages and maybe a group program. A "good" KPI here means three things. It is countable without guessing, it moves before your revenue moves, and you would do something differently if it went the wrong way for two months running. Anything that fails those tests is a vanity number, and there is a section below on which ones to stop tracking.
One thing to say plainly first. Every figure in this article is a worked example, invented to show the arithmetic. None of them is an industry benchmark, and be suspicious of anyone who quotes you one for coaching, because the published samples are small, self-selected and rarely comparable to your practice. Your own numbers from the last six months are the only benchmark worth having. Coachful facts and pricing here were checked on 15 September 2026.
How do you choose which coaching KPIs to track?
Pick one number for each stage a client passes through, and no more than that. A coaching business has four stages, so four to six live KPIs is usually the right size for a solo practice, and fifteen is a menu to choose from rather than a dashboard to build.
The four stages are the spine of the rest of this article:
- Demand. People who want to talk to you. If this dries up you find out about it three months late.
- Money. What you charge, what you actually collect, and what an hour of delivery is worth.
- Delivery. Whether clients show up and do the work. This is the stage where coaching differs structurally from consulting.
- Retention. Whether they stay, renew, or send someone else.
A useful test before you commit to any KPI: write down, in one sentence, what you would do if it dropped by a quarter for two months. If you cannot finish the sentence, do not track the number. "Instagram followers fell by a quarter" has no ending. "Discovery calls booked fell by a quarter" ends with "I go back to the three referral partners who sent people in the spring."
The other thing separating a KPI from a statistic is a denominator. "Six new clients" tells you nothing. "Six new clients from nineteen discovery calls" tells you where the work is, so every KPI below is written as a ratio wherever a ratio is possible.
Four stages, not fifteen dashboards. A solo practice needs one number per stage, which means four to six live KPIs, and the other nine on this list are there for the month you have a specific question.
Which KPIs tell you where new business comes from?
Four numbers cover the front of the practice: how many people ask, how many get on a call, how many buy, and how many of them came from a person rather than a platform. Track these monthly, because weekly numbers in a small practice bounce around too much to read.
1. New enquiries per month
What it is. Every person who raised a hand in a month, counted once, with a source attached. A contact form, a DM that asked about availability, a referral introduction by email, a lead magnet download that later replied. Count the human, not the click.
When to use it. Always. This is the one number that goes wrong first and shows up in revenue last, because a coaching sale usually takes several weeks from first contact to first payment.
Worked example. March: 22 enquiries. April: 19. May: 11. Nothing in the revenue line changed until July, by which point the cause, a referral partner who stopped running her workshop in April, was three months cold.
The mistake to avoid. Counting traffic instead of people. An enquiry has a name and an email address, so you can go back and ask it what happened. A pageview cannot answer.
2. Enquiry to discovery call rate
What it is. Of the people who raised a hand, how many ended up on a scheduled call. Divide calls booked by enquiries in the same month and write it as a percentage.
When to use it. When enquiries are healthy but nothing is converting. This ratio separates a marketing problem from a booking problem, and they have completely different fixes.
Worked example. 19 enquiries, 8 calls booked, so 42 percent. The eleven who did not book were all people who had been asked to reply with three times that suited them. After switching to a booking link in the first reply, the same practice put 13 of 19 on the calendar the following month.
The mistake to avoid. Counting a call as booked when it never happened. Keep no-shows separate: a low show rate is a reminder problem, a low booking rate is a friction problem.
3. Discovery call to paying client rate
What it is. Of the calls that actually happened, how many ended with money. Completed calls as the denominator, first payments as the numerator, measured in the month the call happened rather than the month the payment landed.
When to use it. Every month, and especially before you raise prices. A rate that is very high is not automatically good news, and there is more on that below.
Worked example. Ten calls held, four became clients, so 40 percent. The coach raised her package from $1,200 to $1,800 and the rate fell to 30 percent: three clients from ten calls. Revenue from the same ten calls went from $4,800 to $5,400, and she got two evenings back.
The mistake to avoid. Treating a falling rate as a failure without checking revenue per call. If yours sits above roughly four in five over a long stretch, the honest question is whether you are under-priced.
4. Referral share of new clients
What it is. The proportion of new clients in a period who came from an existing or former client, or from someone who knows your work. Referrals divided by all new clients.
When to use it. Quarterly. It moves slowly and it is the closest thing a practice has to a satisfaction score that costs nothing to collect.
