Homepage Conversion Checklist: Step-by-Step Guide to Turn Visitors into Customers

Getting website traffic but no demos? Check every homepage section against this list: what it must say, copy formulas and real examples. Free.

MobiWire homepage hero with an illustrative heatmap of typical attention: the most on the primary button and the nav button, then the headline
Summary
Share
JaroUpdated

Hero

0/7

The first screen. It's judged at a glance, so it must say what this is, who it's for and why it matters before anyone reads on, and cost the visitor nothing.

MobiWire's hero: the promise, the nav's own button with a lighter second action beside it, and proof right there.
  • TagPut a short tag above the headline

    1. Pick one kind below, 2 to 5 words.
    2. Set it smaller than the headline.
    Who it's forFor product teams
    What it isWebsite as a service
    Micro proofTrusted by 2,400+ teams

    Why?A label tells people at a glance they're in the right place.

  • H1Pick one headline strategy and keep it short

    1. Pick a strategy (tabs below).
    2. Aim for about 10 words, in customers' words.
    3. Run a 5-second test with 5 people.

    When buyers know the problem and want it gone: the result they get, and the pain they no longer have.

    Pattern{Outcome} without {pain}
    BeforeThe all-in-one platform for modern teams
    AfterClose your books in 2 days, not 2 weeks

    Why?You have a few seconds to make the value clear before people leave.

  • SubheadlineBack the headline with one line under it

    1. Pick a strategy (tabs below).
    2. Keep it to one short line, right under the headline.
    3. Aim for grade 7 in Hemingway Editor.

    When the headline is a claim or a USP: say what it is and for whom.

    Pattern{Category} for {persona} in {niche}
    ExampleODM and JDM partner for US hardware brands, from concept to production.

    Why?Plain copy converts about twice as well as professional-level copy.

  • Message matchMatch the headline to the ad or search people came from

    1. List your top ads and searches.
    2. Reuse their words in the headline or tag.
    3. Give a big campaign its own landing page.
    AdClose your books in 2 days
    HeadlineClose your books in 2 days, not 2 weeks

    Why?Google Ads scores landing pages on relevance to the ad, and so do visitors.

  • CTAOne primary button on the first screen

    1. Put it under the subheadline, in the nav's wording.
    2. Add a lighter second action beside it.
    3. Say what happens next under it.
    4. Complex or expensive offer: explain it before the button.
    Primary pattern[Action verb] + [outcome or free qualifier]: Book a 15-minute demo
    Secondary pattern[Low-commitment verb] + [what they'll see]: See how it works (2 min)

    Why?57% of viewing time is spent on the first screen.

  • Social proofPut one line of proof next to the buttons

    1. Pick a couple of items (tabs below).
    2. One line beside or under the buttons.

    A volume or result that proves scale. MobiWire's, in the screenshot.

    Pattern{Number} {what it counts}
    Example36M devices shipped in 2025 · 1,500+ in-house engineers

    Why?Doubt is highest at the moment of commitment.

  • Hero visualShow the real product, not an illustration

    1. Use a real screen or short video of the product.
    2. Load it eagerly, as AVIF at 2x.

    Why?People judge how a page looks in 50 ms, and skip graphics that look like ads.

Social proof

0/6

Straight under the hero, before any feature: can I trust these people? Each kind of proof answers a different doubt, so use the ones your buyer has.

Numbers, certifications and customer logos, straight after the hero, before any feature.
  • LabelOpen the section with one line and a number

    1. Write a number, then who trusts you.
    2. Put it above the logos and numbers.
    Pattern{Number} + {who trusts you}
    ExampleTrusted by 2,400+ teams including Stripe, Notion and Linear.

    Why?A line on top says what the proof adds up to.

  • NumbersShow a few numbers that prove the promise

    1. Pick the kind that backs your headline (tabs below).
    2. Number first, then what it measures.

    How much you've done: customers, units, countries.

    Examples36M devices shipped · 47 countries · 2,400 teams
    Pattern{Big number} + {what it measures}

    Why?Specific numbers read as fact; round claims read as marketing.

  • LogosShow a handful of logos your buyer recognizes

    1. Pick the kind your buyer cares about (tabs below).
    2. About 4 to 8 per row, still, in one color.

    Who pays you. The strongest kind: companies like the visitor's. MobiWire's Select customers row.

    LabelSelect customers · 5 to 10 year partnerships

    Why?A logo they know borrows its trust; an unknown one adds nothing.

