A/B Test Ideas for Websites: 30 Tests Worth Running Right Now
How to Use This A/B Testing List
⠀
These a/b test ideas are organized by page type and element category. The list is designed to be prioritized, not executed in order. Before running any test, evaluate it against three criteria: how much traffic does this element receive, how significant is the potential conversion impact, and how easy is the test to implement?
High-traffic, high-impact, low-implementation-effort tests come first. Low-traffic, minor-impact tests, regardless of their theoretical appeal, should come last.
A key principle from Blakfy's CRO practice: run tests on your highest-traffic pages first. A 5% conversion improvement on a page receiving 10,000 monthly visitors delivers ten times the value of the same improvement on a page with 1,000 monthly visitors. Traffic volume determines the revenue impact of every test you run.
⠀
Hero Section and Landing Page Tests
⠀
Test 1: Headline framing — outcome vs. problem. Pit a headline that states the customer's desired outcome against one that names the problem being solved. Example A: "Get More Leads from Your Website." Example B: "Stop Losing Customers to a Confusing Website." These consistently produce different responses from different audience segments.
Test 2: Hero layout — two-column vs. centered. Compare a two-column hero (text left, image right) against a centered layout with a full-width background. Impact on visual engagement and conversion varies significantly by industry and audience.
Test 3: Hero image — product screenshot vs. lifestyle photography vs. no image. For SaaS products, a product UI screenshot often outperforms lifestyle photography by demonstrating the product before a visitor commits to learning more. Test this assumption.
Test 4: Single CTA vs. dual CTA. Compare a page with only a primary CTA against one with both a primary CTA ("Book a Call") and a secondary CTA ("See How It Works"). Dual CTAs can capture visitors at different stages of readiness.
Test 5: Social proof placement — hero vs. immediately below hero. Test placing your most persuasive trust signal (client logos, aggregate rating) within the hero section itself against placing it in the first section below the fold.
Test 6: Long landing page vs. short landing page. For cold traffic, compare a long page with full objection handling against a short page that leads directly to the CTA. Results vary dramatically by traffic temperature and offer complexity.
Test 7: Video hero vs. static image hero. An explainer video in the hero section can dramatically increase engagement and conversion for complex products. Test with and without video, measuring conversion rate and time on page.
⠀
CTA Button Tests
⠀
Test 8: CTA copy — action-focused vs. value-focused. "Submit" vs. "Get My Free Audit." "Sign Up" vs. "Start My Free Trial." Value-focused copy consistently outperforms action-focused copy, but the degree of improvement is worth measuring.
Test 9: CTA button color contrast. Test your current button color against a higher-contrast alternative. If your current button blends into the surrounding design, a more contrasting color typically produces significant improvement.
Test 10: First-person CTA language. "Get Your Free Proposal" vs. "Get My Free Proposal." The first-person version typically converts slightly higher by reinforcing personal ownership.
Test 11: CTA size — standard vs. oversized. Test a CTA button 20-30% larger than your current default against the current version. Larger buttons command more visual attention and reduce missed-click rates on mobile.
Test 12: Sticky CTA vs. in-content CTA. For long pages, test a persistent sticky CTA bar against relying only on in-content CTAs. Sticky CTAs ensure the conversion action is always accessible during deep scrolling.
⠀
⠀
⠀
Form Design Tests
⠀
Test 13: Three-field form vs. five-field form. Remove non-essential fields and measure the impact on form completion rate and lead quality. Almost every form reduction test shows higher completion rates; the question is whether lead quality is maintained.
Test 14: Multi-step form vs. single-page form. For forms with five or more fields, compare a multi-step format (breaking fields across two or three steps) against a single-page format. Multi-step forms typically show higher completion rates for longer forms.
Test 15: Submit button copy. "Send Message" vs. "Get My Free Consultation" vs. "Start the Conversation." Test three variants of submit copy and measure both click-through rate and lead quality.
Test 16: Form with trust signals vs. form without. Add a satisfaction guarantee statement, a privacy note ("We never share your details"), or a response time commitment adjacent to your form. Measure the impact on submission rate.