Worked example. Q1: 7 new clients, 2 referred, so 29 percent. Q2 after adding a single question to the end-of-engagement call ("who else is in the spot you were in six months ago?"): 6 new clients, 4 referred. Same client volume, much lower acquisition effort.
The mistake to avoid. Not asking where someone came from, then guessing at the end of the quarter. "How did you hear about me?" with a free text box on your intake form is enough.
Which KPIs tell you whether the money actually works?
Four numbers, and the gap between two of them is where most coaching businesses get into trouble. Revenue booked and cash collected are not the same figure, and a practice can sign a strong month and still be short.
5. Monthly recurring revenue
What it is. The portion of your income that arrives every month without a new sale: retainers, monthly packages, memberships, active subscriptions. One number, counted on the first of the month.
When to use it. Monthly. It is the best predictor of whether next month will be calm, and the number to watch when deciding whether you can afford to turn down bad-fit work.
Worked example. Seven active monthly clients at $400 and one at $600 gives $3,400 in recurring revenue. Two one-off intensives at $1,500 each in the same month take the total to $6,400, but only $3,400 of that is there again in October without you doing anything.
The mistake to avoid. Counting a twelve-week package as recurring revenue. It is not, unless it renews. Divide the package price by its length if you must include it, and be clear that the figure ends on a known date.
6. Average revenue per client
What it is. Total revenue in a period divided by the number of paying clients in that period. Not per enquiry, not per session, per person who paid you.
When to use it. When you feel busy and underpaid. It is the fastest diagnostic for a practice full of small engagements, and it is the number that moves most when you package your work differently.
Worked example. A demo Coachful practice shows $13,155.30 collected from 9 paying clients, or $1,461.70 each. Collecting the same total from 26 clients at around $506 each is roughly three times the delivery work and three times the admin for the same money.
The mistake to avoid. Averaging across a period where one client bought a $6,000 intensive. Read the median beside the mean below about fifteen clients, because one outlier moves the average enough to mislead you.
7. Cash collected versus revenue booked
What it is. Two figures side by side. Booked is the total contract value you signed this month. Collected is what actually arrived in the account this month. The gap is your payment plans, your invoice terms and your failed cards.
When to use it. Monthly, without exception, if you sell payment plans or invoice corporate clients. This is the KPI that explains a good month with a thin bank balance.
Worked example. March: $24,000 booked across five clients, four of them on three-month payment plans. Cash collected in March: $14,600. The remaining $9,400 arrives in April and May, which is fine as long as you knew it and did not plan a tax payment around the booked figure.
The mistake to avoid. Reporting only booked revenue to yourself. It is the flattering number, it is the one that appears in launch recaps, and it is not the number that pays rent. Track both or track collected.
Two numbers, one gap. In the worked example above, $24,000 booked in March produced $14,600 in the bank that month, and a practice that only watches the booked figure is short by $9,400 without knowing why.
8. Revenue per coaching hour
What it is. Revenue in a period divided by the hours you spent delivering in that period. Count session time, plus the preparation and the notes if you want an honest figure.
When to use it. Twice a year, and before you add a new offer. It is the only KPI here that prices your time rather than your product, and group work looks very different on it than in the revenue line.
Worked example. One to one package: $1,800 for twelve sessions of an hour, plus roughly fifteen minutes of prep and notes each, which is 15 hours for $120 an hour. Group program: $600 a head, eight clients, eight ninety-minute calls plus four hours of build, which is 16 hours for $4,800, or $300 an hour. Same calendar, very different arithmetic.
The mistake to avoid. Leaving out the unpaid hours: sales calls, admin, rescheduling, chasing a payment. The honest version divides by every hour the business consumed.
Which KPIs tell you whether clients are doing the work?
Four numbers about delivery, and this is the group most coaches skip. They lead everything in the retention section, because a client who has stopped showing up decided to leave weeks before they tell you.
9. Session attendance rate
What it is. Sessions attended over sessions scheduled, on a rolling three months. Count a cancellation inside the notice window as attended if you want a kinder version, but count a no-show as a no-show.
When to use it. Monthly at practice level, and per client the moment one person drops below the rest.
Worked example. A demo practice shows 150 sessions delivered at 93 percent attendance. The eleven missed sessions are not spread evenly. Nine of them belong to two clients, which is a retention conversation with two people rather than a policy change for everyone.
The mistake to avoid. Reading the practice average and stopping. The average is almost always fine. Sort by client and look at the bottom three.