  • CertificationsShow the certifications your buyer's procurement asks for

    1. Show only what comes up in deals.
    2. Use the official badge, linked to proof.

    Manufacturing and process standards. MobiWire's ISO and IATF row.

    ExamplesISO 9001 · ISO 13485 · IATF 16949

    Why?A missing certificate can end a deal before the first call.

  • StarsShow your rating with its source and count

    1. Use the site your buyers read (tabs below).
    2. Show stars, rating, source and count.
    3. Don't fear a 4.5: a perfect 5.0 is trusted less.

    Where B2B software is compared.

    Example★★★★★ 4.9 on G2 from 380 reviews
    AlsoCapterra · TrustRadius · Product Hunt
    Pattern{Rating} on {source} from {count} reviews

    Why?A rating with its source and count can be checked; bare stars can't.

  • TrustShow there are real people behind the site

    1. Show the team or founders with photos.
    2. Put a real address and contact in the footer.
    3. Link a privacy page, and a security page if you hold customer data.

    Why?People check that a company is real before they trust it.

Problem

0/1

Do they understand my situation? Name the pain before the product, in the words customers use.

One blunt title, then four pains in the buyer's words, each answered.
  • Pain pointsName 2 or 3 pains in your customers' words, each with a number

    1. Copy customers' words from calls and reviews.
    2. Title the section with what's broken.
    3. Pick 2 or 3 pains, each with a number.
    Section title: {direct statement about the broken status quo}Sales forecasting is still guesswork.
    Stat: {number} + {unit}12 hrs/week
    Headline: {specific frustration in their words}You're updating the same spreadsheet for the third time today.
    Description: {what it costs} + {why it stays broken}Your team copies data between 6 tools. That's 12 hours a week of work a script could do.

    Why?40 to 60% of B2B deals end in no decision, so show the cost of waiting.

Solution

0/1

How does this fix my problem? One block per pain, the benefit first.

Each block a capability with the benefit first and something real to look at.
  • FeaturesOne feature block per pain, benefit first

    1. About 3 blocks, one per pain, same order.
    2. Title, one paragraph, a real screenshot.
    3. B2B: integration logos after it.

    When the problem section was strong: answer it head-on.

    Section titleHow {product} fixes this.
    Block title: {feature}: {benefit in their language}Auto-sync: your CRM updates itself.
    Paragraph: {what it does} + {why it matters} + {proof}Connect once, never copy-paste again. Teams save 12 hours a week on average.

    Why?Three claims persuade best; a fourth makes people skeptical.

How it works

0/1

Is this going to be hard? Show the way in as a few short steps.

Numbered steps with how long each takes, the first one free.
  • StepsShow a few numbered steps, the first one free and easy

    1. About 3 steps, each one action.
    2. Make step 1 free and easy.
    3. Close with the time to first result.

    When the steps are simple: say so.

    Section titleThree steps. That's it.
    Step: {number}. {action verb} + {what happens} + ({time})1. Answer a few questions (5 min).
    Closing: {timeframe} + {outcome} + {qualifier}Most teams are up and running in under 10 minutes.

    Why?A short numbered path makes starting feel small.

Testimonials

0/2

Did it work for someone like me? Named people, with a number.

Schema Flow's testimonials: named people, in their words, verified on Trustpilot.
  • TestimonialsShow a few named testimonials, each with a number

    1. Pick customers like your buyer.
    2. Get a number and their full name.
    3. Lead with the result.

    The default: a few cards side by side. Schema Flow's, in the screenshot, are verified on Trustpilot.

    Pattern"{Specific result}" Name, Title, Company
    Example"We cut onboarding from 3 weeks to 2 days." Sarah Kim, VP Ops, Stripe
    Metric: {number} + {what improved} + {timeframe}+247% demo bookings in 90 days

    Why?A named peer with a result proves what an anonymous quote can't.

  • Social proofPut one proof point next to every button

    1. Add one line of proof under each button.

    Why?Doubt is highest at the moment of the click.

Pricing

0/1

Can I afford it, and is it worth it? The price on the page, the choice made easy.

Schema Flow's pricing: two plans side by side, one marked Most popular, the price on the page.
  • PricingShow the price, with one plan marked as recommended

    1. Pick how you sell (tabs below).
    2. Test: can someone pick a plan in 10 seconds?

    2 or 3 plans side by side, the price large, one marked Most popular, only what differs listed, and a risk reversal under them. Schema Flow's, in the screenshot.