⠀
Pricing Page Tests
⠀
Test 17: Three-tier pricing vs. two-tier pricing. Most pricing pages benefit from three tiers that leverage the "Goldilocks effect" — the middle option appears most reasonable between two extremes. Test two-tier against three-tier to quantify the effect for your specific audience.
Test 18: Monthly vs. annual pricing as default display. Show annual pricing by default (with monthly as an option) vs. monthly pricing by default. Annual pricing defaults often increase annual subscription rates while reducing churn.
Test 19: Highlighted "recommended" tier. Test visually highlighting one pricing tier with a "Most Popular" or "Best Value" badge against equal visual treatment for all tiers. Recommendation signals reduce decision anxiety and typically increase conversion on the highlighted plan.
Test 20: Money-back guarantee prominence. Test placing a prominent money-back guarantee statement near your pricing CTA against burying it in small print. Risk-reduction statements adjacent to conversion points frequently produce meaningful lift.
⠀
Trust and Social Proof Tests
⠀
Test 21: Testimonial format — text vs. star rating + text vs. video. Video testimonials typically outperform text testimonials for high-consideration purchases. Test the format that produces the strongest conversion impact for your audience.
Test 22: Case study placement — in-line vs. linked. Test displaying a brief case study summary (challenge, outcome, quote) directly on the service page against linking to a separate case study page. In-line summaries reduce friction but may provide less depth.
Test 23: Number of social proof elements. Test a minimal social proof section (one strong testimonial, one logo) against a comprehensive section (three testimonials, five logos, review aggregate). Sometimes less is more; sometimes accumulation builds stronger trust.
Test 24: Specific numbers vs. rounded numbers. "Helped 847 businesses" vs. "Helped 800+ businesses." Specific numbers typically read as more credible, but the difference can be tested.
⠀
Checkout and E-commerce Tests
⠀
Test 25: Guest checkout as default vs. account creation as default. Presenting guest checkout as the prominent option (with account creation as secondary) against equal prominence for both. Reducing account creation friction typically improves checkout completion.
Test 26: One-step checkout vs. multi-step. For simple purchases with few required fields, a single-page checkout can outperform multi-step. Test the format that minimizes abandonment for your specific product and purchase complexity.
Test 27: Free shipping threshold display. Test prominently displaying the free shipping threshold on product pages ("Add $15 more for free shipping") against not displaying it. This typically increases average order value while reducing cart abandonment.
⠀
⠀
⠀
Navigation and UX Tests
⠀
Test 28: Navigation with CTA button vs. without. Add a contrasting-color CTA button to your header navigation and test against navigation with text links only. Navigation CTAs provide persistent conversion access from every page.
Test 29: Sticky navigation vs. standard navigation. Test a header that follows visitors as they scroll against one that disappears on scroll. Sticky navigation keeps navigation accessible but slightly reduces usable screen space.
Test 30: Chatbot vs. no chatbot. A chat widget on key landing and product pages can capture visitors who have questions before committing. Test conversion rate and lead quality with and without a chat widget on your highest-traffic conversion pages.
⠀
Running Tests Responsibly
⠀
A/B tests produce reliable data only when run correctly. The most common testing mistakes: ending tests too early (before statistical significance), running too many simultaneous tests (interaction effects confound results), and testing on pages with insufficient traffic (too few conversions to reach significance).
Each test needs a minimum of 300-500 conversions per variant to reach 95% statistical significance. For low-traffic pages, this means tests may need to run for six to eight weeks. Patience is a prerequisite for reliable A/B testing.
⠀
Frequently Asked Questions
⠀
How do you prioritize which A/B tests to run?
Use the PIE framework: Potential (how much conversion improvement is possible), Importance (how much traffic and revenue does this page drive), and Ease (how easy is this test to implement). Tests with high scores across all three criteria go first.
How many A/B tests should you run simultaneously?
Run one test per page at a time. Running multiple tests on the same page creates interaction effects that make it impossible to attribute results to individual changes. Across your site, you can run tests on different pages simultaneously since those tests don't interact.
What do you do when an A/B test shows no difference?
A null result is informative — it tells you that this particular change doesn't affect the behavior you're measuring. Document the result and move to your next hypothesis. Null results also prevent you from making changes you believe will help but actually don't, which is as valuable as positive results.