10. Between-session completion rate
What it is. Of the things a client agreed to do between sessions, how many they did. Tasks completed over tasks assigned, per client and across the practice.
When to use it. Weekly if your coaching depends on action between calls, which for most coaching it does. It tells you the work is landing before the client does.
Worked example. Practice average 72 percent. One client at 31 percent for three weeks running. In the session, she says everything is going well. The number says otherwise, and it gives you something concrete to ask about instead of "how are you finding it?"
The mistake to avoid. Assigning so much that the rate has nowhere to go but down. If your completion rate is under half, look at the size of what you are asking for before you look at the client's motivation.
11. Clients who have gone quiet
What it is. A count, not a rate. How many clients have not checked in, completed a task, replied to a message or attended a session in the last fourteen days. One integer, reviewed every Monday.
When to use it. Weekly. It is the most actionable number in this article, because the action is obvious: send those people a message today.
Worked example. A practice with 10 clients shows an 81 percent check-in rate and 2 clients flagged as needing attention. Those two are the entire Monday to-do list. The other eight need nothing.
The mistake to avoid. Turning it into a percentage. "Ninety-two percent of clients are engaged" is comfortable. "Two people have not spoken to me in a fortnight" makes you pick up the phone.
12. Goal completion rate
What it is. Of the goals clients set with you, how many they reach by the date they set. Measured at the end of an engagement rather than continuously.
When to use it. Quarterly. It is the raw material for every case study you will write, and it tells you whether your intake picks the right goals.
Worked example. Of 18 goals set across 6 clients in a quarter, 11 were reached by the target date, 4 were changed partway through and 3 were missed. The four that changed are the interesting ones: all four were set in week one, before the client understood what the real problem was.
The mistake to avoid. Scoring yourself on outcomes you do not control. "Get promoted" depends on other people. Track the behavior goal underneath it and read them together.
The quiet client leaves first. In the worked example above, a practice at 81 percent check-in rate still has 2 people who have said nothing for two weeks, and those 2 are the whole Monday list.
Which KPIs tell you whether clients stay?
Three numbers, reviewed quarterly, that decide whether you are building a practice or running on a treadmill. A business that renews half its clients sells half as much as one that renews none.
13. Renewal rate
What it is. Of the clients who reached the end of an engagement in a period, how many bought again. Include a smaller maintenance package as a renewal; exclude anyone whose engagement is still running.
When to use it. Quarterly, with the denominator written next to it. In a practice with eight endings a quarter, one extra renewal moves the rate by more than twelve points, so the percentage on its own is noisy.
Worked example. Q2: 8 engagements ended, 3 renewed, so 38 percent. All three were offered a specific next package in the second-to-last session. Four of the five who did not renew were never offered anything, because the coach did not want to seem pushy in a final call.
The mistake to avoid. Measuring it without a renewal conversation in the process. A rate you never tried to influence measures luck.
14. Average engagement length
What it is. How many months or sessions the typical client stays, from first payment to last. Use the median, not the mean, because one three-year client distorts a small practice badly.
When to use it. Twice a year. It is the multiplier in lifetime value, and it is the number that tells you how many new clients a year you actually need.
Worked example. Median engagement of 5 months at $400 a month means a client is worth about $2,000. If you want $60,000 a year from one to one work, that is 30 clients a year. Getting the median to 7 months takes the same revenue down to 22 clients a year, which is eight fewer sales conversations.
The mistake to avoid. Counting clients who are still active. They drag the median down because their engagement has not finished. Measure completed relationships only.
15. Client lifetime value
What it is. Average revenue per month per client, multiplied by median engagement length, minus whatever it costs you to deliver and to acquire. For most solo coaches, the delivery cost is your time and the acquisition cost is small, so the gross figure is close enough to be useful.
When to use it. Twice a year, and any time you are deciding whether to spend money on getting clients. It sets the ceiling on what an enquiry is worth.
Worked example. $400 a month, median 5 months, renewal rate 38 percent. The base value is $2,000, and the renewals add roughly $760 on average across the group, so about $2,760 per client. If a workshop costs $300 to run and produces two clients, that is a good trade at almost any attendance number.
The mistake to avoid. Using it as a marketing number. Lifetime value is a planning input, not a thing to be proud of, and inflating it by including one exceptional client is a fast way to overspend on acquisition.
Renewals beat new sales. In the worked example, moving the median engagement from 5 months to 7 takes the annual client target from 30 down to 22 at the same revenue.
How do you track these KPIs in Coachful?
