A draft can look finished while leaving you unsure whether it helps a customer. An information gain score gives that review a shared set of questions. ROI.LIVE's rubric has five dimensions, each scored from zero to two. The ten-point total is an editorial aid, not a Google metric or a prediction of search position. Its value is in explaining why a page earned each mark.

Use the worksheet when you need to give a writer more useful feedback than a feeling that something is missing. Each mark needs a sentence of evidence from the page. A false claim or the wrong reader question still needs fixing, however attractive the total looks.

The ten-point review

Put it to work

Five judgments, each with a reason

Dimension 0 1 2
Reader decision The task is unclear The question is stated but only partly answered The reader can make the intended decision
Useful contribution Repeats alternatives without improving them Adds explanation or organization Adds a substantiated example, tool, comparison or insight that resolves a gap
Evidence Material claims have no support Support exists but needs context Sources, dates, definitions and limitations are adjacent to the claims
Practical use Advice remains general Steps exist but require guesswork Inputs, sequence, exclusions and next action are clear
Editorial clarity Repetition or ambiguity obscures the answer Usable with edits Easy to scan and read without distorting the facts
This ten-point rubric is an editorial aid, not Google's metric. There is no validated pass threshold. Unsupported high-consequence claims need correction regardless of the total.

Editable resource

Use the worksheet

The CSV opens in Excel, Google Sheets or another spreadsheet editor. It contains labeled fields; sample data, where included, is illustrative.

Download the editable worksheet (CSV)

Preview the fields
  • URL
  • Reader decision
  • Review date
  • Reviewer
  • Reader decision 0-2
  • Contribution 0-2
  • Evidence 0-2
  • Practical use 0-2
  • Clarity 0-2
  • Reason and source
  • Required repair
  • Owner
  • Status

Score one page against one purpose

Record the primary question, intended reader and review date. Read the whole page, including the title, image labels, calculator outputs and FAQ. Then examine a few current alternatives. Include a primary source when the subject has one. A top result is useful context, but ranking alone does not establish accuracy.

Score each dimension and quote the passage behind your mark. Let two reviewers score on their own before comparing the largest differences. Keep the disagreement in the worksheet long enough to understand it. The writer may be answering a different customer's question, or assuming something the owner knows but a buyer would not.

There is no validated pass threshold. Use the score to prioritize edits within a comparable batch. A seven-point glossary entry and a seven-point buying guide may need different work. A safety, pricing or product claim without evidence requires correction regardless of the total.

A worked example: a service quote guide

This example is illustrative. A draft says a website costs between two large numbers and recommends asking an agency for a quote. Its reader decision score is one: the subject is clear, but the owner still cannot compare scope. Its contribution score is zero if the same range appears across the alternatives without a useful explanation.

Help the owner compare what they would be paying for. Add a sample scope separating page design, content writing, data migration, integrations, hosting and maintenance, then calculate the total first-year cost. Label any prices as sample inputs, not market averages or an agency offer. That revision can improve practical use without requiring an original survey.

Evidence remains a separate judgment. A calculator with the wrong formula can look useful while misleading the reader. Review the arithmetic before raising the score.

What the score cannot tell you

It cannot establish authorship, plagiarism, indexing, conversion lift or an AI citation probability. Those questions require different evidence. An AI detector's percentage is not a substitute for checking whether the claimed experience happened.

The definition of information gain explains why the editorial concept must be separated from the Google patent. Keep that explanation in any client report that uses the term. Calling a private rubric a search-engine score creates false precision.

Turn the marks into a repair list

A worked example

Turn a vague score into an editable instruction

Vague

Needs more depth.

Concrete

Show which migration work is excluded from the quoted price.

Vague

Improve authority.

Concrete

Link the product rating to the manufacturer's current specification and identify the exact model.

These are examples of editorial instructions. A higher review score does not prove that the edit caused a later traffic change.
Read the full explanation

Assign one owner and one concrete change for each weakness. Replace “needs more depth” with “show which migration work is excluded from the quoted price.” Replace “improve authority” with “link the product rating to the manufacturer's current specification and identify the exact model.”

Rescore the revised page and keep the original row so you can see whether the writer addressed the problem. Track search and business outcomes separately over time. A higher editorial score means the page better meets this review method; it does not prove the edit caused a traffic change. The content audit process applies the same distinction across a larger inventory.

Questions owners ask

Is this score validated against Google rankings?

No. It is an editorial rubric for reviewing usefulness, evidence and clarity. No ranking guarantee or internal Google measurement is implied.

What score is enough to publish?

There is no universal numerical cutoff. Correct material errors, satisfy the intended reader and complete the normal editorial and technical checks. Use the score to prioritize improvements.

How should two reviewers resolve a difference?

Compare the passages and assumptions behind their marks. Agree on the reader and page purpose, then record the reason for any changed score.

Can a short page score well?

Yes. The rubric evaluates whether a page completes its task. Extra length earns no points by itself.

Method

This guide presents ROI.LIVE’s editorial analysis and worked methods. Numerical examples are illustrative unless expressly identified otherwise. No ranking, citation or business outcome is guaranteed.

Substantively revised September 8, 2026. Definitions, calculations, sources and internal destinations were reviewed for this edition.