    Section title: {price}. {promise}.$49/seat/mo. Everything included.
    Plan name: {name} ({who it's for})Pro (for growing teams)
    Risk reversal: {free/safe}. {no downside}. {easy exit}.14-day free trial. No credit card required. Cancel anytime.

    Why?A missing price reads as expensive; a marked plan makes choosing easy.

FAQ

0/1

What's the catch? The questions sales hears, answered before the call.

The questions buyers actually ask, then a way to ask the rest.
  • FAQAnswer the real objections, yes or no first

    1. Collect questions from sales calls and chat.
    2. Answer the most-asked ones, yes or no first.
    3. End with a Still have a question? prompt and the button.
    Answer: {yes or no} + {one sentence} + {link or proof}Can I cancel anytime? Yes. No contracts, no fees. Your data exports in one click.
    ExampleIs my data secure? Yes. SOC 2 compliant, encrypted at rest and in transit.

    Why?Every question answered on the page is one less reason to leave.

Final call to action

0/3

One last push at the bottom, for everyone who read this far.

MobiWire's last push: what to send, an honest answer within 48 hours, and proof under the button.
  • Final CTARestate the promise in new words, with the same button

    1. Pick a heading pattern (tabs below).
    2. Add a one-line risk reversal.
    3. Use the hero's exact button.

    The plain invitation.

    PatternReady to {outcome}?: Ready to launch your website?
    Subline: {risk reversal, one line}Free trial. No credit card. Cancel anytime.

    Why?People who reach the bottom are interested; don't make them scroll back.

  • Lead magnetOffer a lighter step for people not ready

    1. Offer a guide, calculator or newsletter.
    2. Ask for an email only.

    Why?Most visitors aren't ready to talk; a smaller ask keeps them.

  • FormKeep the form to the fields you use

    1. Cut every field nobody uses.
    2. Keep name, email and one question.
    3. Show the calendar right after.

    Why?Unused fields cost sign-ups; dropping one you need costs lead quality.

Lighthouse: speed, accessibility, SEO

0/5

Across every section, the technical side Google measures. Run Lighthouse on mobile and work down its four scores.

Lighthouse on mobile for 3day.studio: all four scores green, and the metrics behind them.
  • PerformancePass Core Web Vitals for real users

    1. Check real-user data in PageSpeed Insights (the top section), not only the lab score.
    2. Fix LCP, then INP, then CLS.
    • LCP (largest element loaded): 2.5 s or less
    • INP (response to a tap or click): 200 ms or less
    • CLS (layout shift): 0.1 or less
    • Measured on 75% of real visits, not one test

    Why?B2B sites that load in 1 second convert about 3x better than at 5.

  • AccessibilityGet Accessibility to 90+: tap targets, contrast, labels

    1. Fix every red item in Accessibility.
    2. Make buttons 44px tall, text 4.5:1 contrast.
    3. Honor reduced motion.

    Why?Aim for 44 to 48px; under 24px fails WCAG.

  • KeyboardUse the page with only a keyboard

    1. Press Tab from the top of the page to the form.
    2. Every link and button gets a visible focus outline.
    3. You can send the form without a mouse.

    Why?If they can't reach the button by keyboard, they can't convert.

  • SEOGet Best Practices and SEO to 100

    1. Fix every red item in both sections.
    2. Check title, description and indexing.

    Why?Both are right-or-wrong lists, so 100 is realistic.

  • Scroll depthCheck how far people scroll

    1. Open PostHog's scroll heatmap.
    2. Move key proof above the 50% line.

    Why?74% of viewing time is in the first two screens.

Measure every change

0/3

A change you can't measure is a guess. Track the result before and after, one change at a time. No tracking yet? Our free website analytics guide sets it up step by step.

  • BaselineRecord the conversion rate before you change anything

    1. Track the lead as an event in Google Analytics or PostHog.
    2. Note visitors, leads and rate for the last 4 weeks.
    Set up the tracking with our step-by-step website analytics guide

    Why?Without a before, you can't tell whether the change helped.

  • Before and afterCompare equal periods, one change at a time

    1. Change one section.
    2. Compare 4 weeks after with the 4 weeks before.
    3. Keep it only if leads went up.
    Set up the tracking with our step-by-step website analytics guide

    Why?Two changes at once can cancel out, and seasons skew short periods.

  • A/B testA/B test only with enough traffic

    1. Work out the sample size you need before you start.
    2. Run full weeks and don't stop early.
    3. Low traffic: test with 5 users instead.

    Why?Fewer than a third of tested ideas actually improve the result.

