There's a long-standing tension in coaching between the practitioners who think measurement is a distraction from the relationship and the academics who think practitioners are running blind. Both have a point, and both miss something.
What most working coaches actually need isn't a research apparatus. It's a small, well-chosen set of measures, each running on its own cadence, that turn the coaching from a series of well-intentioned conversations into something with visible signal. This is what that looks like in practice.
The problem with no measurement
When coaching is unmeasured, four things go wrong. They go wrong slowly enough that experienced coaches sometimes don't notice the drift.
You can't see plateaus. A client who stops progressing in months 3-4 of an engagement looks the same in session as a client who's still moving. Without a goal-progress signal, you assume continued momentum because the conversations stay engaging.
You can't see fit problems early. A client who's losing trust in the work doesn't usually announce it. They show up, they speak less, they say "good session" at the end, and they don't renew. You find out two months later, when they're gone.
You can't see your own patterns. Coaches who don't measure don't know whether their clients tend to plateau in month 3, whether their alliance scores trend down with women clients but not men (or vice versa), whether their session-fit scores drop after a long engagement. The patterns are there; without data, they're invisible.
You can't tell good months from lucky ones. Some quarters are great because the work is great. Some are great because you happened to have unusually motivated clients. Without baseline data, you can't tell.
The problem with too much measurement
The opposite trap — measuring everything — makes it worse. A coach who hands every client a 50-item assessment at intake is signaling something incompatible with most coaching engagements: that the relationship is going to feel like a study. Most clients quietly disengage from that.
The right calibration is to measure enough to surface signal, on cadences that don't dominate the work, with instruments brief enough that clients fill them honestly.
A working measurement stack
A pragmatic measurement stack for solo coaching has four layers, each on a different cadence. You don't need to invent any of these — they're all validated, they're all in active clinical use, and they all run in under a minute per session.
1. Goals — every session
The headline measure of coaching outcome is whether the client's goals are moving.
The right tool is Goal Attainment Scaling. For each goal, you write five anchored levels (−2 to +2), agree them with the client, and score against them every session. The signal is the trajectory across sessions, not the absolute score in any one session.
Cost: 3-5 minutes per session, mostly during the wrap-up.
What it tells you: whether the work is actually moving the things the engagement is supposedly about.
2. Session fit — every session
Pre- and post-session, two ultra-brief checks. Pre is the Outcome Rating Scale — how the client is arriving (well-being, relationships, work, overall). Post is the Session Rating Scale — how the session itself landed (relationship, topics, approach, fit).
The two are deliberately separate because combining them hides whether the session itself added anything beyond how the client was already doing.
Cost: 30 seconds before, 30 seconds after. Client fills both.
What it tells you: whether each session is landing, and how the client is doing in their life independent of the work.
3. Alliance — every fourth session
The Working Alliance Inventory check. Three short questions on bond, shared goals, and useful work together.
Why fourth-session cadence: every-session burns the salience (clients tune out); annual loses early signal. Every fourth session catches alliance ruptures while they're still fresh.
Cost: 30 seconds, every 4-5 sessions.
What it tells you: whether the relationship itself is holding, fragmenting, or strengthening.
4. Domain progress — coach's choice, every 4-8 sessions
A domain-specific instrument that measures the underlying capacity the engagement is built around. Different by domain:
- Life coaching: WHO-5 Well-Being Index. Five short items, scored 0-100, well-validated as a general well-being indicator.
- Career coaching: CAAS — Career Adapt-Abilities Scale. Four dimensions (concern, control, curiosity, confidence), 24 items in the full version, can be shortened.
- Executive coaching: the alliance check is often the primary domain measure, sometimes supplemented with stakeholder-feedback pulses.
Cost: 1-2 minutes, every 4-8 sessions.
What it tells you: whether the underlying capacity (well-being, career adaptability, executive trust) is shifting in the direction the engagement is supposed to move it.
The combined signal
Each layer alone is partial. Together, they triangulate.
A client whose goal scores are moving up, session fit is high, alliance is solid, and domain measure is climbing — the work is doing what it's supposed to.
A client whose goal scores are flat but session fit is high — you have a relationship problem disguising itself as a content problem. The client likes you and likes the sessions but isn't actually changing. This is the classic "comfortable rut."
