Max MorrishThe Playbook

What, chapter six

Pages that leak

ledgerline.example Mock page Accounting, reimagined for what's next. Powerful tools for every kind of business. Watch the tour Our story, slide 1 of 3 1,200 contractors close their month with Ledgerline. Book a demo Pricing: contact sales Book a demo Meta description Job costing and invoicing in one ledger, for contractors. 54 in 100 drift off before they get the offer 9 pulled away by the carousel and a second button 14 leave over doubts nine fields, no price 23 reach the button
Visitors who reach the button 23 in 100

Drivers

Value proposition
Relevance
Clarity
Urgency

Inhibitors

Anxiety
Distraction
A mock landing page for Ledgerline, an invented company. Visitors come in at the top as light. On this version, 54 in 100 drift off before they get the offer, 14 leave over hidden pricing and a nine-field form, 9 are pulled off by a carousel and a second button, and 23 reach the button. Turn the six dials and watch where the light goes. The numbers are an illustration from the teaching model below, not a benchmark.

The donation counter... is the most important asset on the entire site. Treat it as the headline, not a stat in a grid.

Anything that doesn't serve those four outcomes is decoration.

From my brief for the UVA Food Assist web redesign, 22 May 2026.

The idea

I think a page converts when the visitor understands the offer, believes it, and has nothing pulling them away. So I score a page on six things: value, relevance, clarity, anxiety, distraction and urgency. Then I fix the cheapest problem that does the most damage, not the one that's most fun to redesign.

Most of the conversion problems I've looked at weren't weak offers. They were strong offers in the wrong place: the best line on the page sitting in the meta description, the proof number halfway down, three equal buttons where one should lead.

How this model works (a teaching model, not a benchmark)

Each dial runs from 0 to 10. The four drivers decide how many visitors get the offer at all, with value counted twice because nothing else matters if the offer is weak. The two inhibitors then take their share of whoever is left.

drive = (2 × value + relevance + clarity + urgency) ÷ 50
reach the button = drive × (1 − 0.45 × anxiety ÷ 10) × (1 − 0.40 × distraction ÷ 10)

The weights are mine and they are round on purpose. Real pages don't follow a formula, so I'd use this to see which way each factor pushes, and a test on real traffic to find out by how much.

What to test first

Once the page is scored, every fix competes for the same week. I rank them with P.I.E.: potential (how much it could move the number), importance (how much traffic and money go through that part of the page) and ease (how cheap it is to ship). The score is the plain average of the three.

Easier to ship Harder More potential Test these first Say what it is Proof in the hero Four fields One button Show pricing Cut the carousel Add a deadline

Test this first

  1. Rewrite the H1 to say what Ledgerline is 8.7
  2. Move the proof number into the hero 8.0
  3. Cut the form to four fields 7.3
  4. Make one button the primary one 7.3
  5. Show pricing on the page 6.3
  6. Remove the carousel 6.0
  7. Add a real deadline 4.7
Seven fixes for the Ledgerline page, placed where I'd put them: ease across, potential up. Drag a fix, or focus it and use the arrow keys, and set how important each one is. The list re-ranks by (potential + importance + ease) ÷ 3. Rewriting the H1 wins here because it is nearly free and the page is useless while nobody can tell what it sells.

Where I used it

  • Recommendations for a small merchandise start-up, in a five-person group audit. My part was the recommendations, built on the group's findings: fix the foundation, make the difference visible, grow into fans and partners. The first step came straight from the speed audit, which scored PageSpeed 62 with 382 KiB of image savings on the table. Group work, my section
  • The UVA Food Assist web redesign. My part was the website. I made the pounds-of-food counter the hero instead of one stat in a grid, and gave the page one primary call to action with the other ways in set below it. Group work, my section

What I'd do now

Write the teardown down so anyone can run it in half an hour, then run it here on a mock product page for Kiln, an invented coffee roaster.

  1. Read the page the way a visitor arrives: from the ad or the search result that sent them, not from the homepage.
  2. Say the offer in one sentence. If I can't, the page can't either, and that is the clarity score.
  3. Find the best line and the best proof wherever they are, including the meta description and the bottom of the page, and ask why they aren't at the top.
  4. Count the choices above the fold. Every extra one adds decision time (Hick's law), so one button should lead and look like it.
  5. Check the button's size and place. A bigger, closer target is faster to hit (Fitts's law), on a phone most of all.
  6. List every doubt the visitor would have at the moment of the click: price, commitment, what happens to their email. Each one unanswered is anxiety.
  7. Score the six factors out of 10, write the one-line reason next to each, and turn the low scores into fixes.
  8. Rank the fixes with P.I.E. and test the top one before touching the rest.
kilncoffee.example, mock page bag photo Our coffee Morning Roast, 250 g Notes of cocoa, orange and brown sugar. $16 WholeFilterFine SubscribeBuy once Join our list! pop-up after 3 seconds Value, relevance:no roast date Anxiety: shippingshows at checkout Clarity: which grindfor which brewer? Distraction: thepop-up, two equalbuttons
Kiln's mock product page, with the problems marked. The brightest mark is the one I'd fix first.
My scores for the Kiln page, out of 10. Kiln and its page are invented; the scores are my judgement, not data.
FactorScoreWhat I sawFix
Value6The tasting notes are good, under a heading that says "Our coffee".Lead with the coffee's name and the notes.
Relevance4The ad says roasted this week. The page never says when it was roasted.Put the roast date by the price.
Clarity5Three grind buttons and no word on which suits which brewer.Name the brewer on each button.
Urgency3Nothing says when the next roast ships."Roasted Tuesday, shipped Wednesday" is true and urgent.
Anxiety6Shipping only appears at checkout.Show the shipping cost beside the price.
Distraction7A sign-up pop-up after three seconds, and two buttons of equal weight.Hold the pop-up until they scroll; make "Buy once" the quieter button.

The roast date wins. It's one line of copy, it answers the promise the ad made, and it fixes two scores at once. I'd test that before the pop-up, which is a bigger argument with whoever owns the email list.

LIFT and P.I.E. are Chris Goward's, from WiderFunnel. Hick's law (Hick, 1952; Hyman, 1953) and Fitts's law (Fitts, 1954). Conversion-rate optimisation as taught in a digital marketing program, 2026.