> ## Content Index > Fetch the complete content index at: https://conversion.club/llms.txt > Use this file to discover other available public pages before exploring further. # Where Do You Start When Everything on Your Site Seems Worth Optimizing? - URL: https://conversion.club/playbook/where-do-you-start-when-everything-on-your-site-seems-worth-optimizing/ - Published: 2026-10-01T20:17:15.000Z - Updated: 2026-10-19T20:26:06.000Z - Description: With what's broken, then with the step where you lose the most people, as close to payment as possible. The homepage rarely comes first, even if it's the page everyone looks at in meetings. And the first thing broken is often measurement. - Author: Jeremie Werner - Tags: Improve conversion, #playbook, CRO Start with what's broken, then with the step where you lose the most people, as close to payment as possible. The homepage rarely comes first, even if it's the page everyone looks at in meetings. --- ## The benchmark Three numbers per step of the journey are enough: how many visitors go through it, how many leave without continuing, and what a visit that continues is worth. Count them on the product page, the cart, shipping and payment. Read priority in volume lost, not in percentages: losing 40% of 500 people costs less than losing 15% of 20,000. Before you read those numbers, check that they exist. In a recent audit, GA4 showed a cart-to-payment rate of 150% on tablet, 165% on direct traffic and 200% on a subdomain. In another, GA4 reported [twice the revenue of Shopify](https://conversion.club/optimizing-e-commerce-checkout-the-complete-audit/), because two tools were each sending their own purchase event. In both cases, the pre-audit was thrown out the day the problem was found, and the quarter's priority became a [source of truth](https://conversion.club/playbook/shopify-or-ga4-which-number-should-you-show-leadership/). --- ## The rule Fixing comes before improving. A payment method throwing errors, a button you can't read on mobile, a promo code that won't apply: all of these can be fixed without a test and pay off right away. At an accessories brand the authors work with, the first batch of fixes took a matter of hours: 44-pixel tap targets, one hour; payment logos in the cart, thirty minutes; fourteen filters cut down to seven, two hours. And once everything is fixed, there's still work left. At a craft brand, known customers converted at 3% and new visits at 0.3%, with no broken page anywhere. At every step, a little proof and a little guidance were missing. --- ## The steps 1. The buying journey. Go through it from start to finish on a phone, all the way to a real payment. 2. Measurement. Compare GA4 orders with Shopify orders before you read any rate. 3. The defects. [List what doesn't work](https://conversion.club/e-commerce-cro-checklist-194-checkpoints-to-audit-your-store/) and fix it this week, without a test. 4. The losses. Quantify them step by step, in number of people. 5. The step. Pick just one and stick with it for [a quarter](https://conversion.club/playbook/how-many-sessions-do-you-need-to-run-an-a-b-test/). A sentence to say to your leadership, to reuse with your own numbers: > "We're starting with the product page on mobile, because three out of four visits go through it. The homepage comes after, and we'll measure it before we touch it." --- ## The trap In July, an executive summed up what he expected at a delivery meeting: > "Honestly, for me the two priorities are a much clearer homepage with a clickable link." His management tool and his two stores weren't working yet, and the authors said yes to requests like that more than once before learning to answer differently: first what breaks, then what shows. If your ads send traffic to product pages, half of your visitors will never see the homepage. Look at your real landing pages before you start work on the homepage. #### Sources - [Understanding Success Criterion 2.5.5: Target Size (W3C)](https://www.w3.org/WAI/WCAG21/Understanding/target-size.html?ref=conversion.club)