tmj-90 Improve page conversion rates
A page that converts well is one where the visitor's question ("is this for me?") is answered before they think to ask it. Diagnose in order of impact; fix the biggest lever first.
Analysis framework (work top-to-bottom)
1. Value proposition clarity (highest impact)
Can a first-time visitor understand what this is and why they should care within 5 seconds?
Signs it's broken:
- Headline describes the product, not the outcome for the user.
- Subheadline is a marketing tagline, not a one-sentence expansion of the benefit.
- First CTA is below the fold on desktop (1440px).
2. Headline effectiveness
- Specific > clever. "Cut your Shopify return rate by 23%" beats "Returns, reinvented."
- Must match the ad/email/referral that sent the visitor — message mismatch kills conversions.
- A/B test at ≥ 100 visitors per variant before concluding.
3. CTA friction
- One primary action per page. Two CTAs = zero conversions.
- CTA label states what happens next: "Start free trial" not "Submit".
- Reduce fields: every additional form field costs ~10% completion. Ask for email only until you need more.
4. Social proof specificity
- Named testimonials (first name + company) outperform anonymous quotes.
- Numbers anchor trust: "4,200 teams" > "thousands of teams".
- Logo strips: name-recognisable logos; max 6 (more = noise).
5. Objection handling
Map the 5 core objections for the page's conversion goal (sign-up / purchase / demo). Confirm each is addressed somewhere on the page. See landing-page-generator references for the standard objection list.
6. Mobile and speed
- 60%+ of traffic is mobile for most marketing pages. Check layout at 375px.
- LCP > 2.5s is a conversion killer — check Lighthouse and fix the hero image.
Steps
- Identify page type and primary conversion goal. Single goal per analysis — don't try to optimise for multiple actions simultaneously.
- Run the 5-second test. Read only the above-the-fold content. Can you answer: what is this, who is it for, what happens when I click? If not — fix the hero first, everything else is secondary.
- Work the framework in order. Value prop → headline → CTA → social proof → objections → mobile/speed. Don't skip to lower items until the higher ones are solid.
- Produce prioritised recommendations. For each finding: what's broken, why it hurts, what to change, estimated impact (high/medium/low).
- Verify. After implementing: re-run Lighthouse; confirm primary CTA is above the fold; confirm the 5-second test passes.
Review checklist
- Primary CTA above the fold — visible on 1440px without scrolling.
- Headline states user outcome — not product name or tagline.
- One primary action — secondary CTAs are visually de-emphasised.
- Social proof specific — names, companies, numbers.
- 5 objections addressed — mapped and answered somewhere on the page.
- Mobile layout passes — no overflow at 375px; primary CTA still above the fold.
- LCP < 2.5s — Lighthouse desktop.
Rules
- Fix the hero (value prop, headline, CTA) before anything else — it's the highest-leverage real estate.
- Never recommend A/B testing something that obviously needs fixing — fix it first, then test variations.
- Never recommend adding more content to a page as a first CRO move — remove friction first.