Why does a website get traffic but no leads?

Because a visitor has to pass several steps before they get in touch, and at every step some of them drop off. The steps multiply, so a site can lose almost everyone without one step looking terrible. Here’s a site with 2,000 visitors a month, with the steps you can track in Google Analytics or PostHog:

StepDrop-offConversion from start
Visitors0%100% (2,000)
Clicked the main button−90%10% (200)
Started the form−50%5% (100)
Sent the form−40%3% (60)
Booked a demo−50%1.5% (30)
Booked demos2,000 × 10% × 50% × 60% × 50% = 30
Example numbers. Pick the funnel closest to how you sell, then compare it with your own.

Each drop has a usual reason. Few clicks on the main button means the page didn’t say what you do or why to trust you. People who start the form and quit hit too many fields or a reason to doubt. Booked calls that never happen mean the next step felt like a commitment.

So there are two ways to get more leads from the same traffic: fewer steps, and fewer people dropping off at each one. Cut a step that isn’t needed, like a separate contact page when the form could sit on the case study, and nobody can drop off there. Then fix the step with the biggest drop-off.

What happens when you fix one step?

It depends which. In the B2B example, make a strong step better, sending the form from 60% to 70%, and you get 35 booked demos instead of 30. Fix the weakest, the main button from 10% to 20% of visitors, and you get 60. Same traffic, twice the demos, from one change.

That’s why a full redesign often changes nothing: the one broken step is still broken. The job is to find it.

Is your conversion rate actually bad?

Maybe not. It depends on which pages you count:

What you measure Typical rate
Whole site, B2B SaaS, every visitor counted about 1% to 2.5%
Campaign landing pages, SaaS median 3.8% (Unbounce, 41,000+ pages)
Pricing and demo pages well above the site average

A site at 0.8% overall may be fine if most traffic reads the blog. Measure the demo and pricing pages on their own, and compare against your own last quarter first.

Is the traffic real, and is it the right traffic?

Check this before touching a page. Bots and AI crawlers inflate visitor counts, and how-to readers came to learn, not to buy. Look at leads by source and by landing page. If most visits land on blog posts, the homepage isn’t your problem.

Do visitors understand what you do in five seconds?

Show a stranger your first screen for five seconds, then ask what you sell and for whom. “The future of data” fails that test. “Sync Salesforce and HubSpot in real time, without code” passes. Then show the product itself: a screenshot, a short video or a clickable tour beats any illustration.

Do they believe it?

Tech buyers check three things before they talk to anyone:

  • Proof: a customer they recognize and a result with a number, near the button.
  • Fit: the integrations they use, and whether it works with their stack.
  • Risk: a security page or SOC 2 status, and a price signal, even “from $X a month”.

Most buyers have a favorite before they ever contact a vendor, so this proof has to be on the site, not in the sales call.

Is the next step easy?

Ask for three form fields, not nine. Show a calendar right after the form, so people book on the spot. For visitors not ready to talk, offer a lighter step: a free trial if the product is self-serve, a recorded demo if it’s sales-led. And reply within minutes; at a founder-led startup that often means the founder’s own inbox.

The homepage conversion checklist

  • A headline that names the result for a specific buyer
  • One line under it that says what the product is
  • A product screenshot or short video above the fold
  • A customer they recognize, with a number, next to the button
  • Integrations, security and a price signal visible
  • One main button, a three-field form, a calendar right after it
  • A lighter step for visitors not ready to talk
  • The form tracked as a goal, by traffic source

Should you A/B test?

Usually not yet. At a 3% rate, spotting a 20% improvement takes about 13,000 visitors for each version. With less traffic, make one bold change, watch five people use the page, and compare the weeks before and after.

What should you fix first?

The weakest step, and nothing else until it’s fixed. Every new visitor you buy meets the same leak.

Questions founders ask

01

Why is my website getting traffic but no leads?

Because one link between the visit and the request breaks: the wrong people arrive, they don't get what you do in seconds, they don't believe it, or the next step asks too much. The links multiply, so the weakest one sets your result. On B2B sites two are easy to miss: proof (nothing near the button shows it worked for someone like them) and fit (most traffic reads how-to posts and came to learn, not to buy).

02

What is a good conversion rate for a B2B website?

It depends on which pages you count. Whole-site visitor-to-lead rates for B2B SaaS sit around 1% to 2.5%, because every blog reader is counted. Campaign landing pages do better: the median SaaS landing page converts 3.8% in Unbounce's benchmark of 41,000+ pages. Measure your demo and pricing pages on their own, compare against your own last quarter first, and count only sent forms and booked calls as leads.