Most of them are already counted, because they are side effects of running the practice rather than a separate reporting job. Coachful is coaching software for solo coaches and small teams who deliver programs rather than one-off sessions: clients, bookings, payments, programs, habits, check-ins, community, website and email in one subscription. It costs $29 a month on Lite for up to 5 clients, $49 Solo for up to 20, $99 Pro for unlimited clients, $199 Studio for 3 coach seats and $299 Agency for 6, with a 0% platform fee on what you sell and a 7-day trial that requires a card and charges nothing during the trial. It does not have a free plan, it is GDPR-ready but not HIPAA-certified, and it is not an accounting system, so your profit and loss still lives with your bookkeeper.
Here is where each group of numbers lives.
Demand. The Prospects page holds every lead with a date, a source channel and a pipeline stage, which gives you enquiries per month and referral share without a spreadsheet. The Intakes page counts discovery calls with a status on each, so booked, completed, no-show and converted stay separate figures. If you are building the front end of this, our notes on selling coaching packages cover where those leads come from.
Money. The Analytics tab reports revenue collected, paying clients, average per paying client, active subscriptions, sessions delivered and attendance, plus revenue by source. The Payments tab lists invoices with amounts and statuses, which is the collected side of the pair.
Delivery. The Clients page carries task completion, check-in rate, average mood and a Needs Attention group, which is the quiet-client count in the form you want it: a short list of names. Habits and programs give the per-client rates underneath the average.
Retention. Sessions delivered and unscheduled session credits show what is left to deliver on packages already paid for, and session and enrollment history gives you engagement length. Renewal rate stays a judgment call, because only you know whether a smaller follow-on package counts.
Two honest notes. There is no single screen showing all fifteen at once; they sit where they are generated. And for a number nobody is calculating for you, the export is the answer: a full data ZIP with the roster, notes, check-ins, goals, habits, sessions and form responses, so you can do whatever arithmetic you like in a spreadsheet. For the wider market, see coaching practice management software and what coaching software costs.
What does one KPI review afternoon actually look like?
About ninety minutes, once a month, on a fixed date. Long enough to get real numbers, short enough that you will do it again next month, and the fixed date matters more than the length because a review that happens when you feel like it happens twice a year.
Here is a workable shape. Assume a practice with 14 active clients and a mix of packages and monthly retainers.
Minutes 0 to 15, demand. Filter the leads list to last month, count the rows, note the source on each, and write two numbers: enquiries and referral share. Open intakes and write three more: calls booked, calls held, calls converted. Five numbers, no interpretation yet.
Minutes 15 to 35, money. Collected revenue from the payments list, booked revenue from whatever you signed, collected divided by paying clients for average revenue per client, and the sum of active subscriptions and retainers for recurring revenue. Four numbers.
Minutes 35 to 55, delivery. Practice attendance rate and task completion rate, then sort by client and write down the names at the bottom of each. Count the people who have gone quiet. You finish this block with two numbers and a list of names.
Minutes 55 to 70, the only part that matters. Look at the five numbers that changed most since last month and pick one to act on. Not three. One, with a date on it.
Minutes 70 to 90, the names. Message everyone on the quiet list and everyone whose engagement ends in the next six weeks. That second group is your renewal conversation, and doing it in the same sitting is what turns a report into revenue.
Retention does not belong in this afternoon. Renewal rate, median engagement length and lifetime value move too slowly to read monthly, so do those four times a year.
Which KPIs should you stop tracking?
Most of them, and specifically anything with no denominator, no owner and no possible response. Four categories show up constantly in coaching businesses and none of them earns the attention.
Audience size. Followers, subscribers, list size on their own. These are inputs to enquiries, so check them only when enquiries fall. A list of 4,000 producing two enquiries a month is a worse asset than a list of 300 producing six, and only the enquiry number tells you that.
Hours worked. Effort is not an outcome. Revenue per coaching hour is the version with a denominator, and it usually tells you to work fewer hours rather than more.
Session count without attendance. "I delivered 40 sessions" hides the six no-shows and the four that got rescheduled twice. Sessions delivered over sessions scheduled is a real number; sessions delivered on its own is a diary entry.
Anything you cannot collect in under five minutes. A KPI that needs a two-hour reconciliation gets tracked twice and then abandoned, and an abandoned KPI is worse than none because you keep half-remembering the old figure.
One more trap, and it is subtler. Tracking a number you have no lever for creates anxiety without changing anything. Client results are the honest example: you do not control whether someone gets the promotion, and you do control whether they showed up, did the between-session work and set a well-formed goal. Track those three and let the outcome be the outcome.