A client whose goal scores are climbing but session fit is dropping — you're getting results but the client isn't enjoying it. Sustainable for a few months, unsustainable across an engagement. Worth bringing into the alliance check.
A client whose alliance is dropping but goals and session fit are fine — something specific has happened in the relationship that the in-session feedback isn't catching. Time for a direct conversation.
A client whose domain measure is moving but goals and alliance look mediocre — the work is helping in ways neither of you can yet name. This is sometimes a sign that the engagement's stated goals were the wrong ones; the real work is happening elsewhere.
The discipline isn't to read each measure in isolation but to read the combination.
What to do when the signal looks bad
A common worry coaches have when starting to measure: "what if the data tells me my coaching is bad?"
In practice, that's almost never what the data says. Far more often, the data points at one of:
A specific client where something is off — usually one client out of the roster. Almost always traceable to a specific incident or shift, and almost always recoverable when surfaced directly.
A pattern across clients in a specific phase of the engagement — for example, alliance scores reliably dipping in months 3-4. This usually responds to a small process change: a deliberate "are we still working on the right thing?" check at the 12-week mark.
A pattern in a specific domain — your career-coaching engagements move faster than your life-coaching ones, or vice versa. Worth knowing about; rarely worth changing.
The data almost never says "you're a bad coach." It says "here are the three specific things to attend to differently this quarter."
How to introduce measurement to a client
If you're adding measurement to an existing practice, the framing matters. Most clients tolerate it well; some resist if it's introduced as research. The framing that works in our experience:
"I'm going to ask you to fill out two short check-ins each session — about 30 seconds each. They help me notice signals I'd otherwise miss, and they'll show you the arc of how things are moving over time. They're for both of us."
Three things make this work:
- Frame it as for both of you. A client who thinks the measurement is for the coach's research project will fill it in dutifully but disengage from it. A client who thinks the measurement is also for them — to see their own arc — engages.
- Be honest about why. Coaches who say "we just always do this" lose the framing. Coaches who say "I want to know what's actually working" gain trust.
- Show the trend. Most clients have never seen their own well-being or session-fit data over time. Showing them the arc, even briefly at month 3 of the engagement, changes how they hold the work.
Measurement that's not worth doing
Three traps worth flagging:
Personality tests at intake. MBTI, DISC, Big Five. Some coaches love them. The data is rarely actionable, and they signal "this is going to feel like a corporate consult." Skip unless your engagement specifically calls for one.
ROI calculations. Some executive coaching contexts demand them. Outside of that, ROI of coaching is hard to compute honestly and tends to overclaim. If a stakeholder demands it, give them goal-progress data instead — it's more honest and more defensible.
Open-ended satisfaction surveys at the end of an engagement. "How was your experience?" produces nice quotes for marketing and almost no actionable data. Replace with the running ORS+SRS+alliance trend, which is actually informative.
A coach's first quarter with measurement
If you're starting from zero, a reasonable adoption arc:
Month 1. Start ORS+SRS for every client. That's the lowest-friction layer. Just the act of running them changes the alliance even before you analyze the data.
Month 2. Start writing GAS-anchored goals with new clients. Don't try to retrofit existing clients' goals into GAS — wait for natural goal pivots.
Month 3. Start the WAI alliance check at session 4 of each engagement. Run on the every-fourth-session cadence after that.
Month 4-6. Add the domain measure (WHO-5 / CAAS) as appropriate. By now you have enough data on the first three layers to start reading the patterns.
After six months, you'll have an honest picture of your practice for the first time. Most coaches find some patterns to be proud of, some patterns to fix, and a few that surprise them.
How Arcline handles this
Arcline's measurement stack is exactly the four layers above:
- Goals. GAS-anchored, scored every session.
- Session fit. ORS pre-session, SRS post-session, kept separate.
- Alliance. WAI-SR, three questions, every fourth session.
- Domain. WHO-5 / CAAS / WAI as the domain anchor depending on the engagement domain.
Each instrument is presented to the client in plain language; the formal names are visible only as secondary labels. Coaches see the trends on a single roster screen. Clients see their own trends on their My Progress page.
The point isn't to make the practice feel clinical. It's to put a small amount of structure under the relationship so the coach can see what they couldn't otherwise see. Measurement done well makes the work more humane, not less.
If you want to try the stack without committing, a 14-day trial gives you a full coaching cycle with all four layers running, no credit card required.