Translating numbers into something a business acts on

A finding nobody acts on was written for the wrong reader. A report's job is not to prove that the analysis happened; it is to make one person able to decide what to do on Monday.

This page is about the last step of an audit — the translation. The readings themselves live elsewhere: metrics and behaviour for what readers did, content detectors for what is wrong with the text, goals and funnels for what the owner said should happen, and did it work? for whether the last change helped.

What shape does a finding have to take?#

Every line answers four questions: what to fix, why this one, who fixes it, how we will know it helped.

Period: last 7 days · 1,240 visits · source tier: full (behaviour + ranked digest)

#1  /billing/invoices — health 31/100 (readers giving up dominates; 3 dislikes)
    Why now:   412 readers, 58 gave up here — the largest single loss in the set
    Effort:    M (rewrite of one section, no new page)
    Effect:    dead-end rate on this page 14% → under 8% within two weeks
    Owner:     the billing docs owner — rewrite brief, after the journeys are read

#2  "sso saml setup" — unanswered 23×, zero results 11× (two independent signals agree)
    Why now:   no page exists; workspace impact: high
    Effort:    L (new page)
    Effect:    the query stops appearing in unanswered questions within one release
    Owner:     whoever owns the authentication docs

#3  /quickstart — health 44/100 (dislikes dominate; 9 of 11 votes negative)
    Why now:   highest-traffic page in the set; a correctness problem, not a gap
    Effort:    S (accuracy pass on two steps)
    Effect:    assistant satisfaction on this page recovers above the warning line
    Owner:     the quickstart owner

Not this week: 14 further pages below the cut, all under 40 visits in the period.
Excluded as noise: 6 pages scoring badly on fewer than 5 events.

The source tier line on the header is part of the finding, not decoration: it says how much data the run actually had, which is what stops a thin week being read as a strong result — see metrics.

The rules behind that block:

• Absolute counts next to every rate. "14%" is unarguable and unactionable; "58 of 412 readers" is neither.
• Effort as S/M/L, with the reason in parentheses. The reason is what stops an L being negotiated down to an S in a meeting.
• The effect line names one number and a horizon. Two numbers is a wish.
• Every line ends with a named owner, or with an explicit "nobody obvious owns this — here is the work in one sentence". A line without an executor is not a plan, it is a complaint.

Five items is a plan. Twenty is a backlog dump that gets ignored. Everything below the cut is one line with a count, so the reader can see that it was considered and set aside rather than missed.

How do I say it in the reader's terms rather than the metric's?#

The pattern in every row is the same: replace the ratio with the two counts it came from, and say what the reader was trying to do when the number was recorded.

Quote the reader#

A query in someone's own words moves an author more than any aggregate, and they can verify it without rerunning the analysis. Wherever a reader wrote something — a search, a question to the assistant, a piece of feedback — quote it verbatim, with the count of separate visits behind it. User language is the fuller reading; here the quote is simply the cheapest way to make a number believable.

Which conversions must you refuse?#

• Never convert findings into money, pipeline, or deflected support tickets on the owner's behalf. Deflection counts a reader who gave up the same as one who got help. If the owner wants a figure, ask them for their cost per ticket and label the result as their assumption.
• Never restate a buying-stage count as intent or revenue. It is a count of conversations classified from text.
• Never restate someone else's impact estimate in your own units. Pass it through as given, and label your own estimates as yours.
• Never present impressions as an audience size. They are counted per query.
• Never promise a search position. You can promise a better pitch to people already being shown the page — which is a different and much more defensible claim, because impressions, average position and click-through are three separate failures wearing one word, and only the third is a pitch problem (Google Search Central, core updates).

Measured or hypothesis, on every line#

Findings backed by data carry the dated window and the numbers behind them. Findings without carry hypothesis and say what the missing data would have settled. Keep the two visibly separated — a reader who cannot tell which is which discounts both.

"This title is 74 characters" is a fact. "This title is a label, not a search intent" is a hypothesis. Length, duplication, missing frontmatter, heading skips, missing alt text and orphan pages are all verifiable from the text alone and stay facts even with no analytics at all.

Where does the machine format go?#

Underneath. Structured findings exist so that a tool can consume them — the action and constraints fields are instructions for a program, not prose for a person. Never paste that JSON into a reply as "the report": a wall of "severity": "high" reads as a system error and buries the one line that mattered.

When a person asked, answer the way an editor would: the worst problem first, in plain language, with the before and after, and an offer to apply the fixes. If the surface can render changes for approval, that is the answer; the machine format stays under it.

{
  "file": "docs/api/authentication.md",
  "line": 2,
  "severity": "critical",
  "rule": "position-5-20-no-clicks",
  "confidence": "measured",
  "evidence": {
    "window": "2026-07-01..2026-07-29",
    "query": "how to authenticate api requests",
    "position": 8.4,
    "impressions": 1840,
    "clicks": 11
  },
  "found": "Ranks at 8.4 for 'how to authenticate api requests' — 1,840 people saw this result and 11 clicked. The title shown to them is 'Authentication', which shares no words with what they typed.",
  "suggestion": "title: 'How to authenticate API requests | Product' (52 chars, matches the query the page already ranks for). The page is already visible; only the click is missing.",
  "action": "rewrite_title",
  "constraints": { "max_length": 60 }
}

Related#

• From finding to change — what happens to a line of this report once someone accepts it.
• Did it work? — the run that comes back and checks the effect line you promised.
• Choosing a lens — deciding which reading produces the findings in the first place.
• Analytics — where the numbers on the left-hand side of the translation table come from.
• Evidence — the same measured-or-hypothesis distinction, graded across the whole handbook.