If you only track three coaching KPIs, which three?
Discovery call to paying client rate, cash collected, and the count of clients who have gone quiet. Those three cover the front, the middle and the back of the business, they take about ten minutes a month between them, and each one has an obvious action attached when it moves the wrong way.
Three exceptions. If you run a group program or a membership, swap the quiet-client count for between-session completion rate across the cohort, because a group goes quiet collectively and you will not notice person by person. If your work is mostly corporate or invoiced, cash collected needs the booked figure beside it or you will misread every month with a thirty-day invoice in it. And in your first year with fewer than ten clients, ratios are close to meaningless at that sample size, so count raw numbers and wait until you have thirty or forty data points before calculating a percentage.
Once those three are stable, add one at a time, and only when you have a question they cannot answer. "Why is revenue flat when I am busier?" adds average revenue per client. "Why do I keep having to sell?" adds renewal rate and median engagement length. The other twelve are there for when the question arrives, not for a dashboard you check every morning.
Frequently asked questions
What are the most important KPIs for a coaching business?
Discovery call to client conversion rate, cash collected, and the count of clients who have gone quiet in the last two weeks. Those three cover demand, money and delivery, and each has a direct action when it moves. Add average revenue per client when you feel busy and underpaid, and renewal rate when you are always selling.
How many KPIs should a solo coach track?
Four to six, reviewed monthly, plus three retention numbers reviewed quarterly. More than that and the review stops happening. The limit is not attention, it is collection time: a KPI you can read off a screen in thirty seconds survives, and one that needs a spreadsheet reconciliation gets abandoned. Start with three and add one at a time.
How often should I review my coaching KPIs?
Demand and money monthly, on a fixed date. Delivery weekly, but only the quiet-client list, which takes two minutes. Retention quarterly, because renewal rate and median engagement length move too slowly to read monthly and the noise will mislead you. A ninety-minute monthly review plus a two-minute Monday check covers a practice under fifty clients.
What is a good conversion rate for discovery calls?
There is no published figure for coaching that we would stand behind, and anyone quoting one is almost certainly quoting a small self-selected sample from a different price point than yours. Use your own last six months. Direction and revenue per call matter more than the level: a rate that falls from 40 to 30 percent after a price rise can still mean more money from the same calendar.
What is the difference between revenue booked and cash collected?
Booked is the total contract value you signed in a period. Collected is what arrived in the bank. Payment plans, invoice terms and failed cards sit in the gap. In the worked example above, $24,000 booked in March produced $14,600 collected, with $9,400 arriving across April and May. Never plan a tax payment around the booked figure.
How do I calculate client lifetime value for coaching?
Average revenue per client per month, multiplied by your median engagement length in months, then add the effect of renewals. At $400 a month, a 5-month median and a 38 percent renewal rate, that is roughly $2,760. Use the median rather than the mean below about fifteen clients, because one large engagement distorts the average badly. It is a planning input for acquisition spend, not a number to quote.
Should I track client results as a KPI?
Track the behavior underneath the result, not the result itself. Whether a client gets promoted depends on other people; whether they attended, completed the between-session work and set a well-formed goal depends on the coaching. Goal completion rate at the end of an engagement is worth a quarterly look because it feeds your case studies, but read it alongside the goals that got changed partway through, which usually says more about your intake than your coaching.
Do I need software to track coaching KPIs?
No, and a spreadsheet with twelve rows works for a practice under ten clients. Software helps in one specific way: it counts the delivery numbers as a side effect of the work, so attendance, task completion and quiet clients are already there. If you are collecting by hand, put a "how did you hear about me" field on your intake form first, because that one field is the hardest data to reconstruct later.
What KPIs matter most for a group coaching program?
Cohort-level between-session completion rate, call attendance across the cohort, and revenue per delivery hour. Individual quiet-client tracking breaks down in a group because people go quiet collectively. Revenue per delivery hour is the number that makes the case for group work: in the worked example above, a group program produced $300 an hour against $120 for the equivalent one to one package.
Join Coachful now
Most of these fifteen numbers are things your practice already produces and nobody is writing down. Coachful counts the delivery side while you work, keeps demand and payments in one place, and hands you a full export when you want to do the arithmetic yourself.
Start the 7-day trial, run one monthly review inside it, and see how many of the fifteen were already sitting there. If it is not for you, cancel in one click.
