Free tool and Claude skill

The Graded Canvas

Fill your value proposition canvas from evidence you already have. Grade every outcome the way Tony Ulwick does, then split each low score into a message problem or a product problem.

Run it here

Five steps in your browser. Every outcome needs a source, and anything from memory gets counted. Nothing you type leaves this page; it's saved here, so a refresh won't lose your work.

Start below ↓

Have the calls? Use Claude

Paste your call notes, interviews, reviews, case studies and win/loss notes. The skill runs all four passes and cites a source on every line.

Download skill file →

Step 1 of 5

Name what they'd use instead

Before you grade anything, name what a buyer would do if you didn't exist. Every grade later is against one of these.

Used only to check your outcomes for your own product's name.

What buyers use now

Include anything that isn't a product: a spreadsheet, someone doing it by hand, doing nothing.

Step 2 of 5

Write the jobs as outcomes

Write each job as an outcome, in the buyer's words, with no product in it. It should stay true even if you went out of business tomorrow.

  • DirectionMinimize
  • Metricthe time it takes
  • Objectto rebuild the day's route
  • Contextwhen a technician calls in sick

Step 3 of 5

Grade each one twice

How much buyers care, and how well the alternative they use today already delivers it. High, medium or low. No scores.

Step 4 of 5

Split the low scores

For each outcome that matters and scores low, one question. Answer it from won deals, reviews and case studies, the things a buyer could check.

Step 5 of 5

Where the white space is

The Claude skill

Have the calls? Let Claude do the first pass from your sources.

Download skill file →

Why it works this way

Most value proposition canvases get filled in from memory at an offsite, then the photo gets treated as research. This tool runs the passes from the newsletter post "Your positioning is a guess": name what buyers would use instead, write the jobs as outcomes, grade each one twice, and split every low score into a perception gap or a capability gap. Every outcome needs a source, and anything from memory gets counted.

Read the post →

Built on Strategyzer's Value Proposition Canvas, Tony Ulwick's outcome statements and grading, Bob Moesta and Chris Spiek's four forces, and April Dunford's competitive alternatives. The four-pass assembly is my process.