03

How do I calculate my website conversion rate?

Divide leads by visitors over the same period and multiply by 100: 20 leads from 5,000 visitors a month is 0.4%. Count a sent form or a booked call as a lead, not a newsletter signup, and use visitors, not page views. With few leads, read it with care: at 5,000 visitors, one lead more or less moves the rate by 0.02 points.

04

What should a B2B conversion checklist include?

One check per link a visitor has to pass. Fit: where the traffic comes from. Clarity: a headline that names the result for a specific buyer, and the main content on screen within 2.5 seconds on a phone. Proof: a case study with a number, named quotes, reviews, your security status and a price signal near the button. Ease: one main button, three form fields, a fast reply and a smaller step. Then tracking of the main goal. Fix the weakest link first instead of working down the list.

05

How do I get more leads from my website without more traffic?

Fix the weakest link before you buy traffic, because every new visitor meets the same leak. Run a five-second test on the first screen, put proof with a number next to the main button, cut the form to three fields and reply within minutes. In the Lead Response Management study, the odds of qualifying a lead were 21 times lower at 30 minutes than at 5.

06

Why is my bounce rate so high?

Usually because the page doesn't match why people came: a how-to reader landing on a sales page, an ad promising something the headline doesn't say, or a page that loads slowly on a phone. GA4 now reports engagement rate instead, the share of visits that stay 10 seconds, view two pages or convert. Low engagement on blog posts is normal; on your homepage and pricing page it points at fit or clarity.

07

Should I A/B test with low traffic?

Usually not. At 3% conversion, detecting a 20% lift at 95% significance and 80% power takes about 13,000 visitors per version, and at 0.4% about 100,000. Below that, make bold changes, watch five people use the page (five users find about 85% of usability problems, in Nielsen's model) and compare the weeks before and after.

08

What should a SaaS homepage include?

A headline that names the result for a specific buyer, a line under it that says what the product is, one main button, proof near that button (logos, a case study with a number, named quotes), a price signal, and a smaller step such as a guide or a recorded demo for visitors who aren't ready to talk.

Sources (14)

  1. Unbounce: SaaS landing page conversion rateThe median SaaS landing page converts 3.8%
  2. Unbounce: Conversion Benchmark Report 202441,000+ landing pages, a median of 6.6% across industries
  3. Ruler Analytics: Average conversion rate by industryMay 2026, 110M+ sessions: software 7.6%, professional services 6.1%
  4. Nature: Users judge websites in 50 milliseconds (Lindgaard et al.)A first impression forms in 50 milliseconds and colors the rest of the visit, so the first screen has to land at a glance
  5. web.dev: Web VitalsGood Largest Contentful Paint is 2.5 seconds or less at the 75th percentile
  6. Portent: Site speed is (still) impacting your conversion rate2022, 20 sites: B2B pages at 1 second convert about 3x those at 5 seconds
  7. 6sense: Buyer Experience Report 2025Vendor-run survey of about 4,000 buyers: first contact at 61% of the journey, the favorite wins 80%
  8. Clutch survey via Agility PR: Trust issues in the B2B sales funnel2020, 451 B2B buyers: transparent pricing 25%, case studies and testimonials 23%
  9. Lead Response Management study2007, six companies: qualifying odds 21x lower at 30 minutes than at 5
  10. Nielsen Norman Group: Why you only need to test with 5 usersFive users find about 85% of usability problems
  11. Evan Miller: Sample Size CalculatorThe two-proportion formula behind the A/B test numbers
  12. Contentsquare: Conversion rate optimization guideThe CRO process for teams with traffic: identify, analyze, gather data, hypothesize, test, iterate
  13. WiderFunnel: The LIFT modelThe lineage of the links: value, relevance, clarity, anxiety, distraction, urgency
  14. Wikipedia: Conversion rate optimizationThe definition and the conversion rate formula
Summary
Share
JaroUpdated
ToolsWhat is a website

Get a free website roadmap in 3 days

  • Tech stack
  • Timeline
  • Migration
  • SEO
  • Budget

Tell us about your company on a 15-minute call. Three days later you get a free roadmap showing how we would approach your website, brand and ads, based on 10 years and 60+ projects. It's yours to keep, so build it with us or take it anywhere.

  1. 01now
    Pick a time
  2. 0215 min
    Discovery call
  3. 03in 3 days
    Your roadmap
Roadmap preview