River
Y CombinatorBacked by Y Combinator
FREE TEMPLATE

Coaching Program Evaluation Template

Two documents and three sheets that isolate a coaching programme's own share of a business result before the number gets reported.

Free download  ·  No account needed

Business Indicator Correlation  ·  Hollins Systems, six-manager coaching cohort  ·  naive vs. adjusted

Naive credits every improvement in full. Adjusted isolates what coaching can actually claim.

ManagerRaw ChangeIsolationConfidenceAdjustedStatus
Simon Vetch+$84,00050%80%+$33,600Included
Lena Marchetti$0n/an/a$0No change
Aldric Whitfield+$42,00030%90%+$11,340Included
Tobias Reyes+$126,00090%95%$0Excluded
Miriam Solano+$42,00040%70%+$11,760Included
Jonas Brandt-$42,000n/an/a-$42,000Negative, kept

Naive total $294,000 against a $54,000 programme cost is a 444% return. Adjusted total $14,700 against the same cost is -73%. The gap is one excluded claim and one kept decline, not a rounding difference.

Search this and one common result is a calculator. Enter the coaching fee, multiply it by a fixed number, and the percentage comes back the same no matter what happened afterward. The industry's own headline figure works the same way: ICF's 2009 Global Coaching Client Study reports a 700% median company return, two sentences after noting that only 9% of respondents supplied enough data to calculate one at all. Neither number asks what else in the business might explain it before the programme gets the credit.

This pack follows the fix instead of the shortcut. The Phillips ROI Methodology that formula descends from sets its own guiding principles. Isolate a programme's real share before crediting it, treat a measure with no data as flat, and exclude an extreme, unsupported claim instead of rounding it down. A per-participant sheet turns each rule into a column: a competing factor named first, then an isolation estimate, then a confidence percentage, multiplied through. It is the same discipline a fractional role's own commitment log applies to calls, turned here on a business result.

At Hollins Systems, a fictional B2B SaaS company, six coached managers were meant to reduce regrettable attrition on their own teams. Crediting every improvement in full, with no exclusions, prices the programme at a 444% return. One manager's 90%-confident estimate gets excluded, because a same-quarter pay correction he never named explains the result at least as well. A sixth manager's team got worse instead, kept in at its full negative value rather than dropped. The adjusted total is $14,700 against a $54,000 cost, a -73% return the naive number never surfaces.

One manager's math, one excluded claim, and what the report says about both

Simon Vetch's full isolation math, Tobias Reyes's excluded claim, and the Evaluation Report's caveats section, side by side.

Simon Vetch, Manager, Customer Success (East)

Regrettable attrition on his team, baseline window vs. current window

StepCalculationResult
Raw change3 departures to 1, 2 avoided2
Raw dollar value2 × $42,000 standard replacement cost$84,000
Isolation estimateSimon's own estimate of coaching's share50%
After isolation$84,000 × 50%$42,000
Confidence adjustmentSimon's stated confidence in that 50%80%
Adjusted value$42,000 × 80%$33,600

$33,600 counts toward the total, not $84,000. The other $50,400 might be real, but nothing in the record supports crediting it to coaching specifically, so it stays out.

Tobias Reyes, Manager, Sales Development

The largest raw improvement in the cohort, and the one row excluded entirely

FieldValue
Raw change4 departures to 1, 3 avoided
Raw dollar value3 × $42,000 = $126,000
Tobias's isolation estimate90%
Tobias's stated confidence95%
Competing factor Tobias namedNone
Competing factor on recordSales Ops corrected his team's base-salary band the same quarter
Result$0, excluded

A 90% claim at 95% confidence, with a known concurrent pay correction never mentioned, is exactly what the guiding principle behind this method excludes. Not discounted to a smaller number. Zero, with the contradiction named in the report.

Evaluation Report, caveats section

What the naive 444% number leaves out entirely

ManagerFinding
Tobias ReyesExcluded: 90%/95% claim contradicts an unnamed concurrent pay correction
Jonas BrandtKept in full: team's departures rose, -$42,000, no isolation applied to soften it
Lena MarchettiRecorded as zero: no business change, regardless of her positive reaction score

Three of six managers never reach the naive total's four-figure return at all. A report that only shows the adjusted percentage without naming these three specifically is still hiding the same information the naive number hid, just one layer deeper.

What's in the pack

01

Method Note

Fixes the isolation technique and the standard dollar value's source before any participant's data gets read.

02

Evaluation Report

The sponsor-facing summary: the naive total next to the adjusted one, with every excluded or unfavorable finding named specifically.

03

Baseline versus Current

Reaction and self-assessed learning per participant, the same baseline discipline a development plan runs against observable behaviour.

04

Stakeholder Feedback Change

Behaviour change corroborated by the people around each participant, rather than their own account. The mid-engagement version of the same check is a mini 360.

05

Business Indicator Correlation

The business measure, isolated and confidence-adjusted, dollar by dollar. Reconstructing unbilled time is the same discipline applied to a different leak entirely.

How it works

  1. 1

    Open in River, or take it blank

    Open the pack in River and send the baseline data and business metric this programme targeted, or download the Word documents and CSV sheets and fill them in yourself.

  2. 2

    Fix the method before the data

    Name the business measure, confirm its standard dollar value and source, and agree the isolation technique before a single participant's number gets read.

  3. 3

    Isolate every claimed dollar

    For each participant, name competing factors first, then get their isolation estimate and confidence percentage, and multiply the raw change through both before it counts.

  4. 4

    Report the gap, not just the total

    Put the naive total next to the adjusted one in the Evaluation Report, and name every excluded or unfavorable finding specifically rather than smoothing it into one caveat line.

Frequently asked questions

Is this template free?

Yes. The zip is Word documents and CSV sheets, no account and no card. Edit with AI is the optional half: send the baseline data and business metric, and the agent sets up the sheets and holds every claimed dollar to the isolation rule before it counts. The rest sit in the template library.

What format are the downloaded files?

Two Word documents and three CSV sheets in one zip. The sheets carry the columns that matter: raw change, competing factor named, isolation estimate, confidence percentage, and the adjusted dollar value. Open them in Excel, Numbers or Google Sheets, and the documents are .docx.

Why exclude an extreme isolation estimate instead of just discounting it?

Because a claim that high and that confident, with no competing factor named, usually means one existed and did not come up. Discounting it to something safer-looking still gives it partial credit. Excluding it and naming the specific contradiction in the report is what keeps the evaluation honest.

Where does the standard dollar value for the business measure come from?

Whatever figure the organisation already uses for it: an average replacement cost, an average deal size, sourced from whoever owns that number day to day. Where no standard exists, an expert's estimate from the person closest to the measure works, as long as its source is named in the sheet rather than left blank.

What if a participant's business measure didn't move at all?

It gets recorded as zero. The guiding principle behind this method is explicit that a measure with no improvement data is assumed to have produced none, never estimated upward from a participant's otherwise positive reaction or learning scores.

How is this different from a coaching ROI calculator?

A calculator multiplies the fee by a fixed number and returns the same percentage regardless of what happened. This pack asks, participant by participant, what else could explain the change before coaching gets credit for any of it, and excludes a claim that does not hold up.

Isolate the number before it goes to the sponsor

Take the Word documents and CSV sheets blank, or send River the baseline data and business metric and get sheets that already hold every dollar to the isolation rule.

Edit with AI